Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
U.S. Environmental Protection Agency
Region 10
Seattle, WA
-------
INTRODUCTION
In May 2012, the U.S. Environmental Protection Agency (EPA) publicly released the first external
review draft of An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay,
Alaska for review and public comment. External peer review of the May 2012 draft assessment was
coordinated by Versar, an independent contractor. Versar assembled 12 independent experts to serve
as peer reviewers. The selected reviewers were:
• Mr. David Atkins, Watershed Environmental, LLC (expertise in mining and hydrology)
• Mr. Steve Buckley, WHPacific (expertise in mining and seismology)
• Dr. Courtney Carothers (expertise in indigenous Alaskan cultures)
• Dr. Dennis Dauble, Washington State University (expertise in fisheries biology and wildlife
ecology)
• Dr. Gordon Reeves, USDA Pacific Northwest Research Station (expertise in fisheries biology
and aquatic biology)
• Dr. Charles Slaughter, University of Idaho (expertise in hydrology)
• Dr. John Stednick, Colorado State University (expertise in hydrology and biogeochemistry)
• Dr. Roy Stein, Ohio State University (expertise in fisheries and aquatic biology)
• Dr. William Stubblefield, Oregon State University (expertise in aquatic biology and
ecotoxicology)
• Dr. Dirk van Zyl, University of British Columbia (expertise in mining and biogeochemistry)
• Dr. Phyllis Weber Scannell (expertise in aquatic ecology and ecotoxicology)
• Dr. Paul Whitney (expertise in wildlife ecology and ecotoxicology)
After review of the May 2012 draft was complete, Versar compiled detailed comments from each of
the 12 reviewers and provided a Final Peer Review Meeting Summary Report (hereafter, the peer
review report) to EPA in September 2012. This report is available at
http://www.epa.gov/ncea/pdfs/bristolbav/Final-Peer-Review-Report-Bristol-Bav.pdf. EPA
reviewed these detailed comments and, based on both review comments and comments submitted by
the public, developed an improved and expanded draft assessment.
This second external review draft assessment was released to the public in April 2013. EPA asked the
12 independent experts who peer reviewed the May 2012 draft to evaluate how well the April 2013
draft of the assessment reflected that we fully understood and sufficiently responded to the comments
they provided on the previous draft. This evaluation was limited to the comments and suggestions
each reviewer provided in the peer review report submitted to EPA by Versar. EPA suggested that
each peer reviewer conduct this evaluation using the April 2013 draft of the assessment, the specific
comments and suggestions provided by the peer reviewer on the May 2012 draft of the assessment
(as reflected in the peer review report), and the draft response to comments document EPA provided
to each reviewer. The draft response to comments document detailed how EPA responded to specific
comments and suggestions provided in the peer review report.
Each peer reviewer was asked to provide a letter report evaluating how well the April 2013 draft of
the assessment addressed both the key recommendations provided in the peer review report's
Executive Summary and the specific comments provided by that individual reviewer in the peer
review report.
-------
This document includes the Scope of Work that each reviewer received from EPA and presents the
follow-on review materials submitted to EPA by each of the 12 reviewers, organized in alphabetical
order. EPA also is releasing a separate document that includes all peer review comments on both the
May 2012 and April 2013 drafts of the assessment and EPA's responses to these comments. This
Response to Peer Review Comments document is available at
http://cfpub.epa. aov/ncea/bristolbav/recordisplav.cfm9deid=242810. under the Downloads section.
-------
Revised Draft Bristol Bay Assessment
Peer Review Follow-On
Scope of Work
Introduction
The U.S. EPA is developing an assessment of the potential impacts of large-scale surface mining on fish
resources of the Bristol Bay watershed, Alaska. A draft assessment was released for public comment in
May 2012 and was also peer reviewed by twelve expert scientists and engineers. This external peer
review was coordinated by a contractor, Versar, Inc. Versar compiled detailed comments from each of
the twelve reviewers and provided a peer review report to the Agency in September 2012. The EPA has
reviewed these detailed comments and has developed an improved and expanded draft assessment.
This revised draft assessment was released for public comment on April 26, 2013.
EPA is asking the twelve experts who peer reviewed the May 2012 draft to evaluate how well the
revised draft assessment incorporates and responds to the comments they provided on the previous
draft. This evaluation is limited to the individual comments and suggestions each reviewer provided in
the September 2012 final peer review report. The purpose of this scope of work is to outline the specific
tasks requested by the EPA, the timelines for all deliverables, and associated terms and conditions.
Level of Effort Estimate:
Task 1: Conflict of Interest
The Contractor shall complete the conflict of interest form (Attachment A). This form will be considered
confidential and used by EPA staff to identify any potential conflicts of interests that have arisen since
completion of the last peer review report. The conflict of interest form shall be provided to the EPA
Project Officer within 10 calendar days of initiation of this contract.
Task 2: Confirm Receipt of Review Materials
On or about May 20, 2013, the EPA Project Officer will provide the Contractor materials to be used for
the evaluation outlined in this contract. Materials to be provided include the following:
1. Final Peer Review Report, External Peer Review of the EPA's Draft Document, An Assessment of
Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska, dated September 17,
2012. This document is 193 pages.
2. The draft response to comments document detailing how EPA responded to comments provided
by the external peer reviewers. It is anticipated that this draft report will be roughly 300 pages.
Peer Review Follow-On Scope of Work
Page 1
-------
3. The April 2013 version of the draft Bristol Bay Assessment. This draft report is approximately
600 pages.
4. Revised Appendix I (Conventional Water Quality Mitigation Practices for Mine Design,
Construction, Operation and Closure) and new Appendix J (Compensatory Mitigation and Large-
Scale Hardrock Mining in the Bristol Bay Watershed). Combined, the appendices are
approximately 50 pages. Other appendices are not considered part of the evaluation being
conducted under this contract.
Materials will be provided as electronic files via email and in paper copy form via overnight delivery. All
documents will be provided in MS Word (Word 2007; Office 97) format.
The contractor shall confirm receipt of review materials within 72 hours of receipt.
Task 3: Evaluation of review materials
The Contractor (reviewer) shall conduct an evaluation of how well the revised draft assessment
incorporates and responds to comments provided by the Contractor on the previous draft. The
Contractor should note that some comments were considered outside of the scope of the assessment
and are addressed in the EPA's response to comments document only. The EPA suggests that the
contractor conduct this evaluation using the following approach:
1. Using the final peer review report (Document 1 above), review comments and suggestions made
on the May 2012 draft. The Contractor should specifically review key recommendations in the
Executive Summary of the peer review report and the comments and suggestions provided by
the Contractor in Section III (i.e., the Contractor does not need to address comments made by
other peer reviewers).
2. Review the EPA's response to comments document (Document 2 above) that details how EPA
responded to specific comments and suggestions provided in the peer review report. The EPA's
response to comments document is structured the same as the peer review report.
3. Review the revised draft Bristol Bay Assessment (Documents 3 and 4 above) to assess how EPA's
responses to comments and suggestions were implemented in the revised draft.
The Contractor shall provide a letter report evaluating how well the revised draft Bristol Bay Assessment
addresses both the key recommendations provided in the Executive Summary of the peer review report
and the specific comments provided by the Contractor in the peer review report.
The Contractor shall deliver their evaluation to the EPA Project Officer within 30 calendar days of EPA
providing all review documents. Evaluations shall be delivered electronically via email.
Peer Review Follow-On Scope of Work
Page 2
-------
Review of Contractor Deliverables
Deliverable 1: The Contractor shall provide a completed and signed conflict of interest statement to the
EPA Project Officer within 10 days of the initiation of this contract.
Deliverable 2: The Contractor shall confirm receipt of EPA-provided review materials within 72 hours of
receipt.
Deliverable 3: The Contractor shall provide a summary evaluation of EPA's response to the September
2012 peer review report in letter form within 30 days of receipt of EPA-provided materials.
Use of Deliverables By the EPA
The EPA plans to publicly release a comprehensive response to comments document upon release of
the final Bristol Bay Assessment. The EPA also plans to release the letter evaluation provided by the
Contractor as part of this contract upon release of the final Bristol Bay Assessment. Release of the final
Bristol Bay Assessment is anticipated in 2013.
Confidentiality and Non-Disclosure
Materials provided by the EPA to the Contractor are to be considered confidential until released by the
EPA. Materials in paper form may not be copied, shared, or otherwise distributed. Electronic files may
not be forwarded or distributed. The EPA realizes that the peer review report (Document 1) was made
publicly available in November 2012 and the revised draft assessment (Documents 3 and 4) were made
available for public comment in April 2013. The EPA prefers that the Contractor direct others interested
in these two documents to the EPA's Bristol Bay website (www.epa.gov/BristolBav) to obtain copies.
EPA's draft response to comments document (Document 2) will be released to the public when finalized
and at the time that the final assessment is completed. The Contractor shall take responsible steps to
insure the confidentiality of materials provided by the EPA for this evaluation.
The Contractor shall not discuss their evaluation of the revised draft assessment with members of the
press or public, or in meetings, workshops, or conferences. This restriction shall remain in effect until
publication of the final EPA assessment, or the end of calendar year 2013, whichever comes first.
The Contractor is free to consult with colleagues on technical issues raised in the report provided that
such consultation is limited to discussions of the revised draft assessment (Documents 3 and 4).
Expenses
No travel is associated with the work outlined in this contract. It is not expected that the Contractor
shall incur any expenses associated with completion of the work outlined in this scope of work.
Peer Review Follow-On Scope of Work
Page 3
-------
EPA Contacts
EPA Project Officer: Ms. Jenny Thomas, U.S. Environmental Protection Agency, Office of Water, 1200
Pennsylvania Avenue, NW., Washington, DC 20460. 202-564-4524, thomas.iennvffiepa.gov.
Peer Review Follow-On Scope of Work
Page 4
-------
Mr. David Atkins, Watershed Environmental, LLC
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: David Atkins [|
Sent: Tuesday, Duly 16, 2013 5:06 PM
To: Thomas, Denny
Subject: review comments...
Dear Denny,
Please find my review comments attached (underlined in text). Thanks for
understanding my need to delay delivery and let me know if you have any
questions.
Best regards,
David
-------
Dr. David Atkins
SUMMARY OF KEY RECOMMENDATIONS FROM PEER
REVIEWERS
This section summarizes the significant general recommendations put forth by the peer
reviewers regarding EPA's draft assessment. In developing these recommendations, peer
reviewers provided input on three major areas of the assessment: (1) scope, (2) technical
content and (3) editorial suggestions. Reviewers also identified research needs for EPA to
consider. Please note that this summary of peer review comments did not reflect a
consensus or group perspective, but was compiled from a discussion of individual peer
reviewer recommendations. Additional details, including references cited, can be found
in the reviewers' individual comments in Section III.
Scope of the Document:
• Articulate the purpose of the document more clearly via a primer on the Ecological
Risk Assessment process. If the purpose of the assessment is to inform EPA as the
decision maker, then the level of detail should correspond to this purpose. The
authors should justify and explain what level of detail is required.
RESPONSE: Additional information on both the purpose of the assessment and
ecological risk assessment (ERA) in general has been added to Chapters 1 and 2, as
well as the Executive Summary. Section 1.2 includes information about the use of
the assessment. The assessment has been reorganized into two major sections
(problem formulation, risk analysis and characterization) to clarify where different
chapters fall in the typical ERA process.
Response: This response is adequate.
• Include a statement upfront about the role of risk managers and other audiences, such
as project managers/engineers, regulators, mine owners/operators. Knowing their role
ensures inclusion of information necessary for any risk assessment by (1) describing
the need for a risk assessment, (2) listing those decisions influenced, and (3)
characterizing what risk managers require from the risk assessment.
RESPONSE: Section 1.2 of the revised assessment discusses the use of the
assessment.
Response: This response is adequate.
• Explain why the scope for human and wildlife impacts was limited to fish-mediated
effects, as well as why fish-mediated effects on humans were limited to Alaska
Native cultures. Reviewing effects beyond fish-mediated ones (e.g., potential for
complete loss of the subsistence way of life) would improve the assessment.
RESPONSE: The scope of the assessment has been clarified throughout the
document, particularly in Chapters 1 and 2. Throughout the assessment we
acknowledge that direct effects of large-scale mining on wildlife and Alaska Native
-------
Dr. David Atkins
cultures may be significant, but that these direct effects are outside the scope of the
current assessment.
Response: This response is adequate.
• Be more consistent throughout the document in terms of the level of detail provided
for the different scenarios and stressors. For example, the document has devoted 36
pages to the discussion of catastrophic Tailings Storage Facility (TSF) failure, while
sections on the pipeline, water treatment, and road/culvert failures are brief. Indeed,
the long discussion on the TSF failure belies a certainty and understanding of dam
failure dynamics that is inaccurate.
RESPONSE: The final document includes more failure scenarios (e.g., diesel
pipeline failure, wastewater treatment plant failure, and refined seepage scenarios)
in Chapter 8. It also explains why these specific failure scenarios were chosen, and
discusses these scenarios in greater detail than the previous draft (i. e., to more
closely match the level of detail originally provided only for the TSF failure
scenario). Also see detailed responses to comments on Peer Review Question 5.
Response: This response is adequate.
Technical Content:
Mine Scenario
• Consider the document to be a screening-level assessment of all potential stressors.
Focusing on failure mode overemphasizes catastrophic events (e.g., TSF failing),
rather than considering all potential stressors, such as holding mine owners strictly
accountable for their day-to-day activities with regard to best practices.
RESPONSE: Additional information on the purpose and scope of the assessment
has been added to Chapters 1 and 2. A screening of all potential stressors,
including individual chemicals, is presented in Section 6.4.2. Also see detailed
responses to Peer Review Question 2 on the use of "best practices" and responses to
Peer Review Question 5 on failure scenarios.
Response: This response is adequate.
• Reexamine the document's use of historical data and case studies to describe and
estimate the risk of failure for certain mine facilities (including the TSF, pipeline,
water treatment, etc.), as these examples from extant mines may not be an appropriate
analog for a new mine in the Bristol Bay watershed.
RESPONSE: The TSF failure range was, and still is, based on design goals, not the
historical data. The historical TSF failure data are provided as background The
pipeline failure rates are based on the most relevant historical data from the
petroleum industry. They are directly relevant to the diesel pipeline, and
experiences at the Alumbrera mine (described in the previous draft) and the
-------
Dr. David Atkins
Antamina and Bingham Canyon mines (added to this draft) suggest that they also
are relevant to the product concentrate pipeline. Water treatment failure rates were
not quantified. However, recent reviews cited in the revised draft indicate that water
collection and treatment failures have been reported at nearly all analogous mines
in the U.S. The estimation of culvert failure frequencies has been revised and is
now based on only recent literature (2002 and later). We believe that these
estimates are appropriate.
Response: This response is adequate.
• Expand the discussion on the use of "best" management practices, as the document
states that the mine scenario employs "good," but not necessarily "best" practice. For
a mine developed in the Bristol Bay watershed, only "best" practice likely would be
appropriate and anything less may not be permitted. Even so, without a track record
of "best" practice (e.g., new technologies), we cannot assume that technology, by
itself without appropriate operational management controls, can always mitigate risk.
RESPONSE: The term "best management practices" is a term generally applied to
specific measures for managing non-point source runofffrom storm water (40
CFR Part 130.2(m)). Measures for minimizing and controlling sources ofpollution
in other situations often are referred to as best practices, state of the practice, good
practice, conventional, or simply mitigation measures. We assume that these types
of measures would be applied throughout a mine as it is constructed, operated,
closed, and post-closure, and have used the term "conventional modern"
throughout the assessment to refer to these measures. To remove any ambiguity
related to the subjectiveness of terms "good" or "best", we have removed them in
the revision and have provided definitions for relevant terms used in Box 4-1.
Response: This response is adequate.
• Adopt a broader range of mine scenarios (not only minimum and maximum) so as to
bound potential impacts, especially at smaller mine sizes (e.g., 50th percentile).
Underground mine development, with its different impacts, also should be considered
and included in the assessment.
RESPONSE: A third mine size scenario (250 million tons) has been added to the
assessment, to represent the worldwide median sized porphyry copper mine (based
on Singer et al. 2008).
Response: This response is adequate.
• Based on the hypothetical mine scenario, perpetual management of the geotechnical
integrity of the waste rock and tailings storage facilities, as well as perpetual water
treatment and monitoring, will most likely be necessary (i.e., a "walk away" closure
scenario after mining ends may not be possible). Therefore, emphasize how
monitoring and management of the geotechnical integrity of waste rocks and tailing
storage facilities should continue "In Perpetuity" (i.e., for at least tens of thousands of
years). Discuss what conditions would need to be met to allow "walk away" closure
-------
Dr. David Atkins
in the Bristol Bay environment gaining insight into these observations from mines
where perpetual treatment and monitoring are ongoing (e.g., the Equity Silver Mine
in British Columbia).
RESPONSE: The conditions for closure and the potential needfor perpetual site
management are discussed in general terms in the revised assessment. The primary
condition assumed to be required is water chemistry that meets all criteria and
permit conditions and that is stable or improving. However, even though there are
some facilities with "perpetual treatment" conditions in place, there is obviously no
information about how these facilities perform over very long periods of time.
Response: This response is adequate.
• Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting,
overflight, etc.) were managed. This includes roads, airstrips, helipads, camps, fuel
dumps, and ATV trails that have already been developed or imposed on the
watershed, and what "mitigation" already has been undertaken on those sites. Assess
the consequences/impacts of these activities in the Cumulative Risks section.
RESPONSE: The effects of exploratory activities are outside of the scope of this
assessment.
Response: This response is adequate.
Risks to Salmonid Fish
• Place potential mining impacts in the context of the entire Bristol Bay watershed by
emphasizing the relative magnitude of impacts. For example, of the total salmon
habitat, assess the proportion lost due to mining. Further, reflect on the non-linear
nature of the relationship between habitat and salmon production; 5% of the habitat
could be critical and thus responsible for 20% or more of salmon recruitment.
Intrinsic potential, which measures the ability of particular habitats to support fishes,
would lend credibility to this analysis.
RESPONSE: We are unable to build a complete Intrinsic Potential (IP) model, as
this would require validation and more elaborate construction of metrics
appropriate to this region. Our preliminary characterization provides the building
blocks for assessing the distribution of key habitat-forming and constraining
features across these watersheds. We now include a characterization of the major
drivers of habitat potential across the watershed and place the mine-site specific
effects in this context (Chapters 3, 7, and 10).
Response: This response is adequate.
• Include a section on the impact of Global Climate Change with explicit reference to a
monitoring program that will allow scientists, if the mine is built, to distinguish
between effects of climate change and mining effects on the physical and biological
components of this ecosystem.
-------
Dr. David Atkins
RESPONSE: Climate change projections and potential impacts are now included
in Chapter 3, and as important external factors in the risk analyses presented in
Chapters 7, 9,10, and 14. Development of a monitoring program to distinguish
between mining and climate change effects is outside of the scope of the
assessment.
Response: This response is adequate.
• Explicitly recognize that the transportation corridor and all associated ancillary
development, including future resource developments made possible by the initial
mining project, will necessarily and inevitably have impacts (hydrologic, noise, dust,
emissions, etc.). These impacts will vary in duration, intensity, severity, relative
importance, spatial dispersion, and inevitably expand geographically through time
with further "development." These impacts should be incorporated into the
Cumulative Risks section.
RESPONSE: The cumulative risk section (Chapter 13) has been expanded to
include the multiple transportation corridors, ancillary mining development and
secondary development associated with multiple mines in a qualitative discussion.
The issues addressed in the assessment of the transportation corridor (Chapter 10)
have also been expanded to include chemical spills, dust, invasive species, and road
treatment salts.
Response: This response is adequate.
• Incorporate current research findings into stream crossing and culvert-design
practices (e.g., arch culverts, bridges, etc.).
RESPONSE: We describe current culvert design practices in a box titled "Culvert
Mitigation" in Chapter 10.
Response: This response is adequate.
• Recognize in the assessment that risk and impact are not equivalent. Risk may be low,
but the potential impact could be huge (e.g., in the case of a TSF failure).
RESPONSE: Risk has been defined in many ways, even by risk assessors. The
commenter seems to define risk as probability. To avoid that potential source of
confusion, we use the term "probability"for that concept. Similarly, the
commenter seems to use "impact" where we use "effect" or "magnitude of effect".
We use "risk" to refer to both concepts combined—that is, an event or effect and its
probability).
Response: This response is adequate.
• Recognize and justify chronic behavioral endpoints, such as those potentially
affecting survival and long-term success of fish populations.
RESPONSE: The chronic behavioral effects of copper on salmonids, the primary
endpoint of concern, were described in Chapter 5 and are now described in Chapter
8. Although those effects occur at lower levels of copper than conventional
-------
Dr. David Atkins
survival, growth and reproduction endpoints for salmonids, they are less sensitive
than the conventional endpoints for aquatic invertebrates.
Response: This response is adequate.
Wildlife
• Recognize that the draft assessment did not account for all levels of ecology, such as the
individual (e.g., a bald eagle nest), population, community, ecosystem, and landscape levels.
Fold other levels of organization into the stressors assessment where appropriate or justify a
more limited approach.
RESPONSE: As is appropriate for an ecological risk assessment (as opposed to an
environmental impact assessment), this assessment focuses on a specific, limited set of
endpoints as defined in Chapter 5. We have added text in Chapters 2 and 5 to explain both
why these endpoints were selected, and that responses other than those considered in the
assessment, at multiple levels of ecological organization, are likely but are outside the
scope of the assessment.
Response: This response is adequate.
• Discuss in the document fishes other than salmonids The assessment focuses on risks to
sockeye salmon in the Bristol Bay watershed (and also considers anadromous salmonids,
rainbow trout, and Dolly Varden), but does not account for potential impacts to other
members of the resident fish community. Further, primary and secondary production,
including nutrient flux was not addressed. Expanding the assessment to consider other levels
of organization, including direct as well as indirect effects on wildlife and other fish, would
provide additional context in the assessment of mine-related impacts.
RESPONSE: See response to comment above; we also incorporated additional
information from Appendices A, B, and C into the Chapter 5 text, to provide additional
detail on the area's biota. We chose our endpoints for reasons described in Chapters 2 and
5. Other endpoints, including indirect effects on fish and wildlife, are now discussed more
explicitly, but are generally considered outside the scope of the assessment.
Response: This response is adequate.
Human Cultures
• Use case histories to provide insight and anticipate mining impacts on Alaska Natives
(e.g., those exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil
and gas development in the North Slope region, and social impacts related to mining
development in Alaska).
RESPONSE: Examples from applicable case studies, including the Exxon Valdez
oil spill, are cited in Chapter 12 of the revised assessment.
Response: This response is adequate.
-------
Dr. David Atkins
• As noted above (Scope of the Document), clarify why the scope was limited to fish-
mediated effects. The potential direct and indirect impacts for human cultures extend
far beyond fish-mediated impacts (e.g., potential complete loss of the subsistence way
of life). The rationale for this narrow focus should be fully explained. In addition, a
clear explanation should be given for why fish-mediated human impacts focused only
on Alaska Native cultures.
RESPONSE: The assessment focuses on a specific, limited set of endpoints as
defined in Chapter 5. We have added text to explain both why these endpoints were
selected, and that responses other than those considered in the assessment are
likely but are outside the scope of the assessment. The assessment was expanded
(Chapters 5 and 12) to acknowledge that there are a wide range of potential direct
and indirect impacts to indigenous culture, but they are outside of the scope of this
assessment. The discussion of potential effects to indigenous cultures was expanded
to explain that a loss of subsistence resources would extend beyond a loss offood
resources to social, cultural, and spiritual disruption. The text has been expanded
to acknowledge the strong cultural ties of many non-Alaska Natives to the region,
and potential effects on all residents from loss of a subsistence way of life.
However, the focus of the assessment remains on effects on indigenous cultures
resulting from effects on salmon.
Response: This response is adequate.
Water Balance/Hydrology
• Better characterize water resources and assess the potential effect of mine
development on these resources by (1) generating a diagram similar to the conceptual
models beginning on page 3-7 to illustrate the potential effects of mine construction
and operation on surface- and ground-water hydrology; (2) developing a quantitative
water balance and identifying water gains and losses; (3) identifying seasonality of
hydrologic processes, including frozen soils and their associated values (e.g., mm/yr)
for each component of the water balance; (4) incorporating these processes into a
landscape characterization; (5) evaluating how global climate change will influence
these hydrologic processes and rates; and 6) using this characterization to
demonstrate the expected hydrologic modification associated with the mine scenarios
and infrastructure development.
RESPONSE: The original Figure 4-9 (new Figure 6-5) has been revised to more
clearly show water management in the assessment's mine scenarios. In addition,
three schematics illustrating water flows under each of the mine size scenarios
(Figures 6-8 through 6-10) have been added to Chapter 6, as have quantitative
water balances for each mine size scenarios. A qualitative discussion of climate
change is included in Chapters 3 (Section 3.8) and 14 (Box 14-2).
Response: This response is adequate.
• Demonstrate the interconnectedness of groundwater, surface water, hyporheic zone,
and its importance to fish habitat. Address how interconnectedness changes over time
-------
Dr. David Atkins
- seasonally, and with varying weather (e.g., wet vs. dry summers or years, and over
the long term as climate changes).
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially
and temporally uniform manner, but do include examples of known points of high
connectivity (Chapter 7) and qualitatively discuss the potential role of climate
change (Chapter 3).
Response: This response is adequate.
• Provide information on all rivers, including ephemeral and intermittent streams, and
first-order to main-stem streams that could be potentially influenced by the proposed
mine, its ancillary facilities, and the transportation corridor.
RESPONSE: Due to lack of consistent coverage, we rely on the NHD hydrography
layer in this analysis, and can only address ephemeral and intermittent streams
qualitatively (Chapter 7).
Response: This response is adequate.
• Emphasize the importance of a thorough characterization of the leaching potential of
acid-generating and non-acid generating waste rock and tailings, given the low
buffering capacity and mineral content in the streams and wetlands that could receive
runoff and treated water from the proposed mine. Recognize that collection and
treatment of runoff and leachate generated will be critical to maintain baseline water
chemistry in these streams and wetlands.
RESPONSE: We agree that these are important issues, and the discussion of
leachate from waste rocks and tailings has been expanded in the revised assessment
(Chapter 8).
Response: This response is adequate.
Geochemi stry/Metal s
• Reference the most current geochemistry data on potentially acid-generating, non-acid
generating, and metal leaching so as to describe any potential effects of seepage and changes
to surface- and ground-water quality via non-catastrophic failure.
RESPONSE: We used the geochemistry data in PLP's Environmental Baseline Document,
as summarized by the USGS in Appendix H. The effects of seepage on water quality are
analyzed in Chapter 8 of the revised assessment.
Response: This response is adequate.
• Explain how contaminants/metals were selected (and others ignored) by EPA as causes for
concern. Information should be included on additional metals and their toxicity so as to
assess impacts of potential leachates. The Pebble Limited Partnership baseline document
presented additional metals that might be useful to include in the assessment.
RESPONSE: The revised assessment describes the selection of contaminants and other
stressors of concern in Section 6.4.2. Additional metals, process chemicals and dissolved
-------
Dr. David Atkins
solids are now included.
Response: This response is adequate.
Mitigation Measures
• Incorporate the critical mitigation information from Appendix I into the main report's
mine scenarios. Include standard mitigation measures that could provide insight into
how well they might work in this context. If this information is not included in the
main report, then justify its absence.
RESPONSE: Mitigation measures incorporated into design and operation to
minimize potential impacts were included in the assessment, as were some
reclamation measures for closure; these measures are made clearer in the revised
assessment. These mitigation measures were a sub-set of those presented in
Appendix I. The assessment assumes that measures chosen for the scenarios would
be effective. Mitigation to compensate for effects on aquatic resources that cannot
be avoided or minimized by mine design and operation would be addressed through
a regulatory process that is beyond the scope of this assessment. Nevertheless, in
response to public and peer comments we have included a discussion of
compensatory mitigation in Appendix J of the revised assessment.
Response: This response is adequate.
• Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in
the main report, as they ultimately influence the range of mining impacts and consider
time frames of mitigation or reclamation measures (e.g., immediate response, long-
term reclamation).
RESPONSE: See response to previous comment. Mitigation measures are
discussed at greater length in the revised assessment report (e.g., Chapter 4 and
Appendix J).
Response: This response is adequate.
Uncertainties and Limitations
• Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of uncertainty
and (2) assessing the certainty of some mine impacts. Discuss data limitations in the
context of uncertainty.
RESPONSE: The individual analysis chapters and the revised Integrated Risk
Characterization (Chapter 14) discuss certainties and data limitations to a greater
extent, as suggested
Response: This response is adequate.
• Articulate early in the document how much uncertainty is acceptable. The assessment
provides little insight with respect to the decisions the document is intended to
support.
-------
Dr. David Atkins
RESPONSE: Acceptable levels of uncertainty can be defined prior to an
assessment if a decision and a decision maker are identified and if data will be
collected by a specified design to implement a specified model, as described in the
EPA's Data Quality Objectives process, However, because this assessment is based
on available data and is intended as a background scientific document rather than
a decision document, it is not possible to specify the amount of uncertainty that is
acceptable. Rather, the available data determine the uncertainty and if the
assessment is subsequently used to inform a decision, the decision maker must
determine whether the level of uncertainty is acceptable.
Response: This response is adequate.
Editorial Suggestions:
• The title of the document leads one to believe that the assessment addresses the entire
Bristol Bay watershed; rather, the report deals with two major rivers and their
watersheds, the Nushagak and Kvichak. Thus, the title should be changed to reflect
the emphasis on these two rivers and their watersheds. A possible title may be "An
Examination (or identification) of the Potential Impacts of Mining and Mining
Associated Activities on Salmon Ecosystems in the Nushagak River and Kvichak
River watersheds, Bristol Bay."
RESPONSE: The assessment addresses multiple scales: the Bristol Bay watershed,
the Nushagak and Kvichak River watersheds, the watersheds of the three streams
draining the Pebble deposit, and the watersheds crossed by the transportation
corridor. These multiple scales, and how they are used throughout the assessment,
are described more clearly in the revision (Chapter 2).
Response: This response is adequate.
• Revise the Executive Summary to more precisely reflect the findings in the
document.
RESPONSE: The Executive Summary has been rewritten to reflect the revised
assessment findings.
Response: This response is adequate.
• The appendices contain detailed and useful information that should be summarized
and included in the main document (e.g., Appendix E: Economics, Appendix G: Road
and Pipelines, and Appendix I: Mitigation). Additionally, consider expanding the
preface to include information on the use of the appendices. If the information is not
included in the main report, then justify its absence.
RESPONSE: More information from the appendices was brought forward into
appropriate chapters of the revised report. The purpose of the appendices—to
provide the detailed background characterization necessary for the ecological risk
assessment—has also been clarified in Chapter 2. The document no longer
-------
Dr. David Atkins
contains a preface because that material has been incorporated into Chapters 1 and
2.
Response: This response is adequate.
• Discuss in more detail the instructive and well-thought-out conceptual models (pages
3-7 to 3-11) illustrating the impacts of mining on Bristol Bay ecosystem processes.
Also, consider expanding the conceptual models to include wildlife, fish-wildlife
interactions, vegetation/terrestrial habitat, and hydrologic processes. Allow them to
guide the text because they appear detailed and complete.
RESPONSE: Additional information on the use of conceptual models throughout
the assessment has been incorporated into Chapter 2. The more comprehensive
conceptual models presented in Chapter 6 (Chapter 3 in the first draft) have been
broken into their relevant component parts throughout the risk analysis and
characterization chapters, to better frame the specific pathways addressed in each
chapter. Additional conceptual models considering impacts on wildlife, Alaska
Native populations, and cumulative effects of multiple mines have been added to
Chapters 12 and 13.
Response: This response is adequate.
• Incorporate the information contained in the conceptual models into a formal
framework, such as a Bayesian or other decision-analysis models.
RESPONSE: This is an excellent suggestion for future efforts, but is beyond the
scope of the current assessment.
Creating a Bayesian Belief Network would require that the Agency convene experts
to subjectively estimate the probabilities of each transition in the conceptual
models. In contrast, this assessment is intended to elucidate the risks from potential
mining based on available data and analyses of those data.
A Decision Analysis would require that alternative outcomes be specified, the utility
of each outcome for a decision maker be defined and the probabilities of each
outcome be estimated for each possible decision so that the expected utilities of
each outcome can be calculated Because this assessment is not a decision
document, these requirements are not feasible or appropriate.
Response: This response is adequate.
• Generate a standard operating protocol for significant figures and use it throughout
the document.
RESPONSE: The authors have carefully addressed this issue. Numbers from the
literature or from the PLPEBD retain the number of significant figures in the
original. Numbers derived for this assessment have the appropriate number of
significant figures given the precision of the input data and uncertainties due to
modeling and extrapolation.
Response: This response is adequate.
-------
Dr. David Atkins
• Remove all references to Mount St. Helens as a surrogate for a TSF failure. Using a
non-human-caused release of material into the ecosystem as an analogue for a mine
failure is not comparable in terms of likelihood or risk for a human-caused release. It
would be more appropriate to extrapolate from the impacts of known mine failures.
RESPONSE: We are puzzled by this comment. The Mount Saint Helens data were
used strictly to address the rate of benthic habitat recovery from a massive
deposition offine mineral particles. The hydrological processes that determine the
recovery of substrate texture and the requirements offish or aquatic invertebrates
are not known to depend on whether mineral particles were from a natural event or
an anthropogenic event. We have reviewed the literature on known mine failures.
They studied tailings spills in terms of toxicity but not in terms ofphysical habitat
effects, which is why we used Mount Saint Helens data. Nevertheless, we have
removed references to Mount St. Helens in the revised assessment to eliminate
concern.
Response: This response is adequate.
• Ensure that the draft assessment remains part of the public record, allowing the
document history to remain intact.
RESPONSE: All drafts of the watershed assessment will remain part of the public
record.
Response: This response is adequate.
Research Needs:
• What are the acute and chronic impacts of mixtures of contaminants, including
metals, acid mine drainage, etc., on the fauna and flora of the Nushagak River and
Kvichak River watersheds? What species are most sensitive and might surrogate
species exist for those for which we do not have data? Review the European literature
and regulatory requirements for additional data.
RESPONSE: The acute and chronic impacts of contaminant mixtures, including
metals and acid mine drainage (i. e., metals in low pH-waters) were addressed using
concentration additivity models in the leachate chemistry tables in Chapters 5 and 6
(now Chapters 8 and 11). Additional toxicity data were obtained by searches of the
EU and OECD database eChem, the EPA '.s ECOTOX and the Environment
Canada site. More metals are now included. In general, metals are most toxic to
aquatic arthropods rather than fish, as discussed for copper.
Response: This response is adequate.
• Can an inventory of nutrients, total organic carbon, and dissolved organic carbon
inputs to aquatic environments be developed that demonstrates their relative
magnitude and spatial variation from headwaters to Bristol Bay? What is the relative
importance of marine-derived nutrients relative to other nutrients from watershed and
terrestrial sources? What is the current atmospheric input of nutrients?
-------
Dr. David Atkins
RESPONSE: These data would be very useful in the risk assessment, but are not
currently available for the Bristol Bay region. We agree this is a research need.
Response: This response is adequate.
• What are the locations of subsistence areas and can these areas be characterized and
differentiated by collecting local environmental and ecological knowledge (e.g., fish
overwintering areas, climate change, ecological shifts, etc.)?
RESPONSE: The revised assessment incorporated current data on subsistence use
areas available from ADF&G. EPA acknowledges that these data are incomplete
and would encourage additional collection of subsistence data and Traditional
Ecological Knowledge.
Response: This response is adequate.
• What impact might mining have on other important wildlife species in the basin (e.g.,
freshwater seals in Iliamna Lake)?
RESPONSE: The scope of the assessment is focused on potential risks to salmon
from large-scale mining and salmon-mediated effects to indigenous culture and
wildlife. Direct effects on wildlife from large-scale mining are likely to be important
and Appendix C (now a stand-alone US Fish and Wildlife report) provides useful
information for a future evaluation of direct effects on wildlife from large-scale
mining. We agree that this is an important area for future research.
Response: This response is adequate.
• What is the comprehensive hydrologic regime of the specific project mining area, and
the broader watershed system as characterized by baseline monitoring, spatial
distribution, and quantitative flow of surface- and ground-waters?
RESPONSE: Comprehensive spatial estimates of mean annual flow are now
presented in Chapter 3. Quantification of spatial and temporal patterns of
groundwater flows is an acknowledged highly desirable product, but it not feasible
within the scope of this assessment. Results of an independent groundwater-surface
water modeling effort are described in Chapter 7.
Response: This response is adequate.
• What is the cumulative impact of commercial fisheries on the Bristol Bay watershed,
especially in an ecosystem context as related to marine-derived nutrient and energy
flow? Acknowledge that commercial fishing has had an impact on the amount of
marine-derived nutrients returned to the watersheds.
RESPONSE: The impact of commercial fisheries on the watershed is not within the
scope of this assessment. Information on commercial fisheries management has
been added in Box 5-2. However, the purpose of this assessment is not to assess the
relative effects ofpotential mining and commercial fishing—it is to evaluate
-------
Dr. David Atkins
potential effects on endpoints if a mine were to be developed, given existing
conditions and activities in the region.
Response: This response is adequate.
-------
Dr. David Atkins
WRITTEN PEER REVIEW COMMENTS
1. GENERAL IMPRESSIONS
The Bristol Bay Watershed Assessment (the Assessment) presents a comprehensive
overview of current conditions in the watershed and establishes the uniqueness and global
importance of the area to global salmon ecology (e.g., the report states that nearly 50% of
the global sockeye salmon population comes from Bristol Bay and nearly 50% of the
salmon in Bristol Bay come from the Nushagak and Kvichak Rivers, which encompass
nearly half of the watershed area). The report also describes in detail the importance of
the fishery to Native Alaska cultures, the importance and uniqueness of subsistence
activities, and the scale of the commercial fishery. Furthermore, the report also outlines
the reliance of the local economy on the salmon fishery.
RESPONSE: No change suggested or required.
Response: None.
There is no question that a mine, especially of the type and magnitude analyzed in the
Assessment, could have significant impacts and that if these impacts are not or cannot be
properly managed and/or mitigated, the consequences could be profound. The
Assessment presents a mining scenario based on preliminary documents prepared for the
Pebble Project, which sets out a conventional approach for development of a very large
mine that includes open-pit and block-cave underground mining methods and
conventional waste rock and tailings management. Development of the mine as proposed
would eliminate streams and wetlands in the project area permanently. The importance of
this impact is not put in context of the watershed as a whole, so it is not possible to
determine the magnitude of the risk to salmon. The Assessment also did not consider
whether there are any methods that could effectively minimize, mitigate or compensate
for these impacts.
RESPONSE: A characterization of the landscape factors influencing salmon habitat
potential is now included to provide context for the stream habitat impacts described in
the document (Chapter 3). The assessment describes the magnitude of risks to salmon
habitat. Due to lack of knowledge of limiting factors, ascribing comprehensive risks to
salmon populations is not feasible in this assessment. Mitigation to compensate for
effects on aquatic resources that cannot be avoided or minimized by mine design and
operation would be addressed through a regulatory process that is beyond the scope of
this assessment. Nevertheless, in response to public and peer review comments we have
included a discussion of compensatory mitigation in Appendix J of the revised
assessment.
Response: This response is adequate.
The Assessment also focuses on the risk of failure of the tailings storage facility, a low
probability, but high impact scenario. The Assessment further describes the potential for
long-term acid and metals production from waste rock and the necessity for water
-------
Dr. David Atkins
treatment. Under the mining scenario as described, perpetual management of the
geotechnical integrity of the waste rock and tailings storage facilities and perpetual water
treatment could be necessary. In addition, failure is always a possibility, albeit a
possibility that is difficult to quantify with any degree of certainty as explained in the
Assessment. The Assessment also does not consider alternative engineering strategies (so
called 'best practice' approaches) that could lessen the risk of failure and possibly the
necessity for perpetual management and water treatment. As such, the report could be
considered a screening level assessment that presents the likelihood of occurrence and
corresponding consequences of failures under the presented development scenario, but
does not describe the magnitude of risk to salmon.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. Measures for minimizing and controlling
sources of pollution, outside of stormwater requirements, may be referred to as best
practices, state of the practice, good practice, conventional, or simply mitigation
measures. We have added a text box in the revision (Chapter 4) to discuss terms.
Mitigation measures considered feasible, appropriate, and 'permittable' (as per
Ghaffari et al. 2011) were considered in the assessment, and these are measures
common to other copper porphyry mines. Evaluation of alternative strategies (e.g.,
other options presented in Appendix I) is outside the scope of this assessment, but such
evaluation should be part of the permitting process for a specific mining plan. The
assessment describes the magnitude of risks to salmon habitat. Due to lack of
knowledge of limiting factors, production, and demographics, ascribing comprehensive
risks to salmon populations is not feasible for this assessment.
Response: I concur with this response. It does highlight information on alternative
strategies that may come from a specific mining plan as it is assessed in the permitting
process.
Question 1. The EPA 's assessment focused on identifying the impacts of
potential future large-scale mining to the fish habitat and populations in these
watersheds. The assessment brought together information to characterize the
ecological, geological, and cultural resources of the Nushagak and Kvichak
watersheds. Did this characterization provide appropriate background
information for the assessment? Was this characterization accurate? Were any
significant literature, reports, or data missed that would be useful to complete
this characterization, and if so what are they?
Based on my general understanding of the watersheds, I consider the general background
information presented in the Assessment accurate and sufficiently complete for the
endpoints of this watershed assessment in the following areas:
• General view of Pacific salmon populations
• General view of resident (non-anadromous) fish
• Wildlife populations
• Native cultures
-------
Dr. David Atkins
RESPONSE: No change suggested or required.
Response: None required.
The Assessment also describes the current economics of the watershed, including
commercial and sport fishing and subsistence activities.
RESPONSE: No change suggested or required.
Response: None required.
Additionally, the report highlights several general aspects of the area that make the
fishery unique in both its abundance and diversity:
• The unique hydrology of the area (strong groundwater and surface water
interaction) that contributes to stable flows and temperatures favorable for salmon
reproduction.
• The importance of anadromous fish in transferring marine-derived nutrients to
upland areas and thus providing nutrients to areas that would naturally be nutrient
poor.
• The lack of roads and infrastructure that make the area unique as one of the few
intact ecosystems remaining in the world, and possibly unique for this type of
fishery.
RESPONSE: No change suggested or required.
Response: None required.
It would be helpful in the background section to better describe the uniqueness of the
Bristol Bay watershed ecosystem in the Pacific Northwest. This could include a
description of other similar ecosystems in the region that have undergone development
and documentation of any changes in fish populations associated with this development.
The Assessment does mention the Fraser River as an analogue, but the scale of
development in this watershed, and even the success of the salmon fishery, seems to be a
point of contention, with some saying mining and fish coexist, and other saying the
impacts are severe.
RESPONSE: The unique conservation value of Bristol Bay fisheries is now discussed
in Chapter 5.
Response: The uniqueness of the watershed is adequately described in Chapter 5.
Organizing this information in terms of endpoints also better frames the context of the
assessment.
It would also be helpful to better explain fish resources in the proposed project area in
comparison to other areas within the watershed. I understand some of the necessary data
may not be available for the project area. It would be helpful to know, however, if the
habitat in the project area is typical, exceptional, or inferior to that in other areas of the
watershed.
-------
Dr. David Atkins
RESPONSE: We now include figures showing reported salmon species distributions
and salmon diversity by HUC-12 watershed, across the Nushagak andKvichak River
watersheds (Figure 5-3; Figures 5-4 through 5-8). It is informative to note that
salmonid diversity is relatively high in the project area. Information on population
sizes and vital rates are limited for the region, but are reported where known. In
addition, we include summary statistics and figures of stream and valley characteristics
across the assessment area (Section 3.4), and compare stream attributes in the project
area to those of the larger watersheds (Section 7.2.1). These results generally illustrate
that the project area contains streams of a size and gradient well within the range of
suitability for salmon, as amply demonstrated by the distribution of spawning and
rearing salmon within the project area streams (Figures 5-4 through 5-8).
Response: The information presented in these sections adequately addresses the concern.
Regarding geological resources, the report describes the Pebble deposit and five other
mineral deposits in the Nushagak and Kvichak watersheds. It would be helpful to know if
there are other mineral resources or oil and gas resources in the Bristol Bay watershed as
a whole that could also be exploited. It would also be helpful to describe the portion of
the watershed that is off-limits to development due to park and protected area status vs.
those lands that are open to mineral development.
RESPONSE: The scope of the assessment was to evaluate the potential impacts from
large-scale mining on salmon resources; thus, consideration of prospective oil or gas
development in the area was outside the scope. The mineral resources identified in the
assessment are those in the Bristol Bay watershed that have had some level of
identification or exploration at this time. Mine claims within the Nushagak and
Kvichak River watersheds are shown in Figure 13-1 and discussed in greater detail
throughout Chapter 13. The assessment assumes that mining would occur on lands
open to mineral development.
Protected areas within the Bristol Bay watershed and the Nushagak and Kvichak River
watersheds are shown in Figures 2-3 and 2-4. We have clarified in the text that the
Nushagak and Kvichak River watersheds represent the least-protected area of the
Bristol Bay watershed. Other state documents exist that map out areas in the Bristol
Bay watershed off-limits to development.
Response: The information presented in these sections and figures adequately addresses
this concern.
Question 2. A formal mine plan or application is not available for the porphyry
copper deposits in the Bristol Bay watershed. EPA developed a hypothetical
mine scenario for its risk assessment, based largely on a plan published by
Northern Dynasty Minerals. Given the type and location of copper deposits in
the watershed, was this hypothetical mine scenario realistic and sufficient for
the assessment? Has EPA appropriately bounded the magnitude of potential
mine activities with the minimum and maximum mine sizes used in the
-------
Dr. David Atkins
scenario? Are there significant literature, reports, or data not referenced that
would be useful to refine the mine scenario, and if so what are they?
The hypothetical mining scenario presented in the Assessment is based on a "Preliminary
Assessment Technical Report" of the Pebble deposit prepared for Northern Dynasty
Minerals by Wardrop (referred to as Ghaffari et al. 2011), in conformance with Canadian
National Instrument 43-101 (NI43-101) which is used to set standards for public
disclosure of scientific and technical information about mineral projects of companies on
bourses supervised by the Canadian Securities Administrators. By most accounts, the
Pebble deposit is a world-class deposit and the Wardrop report counts nearly 11 billion
tonnes of total resource. It is unlikely that all the ore currently identified would be mined,
so 11 billion tonnes would be an upper bound for this particular deposit. It is also certain
that exploiting the Pebble deposit would have to be at a scale large enough to justify the
capital investment to build an infrastructure in such a remote area. Although the
Assessment is ostensibly about any mining development in the Bristol Bay watershed, the
use of the Wardrop scenario for Pebble effectively makes the report an assessment of
mining the Pebble deposit.
RESPONSE: The purpose of the assessment is to estimate potential impacts of large-
scale surface porphyry copper mining on salmon ecosystems in the Bristol Bay
watershed. The preliminary plan for mining the Pebble deposit was used as the basis
for the assessment because that deposit is the most likely to advance in the near term
Also, the Agency believes that mining of other porphyry copper deposits in the
watershed would proceed with a similar approach, since the scenarios used are similar
to what has been done at other porphyry copper deposits. Therefore, it is appropriate to
use Northern Dynasty Mineral's 2011 plan for the Pebble deposit (Ghaffari et al. 2011)
as the basis for the scenarios; however, a final mining plan may differ from what is
presented in Ghaffari et al. (2011). Chapter 13 of the revised assessment also considers
the potential cumulative effects of additional smaller copper porphyry mines in the
watershed No change suggested or required.
Response: The response adequately addresses the concern.
The question then becomes what size mine is feasible from a technical and economic
point of view. The Pebble deposit mine plan, as presented in the Wardrop report, outlines
three scenarios:
• An "investment decision case" for a 25-year mine life that would mine 2 billion
tonnes of ore;
• A "reference case" for a 45-year mine life that would mine 3.8 billion tonnes of
ore; and
• A "resource case" for a 78-year mine life that would mine 6.5 billion tonnes of
ore, or 55% of the total measured, indicated and inferred resource.
The Assessment chose minimum and maximum mine sizes of 2 billion and 6.5 billion
tonnes of ore, respectively. Thus, the resource estimate used for the Assessment is the
same as that for the two end members presented by Wardrop. This would make the mine
one of the largest in the world, exceeding the size of the 10* percentile of global
-------
Dr. David Atkins
porphyry copper deposits by an order of magnitude (see Appendix H of the Assessment).
Mines that ultimately become this size usually expand by increments, as exploration
discovers new ore zones and expansion permits are granted.
RESPONSE: Yes, the scenarios represent large-scale mines. The purpose of the
assessment is to estimate potential impacts of large-scale surface porphyry copper
mining on salmon ecosystems in the Bristol Bay watershed, so large mine sizes are
appropriate. It is quite likely that large mines would be created in increments, but this
would not influence our assessment, as we have evaluated impacts based on volumes of
material released in the event of failures or accidents and on material processed as
proposed in Ghaffari et al. (2011) as reasonable for a deposit of this size, regardless of
the time period for mine operation. However, we have included a third, smaller mine in
our revision to represent the median-sized porphyry copper mine on a worldwide basis
(250 million tons).
Response: This discussion and inclusion of the 250 million ton scenario is adequate.
The Wardrop report further delineates Pebble West as a low-grade deposit near the
surface that would most efficiently be mined using open-pit methods, with Pebble East as
a deeper, higher-grade deposit that would most efficiently be mined using underground
methods (specifically block-caving). Mine facilities, as outlined in the Wardrop report,
would include:
• Open-pit mining utilizing conventional drill, blast and truck-haul methods for
near-surface deposits.
• Underground, block-cave methods for deeper deposits.
• A process plant with throughput of 200,000 tonnes/day that utilizes conventional
crush-grid-float technology with secondary gold recovery.
• Other mine-site facilities, including:
o Tailings storage.
o Waste rock storage (the estimated waste/ore strip ratio is 2:1).
o A natural-gas fired power plant,
o Shop, office, and camp buildings.
o Pipelines to ship ore concentrate slurry to the port facility; return water
from the tailings slurry after separation at the port facility; and fuel.
RESPONSE: No change suggested or required.
Response: None.
This mining and ore processing approach is conventional, and the Assessment includes
these elements. A mine developer may present alternative plans that could vary or alter
how the mine is developed, but the fundamental components would most likely remain
the same.
RESPONSE: The EPA agrees with this comment. No change suggested or required.
Response: None.
-------
Dr. David Atkins
Because the Assessment is presented as a general assessment of mining risks and impacts
in Bristol Bay and not a specific analysis of the Pebble Project, reliance on the scenario
presented in Wardrop makes the assessment overly specific. Further, Chapter 7 provides
more specific information on "Cumulative and Watershed-Scale Effects of Multiple
Mines," which presents analysis of potential impacts from mining five additional deposits
in various stages of development (presumably from early exploration to pre-feasibility).
The information presented in Chapter 7 seems more like another mining scenario than a
cumulative impacts assessment. Therefore, I would suggest a broader range of potential
mining scenarios be organized as follows, with the detail of assessment necessarily
becoming more speculative with each subsequent scenario in the list (due to the lack of
geologic and engineering information on the other deposits):
• Development of one, average-sized porphyry copper deposit (50th percentile or
250 million tonnes of ore as described in Appendix H) in the location of the
Pebble deposit.
• Development of a mega-mine in the location of the Pebble deposit (of the range
between 2 and 6.5 billion tons of ore) that may develop after multiple expansion
and permitting cycles.
• Development of a mining district consisting of an average-sized Pebble mine and
other potential mines (i.e., those presented in Chapter 7).
• Maximum development of all identified potential resources to their most likely
ultimate extent.
Considering this broader range of scenarios would help the reader to better understand
the range of potential risks and impacts.
RESPONSE: The Pebble deposit is located in the watershed of interest, the deposit is
similar to other copper porphyry deposits in the world, and components of the
scenarios are common and anticipated for any such deposit of this type; thus, we feel
that use of the Pebble deposit characteristics and location is appropriate. The revised
assessment includes an additional mine size scenario (Pebble 0.25), representing the
worldwide median size porphyry copper mine (Singer et al. 2008). The revised
assessment expands the cumulative impacts discussion (Chapter 13) further by
including transportation corridors and secondary impacts.
Response: This discussion adequately addresses the concern.
Question 3. EPA assumed two potential modes for mining operations: a no-
failure mode of operation and a mode involving one or more types offailures. Is
the no-failure mode of operation adequately described? Are engineering and
mitigation practices sufficiently detailed, reasonable, and consistent? Are
significant literature, reports, or data not referenced that would be useful to
refine these scenarios, and if so what are they?
The no-failure scenario attempts to quantify the impacts from developing the footprint of
the project alone. In reality, various failures and accidents inevitably occur, and they may
have a range of impacts from inconsequential to large. So this scenario is presented to
-------
Dr. David Atkins
describe the minimum impact that could be expected from project development assuming
everything works as planned.
RESPONSE: This comment is a correct interpretation of our "no-failure" scenario.
Because this distinction was not clear to many other readers, the revision no longer
uses the term "no failure". The no-failure scenario from the draft assessment has been
changed to a chapter on the effects of the footprint of a mining operation, without
regard for operational problems (Chapter 7).
Response: I concur with the modification to describing 'footprint' impacts.
The mine will, by necessity, remove those streams and wetlands that are beneath the pit,
waste rock, tailings and processing plant development areas. There should be some
flexibility in siting facilities other than the pit or underground workings. For the 'no-
failure' scenario, the Assessment presents lengths of stream and areas of wetlands that
would be lost due to physical displacement of the aquatic resources from mine
development and reduction in flows from mine water management. The assessment
presents the following resources that would be lost and that have been shown to be
spawning or rearing habitat for coho, Chinook, and sockeye salmon, or have resident
populations of rainbow trout and Dolly Varden:
25-year scenario
78-year scenario
Eliminated or blocked streams
87.5
141.4
(km)
Reduced flow (>20%; km)
2
10
Eliminated wetlands (km2)
10.2
17.3
Given the range of uncertainty with the proposed mine plan, presenting stream lengths
and wetland areas to the tenth place implies unrealistic accuracy. Significant figures
should be checked and consistent throughout the document, and ranges should be
presented if known (e.g., results for the pits could be presented with more accuracy since
we know where they will be, whereas other facilities that could be located in different
areas should be presented with an appropriate range of uncertainty).
RESPONSE: The authors have carefully addressed this issue. Numbers from the
literature or from the PLPEBD retain the number of significant figures in the
original. Numbers derived for this assessment have the appropriate number of
significant figures given the precision of the input data and uncertainties due to
modeling and extrapolation.
Response: The discussion and modification adequately address the concern.
The impacts as presented appear substantial, mainly because of the very large nature of
the project. However, it would be helpful to describe the significance of this loss,
specifically with regard to the following questions:
• What impact would the loss to streams and wetlands have on the fishery within
the Nushagak and Kvichak basins?
-------
Dr. David Atkins
RESPONSE: Impacts of habitat loss and alteration are very difficult to quantify
given the lack of information on limiting factors, production and capacity
estimates. We were unable to comprehensively evaluate impacts at the population
level, except for the most severe cases where total losses of runs could be
reasonably assumed.
Response: This remains a significant gap in the assessment that will require further data
collection and interpretation.
• Is this loss significant in comparison to the fishery as a whole?
RESPONSE: Losses of streams and wetlands under the mine footprint could not be
related to the fishery due to reasons listed above. For the TSF failure scenario that
completely eliminates or blocks access to suitable habitat in the North Fork Koktuli
River, we estimate that the entire Koktuli portion of the run (-28% of Nushagak
escapement) could be lost. Higher proportional losses would occur if significant
downstream effects occurred due to transport of toxic tailings fines beyond the
Koktuli as modeled under the Pebble 2.0 TSF failure.
Response: Same as above.
• Are there local communities that could be affected by this specific loss?
RESPONSE: Wildlife and resident fish communities would be affected by
reductions in spawning salmon. Local communities would also be affected by the
reduction, which is now discussed in Chapters 12 and 13 (e.g., Chapter 13 now
contains tables which refer to specific subsistence resources used by individual
communities).
Response: This discussion is adequate.
• Is fragmentation of the resource from this loss a significant impact (i.e., are there
stocks that are unique to the project area)?
RESPONSE: Stock structure and genetic diversity are not well known at the project
scale, but based on evidence from other parts of Bristol Bay watersheds, local
adaptation is highly likely. Discussion of fragmentation effects is now included in
Chapters 7 and 10.
Response: This discussion is adequate.
There is no discussion of engineering and mitigation practices in this section. The
responsible regulatory authority would require the project proponent to present a
mitigation plan to compensate for these impacts before permitting. Measures would
include minimization of impact through facility siting, reclamation if possible, and
compensation if reclamation were not feasible. A thorough analysis of possible mitigation
approaches and the likelihood of their success are necessary to fully evaluate impacts
from the 'no-failure' scenario.
RESPONSE: Mitigation measures for design and operation are more clearly called out
in the revised assessment. While measures chosen here may differ from what is
-------
Dr. David Atkins
required during the regulatory process, the assessment is not a mining plan and not an
evaluation of a mining plan. The assessment assumes that measures chosen for the
scenarios would be as effective as possible and examines only accidental failures rather
than a failure to choose a proper mitigation measure. Mitigation to compensate for
effects on aquatic resources that cannot be avoided or minimized by mine design and
operation would be addressed through a regulatory process that is beyond the scope of
this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised
assessment.
Response: This response highlights the fact that the assessment is not an evaluation of a
mining plan, but rather more general. As such, the new material is acceptable and the
revisions are appropriate for the intention of the document.
Questic Are the potential risks to salmonid fish due to habitat loss and
modification and changes in hydrology and water quality appropriately
characterized and described for the no-failure mode of operation? Does the
assessment appropriately describe the scale and extent of risks to salmonid fish
due to operation of a transportation corridor under the no-failure mode of
operation ?
For the no-failure mode of mine operation, the risks to salmonid fish due to habitat loss
and modification in the vicinity of the project are described in terms of loss of lengths of
stream or areas of wetlands. Project proponents state that the mine will only impact a
very small fraction of the watershed (under a no-failure scenario). It is important to
establish whether the modeled impact (e.g., the loss of 87.5 km of streams) is significant,
both in terms of the absolute impact, as well as the effect on ecosystem fragmentation.
RESPONSE: Footprint effects on habitat loss are now characterized in relation to the
distribution of habitat conditions throughout the larger watersheds. Fragmentation
effects are not anticipated at the mine site, apart from blockage of headwater streams
as described, but are anticipated in the case of the transportation corridor (Chapter 10)
and TSFfailure (Chapter 9).
Response: This response is adequate.
In addition, project proponents often state they will preserve and even improve the
fishery. As mentioned in the answer to the previous question, it would be helpful to know
what kinds of mitigation efforts could be employed - minimization, reclamation and
compensation - and have some assessment of the potential effectiveness.
RESPONSE: Mitigation measures, including wastewater treatment and closure and
post-closure monitoring and maintenance, were included in discussion of the mine
scenarios in the draft assessment, and this discussion has been expanded in the revised
assessment. The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine (they are a subset of
options presented in Appendix I), all of which were presented as appropriate for the
Pebble deposit in Ghaffari et al. (2011). Reclamation is not mitigation, but the revised
-------
Dr. David Atkins
assessment includes also some suggested measures to be used in closure/post-closure to
reclaim the disturbed areas. Mitigation to compensate for effects on aquatic resources
that cannot be avoided or minimized by mine design and operation would be addressed
through a regulatory process that is beyond the scope of this assessment. Nevertheless,
in response to public and peer comments we have included a discussion of
compensatory mitigation in Appendix J of the revised assessment.
Response: This response is appropriate and highlights the strengths of the regulatory
process that would assess compensation for effects that cannot be avoided or minimized.
The expanded discussion of mitigation is appropriate for the context of the assessment. It
does not address whether mitigation could improve the fishery.
The Assessment determines that construction of the transportation corridor could alter the
habitat, chemistry, and the migration path across the corridor for the over 30 streams that
the corridor will cross or come near. The report further states that the corridor could
affect 270 km of streams below the corridor and 240 km of streams above, but that there
is no way to assess the magnitude. Therefore, the impacts of the corridor on fish
populations are unknown, and this impact is not described in a way that can allow a
reviewer to draw any conclusion.
RESPONSE: The revised assessment states that "the exact magnitudes of changes in
fish productivity, abundance and diversity cannot be estimated at this time," but
summarizes the species, abundances, and distributions that would potentially be
affected. Also, the assessment concludes that, assuming typical maintenance practices
after mine operations, approximately 15 of 32 culverted streams with restricted
upstream habitat would be entirely or in part blocked at any time. "As a result,
salmonidpassage—and ultimately production—would be reduced in these streams,
and they would likely not be able to support long-term populations of resident species
such as rainbow trout or Dolly Varden."
Response: This response is appropriate. I am not an expert in culvert design and
maintenance, so have no basis to evaluate the appropriateness of the estimate of the
number of culverts that could be blocked at any time after mine operations cease.
Further, the references for road design and construction practices seem to be more
representative of forest and rangeland roads than the type of road that would likely be
constructed for this type of project. It would be helpful to cite experience from other
transportation corridors constructed for mining and oil and gas projects and developed
recently in Alaska.
RESPONSE: Because the proposed mining would take place in an undeveloped area,
the literature is necessarily from areas outside of Bristol Bay. Further, we found no
data concerning the performance of culverts for mining or oil and gas projects in the
region. However, to the extent possible we used examples from representative
environments. The failure frequencies cited in the assessment are not restricted to
forest roads. One of the papers usedfor general information (Furniss et al. 1991)
focuses on forest and rangeland roads, but it is a seminal publication on the potential
effects of roads, particularly as they relate to salmon. The general conclusions of that
-------
Dr. David Atkins
paper should be applicable to the transportation corridor considered in the assessment.
Information on current design standards is now included within throughout Chapter
10, and relies on recent literature.
Response: The revised section is appropriate for the scope of the assessment. Other
designs may be implemented that could result in more favorable performance, but the
specific designs would have to be reviewed in detail and appropriate existing analogs
would have to be selected for performance evaluation.
Question 5. Do the failures outlined in the assessment reasonably represent
potential system failures that could occur at a mine of the type and size outlined
in the mine scenario? Is there a significant type of failure that is not described?
Are the probabilities and risks offailures estimated appropriately? Is
appropriate information from existing mines used to identify and estimate types
and specific failure risks? If not, which existing mines might be relevant for
estimating potential mining activities in the Bristol Bay watershed?
The Assessment focuses on some low probability, high impact failures (e.g., TSF failure),
and presents summaries of failures at existing mines. The majority of the focus is on
catastrophic failures, such as TSF, pipeline, water collection and treatment, and road and
culvert. Anecdotal information regarding mine failures is numerous, but often not well
documented, so it is difficult to get information on the details of failures of other projects.
It is also difficult to extrapolate the probability of failure from one site to the next, and
the report stresses the wide range of uncertainty, depending on design and environment.
Without a more detailed understanding of the mine plan and associated engineering, as
well as additional detailed analysis, it is difficult to determine if the failure probability
estimates presented in the Assessment are reasonable.
RESPONSE: The authors concur with the commenter that it can be "difficult to get
information on the details offailures of other projects". The statistics for historic
tailings dam failures are derivedfrom the largest available database and include many
tens of thousands of dam-years. The pipeline failure data cover millions of kilometer-
years ofpipeline experience. The data on failures of water collection and treatment
systems and of culverts are less extensive. We also recognize that even with detailed
engineering and design information, the prediction of failure probabilities is extremely
difficult. Finally, since all of these low-probability failures are statistical phenomena,
the actual experience at any one site could be vastly different than another similar site,
even when the failure probabilities have the same distribution.
Response: This response is adequate.
The focus on catastrophic failures also takes attention away from what is probably a more
likely scenario. Every project is subject to accidents and smaller, non-catastrophic
failures that have varying degrees of consequence. Sometimes these failures are easily
identified and fixed and other times they can go un-noticed for periods of time.
-------
Dr. David Atkins
RESPONSE: Because the number ofpotential failures is extremely large, it is
necessary to choose a representative set of failure scenarios. The revised assessment
includes more failure scenarios (e.g., diesel pipeline failure, quantitative wastewater
treatment plant failure, truck accidents and spills, and refined leachate seepage
scenarios) and explains why these particular failure scenarios were chosen.
Response: This response is adequate.
It would be helpful to describe some smaller-scale failures that have occurred at mine
sites. A partial list includes: accidents and spills along the transportation corridor or
within the mine site; unanticipated seepage of contaminated water that may be difficult to
detect, collect and treat; movement of water along preferential flow pathways that are
difficult to characterize; temporary failure of water collection and treatment systems;
mistakes in engineering analysis that underestimate the volume of water that must be
collected and treated or overestimate the volume of water available for use; and designing
based on incomplete data and understanding of climate conditions.
RESPONSE: There is a wide variety of failures that could occur, including those
provided by the commenter. Because the number ofpotential failures is extremely
large, it is necessary to choose a representative set of failure scenarios. The revised
assessment includes more failure scenarios (e.g., diesel pipeline failure, quantitative
wastewater treatment plant failure, truck accidents and spills, and refined seepage
scenarios) and explains why these particular failure scenarios were chosen.
Response: This response is adequate.
Question i. Does the assessment appropriately characterize risks to salmonid
fish due to a potential failure of water and leachate collection and treatment
from the mine site? If not, what suggestions do you have for improving this part
of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
Water treatment failures of varying scale occur at virtually every site that treats water,
and mine sites are no exception. The risk of failure of water treatment described in the
assessment is useful as background, but as the report states, the risk is highly uncertain. A
non-catastrophic water treatment system failure is fairly likely to occur at some point
during the mine life, and, hence, requires a detailed assessment. The treatment in the
Assessment is cursory (less than one page). This type of failure is much more likely than
a TSF failure (which receives more than 20 pages of analysis), and therefore requires a
much more thorough treatment given the probability of occurrence and likelihood of
impact to salmon species.
RESPONSE: The wastewater treatment failure scenario has been expanded and is now
detailed and quantified in Chapter 8 of the revised assessment. Additionally, the non-
catastrophic failures of seepage collection from the TSF and waste rock piles also have
been included as scenarios with new and refined analyses.
-------
Dr. David Atkins
Response: This response is adequate.
The background water chemistry indicates mineral concentrations are very low.
Therefore, water treatment will be challenging if background conditions are to be met.
Treatment will be especially challenging given the sensitivity of the species of concern to
concentrations of copper, for instance, as well as the sensitivity to temperature that may
be difficult to match in a water treatment system.
RESPONSE: A mine would not necessarily have to match the low background
concentrations, but would at minimum be held to the water quality standards that
protect the existing and designated uses of the waterbodies where a discharge might
occur. In the Bristol Bay area, all waterbodies are protected for all uses and permit
limitations would be derived to be protective of the most stringent applicable Alaska
Water Quality Standards (WQS'). In addition, we believe that compliance with the more
recent Federal Water Quality Criterion for copper would be necessary to protect
aquatic life.
Response: This response is adequate.
During mine operation, a lapse in treatment would likely be identified and addressed
quickly. This type of treatment failure is ephemeral and would likely have a short-term
impact on the fishery, depending on the time of year of occurrence. It is likely that any
impacts to the fishery could recover in subsequent years after the problem is fixed. The
site will require water treatment long after closure, possibly in perpetuity. This period is
more problematic, as a water treatment failure could go unnoticed for some time or the
resources may not be available to correct it quickly, depending on how long after closure
the failure occurs and the stewardship of the treatment system.
RESPONSE: A failure during operation would be corrected quickly only if it did not
require extensive repairs or the manufacture and import of replacement equipment.
The EPA agrees that water collection and treatment failures after mine closure would
be less likely to be corrected in a timely manner. In addition, events after closure, such
as filling of the pit, would affect water quantities and qualities in ways that would
affect treatment success. A discussion of this issue has been added to Chapter 8 of the
revised assessment. A discussion of financial assurance has been added in Box 4-3.
Response: This response is adequate.
Question 7. Does the assessment appropriately characterize risks to salmonid
fish due to culvert failures along the transportation corridor? If not, what
suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
-------
Dr. David Atkins
The description of culvert failure is necessarily general because there are currently no
designs. The general data on culvert failures presented for the types of culverts described
in the references cited (principally for forest and range land) indicate a high probability of
failure (30-66% failure rate). It is probable, however, that the transportation corridor for
the project would be constructed to a higher standard than most of the roads included in
these papers. It would be helpful to know if similar data are available for highly
engineered roads of the type likely to be built for the project.
RESPONSE: The assessment assumes modern mining technology and operations. We
did not find information explicitly from highly engineered roads, but to the extent
possible we used recent literature from representative environments. The failure
frequencies cited in the revised assessment are from modern roads and are not
restricted to forest or rangeland roads. Information on current design standards is now
included in text boxes throughout Chapter 10
Response: This response is adequate.
Question 8. Does the assessment appropriately characterize risks to salmonid
fish due to pipeline failures? If not, what suggestions do you have for improving
this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are
they?
The discussion of pipeline failures is based on published failure rates, principally for oil
and gas pipelines. This analysis results in a pipeline failure rate of one per 1,000 km per
annum. This is a pretty generic number that does not consider actual pipeline design.
Rather it indicates that pipelines designed using standard practices do fail with a fairly
high frequency. The Assessment does not apply this failure rate to the gas and diesel
pipelines because "they are not particularly associated with mining." Without the mine,
there would be no pipeline. So given that this rate of failure is quantifiable based on good
data and that the pipeline would be built to serve the project, this risk should be
considered.
RESPONSE: The assessment assumes that pipeline design follows standard ASME
practices. A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11 in the revised assessment. A gas leak is considered but is not analyzed
because of the lack of significant causal linkage to fish production.
Response: This response is adequate.
A concentrate pipeline spill would have differing impacts depending on when and where
the spill occurred, with deposition in Lake Hiamna likely being the worst outcome. As
noted in the report, it is likely that a pipeline spill would be detected rapidly and that the
volume of the spill would be limited and amenable to remediation. A better description of
how concentrate pipeline failures have occurred would be helpful to better understand the
risk for this project (e.g., the July 2012 Antamina concentrate pipeline failure, although
-------
Dr. David Atkins
this pipeline would operate under a much different pressure regime due to extreme
altitude change).
RESPONSE: Consolidated, statistically representative data on concentrate pipeline
failures are not readily available, although anecdotal evidence from some case studies
can be found. The pipeline failure statistics reviewedfor the assessment come primarily
from oil and gas pipelines, but also include some water and hazardous liquid pipelines.
The performance of mining concentrate pipelines is not expected to be better than the
performance of oil pipelines, because concentrate pipelines would be expected to be
more susceptible to internal corrosion and abrasion. The 2012 Antamina concentrate
pipeline failure in Peru was reportedly caused by the rupture of a pipe elbow in a valve
station. The regulatory, geographic, and operating conditions of the Antamina pipeline
may differ greatly from those of the concentrate pipeline in the assessment scenarios. A
discussion of causes and probabilities ofpipeline failures is included in Section 11.1 of
the revised assessment. The revised assessment also includes discussions of concentrate
spills at the Bingham Canyon, Utah, and Alumbrera, Argentina, copper mines (Section
11.3.4.2).
Response: This response is adequate.
Question 9. Does the assessment appropriately characterize risks to salmonid
fish due to a potential tailings dam failure? If not, what suggestions do you have
for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so
what are they?
The Assessment generically describes tailings dam failures and the potential impact in
detail. It also uses some site-specific information on tailings supernatant and humidity
cell leachate. There is no question that a tailings dam failure would be catastrophic for the
fishery and the project, and although low probability, is the single largest risk to the
fishery. A tailings dam failure could harm a very large area of the watershed for a very
long period of time and could require a massive and expensive remediation effort.
RESPONSE: Agreed. No change suggested or required.
The tailings deposition and storage methods outlined in the Wardrop NI 43-101 report
and presented in the Assessment are conventional for the industry and comply with
Alaska State regulations. Because of the dire consequences of a failure in this highly
sensitive and unique environment, it would be necessary to employ state of the art
methods for tailings management and go 'beyond compliance' when designing and
constructing this facility. This may include employing methods that are novel, incur
significant additional cost for construction, and lead to a more stable and lower
maintenance facility in the long term, such as dry stack or paste rock tailings (blending
waste rock in with tailings in the impoundment to provide extra geotechnical stability).
These methods, however, are not common practice and in some instances are still under
development.
-------
Dr. David Atkins
RESPONSE: The assessment addresses state of practice methods to identify potential
risks that could result from such practices. The EPA agrees that there is a possibility
that a mining proponent could design and propose practices that go beyond the state of
practice in order to reduce potential risks.
Response: This response is adequate.
Question 10. Does the assessment appropriately characterize risks to wildlife
and human cultures due to risks to fish? If not, what suggestions do you have
for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so
what are they?
The assessment does a good job analyzing the importance of fish resources to other
wildlife and to Alaska Native communities. The lack of site-specific information in the
report results in only very general conclusions that there 'would be some effect.' Of
course, wildlife in the project area and any traditional use of these lands would be
affected by project construction under the no-failure scenario. However, due to the lack
of information, it is unclear if this is an area rich in other wildlife or if there are
traditional native land users that rely on the area. Although the conclusion of this section
is necessarily general, it would be helpful to have more detailed characterization of
wildlife and native use in the project area.
RESPONSE: More information about subsistence harvest use in the area of the mine
scenarios and other areas of the watershed that may be affected by mining activities
has been added to the report. Because the scope of the assessment includes potential
effects to Alaska Native communities related to salmonids, this has been the focus of
the evaluation of potential effects on indigenous cultures. However, EPA recognizes
that there are potential effects due to loss of subsistence harvest areas for other species,
as well as potential effects on non-Alaska Natives who practice a subsistence way of
life. This is clarified in the revised assessment (particularly in Chapters 12 and 13).
As the commenter notes, there is limited information about the use of specific mining
claim areas by wildlife. The Pebble Limited Partnership Environmental Baseline Data
Report includes data on the presence/absence of wildlife species around the Pebble
claim, but we are not aware of any data on abundance of wildlife, so a more detailed
characterization of a specific mine claim area may not be possible. There is some
information about subsistence use of the mining claim areas which has been published
by Alaska Department of Fish and Game. General areas of wildlife harvests have been
added as a figure to the revised assessment.
Response: This response is adequate.
Under the failure scenario, a tailings dam failure, in particular, would be catastrophic for
wildlife and Alaska Native communities that use the area.
-------
Dr. David Atkins
RESPONSE: A new Figure 5-2 illustrates that subsistence use of fish is extensive in
areas downstream of the TSF in the Mulchatna and Nushagak Rivers, and this fact is
now referenced. Effects of a TSF failure on wildlife are now discussed in Section 12.1.
Response: This response is adequate.
Question 11. Does the assessment appropriately describe the potential for
cumulative risks from multiple mines? If not, what suggestions do you have for
improving this part of the assessment?
According to the Assessment, cumulative risks result from the potential development of
at least five additional prospects: Humble, Big Chunk, Groundhog, Sill, and 38 Zone.
Exploiting these prospects would amount to development of a mining district (see
discussion for Question 2 in regards to appropriateness of the mining scenario).
RESPONSE: No change suggested or required.
The Assessment quantifies the loss of stream lengths and wetland areas that potentially
support salmon and resident fish from the development of these projects under a 'no-
failure' scenario. The assessment is highly speculative given that mine development plans
are not available for these prospects.
RESPONSE: Cumulative impacts assessments evaluate past, present and reasonably
foreseeable future actions that are temporally and spatially linked to the project under
consideration. The potential mines in the revised assessment are reasonably
foreseeable based on State of Alaska planning documents and industry exploration
activities. We have expanded the discussion of cumulative impacts in Chapter 13.
Response: This response is adequate.
As with the Pebble scenario, it would be helpful to put this loss of resource in perspective
in terms of the fish resources as a whole. It would also be helpful to describe any
mitigation measures that are feasible to offset the impact of loss of streams and wetlands.
Furthermore, it would be helpful to better understand the role these developments could
have in further fragmenting salmon populations.
RESPONSE: Due to lack of comprehensive estimates of limiting factors across the
impacted watersheds, population level effects could not be quantitatively estimated
except for the most severe cases where total losses of runs could be reasonably
assumed Our ability to estimate population level effects was limited to situations that
were assumed to completely eliminate habitat productivity and capacity in an entire
watershedfor which estimates of escapement could be inferred. For this assessment,
these conditions are only met in the TSF failure scenario that completely eliminates
and blocks access to suitable habitat in the North Fork Koktuli River. In that case, we
estimate that the entire Koktuli portion of the run (-28% of Nushagak escapement)
could be lost. Higher proportional losses would occur if significant downstream effects
occurred due to transport of toxic tailings fines beyond the Koktuli as modeled under
the full TSF failure.
-------
Dr. David Atkins
Compensatory mitigation is governed by a regulatory process outside the scope of this
assessment but will be an important part of any permitting process. Nevertheless, in
response to public and peer comments we have included a discussion of compensatory
mitigation in Appendix J of the revised assessment.
We have expanded the cumulative effects discussion (Chapter 13) to give a broader
description of streams, wetlands and fish at risk and have incorporated a discussion of
the possible effects on biological complexity and fragmenting salmon populations.
Response: This response is adequate and points out the additional analyses that would be
required during the permitting process.
The following potential subsidiary impacts from development of a mining district of this
scale should also be described in more detail or at least mentioned:
• The extensive road network required to support mines in the area and the
attendant development associated with this network.
• The camps associated with the project, in migration of workers to the project
areas, and the demand for resources to be imported from outside the area.
• Invasive species that may follow this scale of development.
RESPONSE: The discussion about the cumulative effects of roads (including invasive
species), secondary development, and mining camps has been expanded in Chapter 13.
Response: This response is adequate.
Questio . Are there reasonable mitigation measures that would reduce or
minimize the mining risks and impacts beyond those already described in the
assessment? What are those measures and how should they be integrated into
the assessment? Realizing that there are practical issues associated with
implementation, what is the likelihood of success of those measures?
The Assessment describes what is considered to be conventional 'good' mining practice,
but does not adequately describe and assess mitigation measures that could be required
by the permitting and regulatory process. A thorough analysis of possible mitigation
measures as employed for other mining projects and the likelihood that they could be
successful in this environment would be necessary.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. The terms are qualitative when generally
interpreted, or have a regulatory meaning. The term "best management practices" is a
term generally applied to specific measures for managing non-point source runoff
from stormwater (40 CFR Part 130.2(m)). Measures for minimizing and controlling
sources ofpollution in other situations are referred to as best practices, state of the
practice, good practice, conventional, or simply mitigation measures. We have added a
text box in the revision Chapter 4 to discuss terms. Mitigation measures considered
-------
Dr. David Atkins
feasible, appropriate, and 'permittable' (as per Ghaffari et al. 2011), were considered
in the assessment, and these are measures common to other copper porphyry mines.
Evaluation of alternative strategies (e.g., other options presented in Appendix Ifor the
mitigation of the same issue) should be part of the permitting process for a specific
mining plan. The State of Alaska does have statutory and/or regulatory requirements
for an approved Plan of Operations (11 11C 86.800), a Reclamation Plan (Alaska
Statute (AS) 27.19.30) and appropriate Financial Assurance (AS 27.19.040) and the
revised Chapter 4 notes these requirements.
Response: This response is adequate.
It is highly likely that for mines located in the Bristol Bay watershed, conventional
engineering practices would not be sufficient. Therefore, it is important to consider
mitigation on numerous fronts when determining the viability of the project. A section on
innovative and state-of-the-art approaches for both mitigation and construction of mine
facilities would be helpful to better understand if risks can be minimized or eliminated
given sufficient funds.
RESPONSE: The EPA agrees generally that conventional engineering practices may
not be sufficient, but it may not be appropriate to test innovative approaches in a
watershed such as the Bristol Bay watershed. The EPA is not aware of innovative
mitigation measures that have sufficient history to be applicable to this location.
Response: This area requires more research and would benefit from a survey of current
industry practice.
Under the no-failure scenario, the footprint of the mine (open pit, block-cave subsidence
zone, waste rock and tailings areas) will by necessity destroy habitat. There may be ways
to create equivalent habitat to compensate for lost habitat in areas within the watershed
that are currently not productive for fish. This form of mitigation may work for resident
fish, but it is unclear if it would work for anadromous fish that return to very specific
locations to spawn.
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be
avoided or minimized by mine design and operation would be addressed through a
regulatory process that is beyond the scope of this assessment. Nevertheless, in
response to public and peer comments we have included a discussion of compensatory
mitigation in Appendix J of the revised assessment. Appendix J includes a discussion
of the challenges of creating equivalent habitat for anadromous and resident fish and
why this would be particularly challenging in the context of the Bristol Bay watershed
Response: This response is adequate.
It is also becoming common practice to offset impacts from project development with
preservation of equivalent habitat areas that are also at risk from development
(http://bbop.forest-trends.org/). It is unclear if this is a feasible consideration for this
project as this could involve allowing one development (e.g., Pebble), while potentially
taking away the development rights of others (presumably for proper compensation).
-------
Dr. David Atkins
RESPONSE: See response to above comment. The potential efficacy of using habitat
preservation as a form of compensatory mitigation for impacts in the Bristol Bay
watershed is discussed in Appendix J.
Response: This response is adequate.
Question 13. Does the assessment identify and evaluate the uncertainties
associated with the identified risks?
The Assessment states: the 'range of failures is wide, and the probability of occurrence of
any of them cannot be estimated from available data.' Uncertainty is addressed
throughout the report, typically with a qualitative discussion. There is a high degree of
uncertainty with respect to how the mine would be developed, operated and closed, as
well as how any impacts would be mitigated. This large uncertainty makes assessing risk
difficult.
RESPONSE: No change suggested or required.
Response: None.
Question 14. Are there any other comments concerning the assessment, which
have not yet been addressed by the charge questions, which panel members
would like to provide?
Long-term risks from development of an open pit have not been characterized. It is
difficult to predict the chemistry of the lake that will form in the open pit, but there is
some potential that water quality will be poor, which may be exacerbated by pit
backfilling with waste rock. The pit lake could impact waterfowl and may have some
impact on groundwater if there is outflow when the lake reaches an equilibrium level.
RESPONSE: The potential pit was not assessed because it is not anticipated that
salmon or other fish (the primary assessment endpoint) would reach it. Therefore, it is
out of scope. However, a pit lake as a potential source of water to streams is briefly
considered in the revised Chapter 8. The scenarios no longer include the placement of
acid generating waste rock in the pit at closure
Response: This response is adequate.
3. SPECIFIC OBSERVATIONS
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
1. Global: All significant figures should be reviewed to make sure they are reflective of
the level of uncertainty (i.e., using an estimate of 141.4 km of streams eliminated
when this value is probably realistically +/- 50%).
-------
Dr. David Atkins
RESPONSE: All significant figures in calculations and measurements made by
EPA have been reviewed and changed to reflect the uncertainty of the analysis. For
example km of stream is reported to the km. Significant figures from cited
documents are as reported in the document.
Response: This response is adequate.
2. Global: Many references cited in the text are not included in the reference list.
RESPONSE: References have been updated.
Response: This response is adequate.
3. Global: The executive summary, main report, and appendices, in many instances,
present different information with sometimes different implications. These three
levels of detail of information should be more cohesive.
RESPONSE: We have re-written the executive summary and main report to make
them more consistent with each other and the appendices.
Response: This response is adequate.
4. Page ES-24 (P2): Suggest changing 'Cumulative Risks' to 'Cumulative Effects of
Multiple Mine Development.'
RESPONSE: The title of the chapter (now Chapter 13 was changed.
Response: This response is adequate.
5. Page 3-2 (PI): The justification for excluding ancillary development from the
assessment should be better explained. In some instances, opening up an area for
natural resource development has had as much or more impact on the environment
and ecosystems as the development itself (for example, oil and gas development in
some areas of the Amazon Basin)
RESPONSE: Discussion of ancillary mining infrastructure (in addition to the pit,
TSFs, waste rock piles, roads and pipelines) has been added to Box 6-1, and
discussion of induced development (development which follows initial development
in an unimproved setting) has been added to Chapter 13.
Response: This response is adequate.
6. Page 3-7 to 3-11: The conceptual models are quite helpful, but are not referenced or
utilized sufficiently when discussing impacts. It would be helpful to more fully
incorporate them into the assessment.
RESPONSE: Sub-components of the conceptual models have been included at the
outset of the risk analysis chapters, to better frame the pathways evaluated in each
chapter.
-------
Dr. David Atkins
Response: This response is adequate.
7. Page 4-1 (PI, L9): Why does the assessment describe current 'good' and not 'best'
practice? The rationale for this decision needs to be described. In addition, it is likely
that anything other than 'best' practice would not be permitted in this context.
RESPONSE: With regard to the terminology of "best", "good", or other terms for
the practices used, what was intended to be conveyed in the assessment is that we
assumed modern mining technology and operations. The terms are qualitative
when generally interpreted, or have a regulatory meaning. The term "best
management practices" is a term generally applied to specific measures for
managing non-point source runoff from storm water (40 CFR Part 130.2(m)).
Measures for minimizing and controlling sources ofpollution in other situations
are referred to as best practices, state of the practice, good practice, conventional,
or simply mitigation measures. We have added a text box in the revision Chapter 4
to discuss terms. Mitigation measures considered feasible, appropriate, and
'permittable' (as per Ghaffari et al. 2011) were considered in the assessment, and
these are measures common to other copper porphyry mines.
Response: This response is adequate.
8. Page 4-11 (P3, L10): The liner lifetime is quite low, and given the importance of this
assumption, would warrant more than a personal communication that does not appear
in the references (North pers. comm.).
RESPONSE: The discussion of liners has been expanded and moved to Chapter 4
(Section 4.2.3.4) in the revision. The personal communication reference has been
removed.
Response: This response is adequate.
9. Page 4-21 (PI, L4): Is the assumption about the TSF locations from the authors or
from the Wardrop 43-101 report?
RESPONSE: The location of the TSFs is a combination of alternative sites
described in Ghaffari et al. 2011 (the Wardrop report), in the NDM 2006 Water
Rights Application (see reference list for Chapter 6), and our knowledge of site
characteristics suitable for tailings impoundments.
Response: This response is adequate.
10. Page 4-26 (S4.3.7-Water Management): It would help to provide appropriate ranges
for numbers (e.g., the precipitation at the mine and TDF is 803 and 804 mm/yr,
respectively, which implies an unrealistic degree of certainty). It would also be
helpful to include a diagram. I am uncertain why 'cooling tower' water losses would
be included in the mine water balance since power generation would likely be at a
remote location and other impacts from power generation are not considered in this
-------
Dr. David Atkins
assessment.
RESPONSE: In general, we have presented our best estimates for the parameter
values discussed in the assessment and have used these estimates in our analyses
and calculations. A full uncertainty analysis is beyond the scope of this assessment,
so we have provided ranges only for critical parameters such as the mine size. We
followed the Ghaffari et al. (2011) mine plan in placing power generation on the
site, which is the reason for the natural gas pipeline. Therefore, water use for
cooling is appropriate.
Net precipitation at the mine site has been recalculated. The monthly mean flows
for each gage were summed across the year, producing an area-weighted average
of860 mm/yr. Ghaffari et al. (2011) describe the construction of a combined-cycle
natural gas-fired gas turbine power plant at the mine site and estimate the cooling
tower evaporative and drift losses from the plant. The cooling tower losses were
included because they would constitute a substantial consumptive water use.
Response: This response is adequate.
11. Page 4-31 (P3, L6): How was the filling time of 100 to 300 years for the pit
estimated? What constitutes full (e.g., within x% of the pre-mining water table)?
RESPONSE: The filling time was calculated from the estimated inflow rate to the
empty pit. Most of the groundwater infiltration would come from the uppermost 100
m and the direct precipitation rate would be constant, so the filling rate would not
decrease significantly until the water level was close to the surface compared to the
pit depths. We recognize that as the water level approached the surface, the cone of
depression would shrink and the filling rate would decrease. We have not attempted
to refine our analysis to account for this decrease in rate and the available data do
not merit such a refinement.
The filling time was simply calculated by dividing the pit volume by the inflow rate.
In practice, when the pit was full the pit level would be maintained below the
overflow level by pumping if treatment were required or be allowed to establish a
natural outlet at its low point if treatment was not necessary.
Response: This response is adequate.
12. Page 4-52: The PMP is based on the Miller (1963) reference (not included in the
reference list). How was it estimated?
RESPONSE: The PMP was based on Technical Paper No. 47
(httv://www.nws.noaa.eov/oh/hdsc/PMP documents/TP47.pdf). This PMP was
used as the precipitation input to a Type-1 (for Alaska) HEC-HMS simulation to
calculate a runoff hydrograph. 291 cms was determined to be the peak discharge
from this 24-hour design storm.
Response: This response is adequate.
-------
Dr. David Atkins
13. Page 5-22: Should refer to 'recapture efficiency' rather than 'recovery rate'. How
were the values of 16% and 63% derived?
RESPONSE: The words "recovery rate" were meant to be descriptive and not as a
defined technical term. There is no standard term in widespread usage for the
parameter reported (i.e. the percentage of excess water available for reintroduction
into the streams versus the total amount captured by the mine processes). In the
revised assessment, we have changed the wording to "reintroduced" (Table 6-3,
Section 6.1.2.5). The reintroduction percentages were calculated as part of the
water balance. The water balance totaled all the sources of water captured per year
and then subtracted all of the annual consumptive water losses. The remaining
annual volume of water is excess water that the mine does not need for operations
and which is available for reintroduction to the streams. The ratio of the annual
reintroduced volume to the annual captured volume yields the reintroduction
percentage.
Response: This response is adequate.
14. Page 5-23: The mean annual unit runoff values are not reproducible from the values
given for drainage area and measured mean annual flow.
RESPONSE: These values have been corrected.
Response: This response is adequate.
15. Page 5-32 to 5-39: Tables showing flow changes for different mine sizes and figures
for minimum mine size are difficult to interpret. A different presentation method
and/or narrative description would help.
RESPONSE: Tables, figures, and text for this section have been revised
Response: This response is adequate.
16. Page 5-45 (S5.2.3): The preceding section (Section 5.2.2) focuses on 'Effects of
Downstream Flow Changes' and Section 5.2.3 focuses on wastewater treatment. It is
not clear why this is the sole focus of Section 5.2.3. In addition, only a short
paragraph is included in this section. Certainly there are other possible risks beyond
water treatment, and even this discussion is too cursory given the importance of the
issue.
RESPONSE: Discussion of water treatment has been expanded in Chapter 8 (a
new chapter in the revised assessment).
Response: This response is adequate.
17. Page 5-45 (S5.2.4): This section states a number of assumptions that could result in
under or over estimating impacts on stream flow from the mine footprint. This section
leaves the impression there is a lot of uncertainty, both with assumptions behind the
estimation and with how successful any attempts at mitigation may be. Therefore, we
-------
Dr. David Atkins
are left with very large error bars on estimates that should be reflected in the numbers
presented for loss of length of streams and areas of wetlands.
RESPONSE: We have improved the rigor of the analyses regarding the impact of
mining on streamflow. We have attempted to be very explicit about our assumptions
and approaches, and we believe the analyses are appropriate and defensible.
Response: This response is adequate.
18. Page 5-46 (P3): This paragraph discusses the possibility that estimates of stream
length blocked by mine construction may be overestimated if engineered diversion
channels are successful. This is an important form of mitigation that needs to be
evaluated further. It would be helpful to evaluate mitigation efforts in similar types of
systems to determine if reconstructing streams is feasible and could be successful.
RESPONSE: Appendix J includes a discussion of the efficacy of constructed
spawning channels. There is little evidence in the scientific literature to suggest
that such channels are effective at creating suitable spawning habitat.
Furthermore, there is nothing in the scientific literature that suggests salmon
streams could be successfully reconstructed in landscapes where significant
alternations in area soils, hydrology and groundwater flow paths have occurred,
such as would be the case with the mine scenarios described in the assessment.
Response: This response is adequate.
19. Page 5-59 (Bullet 4): Whole effluent toxicity (WET) testing and downstream biotic
community monitoring would likely be part of any discharge permit. This
requirement would not preclude developing a better understanding of protective
discharge chemistry and temperature requirements before permitting, especially given
the quality of the receiving water.
RESPONSE: Appropriate WET testing and downstream biotic monitoring would be
desirable if a mine were permitted. Additional studies of site specific chemistry and
temperature tolerances may also be desirable. No change required.
20. Page 5-60 (S5.4.2): The statement that mine traffic will not be a large enough volume
to affect runoff needs support. Do we know the road will only be used for mine
traffic? Can we estimate the volume of mine traffic for a mine of this size and then
look at runoff from an analog system?
RESPONSE: This statement has been deleted from the assessment. However, we
found no analogous data that would allow us to quantify runoff.
Response: This response is adequate.
21. Page 5-60 (S5.4.4.1): How was the number of stream crossings determined (e.g., what
criteria were used to define a stream vs. a channel)?
-------
Dr. David Atkins
RESPONSE: All USGS designated streams were included. No artificial channels
were identified and all natural channels are streams.
Response: This response is adequate.
22. Page 5-65 to 5-68 (S5.4.8.2): Text states that 240 km of stream upstream of the
transportation corridor has a gradient greater than 10% and, therefore, is likely to
support fish. Should this be less than 10% (as stated in the header to Table 5-22 and
Table 6-9)? If so, how was the <10% value chosen?
RESPONSE: The comment is correct, though in the revised assessment we use
<12% as the criterion and cite a relevant source.
Response: This response is adequate.
23. Page 5-74 (S5.5): Salmon-mediated effects on wildlife seem under-analyzed in the
report, particularly when compared to the information presented in Appendix C.
RESPONSE: The discussion of potential salmon-mediated effects on wildlife has
been expanded (Chapter 12).
Response: This response is adequate.
24. Page 5-75 (S5.6): As above, salmon-mediated effects on Alaska Native Cultures
seems under-analyzed in the report, particularly when compared to the information
presented in Appendix D.
RESPONSE: The discussion of potential salmon-mediated effects on Alaska Native
cultures has been expanded (Chapter 12).
Response: This response is adequate.
25. Page 6-3 (S6.1.2.1): I don't find the Mt. St. Helens analogy useful.
RESPONSE: It has been removed
Response: This response is adequate.
26. Page 6-9 (PI, LI): Why would 'present' resident and anadromous fish not suffer
habitat loss in the event of a TSF failure? Are they upstream of the area inundated?
RESPONSE: Yes, these populations are upstream of the TSF-affected area.
However, they may be impacted by barriers to seasonal movement or fragmented by
loss of connectivity. This has been clarified in the text.
Response: This response is adequate.
27. Page 6-28 (S6.1.5): The 'Weighing Lines of Evidence' section is not well developed
and not particularly useful. It also does not inform the risk characterization and
-------
Dr. David Atkins
uncertainty discussion.
RESPONSE: The EPA believes that it is important to weigh all relevant lines of
evidence. An explicit weighing allows readers and reviewers to see what was
weighed and how it was weighted. Such transparency is desirable in general.
Therefore, the weighing of evidence has been expanded and the method is better
explained.
Response: This response is adequate.
28. Page 6-30 (PI, L2): The statement that risks of failure of the gas and diesel pipelines
are not considered because they are not particularly associated with mining makes no
sense. Without the mine, there would be no need for the pipeline. And if the types of
failures and risks are well known, then this is one of the areas that could actually be
assessed with some degree of certainty.
RESPONSE: Diesel pipeline failure has been added A gas pipeline is still not
assessed because it is judged to not pose a potentially significant risk to fish (see
Chapter 11).
Response: This response is adequate.
29. Page 6-36 (S6.3, P2): Designating closure as 'premature closure,' 'planned closure,'
and 'perpetuity' with water treatment ceasing immediately, continuing until permits
are exhausted or water is nontoxic, or until institutions fail does not seem reasonable.
Any of these closure scenarios would be planned, and would involve regulatory
compliance reviews, bonding, etc. Walking away without continuing to collect and
treat water would be an unlikely scenario. The issue of treatment in perpetuity is a
larger issue that needs to be treated in detail.
RESPONSE: The premature closure scenario is intended to address the possibility
that the closure plan is not followed. It highlights the needfor adequate bonding
and for plans that include closure by the bond-holding agency.
Response: This response is adequate.
30. Page 6-37 (P4, L4): The report states that (acid generating) waste rock could be left in
place in the event of premature closure. This scenario should be addressed in the mine
closure plan and during the closure bonding process.
RESPONSE: The premature closure scenario is intended to address the possibility
that the closure plan is not followed. It highlights the needfor adequate bonding
and for plans that include closure by the bond-holding agency.
Response: This response is adequate.
31. Page 8-1 to 8-2: Given that this chapter discusses overall risk to salmon, it would be
helpful to put the estimates of km lost in terms of the total stream or watershed
available habitat. We need some context and metric for assessing significance.
-------
Dr. David Atkins
RESPONSE: In Table 2-1, we have explicitly expressed the relative areas of each
spatial scale as % of the scale above it (e.g., the % of the entire Bristol Bay
watershed made up of the Nushagak and Kvichak River watersheds, the % of the
Nushagak and Kvichak River watersheds made up of the mine scenario watersheds,
etc.).
Response: This response is adequate.
32. Page 8-2 (S8.1.2): The focus on a few types of catastrophic failures does not reflect
the current typical mining scenario. Based on experience at other mines, it is more
likely that smaller-impact failures and accidents would occur during the mine life. It
would be helpful to use some current case studies to illustrate this point.
RESPONSE: The number of possible types, magnitudes and combinations of
failures and accidents is virtually infinite. Therefore, we used a bounding strategy.
The potential effects are bounded by those of routine operations and a realistic
severe failure or accident. We also cite case studies from recent reviews of mining
failures and individual failure cases to indicate some of the possibilities.
Response: This response is adequate.
33. Page 8-3 (T8-1): Why would most concentrate pipeline failures occur between stream
and wetland crossings?
RESPONSE: Most of the transportation corridor is not on or adjacent to a stream
or wetland. Therefore, most failures would occur in those areas between stream
and wetland crossings.
Response: This response is adequate.
-------
Mr. Steve Buckley, WHPacific
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Buckley, Steve rmailto:SBucklev@whpacific.coml
Sent: Thursday, July 11, 2013 6:25 PM
To: Thomas, Jenny
Subject: RE: Bristol Bay Peer Review - Confidentiality Reminder
Thank you Jenny
Please find attached my follow-on review.
My comments are in red.
Let me know if you have any questions or need any more information.
Thanks
Steve Buckley
-------
Mr. Steve Buckley
DRAFT
EPA Response to Peer Review Report
External Peer Review of EPA's Draft Document
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
11 June 2013
U.S. Environmental Protection Agency
Seattle, WA
DRAFT - EPA Use Only - Do Not Distribute, Cite or Quote - DRAFT
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS
An Assessment of Potential Mining Impacts on Salmon Ecosystems
of Bristol Bay, Alaska
OVERVIEW
In May 2012, the U.S. Environmental Protection Agency (EPA) released the draft assessment
entitled An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay,
Alaska (hereafter, the assessment). Because this document is considered a highly influential
scientific assessment, it has strict requirements for peer re\ iew as laid out in the EPA 's Peer
Review Handbook (3rd Edition). These requirements included peer review by external expert
reviewers and documentation of how peer review comments are incorporated into the revised
work product.
From May to August 2012, this draft assessment was evaluated by iweh e external expert
reviewers; comments from these reviewers were incorporated into a final peer report, which was
submitted to the EPA in September 2012. This document details the EPA's draft responses to the
peer review comments received on the May 2012 draft of the report, as provided in the final peer
review report. Please note that this i s a draft document, and that it should not be distributed
beyond the EPA and the peer reviewers. It is considered deliberative material and intended for
internal use only, as the assessment continues to undergo re\ ision Once the assessment has been
finalized, a final draft of this document will be completed and released to the public.
STRUCTURE OF I III. DRAFT RESPONSE TO COMMENTS DOCUMENT
This draft response to comments document follows the structure of the final peer review report
for the assessment It is organized into two main sections:
(1) Summary of key Recommendations IVom Peer Reviewers. This section details
general big-picture issues raised by the peer reviewers in the August 2012 peer review
meeting
(2) Written Peer Review Comments. This section details each reviewer's individual
comments on the assessment, organized according to general impressions, responses to
charge questions, and specific observations.
Within each section, the original text from the final peer review report is included here, followed
by EPA responses to the comments contained in that text (in bold italics). The EPA considered
all comments provided in the final peer review report, although not all comments resulted in
changes to the draft assessment. The EPA responses provided in this document explain either
how the assessment was revised to address the comment, or why the assessment was not changed
in response to the comment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
Page 2
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
ORGANIZATION OF I II I APRIL 2013 DRAFT ASSESSMENT
In response to several comments, the revised April 2013 draft of the assessment has been
reorganized to highlight its foundation as an ecological risk assessment, improve overall clarity,
and reduce redundancy. A cross-walk between chapters of the May 2012 and April 2013 drafts is
provided below to help illustrate where material from the May 2012 can be found in the revised
assessment.
May 2012 Draft
1. Introduction
2. Characterization of Current Condition
3. Problem Formulation
4. Mining Background and Scenario
5. Risk Assessment; No Failure
6. Risk Assessment: FjiL;ip
7. Cumulative and Wntei\lvc:-Sr 6. Mine Scenarios
7. Mine Footprint
5. VVjter Collection, Treatment, and Discharge
9. I.:ilii . . Dam Failure
10. Iiansportation Corridor
11. Pipeline Failures
12. Fish-Mediated Effects
v 13. Cumulative Effects of Large-Scale Mining
-> 14. Integrated Risk Characterization
—> 15. References
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 3
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
SUMMARY OF KEY RECOMMENDATIONS FROM PEER
REVIEWERS
This section summarizes the significant general recommendations put forth by the peer reviewers
regarding EPA's draft assessment. In developing these recommendations, peer reviewers
provided input on three major areas of the assessment: (1) scope, (2) technical content and (3)
editorial suggestions. Reviewers also identified research needs for EPA to consider. Please note
that this summary of peer review comments did not reflect a consensus or group perspective, but
was compiled from a discussion of individual peer reviewer recommendations. Additional
details, including references cited, can be found in the reviewers' individual comments in
Section III.
Scope of the Document:
• Articulate the purpose of the document more clearly via a primer on the Ecological Risk
Assessment process. If the purpose of the assessment is to inform EPA as the decision maker,
then the level of detail should correspond to this purpose. The authors should justify and
explain what level of detail is required.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (ERA) in general has been added to Chapters 1 and 2, as well as the
Executive Summary. Section 1.2 includes information about the use of the assessment.
The assessment has been reorganized into two major sections (problem formulation, risk
analysis and characterization) to clarify where different chapters fall in the typical ERA
process.
The reorganization of the report clarifies the purpose of the assessment and includes
information on the intended audience. It also explains the ERA process much better than the
initial draft. The overall organization of the assessment has been improved by identifying the
assessment endpoints first, before outlining the potential risks. In addition, the elimination of
the No Failure/Failure sections in the draft report helps avoid confusion on what type of
system failures are being assessed.
• Include a statement upfront about the role of risk managers and other audiences, such as
project managers/engineers, regulators, mine owners/operators. Knowing their role ensures
inclusion of information necessary for any risk assessment by (1) describing the need for a
risk assessment, (2) listing those decisions influenced, and (3) characterizing what risk
managers require from the risk assessment.
RESPONSE: Section 1.2 of the revised assessment discusses the use of the assessment.
Section 1.2 discusses the use of the assessment in greater detail than the initial draft. This
information is also treated more thoroughly in the Executive Summary.
• Explain why the scope for human and wildlife impacts was limited to fish-mediated effects,
as well as why fish-mediated effects on humans were limited to Alaska Native cultures.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 4
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Reviewing effects beyond fish-mediated ones (e.g., potential for complete loss of the
subsistence way of life) would improve the assessment.
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. Throughout the assessment we acknowledge that direct
effects of large-scale mining on wildlife and Alaska Native cultures may be significant, but
that these direct effects are outside the scope of the current assessment.
The scope of the assessment is outlined in the Executive Summary and explains which
potential impacts are considered and which are not and why.
• Be more consistent throughout the document in terms of the level of detail provided for the
different scenarios and stressors. For example, the document has devoted 36 pages to the
discussion of catastrophic Tailings Storage Facility (TSF) failure, while sections on the
pipeline, water treatment, and road/culvert failures are brief. Indeed, the long discussion on
the TSF failure belies a certainty and understanding of dam failure dynamics that is
inaccurate.
RESPONSE: The final document includes more failure scenarios (e.g., diesel pipeline
failure, wastewater treatment plant failure, and refined seepage scenarios) in Chapter 8. It
also explains why these specific failure scenarios were chosen, and discusses these
scenarios in greater detail than the previous draft (i.e., to more closely match the level of
detail originally provided only for the TSF failure scenario). Also see detailed responses to
comments on Peer Review Question 5.
The inclusion of additional system failure scenarios does indeed more closely match the level
of detail of the TSF failure scenario. However, the level of detail does still seem to imply a
certainty and understanding of the impacts of a tailings dam failure which are simply beyond
the scope of the assessment.
Technical Content:
Mine Scenario
• Consider the document to be a screening-level assessment of all potential stressors. Focusing
on failure mode overemphasizes catastrophic events (e.g., TSF failing), rather than
considering all potential stressors, such as holding mine owners strictly accountable for their
day-to-day activities with regard to best practices.
RESPONSE: Additional information on the purpose and scope of the assessment has been
added to Chapters 1 and 2. A screening of all potential stressors, including individual
chemicals, is presented in Section 6.4.2. Also see detailed responses to Peer Review
Question 2 on the use of "best practices" and responses to Peer Review Question 5 on
failure scenarios.
There still seems to be room for improvement on the organization of the assessment in trying
to identify the potential stressors from catastrophic events versus normal operations.
Chapters 7-11 are better organized and help to outline these various system failure scenarios.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 5
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
• Reexamine the document's use of historical data and case studies to describe and estimate the
risk of failure for certain mine facilities (including the TSF, pipeline, water treatment, etc.),
as these examples from extant mines may not be an appropriate analog for a new mine in the
Bristol Bay watershed.
RESPONSE: The TSF failure range was, and still is, based on design goals, not the
historical data. The historical TSF failure data are provided as background. The pipeline
failure rates are based on the most relevant historical data from the petroleum industry.
They are directly relevant to the dieselpipeline, and experiences at the Alumbrera mine
(described in the previous draft) and the Antamina and Bingham Canyon mines (added to
this draft) suggest that they also are relevant to the product concentrate pipeline. Water
treatment failure rates were not quantified. Howe\>er, recent reviews cited in the revised
draft indicate that water collection and treatment failures have been reported at nearly all
analogous mines in the U.S. The estimation of culvert failure frequencies has been revised
and is now based on only recent literature (2002 and later). We believe that these estimates
are appropriate.
Culvert failure frequency based primarily on large woody debris loading in forest streams is
probably not appropriate information to accurately estimate failure frequency along the
transportation corridor. Ice and floods would probably have more impacts on stream
crossings. There is not much discussion of other crossing methods other than bridges which
might reduce the impacts of stream crossings on fish in Box 10.2 Culvert Mitigation.
• Expand the discussion on the use of "best" management practices, as the document states that
the mine scenario employs "good," but not necessarily "best" practice. For a mine developed
in the Bristol Bay watershed, only "best" practice likely would be appropriate and anything
less may not be permitted. Even so, without a track record of "best" practice (e.g., new
technologies), we cannot assume that technology, by itself without appropriate operational
management controls, can always mitigate risk.
RESPONSE: The term "best management practices" is a term generally applied to
specific measures for managing non-point source runofffrom storm water (40 CFR Part
130.2(m)). Measures for minimizing and controlling sources ofpollution in other
situations often are referred to as best practices, state of the practice, good practice,
conventional, or simply mitigation measures. We assume that these types of measures
would be applied throughout a mine as it is constructed, operated, closed, and post-closure,
and have used the term "conventional modern " throughout the assessment to refer to these
measures. To remove any ambiguity related to the subjectiveness of terms "good" or
"best", we have removed them in the revision and have provided definitions for relevant
terms used in Box 4-1.
These changes help reduce confusion about "good" and "best" management practices.
• Adopt a broader range of mine scenarios (not only minimum and maximum) so as to bound
potential impacts, especially at smaller mine sizes (e.g., 50th percentile). Underground mine
development, with its different impacts, also should be considered and included in the
assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
Page 6
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: A third mine size scenario (250 million tons) has been added to the
assessment, to represent the worldwide median sized porphyry copper mine (based on
Singer et ah 2008).
The addition of the .25 hypothetical mine scenario helps bracket the potential impacts
described in the assessment and there is a short discussion about the potential impacts from
large-scale underground mine development.
• Based on the hypothetical mine scenario, perpetual management of the geotechnical integrity
of the waste rock and tailings storage facilities, as well as perpetual water treatment and
monitoring, will most likely be necessary (i.e., a "walk away" closure scenario after mining
ends may not be possible). Therefore, emphasize how monitoring and management of the
geotechnical integrity of waste rocks and tailing storage facilities should continue "In
Perpetuity" (i.e., for at least tens of thousands of years). Discuss what conditions would need
to be met to allow "walk away" closure in the Bristol Bay environment gaining insight into
these observations from mines where perpetual treatment and monitoring are ongoing (e.g.,
the Equity Silver Mine in British Columbia).
RESPONSE: The conditions for closure and the potential needfor perpetual site
management are discussed in general terms in the revised assessment. The primary
condition assumed to be required is water chemistry that meets all criteria and permit
conditions and that is stable or improving. However, even though there are some facilities
with "perpetual treatment" conditions in place, there is obviously no information about
how these facilities perform over very long periods of time.
The revised assessment clarifies the potential conditions at closure and some of the water
treatment scenarios which could take place.
• Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting, overflight,
etc.) were managed. This includes roads, airstrips, helipads, camps, fuel dumps, and ATV
trails that have already been developed or imposed on the watershed, and what "mitigation"
already has been undertaken on those sites. Assess the consequences/impacts of these
activities in the Cumulative Risks section.
RESPONSE: The effects of exploratory activities are outside of the scope of this
assessment.
Risks to Salmonid Fish
• Place potential mining impacts in the context of the entire Bristol Bay watershed by
emphasizing the relative magnitude of impacts. For example, of the total salmon habitat,
assess the proportion lost due to mining. Further, reflect on the non-linear nature of the
relationship between habitat and salmon production; 5% of the habitat could be critical and
thus responsible for 20% or more of salmon recruitment. Intrinsic potential, which measures
the ability of particular habitats to support fishes, would lend credibility to this analysis.
RESPONSE: We are unable to build a complete Intrinsic Potential (IP) model, as this
would require validation and more elaborate construction of metrics appropriate to this
region. Our preliminary characterization provides the building blocks for assessing the
distribution of key habitat-forming and constraining features across these watersheds. We
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
Page 7
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
now include a characterization of the major drivers of habitat potential across the
watershed and place the mine-site specific effects in this context (Chapters 3, 7, and 10).
The revised assessment does a better job of providing some context to the potential impacts
of the various mine scenarios in Chapter 7.
• Include a section on the impact of Global Climate Change with explicit reference to a
monitoring program that will allow scientists, if the mine is built, to distinguish between
effects of climate change and mining effects on the physical and biological components of
this ecosystem.
RESPONSE: Climate change projections and potential impacts are now included in
Chapter 3, and as important external factors in the risk analyses presented in Chapters 7,
9,10, and 14. Development of a monitoring program to distinguish between mining and
climate change effects is outside of the scope of the assessment.
The revised draft includes several new sections dealing with the potential role of climate
change.
• Explicitly recognize that the transportation corridor and all associated ancillary development,
including future resource developments made possible by the initial mining project, will
necessarily and inevitably have impacts (hydrologic, noise, dust, emissions, etc.). These
impacts will vary in duration, intensity, severity, relative importance, spatial dispersion, and
inevitably expand geographically through time with further "development." These impacts
should be incorporated into the Cumulative Risks section.
RESPONSE: The cumulative risk section (Chapter 13) has been expanded to include the
multiple transportation corridors, ancillary mining development and secondary
development associated with multiple mines in a qualitative discussion. The issues
addressed in the assessment of the transportation corridor (Chapter 10) have also been
expanded to include chemical spills, dust, invasive species, and road treatment salts.
The reorganization of these chapters provides additional information on potential impacts
from the transportation corridor, and reads more clearly than the initial assessment.
• Incorporate current research findings into stream crossing and culvert-design practices (e.g.,
arch culverts, bridges, etc.).
RESPONSE: We describe current culvert design practices in a box titled "Culvert
Mitigation " in Chapter 10.
There is not much discussion of other crossing methods other than bridges which might
reduce the impacts of stream crossings on fish in Box 10.2 Culvert Mitigation.
• Recognize in the assessment that risk and impact are not equivalent. Risk may be low, but the
potential impact could be huge (e.g., in the case of a TSF failure).
RESPONSE: Risk has been defined in many ways, even by risk assessors. The commenter
seems to define risk as probability. To avoid that potential source of confusion, we use the
term "probability"for that concept. Similarly, the commenter seems to use "impact"
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
Page 8
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
where we use "effect" or "magnitude of effect". We use "risk" to refer to both concepts
combined—that is, an event or effect and its probability).
This seems to be a reasonable way to deal with the confusion of terms.
• Recognize and justify chronic behavioral endpoints, such as those potentially affecting
survival and long-term success of fish populations.
RESPONSE: The chronic behavioral effects of copper on salmonids, the primary endpoint
of concern, were described in Chapter 5 and are now described in Chapter 8. Although
those effects occur at lower levels of copper than conventional survival, growth and
reproduction endpoints for salmonids, they are less sensitive than the conventional
endpoints for aquatic invertebrates.
The new format treats these issues more clearly.
Wildlife
• Recognize that the draft assessment did not account for all levels of ecology, such as the
individual (e.g., a bald eagle nest), population, community, ecosystem, and landscape levels.
Fold other levels of organization into the stressors assessment where appropriate or justify a
more limited approach.
RESPONSE: As is appropriate for an ecological risk assessment (as opposed to an
environmental impact assessment), this assessment focuses on a specific, limited set of
endpoints as defined in Chapter 5. We have added text in Chapters 2 and 5 to explain both
why these endpoints were selected, and that responses other than those considered in the
assessment, at multiple levels of ecological organization, are likely but are outside the
scope of the assessment.
The revision explains the ERA process much better and outlines the decision to use the given
assessment endpoints as well as the limitations of the scope of the assessment.
• Discuss in the document fishes other than salmonids The assessment focuses on risks to
sockeye salmon in the Bristol Bay watershed (and also considers anadromous salmonids,
rainbow trout, and Dolly Varden), but does not account for potential impacts to other
members of the resident fish community. Further, primary and secondary production,
including nutrient flux was not addressed. Expanding the assessment to consider other levels
of organization, including direct as well as indirect effects on wildlife and other fish, would
provide additional context in the assessment of mine-related impacts.
RESPONSE: See response to comment above; we also incorporated additional
information from Appendices A, B, and C into the Chapter 5 text, to provide additional
detail on the area's biota. We chose our endpoints for reasons described in Chapters 2 and
5. Other endpoints, including indirect effects on fish and wildlife, are now discussed more
explicitly, but are generally considered outside the scope of the assessment.
These changes improve the organization of the document and bring information from some
of the Appendices forward into the document.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
Page 9
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Human Cultures
• Use case histories to provide insight and anticipate mining impacts on Alaska Natives (e.g.,
those exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil and gas
development in the North Slope region, and social impacts related to mining development in
Alaska).
RESPONSE: Examples from applicable case studies, including the Exxon Valdez oil spill,
are cited in Chapter 12 of the revised assessment.
• As noted above (Scope of the Document), clarify w h\ the scope was limited to fish-mediated
effects. The potential direct and indirect impacts for human cultures extend far beyond fish-
mediated impacts (e.g., potential complete loss of the subsistence way of life). The rationale
for this narrow focus should be fully explained In addition, a clear explanation should be
given for why fish-mediated human impacts focused only 011 Alaska Native cultures.
RESPONSE: The assessment focuses on a specific, limited set of end/mints as defined in
Chapter 5. We have added text to explain both why these endpoints were selected, and that
responses other than those considered in the assessment are likely but are outside the
scope of the assessment. The assessment was expanded (C hapters 5 and 12) to
acknowledge that there are a wide range of potential direct and indirect impacts to
indigenous culture, but they are outside of the scope of this assessment. The discussion of
potential effects to indigenous cultures was expanded to explain that a loss of subsistence
resources would extend beyond a loss of food resources to social, cultural, and spiritual
disruption. The text has been expanded to acknowledge the strong cultural ties of many
non-Alaska Natives to the region, and potential effects on all residents from loss of a
subsistence way of life. However, the focus of the assessment remains on effects on
indigenous cultures resulting from effects on salmon.
Water lialance I Ivdrolouv
• Better characterize water resources and assess the potential effect of mine development on
these resources by (I ) generating a diagram similar to the conceptual models beginning on
page 3-7 to illustrate the potential effects of mine construction and operation on surface- and
ground-water hydrology: (2) developing a quantitative water balance and identifying water
gains and losses. (3) identifying seasonality of hydrologic processes, including frozen soils
and their associated \ allies (e.g., mm/yr) for each component of the water balance; (4)
incorporating these processes into a landscape characterization; (5) evaluating how global
climate change will influence these hydrologic processes and rates; and 6) using this
characterization to demonstrate the expected hydrologic modification associated with the
mine scenarios and infrastructure development.
RESPONSE: The original Figure 4-9 (new Figure 6-5) has been revised to more clearly
show water management in the assessment's mine scenarios. In addition, three schematics
illustrating water flows under each of the mine size scenarios (Figures 6-8 through 6-10)
have been added to Chapter 6, as have quantitative water balances for each mine size
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 10
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
scenarios. A qualitative discussion of climate change is included in Chapters 3 (Section
3.8) and 14 (Box 14-2).
It seems that some additional explanation of the water balance results would be appropriate.
Perhaps an additional diagram which is somewhere between the simplified cartoon in Figure
6.5 and the detailed Tables such as 6.8 would help the reader visualize the results of the
water balance model.
• Demonstrate the interconnectedness of groundwater, surface water, hyporheic zone, and its
importance to fish habitat. Address how interconnectedness changes over time - seasonally,
and with varying weather (e.g., wet vs. dry summers or years, and over the long term as
climate changes).
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
(Chapter 7) and qualitatively discuss the potential role of climate change (Chapter 3).
Figure 7.14 helps provide some additional information on surface and groundwater
interconnectivity and fish habitat. Given the numerous locations where the document
mentions these features as being critical to the understanding of the watersheds it seems that
a more detailed treatment of these factors is warranted.
• Provide information on all rivers, including ephemeral and intermittent streams, and first-
order to main-stem streams that could be potentially influenced by the proposed mine, its
ancillary facilities, and the transportation corridor.
RESPONSE: Due to lack of consistent coverage, we rely on the NHD hydrography layer in
this analysis, and can only address ephemeral and intermittent streams qualitatively
(Chapter 7).
Use of the NHD layer seems appropriate at this level of assessment. Hopefully more detailed
information on stream character and location will be developed in the future.
• Emphasize the importance of a thorough characterization of the leaching potential of acid-
generating and non-acid generating waste rock and tailings, given the low buffering capacity
and mineral content in the streams and wetlands that could receive runoff and treated water
from the proposed mine. Recognize that collection and treatment of runoff and leachate
generated will be critical to maintain baseline water chemistry in these streams and wetlands.
RESPONSE: We agree that these are important issues, and the discussion of leachate
from waste rocks and tailings has been expanded in the revised assessment (Chapter 8).
Chapter 8 is better organized, however it seems that some additional diagrams would help the
reader visualize the information contained in the Tables 8.1 to 8.3 i.e. where and what does
this information potentially impact.
Geochemi stry/Metal s
• Reference the most current geochemistry data on potentially acid-generating, non-acid
generating, and metal leaching so as to describe any potential effects of seepage and changes
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 11
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
to surface- and ground-water quality via non-catastrophic failure.
RESPONSE: We used the geochemistry data in PLP's Environmental Baseline Document,
as summarized by the USGS in Appendix H. The effects of seepage on water quality are
analyzed in Chapter 8 of the revised assessment.
• Explain how contaminants/metals were selected (and others ignored) by EPA as causes for
concern. Information should be included on additional metals and their toxicity so as to
assess impacts of potential leachates. The Pebble Limited Partnership baseline document
presented additional metals that might be useful to include in the assessment.
RESPONSE: The revised assessment describes the selection of contaminants and other
stressors of concern in Section 6.4.2. Additional metals, process chemicals and dissolved
solids are now included.
Mitigation Measures
• Incorporate the critical mitigation information from Appendix I into the main report's mine
scenarios. Include standard mitigation measures that could provide insight into how well they
might work in this context. If this information is not included in the main report, then justify
its absence.
RESPONSE: Mitigation measures incorporated into design and operation to minimize
potential impacts were included in the assessment, as were some reclamation measures for
closure; these measures are made clearer in the revised assessment. These mitigation
measures were a sub-set of those presented in Appendix I. The assessment assumes that
measures chosen for the scenarios would be effective. Mitigation to compensate for effects
on aquatic resources that cannot be avoided or minimized by mine design and operation
would be addressed through a regulatory process that is beyond the scope of this
assessment. Nevertheless, in response to public and peer comments we have included a
discussion of compensatory mitigation in Appendix J of the revised assessment.
The addition of the information in Appendix J helps to explain where and when mitigation
might enter into the regulatory process.
• Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in the main
report, as they ultimately influence the range of mining impacts and consider time frames of
mitigation or reclamation measures (e.g., immediate response, long-term reclamation).
RESPONSE: See response to previous comment. Mitigation measures are discussed at
greater length in the revised assessment report (e.g., Chapter 4 and Appendix J).
The addition of the information in Appendix J helps to explain where and when mitigation
might enter into the regulatory process.
Uncertainties and Limitations
• Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of uncertainty and (2)
assessing the certainty of some mine impacts. Discuss data limitations in the context of
uncertainty.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 12
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The individual analysis chapters and the revised Integrated Risk
Characterization (Chapter 14) discuss certainties and data limitations to a greater extent,
as suggested.
This is a much better way to outline the uncertainty and data limitations.
• Articulate early in the document how much uncertainty is acceptable. The assessment
provides little insight with respect to the decisions the document is intended to support.
RESPONSE: Acceptable levels of uncertainty can be defined prior to an assessment if a
decision and a decision maker are identified and if data will be collected by a specified
design to implement a specified model, as described in the EPA '.s Data Quality Objectives
process, However, because this assessment is based on available data and is intended as a
background scientific document rather than a decision document, it is not possible to
specify the amount of uncertainty that is acceptable. Rather, the available data determine
the uncertainty and if the assessment is subsequently used to inform a decision, the
decision maker must determine whether the level of uncertainty is acceptable.
The revised draft does a better job in the Executive Summary identifying the use of the
assessment and the uncertainties and limitations of the assessment.
Editorial Suggestions:
• The title of the document leads one to believe that the assessment addresses the entire Bristol
Bay watershed; rather, the report deals with two major rivers and their watersheds, the
Nushagak and Kvichak. Thus, the title should be changed to reflect the emphasis on these
two rivers and their watersheds. A possible title may be "An Examination (or identification)
of the Potential Impacts of Mining and Mining Associated Activities on Salmon Ecosystems
in the Nushagak River and Kvichak River watersheds, Bristol Bay."
RESPONSE: The assessment addresses multiple scales: the Bristol Bay watershed, the
Nushagak and Kvichak River watersheds, the watersheds of the three streams draining the
Pebble deposit, and the watersheds crossed by the transportation corridor. These multiple
scales, and how they are used throughout the assessment, are described more clearly in the
revision (Chapter 2).
• Revise the Executive Summary to more precisely reflect the findings in the document.
RESPONSE: The Executive Summary has been rewritten to reflect the revised assessment
findings.
• The appendices contain detailed and useful information that should be summarized and
included in the main document (e.g., Appendix E: Economics, Appendix G: Road and
Pipelines, and Appendix I: Mitigation). Additionally, consider expanding the preface to
include information on the use of the appendices. If the information is not included in the
main report, then justify its absence.
RESPONSE: More information from the appendices was brought forward into
appropriate chapters of the revised report. The purpose of the appendices—to provide the
detailed background characterization necessary for the ecological risk assessment—has
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 13
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
also been clarified in Chapter 2. The document no longer contains a preface because that
material has been incorporated into Chapters 1 and 2.
• Discuss in more detail the instructive and well-thought-out conceptual models (pages 3-7 to
3-11) illustrating the impacts of mining on Bristol Bay ecosystem processes. Also, consider
expanding the conceptual models to include wildlife, fish-wildlife interactions,
vegetation/terrestrial habitat, and hydrologic processes. Allow them to guide the text because
they appear detailed and complete.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual
models presented in Chapter 6 (Chapter 3 in the first draft) have been broken into their
relevant component parts throughout the risk analysis and characterization chapters, to
better frame the specific pathways addressed in each chapter. Additional conceptual
models considering impacts on wildlife, Alaska Nativepopulations, and cumulative effects
of multiple mines have been added to Chapters 12 and 13.
• Incorporate the information contained in the conceptual models into a formal framework,
such as a Bayesian or other decision-analysis models
RESPONSE: This is an excellent suggestion for future efforts, but is beyond the scope of
the current assessment.
Creating a Bayesian Belief Network would require that the. \gency convene experts to
subjectively estimate the probabilities of each transition in the conceptual models. In
contrast, this assessment is intended to elucidate the risks from potential mining based on
available data and analyses of those data.
A Decision Analysis would require that alternative outcomes be specified, the utility of
each outcome for a decision maker be defined and the probabilities of each outcome be
estimated for each possible decision so that the expected utilities of each outcome can be
calculated. Because this assessment is not a decision document, these requirements are not
feasible or appropriate.
• Generate a standard operating protocol for significant figures and use it throughout the
document
RESPONSE: The authors have carefully addressed this issue. Numbers from the literature
or from the PLP Elil) retain the number of significant figures in the original. Numbers
derivedfor this assessment have the appropriate number of significant figures given the
precision of the input data and uncertainties due to modeling and extrapolation.
• Remove all references to Mount St. Helens as a surrogate for a TSF failure. Using a non-
human-caused release of material into the ecosystem as an analogue for a mine failure is not
comparable in terms of likelihood or risk for a human-caused release. It would be more
appropriate to extrapolate from the impacts of known mine failures.
RESPONSE: We are puzzled by this comment. The Mount Saint Helens data were used
strictly to address the rate of benthic habitat recovery from a massive deposition of fine
mineral particles. The hydrological processes that determine the recovery of substrate
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 14
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
texture and the requirements of fish or aquatic invertebrates are not known to depend on
whether mineral particles were from a natural event or an anthropogenic event. We have
reviewed the literature on known mine failures. They studied tailings spills in terms of
toxicity but not in terms of physical habitat effects, which is why we used Mount Saint
Helens data. Nevertheless, we have removed references to Mount St. Helens in the revised
assessment to eliminate concern.
• Ensure that the draft assessment remains part of the public record, allowing the document
history to remain intact.
RESPONSE: All drafts of the watershed assessment will remain part of the public record.
Research Needs:
• What are the acute and chronic impacts of mixtures of contamiminis, including metals, acid
mine drainage, etc., on the fauna and flora of ilie Nusliagak Ri\ er and Kvichak River
watersheds? What species are most sensili\ e and might surrogate species exist for those for
which we do not have data? Review the European literature and regulatory requirements for
additional data.
RESPONSE: The acute and chronic impacts of contaminant mixtures, including metals
and acid mine drainage (i.e., metals in low pH-waters) were addressed using concentration
additivity models in the leachate chemistry tables in Chapters 5 and 6 (now Chapters 8 and
11). Additional toxicity data were obtained by searches of the EU and OECD database
eChem, the EP. I \ ECi)TO\ and the Environment Canada site. More metals are now
included. In general, metals are most toxic to aquatic arthropods rather than fish, as
discussedfor copper.
• Can an in\eniory of nutrients, total organic carbon. and dissolved organic carbon inputs to
aquatic en\ ironmeiits he de\ eloped that demonstrates their relative magnitude and spatial
variation from headwaters to liristol Bay' \\ hat is the relative importance of marine-derived
nutrients relative to other nutrients from watershed and terrestrial sources? What is the
current atmospheric input of nutrients '
RESPO NSE: These data would be very useful in the risk assessment, but are not currently
available for the Bristol Hay region. We agree this is a research need
• What are the locations of subsistence areas and can these areas be characterized and
differentiated by collecting local environmental and ecological knowledge (e.g., fish
overwintering areas, climate change, ecological shifts, etc.)?
RESPONSE: The revised assessment incorporated current data on subsistence use areas
available from ADF&G. EPA acknowledges that these data are incomplete and would
encourage additional collection of subsistence data and Traditional Ecological Knowledge.
• What impact might mining have on other important wildlife species in the basin (e.g.,
freshwater seals in Iliamna Lake)?
RESPONSE: The scope of the assessment is focused on potential risks to salmon from
large-scale mining and salmon-mediated effects to indigenous culture and wildlife. Direct
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 15
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
effects on wildlife from large-scale mining are likely to be important and Appendix C (now
a stand-alone US Fish and Wildlife report) provides useful information for a future
evaluation of direct effects on wildlife from large-scale mining. We agree that this is an
important area for future research.
• What is the comprehensive hydrologic regime of the specific project mining area, and the
broader watershed system as characterized by baseline monitoring, spatial distribution, and
quantitative flow of surface- and ground-waters?
RESPONSE: Comprehensive spatial estimates of mean annual flow are now presented in
Chapter 3. Quantification of spatial and temporal patterns of groundwater flows is an
acknowledged highly desirable product, but it not feasible within the scope of this
assessment. Results of an independent groundwater-surface water modeling effort are
described in Chapter 7.
• What is the cumulative impact of commercial fisheries on the Bristol liay watershed,
especially in an ecosystem context as related lo marine-derived nutrient and energy flow?
Acknowledge that commercial fishing has had an impact on the amount of marine-derived
nutrients returned to the watersheds.
RESPONSE: The impact of commercial fisheries on the watershed is not within the scope
of this assessment. Information on commercial fisheries management has been added in
Box 5-2. However, the purpose of this assessment is not to assess the relative effects of
potential mining and commercial fishing—it is to evaluate potential effects on endpoints if
a mine were to be developed, given existing conditions and activities in the region.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 16
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
WRITTEN PEER REVIEW COMMENTS
1. GENERAL IMPRESSIONS
David A. Atkins. M.S.
The Bristol Bay Watershed Assessment (the Assessment) presents a comprehensive overview of
current conditions in the watershed and establishes the uniqueness and global importance of the
area to global salmon ecology (e.g., the report states that nearly 50% of the global sockeye
salmon population comes from Bristol Bay and nearly 50% of the salmon in Bristol Bay come
from the Nushagak and Kvichak Rivers, which encompass nearly half of the watershed area).
The report also describes in detail the importance of the fishery lo Native Alaska cultures, the
importance and uniqueness of subsistence activities, and the scale of the commercial fishery.
Furthermore, the report also outlines the reliance of the local economy on the salmon fishery.
RESPONSE: No change suggested or required.
There is no question that a mine, especially of the type and magnitude analyzed in the
Assessment, could have significanl impacts and that if these impacts are not or cannot be
properly managed and/or mitigated, the consequences could he profound. The Assessment
presents a mining scenario based on preliminary documents prepared for the Pebble Project,
which sets out a conventional approach for development of a \ cry large mine that includes open-
pit and block-ca\ e underground mining methods and coin entional waste rock and tailings
management. De\ elopment of the mine as proposed would eliminate streams and wetlands in the
project area permanently. The importance of this impact is not put in context of the watershed as
a whole, so it is not possible to determine the magnitude of the risk to salmon. The Assessment
also did not consider whether there are any methods that could effectively minimize, mitigate or
compensate for these impacts.
RESPONSE: A characterization of the landscape factors influencing salmon habitat potential
is now included to provide context for the stream habitat impacts described in the document
(Chapter 3). The assessment describes the magnitude of risks to salmon habitat. Due to lack
of knowledge of limiting factors, ascribing comprehensive risks to salmon populations is not
feasible in this assessment. Mitigation to compensate for effects on aquatic resources that
cannot be avoided or minimized by mine design and operation would be addressed through a
regulatory process that is beyond the scope of this assessment. Nevertheless, in response to
public and peer review comments we have included a discussion of compensatory mitigation in
Appendix J of the revised assessment.
The Assessment also focuses on the risk of failure of the tailings storage facility, a low
probability, but high impact scenario. The Assessment further describes the potential for long-
term acid and metals production from waste rock and the necessity for water treatment. Under
the mining scenario as described, perpetual management of the geotechnical integrity of the
waste rock and tailings storage facilities and perpetual water treatment could be necessary. In
addition, failure is always a possibility, albeit a possibility that is difficult to quantify with any
degree of certainty as explained in the Assessment. The Assessment also does not consider
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 17
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
alternative engineering strategies (so called 'best practice' approaches) that could lessen the risk
of failure and possibly the necessity for perpetual management and water treatment. As such, the
report could be considered a screening level assessment that presents the likelihood of
occurrence and corresponding consequences of failures under the presented development
scenario, but does not describe the magnitude of risk to salmon.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed modern
mining technology and operations. Measures for minimizing and controlling sources of
pollution, outside of stormwater requirements, may be referred to as best practices, state of the
practice, good practice, conventional, or simply mitigation measures. We have added a text
box in the revision (Chapter 4) to discuss terms. Mitigation measures considered feasible,
appropriate, and 'permittable' (as per Ghaffari et al. 2011) were considered in the assessment,
and these are measures common to other copper porphyry mines. Evaluation of alternative
strategies (e.g., other options presented in Appendix I) is outside the scope of this assessment,
but such evaluation should be part of the permitting process for a specific mining plan. The
assessment describes the magnitude of risks to salmon habitat. Due to lack of knowledge of
limiting factors, production, and demographics, ascribing comprehensive risks to salmon
populations is not feasible for this assessment.
Steve Buckley, M.S., CPG
The assessment attempts to evaluate the potential impacts of mining development in the
Nushagak and Kvichak watersheds. The main deficiency in the assessment is that it uses only
two hypothetical mine scenarios to bracket the potential impacts of mining activities on the
ecological resources in the watershed. Both of these mine scenarios are larger than the 90th
percentile of all porphyry copper deposits in the world. In order to properly assess the potential
effects of mining activities, in the absence of any specific mining proposal, a minimum mine
scenario on the order of the 50th percentile of worldwide porphyry copper deposits would be
more appropriate. Three or four mine scenarios would allow for a broad range of analysis, and
the reader would be able to put the potential impacts of mining development in wider
perspective.
RESPONSE: A third mine size scenario (Pebble 0.25, 250 million tons) is included in the
revised assessment, to represent the worldwide median size porphyry copper mine (Singer et al.
2008).
This is a major improvement in the assessment.
A large part of the assessment provides information related to catastrophic potential system
failures such as tailings dam failures and pipeline ruptures. There is inadequate information on,
and analysis of, potential mitigation measures at the early stages of mine development, which
would attempt to reduce the impacts of mining activities on fish and water quality. The bulk of
the document is dedicated to evaluating the impacts of tailings dam failure on aquatic resources
and yet in Chapter 4, the assessment provides a probability of tailings dam failure at 1 in every
2,000 mine years.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 18
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Mitigation measures were included in the draft assessment and are more clearly
identified in the revision. Analysis of alternative mitigation measures would be part of a
permitting process and is outside the scope of this assessment. A discussion of compensatory
mitigation in the Bristol Bay watershed has been added as Appendix J in the revised
assessment.
While failures of a TSF might be rare, they do happen, their effects may be very damaging,
and they could be devastating to local communities; thus, the assessment evaluates what risks
might be evident should such an event occur. The revised assessment also expands the
evaluation of risks for some lesser magnitude, but higher probability, events.
The reorganization of the assessment improves the discussion about when and where mitigation
measures might enter the regulatory process and the addition of several additional system
failures brings more balance to the TSF failure analysis. The level of detail of the TSF analysis
still might present the reader with the sense that the impacts from this type of system failure can
be modeled or predicted beyond actuality.
The assessment identifies the interconnectivity of groundwater, surface water, and fish habitat as
being a major component of the quality of the fishery in the watershed yet puts relatively little
effort into the analysis of the detailed relationships between groundwater, surface water, water
quality, and fish habitat, even though this is likely the most important factor in assessing the
potential impacts of mining activities on the fisheries in the watershed.
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
and modeled locations of high groundwater-surface water interaction (Chapter 7)
The addition of Figure 7.14 improves the understanding of some of these relationships. More
effort could be put into bringing this information forward given its importance to understanding
the watershed and fish habitat.
Additional mine scenarios and a more detailed investigation of the geomorphology, surface, and
groundwater hydrology and their relation to fish habitat would provide the reader with a more
accurate and more useful scope of analysis.
RESPONSE: We now describe the broad geomorphic context for stream habitat in the
Nushagak and Kvichak River watersheds by characterizing gradient and watershed terrain
(Chapter 3). The revised scenarios include an additional mine size (representative of the
worldwide median size).The revision includes more failure scenarios (e.g., diesel pipeline
failure, quantitative water treatment failure, and refined seepage scenarios) and explains why
these particular failure scenarios were chosen.
The addition of the stream characterization in Chapter 3 and the analysis of the 0.25 mine
scenario are major improvements.
Courtney Car others, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 19
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Synopsis: EPA's draft document examines the potential impacts of large-scale mining
development on the quality, quantity, and genetic diversity of salmonid fish species in the
Nushagak River and Kvichak River watersheds of Bristol Bay, Alaska. To the extent that both
wildlife and Alaska Native communities in the region depend upon salmonids, fish-mediated
impacts to these other "endpoints of interest" are also explored. A hypothetical mining scenario,
informed by current exploration, planning, and study in the Pebble deposit area, is described
using minimum and maximum estimates for mine production and includes the construction of a
transportation corridor to Cook Inlet. Even in the absence of any failures or accidents,
construction and operation of such a mine would have significant impacts to salmonids in stream
systems proximate to the mine footprint with some related impacts to wildlife and human
communities. At least one or more accidents or failures are expected to occur over the long
lifetime of the mine. Immediate and long-term severe impacts to salmonids are expected to occur
with any significant failure, with relatedly pronounced impacts lo wildlife and Alaska Native
communities in the region. Multiple mines in the region would amplify these impacts.
RESPONSE: No change suggested or required.
General impressions: Overall, the main report is well-written and presents information in
multiple ways, including: narrative, conceptual models, images, figures, and tallies. The report
synthesizes a large amount of information, much of which is described in detail in the report's
appendices. The report highlights the unique characteristics of this watershed: incredibly
productive and sustainable salmon fisheries, relatively little large-scale modification of the
natural environment, and active subsistence-based indigenous cultures still occupying their
homelands and many still using their Native language Making central these features of the
watershed, the tone of the report suggests that some negati\ e impacts to salmonids, wildlife, and
Alaska Native cultures are necessarily expected to accompany any large-scale mining
development and operation in this region.
RESPOXSE: Xo change suggested or required.
The document should pro\ ide a clear articulation of the scope of human impacts considered in
this assessment The main report considers only fish-mediated impacts to Alaska Native cultures.
The restriction of scope to only lish-mediated impacts should be further clarified. A host of
social, cultural, and economic impacts would accompany large-scale mining development in this
region. These direct and indirect human impacts, both positive and negative, were the focus of
many public comments on the I -PA draft document, yet they fall outside of the scope of
consideration in this report If the narrowed scope of fish-mediated impacts is justified, these
other impacts should be clearly identified as outside of the scope of this report. At times in the
report (e.g., p. 5-77), these other impacts are superficially mentioned. Unless a full treatment of
these impacts is included (including a presentation of a large literature explores these impacts
internationally, e.g., Ballard and Banks 2003), this cursory discussion should be removed. If
maintained, the narrow scope should be reiterated throughout the report to remind the reader that
these larger human impacts are not considered.
RESPONSE: The scope of the assessment has been more clearly articulated in Chapter 2,
which also now contains an overview conceptual model diagram demonstrating which
potential sources, stressors, and responses associated with large-scale mining were considered
outside of scope. The fact that direct impacts to Alaska Native cultures are not within the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 20
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
scope of the assessment does not imply that they will not occur or that they are unimportant,
and this has been clarified in Chapters 2 and 12.
The report should articulate more clearly why Alaska Native cultures are the only human groups
included in the assessment of fish-mediated human impacts. The report notes:
"because.. .Alaska Native cultures are intimately connected and dependent upon fish, .. .the
culture and human welfare of indigenous peoples, as affected by changes in the fisheries are
additional endpoints of the assessment" (ES-1-2). This suggests that the limitation of fish-
mediated human considerations to Alaska Native cultures is not due to government-to-
government relationship between tribes and the federal government, nor the special status
afforded by environmental justice concerns, but rather because of their close connections to, and
dependence on fish. Arguably, other human groups also have connections to fish and depend
upon on salmon in this region in various ways, but are excluded from analysis of potential
impact in this report. This comment is not meant to detract from the importance of the focus on
Alaska Native cultures and the primarily indigenous communities in this region for assessing
fish-related impacts. Rather, the comment is made to suggest the inclusion of a clear justification
for this focus, or the broadening of scope to include other human groups who are also connected
to, and dependent upon, salmon in this region (e u . substantial information on the economic
dimensions of salmon resources in this region is summarized in Appendix li, but little is
presented in the main report). Additionally, the assessment of lish-mediated effects to Alaska
Native cultures is primarily focused on subsistence fisheries More discussion of the role of
commercial engagements in salmon fisheries (eg. commercial harvesting, processing,
recreational fishing businesses and employment) in the watershed communities in this region
would be helpful.
RESPONSE: EPA focused on the Alaska Native communities in response to the original
request we received from nine federally recognized tribal governments. The text (Chapter 12)
has been expanded to acknowledge the strong cultural ties of many non-Alaska Natives to the
region, and potential effects on all residents from loss of a subsistence way of life. However,
the focus of the assessment remains on effects on indigenous cultures. The importance of the
commercial fishery to the regional culture has been added to the text.
Dennis I). Dauble, Ph.D.
Overall, the main report and each of the accompanying appendices were well written. I was
unable to identify major inaccuracies or bias in the material as presented. There were
shortcomings in the main report, however. For example, some topics would benefit by being
expanded (Sections 5.6 and 8.7), while others have more detail than appeared necessary (Section
6.1). The assessment effectively addressed three appropriate time periods: (1) operation, (2) post-
closure, and (3) perpetuity. Potential effects are bounded by a minimum and maximum mine
size, which is also appropriate. Inclusion of inference by analogy strengthened the conclusions
reached in the assessment and helped validate results obtained from model predictions.
RESPONSE: Previous Section 5.6 (wildlife and culture) has now been expanded and treated
as a stand-alone chapter (Chapter 12). The summary of risks from the mine scenarios
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 21
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
(previous Section 8.7, now Chapter 14) has been expanded to include fish-mediated risks to
wildlife and culture, and more numerical results are included.
Most figures and tables were useful. The conceptual models and accompanying illustrations of
potential habitat effects (Figs 3-2A and C) are important because they provide a view of
complicated pathways and relationships among potential activities and environmental attributes.
However, these relationships are not revisited in any detail later in the document. I recommend
discussing the conceptual models in more detail in the main report (Section 3.6) and summary
section in Chapter 8.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual models
presented in Chapter 6 (previously in Chapter 3) have been broken into their relevant
component parts throughout the risk analysis and characterization chapters, to better frame
the specific pathways addressed in each chapter.
The Integrated Risk Assessment (Chapter 8) did a creditable job of summarizing habitat losses
and risks from mine operations. What is missing, however, are quantitative descriptions of
habitat lost relative to total habitat available in the larger watershed and individual systems.
Habitat loss should be further discussed in terms of salmonid life stage and productivity (i.e., not
all stream miles are equal).
RESPONSE: Unfortunately, no salmon habitat characterization is available for the region.
The State of Alaska has not even identified all anadromous streams in the region.
Productivity data are not available, even for the streams studied by the PLP. However, the
revised Chapter 14 contains tables summarizing habitat loss in stream lengths and wetland
areas.
If anything, the conclusions could be strengthened The summary of uncertainties and
limitations (Section 8.5) dwells on things that "could not be quantified" due to lack of
information, model limitations, or insufficient resources Thus, this reader was left somewhat in
limbo as to the potential magnitude of effects from mining activities. (Note that this "neutral
voice" is carried throughout the Executive Summary). Many people might interpret such
statements of uncertainty as no proven effect. My point is that probable environmental
consequences of mining activities are much greater than this report alludes to, given that
consequences are likely, even iI"their magnitude is "uncertain."
RESPONSE: We use a neutral voice throughout the document to convey the neutral scientific
perspective of this scientific assessment. We tried to convey the qualitative likelihood of
occurrence when quantitative probabilities were not obtainable. This section has been edited
in Chapter 14 of the revised version to make the relationship between uncertainty and
probability of occurrence clearer.
Section 8.7 is perhaps the most important section of the report. It should be comprehensive, i.e.,
cover all resources and be more quantitative. Missing from the summary were impacts on
wildlife, human culture, resident fish, and other ecological resources. Essential details from
Appendices A, C, E, F, and I, for example, could be synthesized and moved into the main report.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 22
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The summary of risks from the mine scenarios (Chapter 14 in the revised
version) has been expanded to include fish-mediated risks to wildlife and culture and more
numerical results.
Gordon H. Reeves. Ph. IX
The purpose of the report is unclear, which makes it difficult to assess. The report focused on
the potential impact of a hypothetical mine on salmon and salmon habitat in two watersheds in
Bristol Bay, AK. However, it is not clear whether the analysis was intended to be a case study of
the potential impacts of a hypothetical mine under the various scenarios presented or whether the
intent was to develop a framework for assessing mining scenarios These are two very different
objectives, which makes it critical that the purpose be clearly staled in the beginning of the
document so that reviewers and others understand the purpose of the document. There certainly
was much confusion among members of the re\ iew panel and the people who commented on the
report because of this.
RESPONSE: We have clarified the purpose of the assessment in Chapters / and 2.
I think that the credibility of the report could be impro\ ed substantially if the analyses were
formalized and more clearly articulated and defined The authors could consider using a
decision support process, such as a Bayesian approach (see Mareot, B.G., J.D. Steventon, G.D.
Sutherland, and R.K. McCann. 2006. Guidelines for developing and updating Bayesian belief
networks applied to ecological modeling and conservation. Canadian Journal of Forest Research
36: 3063-3074). This would provide more transparency to any analysis and allow others to
better understand how results and conclusions were derived. Also, it would identify critical
relations that should be considered and provide insight about the consequences of not
considering them. This will undoubtedly take additional time and effort, but I believe it would be
well worthwhile l-\amples of where such analysis has been done are in: (1) Armstrup et al.
2008 A liayesian Network Modeling Approach to Forecasting the 21st Century Worldwide
Status of Polar Bears. Pages 213-20S In E.T. DeWeaver et al., editors. Artie Sea Ice Decline:
Observations. Projections, Mechanisms, and Implications. Geophysical Monograph 180.
American Geophysical Union. Washington, D.C.; and (2) Lee, D.C. et al. 1997. Broadscale
Assessment of Aquatic Species and Habitats. Vol. Ill, Chapter 4. U.S. Forest Service, General
Technical Report PWV-GTR—lo.r Portland, Oregon.
RESPONSE: Creating a Bayesian Belief Network would require that the Agency convene
experts to subjectively estimate the probabilities of each transition in the conceptual models.
In contrast, this assessment is intended to elucidate the risks from potential mining based on
available data and analyses of those data.
I thought one of the strongest aspects of the report were the conceptual diagrams of relations
between the various aspects of the development and operation of a mine and the components of
the ecosystem that influence salmon and their habitat (Chapter 3). These diagrams show the
components of the ecosystem, the relation among them, and how mine impacts could potentially
influence given parts of the ecosystem directly or indirectly as a result of cascading effects. They
are a good first step in developing a decision support framework, as suggested in the previous
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 23
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
paragraph. There was, however, little discussion about them in the text and it was not clear if or
how they were used or considered in the analyses. The authors should, at the very least, clearly
identify which parts of the networks were considered and why these particular avenues were
pursued and others were not. This would provide additional insights into potential limitations of
the analyses and results.
RESPONSE: The more comprehensive conceptual models presented in Chapter 6 (previously
in Chapter 3) have been broken into their relevant component parts throughout the risk
analysis and characterization chapters, to better frame the specific pathways addressed in each
chapter.
If this was a case study, the report appeared to have considered a\ ailable literature and reports on
all aspects of the mine, its operation and the parameters lhal could he affected by it. I am not
familiar with this literature so it is not possible for me to comment on the adequacy of the
literature and reports considered. Assumptions about the location and operation of the mine
seemed reasonable and the authors clearly articulated limitations of a\ ailable data and other
information concerning the mine's location and operation. I found the consideration of the mine
during the various phases of development and operation and the discussion about potential
development of other mines in the area particularly insightful Inclusion of experiences from
other mining operations was also helpful in understanding the conclusions about potential
impacts of the mine and its operation o\ er time Additionally, the consideration of the potential
development of other mines in the area was particularly insightful and provided a good picture,
albeit not in depth, of potential cumulati\ e effects on aquatic resources in the Bristol Bay area.
RESPONSE: No changes suggested or required.
Parts of the report 011 the ecology of fish and aquatic ecosystems, road, and culverts - topics that
I am familiar with - were covered \ cry well and the conclusions about potential impacts of the
mine and its operation generally seemed justified The authors presented available data and
information on fish distribution and abundance relative to the presumed location of the various
components of the mine operation. Their analyses were appropriate but rather cursory, which is
not unexpected given the restrictions of time and available data. However, there are some
additional considerations and analyses that could be done, which I think would improve the
report. I identify these in answers to specific charge questions. Limitations of the results were
readily acknowledged Howe\ er. as mentioned above, there are additional limitations that
resulted from only considering selected potential avenues of impacts. These should be discussed
in the revision.
RESPONSE: The discussion of scope (Chapter 2), endpoints (Chapter 5), and uncertainties
(throughout the risk analysis and characterization chapters) has been expanded in the revised
assessment.
The authors do a good job of summarizing the scientific literature on salmon ecosystems, roads,
and culverts. Most of this is from studies in areas outside of Bristol Bay. Interpretations of the
findings were accurate. However, there was no discussion about potential limitations on the
application of the studies to the area being considered. For example, Furniss et al. (1991) deals
with roads in forest and rangeland settings. These are very different environments than Bristol
Bay, which suggests that road impacts will likely differ. Much attention is given to "headwater
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 24
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
streams" and their ecological importance (p. 5-19 - 5-21). Headwater streams for the area of
consideration need to be defined so that appropriateness of the application of the literature can be
better judged.
RESPONSE: Headwater streams in the study area are now more fully described (Chapter 7).
Because the potential mining described in the assessment would take place in an undeveloped
area, much of the literature is necessarily from areas outside of Bristol Bay. However, to the
extent possible we used examples from representative environments. With respect to Furniss et
al. (1991), though it focuses on forest and rangeland roads, it is a seminal publication on the
potential effects of roads, particularly as they relate to salmon. The general conclusions of that
paper should be applicable to the transportation corridor described in the assessment.
A major component that is missing from the report is consideration of the potential impacts of
climate change. Climate change is identified as a facl or in the conceptual model of potential
habitat and water quality effects associated with mine accidents and catastrophic failures (Fig. 3-
2D). However, I believe that it is a key factor thai will have influence in all aspects of the
assessment, not just failures and natural disturbance events (Fig. 3-2C). Il needs to be considered
in other aspects, such as water quality and a\ ailabilily. Climate change should also be included
in any analysis because it will be critical to build it into any monitoring program that is
developed in order to be able to differentiate its impact on salmon and their habitat from
potential impacts of the mine.
RESPONSE: Climate change projections and potential impacts are now included in Chapter 3
(Section 3.8). It is mentioned as an important externa! factor in the risk analyses presented in
Chapters 7, 9, and Hh and the issue is summarized in Box 14-2.
Charles Wesley Slaughter, Ph.D.
The Assessment (Volume I - Main Report) pro\ ides a fairly comprehensive review of fisheries-
driven issues, from the perspecti\ e of salmonids. Appendices (Volumes 2 and 3) are very
informati\ e The high significance of the Bristol Bay watershed, specifically of the Nushagak
and Kvichak ri\ er systems, for commercial fisheries on the global scale and for sport and
subsistence fisheries at the regional and local scales, was appropriately described.
RESPONSE: No changes suggested or required.
The potential risks and impacts are fairly and succinctly stated. Given the extremely long-term
nature of the projected Pebble project, and the irreversible changes which would be imposed to
the region, the risks seem, if anything, understated. I attribute this to the decision to focus this
Assessment on salmon and anadromous fisheries, with less attention on "salmon-mediated"
impacts - i.e., effects on indigenous culture, on wildlife other than salmon, etc.
RESPONSE: No changes suggested or required.
Chapter 2 (Characterization of Current Condition) provides only a superficial overview of the
landscape of the Bristol Bay watersheds; a reader would preferably have access to Wahrhaftig
(1965) or Selkregg (1976), as only two (relatively dated) suggestions, to gain a more
comprehensive understanding of the region.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 25
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Additional information on the physical environment of the region (e.g., geology,
vegetation, etc.), along with an expanded treatment of the regional landscape, has been
incorporated into Chapter 3 (e.g., Figures 3-4 through 3-7). Chapter 3 also includes the
suggested citations.
The "Water Management" section (4.3.7) seems cursory, highly generalized, and optimistic.
Statements such as "uncontrolled runoff would be eliminated"; "water from these upstream
reaches would be diverted around and downstream of the mine where practicable"; and
"Precipitation.. .would be collected and stored..." do not indicate actual (proposed) practices or
techniques, nor inspire confidence that actual runoff e\ cuts during "normal" conditions, let alone
during hydrologic extremes (such as a rain-on-snow e\ en I with underlying soils still frozen),
would be planned for or actually managed adequately
RESPONSE: Water management measures are more clearly described and discussed in
Section 6.1.2.5 of the revised draft, and in sub-sections for the mine components in the
scenarios. The assessment no longer contains a no failure scenario, so complete water
collection is not longer assumed Rather, standard and common practices are incorporated
Perhaps I missed it, but I found no acknowledgment of ilie potential presence of or consequences
of perennially frozen soils - permafrost in the Bristol liny watershed, or more specifically in
the Pebble ore deposit locale or the proposed transportation corridor. Selkregg (1976), Fig. 136,
shows soils of the Pebble locale as INT 2u. INT lu 11YI\ or S()U/2g-HYP - that is, well-
drained gravelly soils (INT) or well-drained acidic soils (SOI ) with interspersed peaty, poorly-
drained shallow discontinuous permafrost. There is abundant literature on the influence of
permafrost on engineered structures, roads, hydrology, etc. F.\ en if the bulk of the terrain
involved in the proposed Pebble mine, road and infrastructure project is founded on well-drained
gravelly soils, any interspersed permafrost-underlain terrain can prove problematic in terms of
landscape stability, potential erosion, and consequent structural, engineering, hydrologic and
water quality issues See Specific ()hser\ ations for a few suggested references in.
RESPOSSE: We have expanded our characterization of the soils and permafrost distribution
in the Bristol Hay watershed in Chapter 3 of the revised assessment. As part of this expansion,
we summarize the nature and distribution of permafrost by physiographic region.
While there is extensive discussion of a proposed transportation corridor, there was no mention
of construction of a major ai rlleld A project of this magnitude would undoubtedly require
development of a facility in close proximity to the mine(s) capable of handling C130 and
commercial jet passenger and cargo traffic, at least to the 737 class, if not 747. I don't know
what the footprint for such an airfield would be, but it would be substantial, and with requisite
roads, fuel handling, etc., would be a major project in itself. This would seem to be a logical
component of a comprehensive assessment of the potential Pebble project.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and construction
and operation of a new airport is considered outside the scope of the assessment. We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
effects if a new airport is proposed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 26
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
As noted in the Executive Summary, the Assessment does NOT address several major
components of the (hypothetical) Pebble project, including electrical generation and
transmission, a deep-water port, or "secondary development" and associated infrastructure which
would follow an initial mining project. A truly comprehensive analysis should incorporate full
analysis of these aspects. This Assessment is thus inadequate in terms of considering potential
broader consequences for the Bristol Bay watershed system.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and we have stated
throughout the text that areas outside of scope may also be important factors.
John D. Stednick, Ph.D.
The purpose of the document is not clearly stated in either the l-\eaili\ e Summary or the
Introduction. Need to specifically identify the document as an en \ iron mental risk assessment.
There is a misconception that it is a CWA Section 404(c) review, ml her than an environmental
risk assessment. The document should have the utility to inform future users of the risk to the
watershed resources from mining activities in the watershed The assessment can be used by
others for decision making purposes, and includes current and appropriate methodologies for all
identified stressors, such that study results can he duplicated And all stressors are evaluated to a
similar level of detail.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (ERA) in general has been added to C liapters 1 and 2, as well as the Executive
Summary. Section 1.2 includes information about the use of the assessment. The assessment
has been reorganized into two major sections (problem formulation, risk analysis and
characterization) to clarify where different chapters fall in the typical ERA process.
The document characterizes the potential en\ ironmeiital effects of an open pit mine over a
copper porphyry complex in southwest Alaska using a hypothetical mine design based on similar
ore deposits and mine complexes elsewhere. Proposed mine activity has been identified by the
Pebble Limited Partnership though Northern Minerals Dynasty and should be cited to improve
applicability of the risk assessment. Furthermore, a wider range of mining scenarios should be
developed and analyzed for environmental risk assessment. Environmental consequences were
estimated by the en\ ironmental risk assessment model approach for both 'no-failure' and
'failure' scenarios The Executive Summary concluded that the effects of mine development
resulted in significant salmon habitat losses. Potential effects on other aquatic species were not
identified. The assessment e\ aluated environmental risks under the development and closure
scenarios using large catastrophic events and did not include smaller, yet more frequent
excursions or system failures. Nor did the assessment look at the full range of mine development
scenarios, specifically what are the risks associated with a smaller underground operation?
RESPONSE: The assessment used the Pebble deposit and its characteristics, as described by
Northern Dynasty Minerals in the Ghaffari et al. (2011) report. That report is cited extensively
in both the original review draft and the revision. A median-sized mine (based on worldwide
mine sizes) has been added to the scenarios in the revised assessment. Because the number of
potential failures is extremely large, it is necessary to choose a representative set offailure
scenarios. The final document includes more failure scenarios (e.g., diesel pipeline failure,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 27
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
quantitative water treatment failure, and refined seepage scenarios) and explains why the
particular failure scenarios were chosen. Underground mining is a potential for any mining
site that has high-quality ore located at depth, but sources of potential impact considered in
scope for the assessment would be common for either a surface or an underground mine (e.g.,
water withdrawal, tailings dam failure, water treatment failure, seepage, etc).
The conclusions of the Executive Summary are strongly worded (e.g., pages ES 13 to 24), yet the
uncertainties presented later in the report make the strong conclusions tenuous. An expanded
discussion of uncertainties and limitations may temper those 'conclusions.'
RESPONSE: Each risk analysis chapter of the revised assessment now includes an
uncertainty section. The Executive Summary has been rewritten to reflect the revised
assessment text.
Site characterization/description of current conditions is too brief More information is needed
for a full site characterization. Any reader unfamiliar with the setting would not fully understand
the physical, biological, or ecological inventories and linkages in the study area. The risk
assessment of failure and no failure are covered in Chapters 5 and 6 with varying levels of detail
and substantiation of conclusions. Statements like "salmon is important in the human diet, thus a
salmon loss affects human health" seem like a weak argument, especially when additional
information in the appendix suggests a larger effect.
RESPONSE: Additional information on the region '.v physical environment has been added to
Chapter 3 (e.g., Figures 3-4 through 3-7), and additional information on the region's
biological communities from Appendices . I through C' has been incorporated into the main
text. The purpose of the appendices—to provide the detailed background characterization
necessary for the ecological risk assessment—has also been clarified in Chapter 2.
The Pebble T.imited Partnership has a large en \ iron mental baseline database (EBD), but does not
appear to be cited or used, Justification for the inclusion or exclusion of these data should be
made Reference is often made to various data, but these data were not presented.
RESPONSE: The EBD was used and cited more than 70 times in the May 2012 review draft
and even more in the revised assessment. Data from the PLPEBD concerning hydrology,
water quality. and biology of the streams on the site and along the transportation corridor
have been extensively incorporated into the assessment in the analyses in Chapters 7 through
11. However, to the extent possible, the assessment relies on peer-reviewed literature.
Review and revise the water balance section, which would include: 1) generating a diagram or
conceptual figure similar to page 3-7 to illustrate the potential effects of mine construction and
operation on surface and groundwater hydrology; 2) developing a quantitative water balance for
surface and groundwater resources; 3) incorporating seasonality (especially assessing the role of
frozen soil); 4) identifying hydrologic processes and their associated values (e.g., mm/yr) for
each component of the water balance in time and space, and then incorporating into a landscape
characterization; 5) demonstrating the interconnectedness of groundwater, surface water, and the
importance to fish habitat and stream productivity; 6) evaluating the influence of global climate
change on these hydrologic processes and rates; and 7) using this characterization demonstrate
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 28
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the expected hydrologic modification associated with the mine scenarios and infrastructure
development and closure scenarios.
RESPONSE:
1) We included schematics to illustrate potential effects of mine construction and operation of
surface hydrology, including effects via groundwater changes (e.g., Figure 6-5, Figures 6-8
through 6-10).
2) Our water balance focuses on surface water hydrology (including interactions with
groundwater). A comprehensive groundwater hydrology water balance is beyond the scope of
the assessment.
3) The core of our analyses is an annual water balance, but uv have maintained the simple
approach to seasonality used in the first draft of the assessment.
4) We have adopted a basic approach to representing the dominant hydrologic processes at the
mine site; a comprehensive representation of all hydrologic pntcesses is beyond the scope of
this assessment.
5) Throughout the assessment, we have identified and quantified the interconnectedness of
surface water, groundwater, and their importance to fish habitat and stream productivity.
6) A section on potential climate change effects has been added to Chapter 3.
7) We have used our updated water balance and hydrologic modeling approaches to estimate
expected responses of mine scenarios. infrastructure development and closure scenarios.
One common theme lhat emerged from the public comment session during the peer review
meeting in Anchorage. Ak was (he questioning ofllie document timing, from draft release to the
public comment period to the unannounced completion of a final document. These concerns
should he addressed in the new document
RESPONSE: This type of contextual information is not directly relevant to the ecological risk
assessment, but clarification of the timing and use of the assessment has been included in
Chapter /.
Roy A. Stein. Ph. D.
Accuracy of Presentation ()\ eral I. I was pleased with the accuracy of the presentation.
Typically, peer-reviewed citations to the scientific literature were cited as supportive
documentation for most all of the factual information (though the well-developed appendices,
e.g., Appendix E: Economics; Appendix I: Mitigation, could be used to far better advantage, see
below). Unfortunately, in the main report, many data are missing, especially with regard to
salmonid populations, their diversity (both across species wand within species across
populations), their relative population sizes, their distribution across the watershed, their vital
rates (i.e., recruitment, growth, and survival across life stages), and to what extent the Pebble
Mine and its associated activities will reduce these populations (for there is no question they will
indeed be reduced through both the mine footprint and all allied operations in the drainage), both
through impacts on individual populations and the overall production of salmonids (and other
fishes) in the Bristol Bay watershed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 29
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: We now include figures showing reported salmon species distributions and
salmon diversity by HUC-12 watersheds across the Nushagak andKvichak River watersheds
(Figures 5-3 through 5-8). Information on population sizes and vital rates are limited for the
region, but are reported where known. Due to lack of comprehensive estimates of limiting
factors across the impacted watersheds, population-level effects could not be quantitatively
estimated except for the most severe cases, where total losses of runs could be reasonably
assumed
Whereas I am relatively confident about accuracy of the fisheries information included, I cannot
comment in detail regarding the accuracy of the mining information or impacts on the Native
Alaskan cultures (though the impact of the mine on this culture was confined to fish-mediated
effects). That a Native Alaskan culture 4,000 years old is in jeopardy bothers me greatly; might
this complete subsistence way of life in the Bristol Bay \\ alershccl he eliminated with the
exploitation of the copper via open-pit mining0 In turn. what impacts i night there be on
subsistence users, other than Native Alaskans ' I a en though these sections seemed reasonably
well presented (with caveats above) and appropriately supported with citations, they do lie
beyond my expertise.
RESPONSE: No changes suggested or required.
My concerns about the document re\ ol\ e around issues that were not considered, i.e., Global
Climate Change, "In Perpetuity" issues, groundwater-surface water exchange issues (owing to
missing information), impacts of Routine Mine Operations in a more realistic setting, the
seemingly undue influence on a failure of the Tailings Storage facility, and other somewhat
more minor issues (see comments below). With any re\ ision. the authors should include this
information by eliminating redundancy (see below), thereby not increasing document length.
RESPONSE: We have thoroughly revised our approach to quantifying hydrologic responses
to the mine scenarios. We explicitly include groundwater-mitigated effects on surface waters.
Climate change projections and potential impacts are now included in Chapter 3 (Section 3.8),
and are considered as important external factors in the risk analyses as summarized in Box
14-2. See responses to the commenteds specific comments and to Dr. Stednick's hydrologic
comments below.
Clarity of Presentation. General I y speaking, I believe that the writing was intelligent, reasonably
insightful, and, more specifically, on task. One significant criticism with regard to the
presentation revolves around the organization of the document. As detailed below, the
organizational scheme lent itself to redundancy, from the Introduction through the various
chapters to the Integrated Risks Characterization chapter. Owing to this redundancy, the report
is likely too long by about 20% and any revision and shortening should serve to improve its
impact on readers.
RESPONSE: The assessment has been reorganized to eliminate redundancy and help clarify
the structure of the document.
The conceptual block and arrow diagrams (pages 3-7 to 3-11) were quite instructive. They
nicely demonstrate the interactions that occur within this mining scenario. The main report
would be much improved if text were to review this set of interactions. Clearly, a tremendous
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 30
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
amount of time, effort, and thought went into generating these diagrams and it is indeed a true
shortcoming of the main report that essentially no text was spent stepping through these
diagrams.
RESPONSE: Additional information on the use of conceptual models in the assessment has
been incorporated into Chapter 2. The more comprehensive conceptual models presented in
Chapter 6 (previously in Chapter 3) have been broken into their relevant component parts
throughout the risk analysis and characterization chapters, to better frame the specific
pathways addressed in each chapter. Additional conceptual models considering impacts on
wildlife, Alaska Native populations, and cumulative effects of multiple mines have been added
to Chapters 12 and 13.
Soundness of Conclusions. The conclusions were well supported, where there were published
data to support them. Many statements that could be interpreted as conclusions were often more
qualitative than desirable in a review document such as this one, ow i nu to the lack of information
(percent of salmonids lost owing to routine mine operations, impacts of mining and the
transportation corridor on wetlands, extent of urouiulwater-surface water disruptions, just to
name a few). Consequently, the soundness of the conclusions are somew hat compromised by a
lack of information.
RESPONSE: No changes suggested or required.
In addition, what would aid readers is a succinct statement of the purpose (risk assessment?,
impact on water quality and then through to fishes and beyond"', etc.) and scope (relatively
narrow impact of the mine on salmonids and ripple effects out from there) of the document early
in the initial chapter In so doing, both re\ iewers and readers will be informed as to the direction
of the document and thus better informed as they move through the document.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (ERA) in general lias been added to Chapters 1 and 2, as well as the Executive
Summary.
Finally, a portion of the public testimony complained about the process, specifically about the
time allowed for document review, the data reviewed, the validity of the hypothetical mine, etc.
Though I found most all comments to be somewhat disingenuous, I still would offer the
following advice: IYo\ ide a section upfront that deals with process issues surrounding the
review, i.e., explaining the constraints under which EPA was operating; without a section like
this, complaints, such as those described above (coming from just one segment of the public),
will go unanswered.
RESPONSE: Chapters 1 and 2 now clarify the purpose of the assessment and document how
public participation was incorporated into the process (e.g., Box 1-1).
William A. Stubblefield. Ph.D.
The document, "An Assessment Of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska," is a well-written, comprehensive document that employs a risk assessment-type
approach to an a priori evaluation of potential environmental effects on the ecosystem and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 31
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
potential receptor species (e.g., salmon) that may be affected by a potential copper mine located
in the Bristol Bay area of Alaska. This document is somewhat unique, in that no actual mine has
been proposed at the location and few site- or project-specific data are available. Therefore, no
specific information about development plans and potential operational and closure activities
associated with the mine are available. Rather, the authors have attempted to develop a
hypothetical mine and attempted to assess possible environmental effects associated with mine
development, operation, and closure. Although interesting, the potential reality of the assessment
is somewhat questionable. It is also unclear why EPA undertook this evaluation, given that a
more realistic assessment could probably have been conducted once an actual mine was
proposed and greater detail about operational parameters available The approach taken in the
document attempted to be comprehensive and evaluated a \ a rich of scenarios that may affect
aquatic resources in the Bristol Bay region. Given the importance of salmon populations in the
area, both from a financial and societal perspective, it is important that a comprehensive
evaluation of potential environmental effects associated with mine de\ elopment and operations
be conducted. The authors have attempted to conduct such a comprehensive evaluation and have
attempted to quantify (to the extent possible) the probability of adverse effects occurring.
Implementation of this approach is proper, and with the correct data, can provide a
comprehensive evaluation of potential environmental effects rnfortunately, because of the
hypothetical nature of the approach employed, the uncertainly associated with the assessment,
and therefore the utility of the assessment, is questionable.
RESPONSE: The EPA respectfully disagrees that the hypothetical nature of the approach
compromises the utility of the assessment.. ill mining plans are hypothetical. They change in
response to the results of assessments, regulatory requirements, public input, and unforeseen
conditions and events. They cease to be hypothetical only after the mine is closed At every step
in the process, assessments of the current plan are useful even though plans will change. This
assessment is based largely on a preliminary plan, published by Northern Dynasty Minerals
(Ghaffari et al. 2011). Although layout of mining components in a future mine plan may differ
sometvhat from the preliminary plan or the EPA scenarios, the main components of mining
would remain the same for open-pit mining (and underground mining would face the same
waste issues).
A variety of uncertainties and data needs were identified as a result of this effort and this alone
may provide sufficient value to justify the document and approach. For example, the authors note
that there is not an abundance of chronic toxicity data considered in deriving the EPA's ambient
water quality criteria lor copper and that there is an uncertainty associated with whether the
biotic ligand model (B l AI) adequately protects species of concern in Bristol Bay. It would seem
appropriate for EPA (perhaps in concert with industry) to develop the data to improve our
understanding of copper toxicity and to ensure that regulatory standards are, in fact, appropriate
for their intended use. A substantial body of data evaluating copper chronic toxicity has been
developed by the copper industry as a result of regulatory requirements driven by the European
REACH regulations. It may be beneficial for EPA to examine these data, thus resulting in a
reduction in any uncertainty associated with the evaluation of environmentally acceptable metals
concentrations. It should also be noted that similar datasets and biotic ligand models exist for
number of other metals that may be of concern at the Bristol Bay site.
RESPONSE: The EPA has examined the EU's 2008 Voluntary Risk Assessment of Copper
(the relevant REACH document). Although they do derive a chronic species sensitivity
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 32
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
distribution, it is because of the way they include and aggregate data, rather than the
generation of new data. In particular, they have no data for sensitive aquatic insects, so the
EU does not resolve that problem. The BLM was used for copper because copper is the
contaminant of greatest concern and because the copper BLM has been approved by the EPA
Office of Water. Other metals with BLMs, such as zinc and nickel, occur at much lower levels
in leachates.
One suggestion that would improve the document is that EPA should include a basic description
of the risk assessment process and the relationship between the risk assessor and the risk
manager, i.e., the decision maker. They must include a discussion of why the assessment is
being conducted, the decisions that will be informed, and u lial information they need from the
risk assessor.
RESPONSE: Additional contextual information for the assessment has been included in
Chapter 1, and additional information on ecological risk assessment has been incorporated
into Chapters 1 and 2. The assessment has also been restructured into problem formulation
and risk analysis and characterization sections, to make the assessment's structure as an
ecological risk assessment clearer.
Taken from the USEPA's Guidelines for Ecological Risk Assessment (EPA630/R.-95/002F; April
1998). Note 2nd sentence re: the role of I lie risk manager
"2.1. THE ROLES OF RISK \ /. I.Y. K ;ERS, RISK , ISSLSSORS, AND INTERESTED
PARTIES IN PTANNING
During the planning dialogne, risk managers and risk assessors each bring important
perspective to the table. Risk managers, charged with protecting human health and the
environment, help ensure that risk assessments provide information relevant to their
decisions by describing why the risk assessment is needed, what decisions it will
influence, and what they want to receive from the risk assessor. It is also helpful for
managers to consider and communicate problems they have encountered in the past when
trying to use risk assessments for decision making.
In turn, risk assessors ensure that scientific information is effectively used to address ecological
and management concerns. Risk assessors describe what they can provide to the risk manager,
where problems are likely to occur, and where uncertainty may be problematic. In addition, risk
assessors may provide insights to risk managers about alternative management options likely to
achieve stated goals because the options are ecologically grounded. "
RESPONSE: Section 1.2 in the revised assessment discusses uses of the assessment.
Dirk van Zvl, Ph.D., P.E.
Planning and designing a large mine, and especially one in a sensitive environmental setting such
as Bristol Bay, involves many iterations before a design evolves that is provided for further
public considerations. The EPA elected to use a design, developed for Northern Dynasty
Minerals Ltd. in a preliminary assessment prepared following the guidance of National
Instrument (NI) 43-101, as the basis for extensive evaluations in their risk assessment. The
resulting risk assessment can be at best characterized as preliminary, screening level, or
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 33
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
conceptual. There are both technical and process issues that must be addressed before this risk
assessment can be considered complete or of sufficient credibility to be the basis for a better
understanding of the impacts of mining in the Bristol Bay watershed.
RESPONSE: The EPA respectfully disagrees that the hypothetical nature of the approach
compromises the utility of the assessment. All mining plans are hypothetical They change in
response to the results of assessments, regulatory requirements, public input, and unforeseen
conditions and events. They cease to be hypothetical only after the mine is closed. At every step
in the process, assessments of the current plan are useful even though plans will change. This
assessment is based largely on a preliminary plan, published by Northern Dynasty Minerals
(Ghaffari et al. 2011). Although layout of mining components in a future mine plan may differ
somewhat from the preliminary plan or the EPA scenarios, the main components of mining
would remain the same for open-pit mining (and underground mining would have the same
waste issues).
With respect to the proposed transportation corridor, we note in the assessment that
"Although this route (the one proposed in the HP. I scenario) is not necessarily the only option
for corridor placement, the assessment of potential environmental risks would not be expected
to change substantially with minor shifts in road alignment.. \ long any feasible route, the
proposed transportation corridor would cross many streams, rivers, wetlands, and extensive
areas with shallow groundwater, including numerous mapped (and likely more unmapped)
tributary streams to Iliamna Lake (Figures 10-1 and 10-2). "
There are a number of items that require specific attention prior lo finalizing the report. While
my comments below pi o\ ide further details, from a global perspective the following aspects
must be addressed:
• A better sense about the range of impacts lYora a mining project that use not only
different technologies but also different lay-out options in its development than that
assumed in the I-PA Assessment:
• More attention lo the use of appropriate order of magnitude numbers reflective of the
quality of data, e g. less accuracy is obtained when 1:62,500 scale vs. 1:12,500 scale
maps are used;
• Correction of errors associated with misquoting and incorrect use of information in the
literature, and
• A critical re\ iew and lew rite of the Executive Summary to reflect the tone, terminology,
information sources and results of the main body of the report. One example of an error
and one of inconsistent terminology are:
o Page ES-1 < >. "Thus, the mine draws on plans published by the Pebble Limited
Partnership (PLP)", this is incorrect as the plans that were used were prepared for
Northern Dynasty Minerals Ltd.
o Page ES-10: ".. .our scenario reflects the general characteristics of mineral
deposits in the watershed, contemporary mining technologies and best
practices..The main body of the report emphasizes on a number of occasions
(such as Page 4-1,4-17) that "Our mine scenario represents current good, but not
necessarily best, mining practices".
My comments contained above and below are based on a single review of the report, i.e.
contractual time constraints were such that I could not afford a second review of the report. It is
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 34
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
therefore possible that there are other errors remaining in the report that I did not observe in my
review. It is therefore recommended that after making these corrections and edits that EPA
subject the report again to a rigorous independent review.
RESPONSE: The scenarios evaluated are meant to represent those expected to be present as
typical for mining porphyry copper deposits of this type. Although layout of mining
components at a site may differ somewhat from what we present in the scenarios, the main
components of mining would remain the same for open-pit mining (and underground mining
would have the same waste issues). Therefore, no change is requiredfor technologies
presented in the original assessment, and we have noted in the assessment that there could be
different layouts than what we have presented.
Errors and inconsistencies in sections of the document are noted and have been corrected in
the revised assessment. With regard to the terminology of "best", "good", or other terms for
the practices used, what was intended to be conveyed is that uv have assumed modern mining
technology and operations. The terms are qualitative when generally interpreted, or have a
regulatory meaning (for example, "best management practices" applies to the setting of
stormwater control, but not specific to mining sites), and thus we have eliminated their use in
the revised assessment. The assessment is being re-reviewed by the externa! expert reviewers.
Phyllis K. Weber Scannell, Ph. I).
My comments on EPA's draft document, An Assessment of I'oiential Mining Impacts on Salmon
Ecosystems of Bristol Bay, Alaska, follow a three-day peer ie\ iew meeting in Anchorage, AK.
On the first day of the meeting, the Peer Review Team heard testimony on the importance of the
resources in the potentially affected area and on possible effects of mineral development on the
fish and wildlife resources and on local residents The issues of mineral development are
complex, particularly with respect to protecting the environment and the interests of local
residents I understand and appreciate the complexity of these issues; however, the charge of the
Peer Re\ iew Team is to review EPA's draft document, An Assessment of Potential Mining
Impacts on Salmon Ecosystems <>J Bristol Bay, Alaska, and offer suggestions to strengthen the
report. My comments, included below. are focused on the accuracy and thoroughness of the
draft document
The document ''An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska" and the accompain ing appendices provide an in-depth and thoroughly documented
description of the environment and resources of the areas under consideration for mineral
development, although not in the entire Bristol Bay region. Appendices A and B are particularly
thorough in describing the salmon and non-salmon fishes in the region; the discussion of species
specific fish sensitivities to certain toxicants adds important information for future consideration
of project development.
RESPONSE: No change suggested or required
The assumptions for developing and operating large porphyry copper mine may not be aligned
with features of a future mining project. Too much emphasis was placed on effects of
catastrophic failures, such as failure of a tailings dam or pipeline, and too little emphasis on the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 35
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
need to identify and control seepage water, run-off from PAG (potentially acid generating) and
NAG (not acid generating) waste rock areas, and water treatment.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
seepage scenarios) and explains why the particular failure scenarios were chosen.
The document discussed effects of dewatering on suppressing stream flows and groundwater
inputs but did not consider effects of the discharge of treated wastewater. The section on
hydrology illustrates the need for more complete hydrologic information before any project
development. The need for bypassing all clean water sources around a development site should
be addressed.
RESPONSE: The revised assessment more clearly presents that dean water would be diverted
around the site, retained in settling ponds, and released following settling and/or treatment, if
required. Discharge of treated wastewater is analyzed and discussed in greater depth in the
new Chapter 8, Water Collection, Treatment, and Discharge. The assessment is based on the
best hydrologic information available for the site; however. additional hydrologic information
may be available and/or acquired for any future mine plan in this watershed'. II e agree that
detailed hydrologic information is critically important for responsible project development. We
have updated our hydrologic analyses to represent the probable influence of mine scenarios
on surface water/groundwater interaction.
As stated in my response lo charge questions. I helie\ e llial the two most important questions for
mineral development in this region are: can a mine he designed and operated for future closure?
and, if not, is it acceptable to develop a large porphyry copper mine in a region of high value
salmon habitat that will essentially require perpetual treatment? These two questions must be
addressed when considering protection of the fish, wildlife, and human resources of the region.
RESPOXSE: II e agree that these are important questions to be addressed but they are risk
management, not risk assessment, questions. The purpose of the assessment is to evaluate
risks to the salmon fishery from large-scale mining. Risk management decisions will be made
during the permitting process. Xo changes to the assessment were made in response to this
comment.
Paul Whitney. Ph.D.
Response (with a wildlife perspective) - The main document is fish centric and it should be,
given the importance of salmon in the Bristol Bay ecosystem. Wildlife (aquatic, wetland and
upland species) and terrestrial resources related to potential mine and haul road impacts are
glossed over. The summary write ups for several species of wildlife (Appendix C) are very good
regarding natural history and some potential impacts. Information in Appendix C tends to focus
on the proposed mine site and less on the proposed haul road and game management units in the
Kenai Mountains.
RESPONSE: Direct effects on wildlife and terrestrial resources are outside the scope of the
assessment, as clarified in Chapter 2. Effects on wildlife are now treated in Chapter 12.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 36
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
USFWS RESPONSE: We acknowledge the comment regarding quality of Appendix C.
Information in Appendix C is intended to focus on the entire Nushagak and Kvichak River
watersheds to the extent that data exist. To the extent that a potential mining-related road is
within the Nushagak and Kvichak River watersheds, information about selected wildlife
species is included in Appendix C. Information about selected wildlife species on the Cook
Inlet side of the Chigmit Mountains is not included in the wildlife report. The Kenai
Mountains are not in the Nushagak and Kvichak River watersheds.
A variety of authors have obviously contributed to the documents and it appears that the
direction given to them or their interpretation of goal statements varies. For example, if one of
the goals of the assessment is to evaluate the risk to wildlife due to risk to fish (Executive
Summary, page 1, last para) it's not clear why so much verbiage in Appendix C (wildlife) is
devoted to species such as caribou that are not closely associated u ith fish. Information in
Appendix C could be used to assess direct impacts if the scope of I lie assessment is expanded.
For example, if the goal is to assess the impact of potential mining on the ecosystem (see
Executive Summary page 1, para 1), the information on caribou in Appendix C is more relevant.
The apparent diversity of goal statements cited in the main assessment gi\ es mixed messages
regarding the clarity of the presentation (see more detailed discussion below )
RESPONSE: As the commenter notes, the scope of the assessment is focused on potential
risks to salmon from large-scale mining and salmon-mediated effects to indigenous culture
and wildlife. EPA agrees with the commenter that direct effects on wildlife are likely to be
important and that Appendix C (now a stand-alone USFWS document) provides useful
information for an evaluation of direct effects on wildlife from large-scale mining. We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
direct effects. The revised assessment acknowledges the potential for direct effects on wildlife
as well as risks due to fish, but states that these effects are outside the scope of the assessment.
USFIIS RESPONSE: The scope of. 1 ppendix C' is broader than that of EPA '.s assessment
because it is a I Sl'tt S document prepared to serve various purposes, including statewide or
regional land use planning, completion of environmental documentation for permitting of
development projects, and activities related to Landscape Conservation Cooperatives in
Alaska, The former Appendix C' is now a separate USFWS report which is cited by the
assessment but is no longer an appendix of the assessment. However, information in the
USFWS report document has been used by EPA to provide a more complete assessment of
overall watershed resources at risk due to potential mining, and to strengthen the assessment
of risks to wildlife from fish-mediated effects of the mine in the revised assessment.
The charge question related to wildlife asks for an evaluation of the risk to wildlife due to the
risk to fish. If the risk to fish cannot be quantified because there is little or no demographic
information, then any evaluation of risk to wildlife can't be quantified and must be qualitative.
Merely stating that a qualitative increased risk for fish will also result in a qualitative increased
risk for wildlife is not adequate. I am not satisfied with such an obvious and general conclusion. I
do not understand why the scope of the main document is limited to an indirect evaluation of
fish-caused risk to wildlife. The following responses to charge questions leans more toward an
ecosystem evaluation that includes, not only risk of fish to wildlife, but also risk of direct
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 37
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
wildlife and vegetation loss to fish and other direct risks to wildlife, such as noise and human
presence.
RESPONSE: EPA acknowledges that there are numerous potential direct risks to wildlife
from large-scale mining. However, this evaluation is outside of the scope of the assessment.
The revised assessment provides a clearer explanation of the reasons for its defined scope.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 38
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
2. RESPONSES TO CHARGE QUESTIONS
Question 1. The EPA's assessment focused on identifying the impacts ofpotential
future large-scale mining to the fish habitat and populations in these watersheds. The
assessment brought together information to characterize the ecological, geological, and
cultural resources of the Nushagak and Kvichak watersheds. Did this characterization
provide appropriate background information for the assessment? Was this
characterization accurate? Were any significant literature, reports, or data missed that
would be useful to complete this characterization, and if so what are they?
David A. Atkins. M.S.
Based on my general understanding of the watersheds. I consider the general background
information presented in the Assessment accurate and sufficiently complete for the endpoints of
this watershed assessment in the following areas
• General view of Pacific salmon populations
• General view of resident (non-anadromous) fish
• Wildlife populations
• Native cultures
RESPONSE: No change suggested or required.
The Assessment also describes the current economics of the watershed, including commercial
and sport fishing and subsistence activities.
RESPONSE: No change suggested or required.
Additionally, the report highlights several general aspects of the area that make the fishery
unique in both its abundance and diversity:
• The unique hydrology of the area (strong groundwater and surface water interaction) that
contributes to stable flows and temperatures favorable for salmon reproduction.
• The importance of anadromous fish in transferring marine-derived nutrients to upland
areas and thus providing nutrients to areas that would naturally be nutrient poor.
• The lack of roads and i n I'm structure that make the area unique as one of the few intact
ecosystems remaining in the world, and possibly unique for this type of fishery.
RESPONSE: No change suggested or required.
It would be helpful in the background section to better describe the uniqueness of the Bristol Bay
watershed ecosystem in the Pacific Northwest. This could include a description of other similar
ecosystems in the region that have undergone development and documentation of any changes in
fish populations associated with this development. The Assessment does mention the Fraser
River as an analogue, but the scale of development in this watershed, and even the success of the
salmon fishery, seems to be a point of contention, with some saying mining and fish coexist, and
other saying the impacts are severe.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 39
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The unique conservation value of Bristol Bay fisheries is now discussed in
Chapter 5.
It would also be helpful to better explain fish resources in the proposed project area in
comparison to other areas within the watershed. I understand some of the necessary data may not
be available for the project area. It would be helpful to know, however, if the habitat in the
project area is typical, exceptional, or inferior to that in other areas of the watershed.
RESPONSE: We now include figures showing reported salmon species distributions and
salmon diversity by HUC-12 watershed, across the Nushagak andKvichak River watersheds
(Figure 5-3; Figures 5-4 through 5-8). It is informative to note that salmonid diversity is
relatively high in the project area. Information on population siz.es and vital rates are limited
for the region, but are reported where known. In addition, uv include summary statistics and
figures of stream and valley characteristics across the assessment area (Section 3.4), and
compare stream attributes in the project area to those of the larger watersheds (Section 7.2.1).
These results generally illustrate that the project area contains streams of a size and gradient
well within the range of suitability for salmon, as amply demonstrated by the distribution of
spawning and rearing salmon within the project area streams (Figures 5-4 through 5-8).
Regarding geological resources, the report describes the IVhhle deposit and fi\ e other mineral
deposits in the Nushagak and Kvichak watersheds. It would he helpful to know if there are other
mineral resources or oil and gas resources in the Bristol Bay watershed as a whole that could also
be exploited. It would also be helpful to describe the portion of the watershed that is off-limits to
development due to park and protected area status \ s those lands that are open to mineral
development.
RESPONSE: The scope of the assessment was to evaluate the potential impacts from large-
scale mining on salmon resources; thus, consideration of prospective oil or gas development
in the area was outside the scope. The mineral resources identified in the assessment are those
in the Bristol Hay watershed that have had some level of identification or exploration at this
time. Mine claims within the Nushagak and Kvichak River watersheds are shown in Figure
13-1 and discussed in greater detail throughout Chapter 13. The assessment assumes that
mining would occur on lands open to mineral development.
Protected areas within the Bristol Kay watershed and the Nushagak andKvichak River
watersheds are shown in Figures 2-3 and 2-4. We have clarified in the text that the Nushagak
and Kvichak River watersheds represent the least-protected area of the Bristol Bay watershed
Other state documents exist that map out areas in the Bristol Bay watershed off-limits to
development.
Steve Buckley, M.S., CPG
The background information presented in the characterization of the ecologic, hydrologic, and
geologic resources is overly broad in scope. Specifically, the descriptions of the relationship
between landforms, streams, and surface water and the interaction with groundwater are
mentioned as very important to fish in the watersheds, yet there is insufficient detail to assess
these interactions and consequently, the characterization of these resources is weak. There is
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 40
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
more detailed information available in the Environmental Baseline Document (EBD) regarding
the relation between landforms, streams, groundwater, and fish habitat in the watershed.
RESPONSE: Descriptions of the region's physical environment have been expanded in
Chapter 3. We provide additional detail on the broad-scale habitat characteristics of the
watersheds, but providing the detail necessary to assess groundwater interactions
comprehensively is beyond the scope of this document, and data are not available to do so.
The revision improves the treatment of this issue but given how critical this information is to
understanding the watershed, fish habitat and the potential impacts from mining, a more
comprehensive analysis will be required at some point.
Courtney Car others, Ph.D.
The background information presented on the ecological and geological resources of the
Nushagak and Kvichak watersheds appears to be appropriate and accurate The report notes that
there is a lack of quantitative data on salmonid populations in this region, a lack of a full
identification and characterization of salmon presence, spawning, and rearing areas, and a lack of
detailed understanding of how local stream and river system features (e.g., temperature, habitat
structure, predator-prey relationships, limiting factors) affect salmonid production in the region.
Further, climate change is noted to be affecting local conditions These unknowns are important
to stress throughout the report.
RESPONSE: Each risk analysis chapter of the revised assessment now includes an
uncertainty section. Climate change is now incorporated more explicitly as an important
external factor that could interact with mining impacts (Box 14-2).
The cultural characterization presented in Appendix D presents detailed information on historical
and contemporary Yup'ik and Dena'ina communities of this region, stressing the centrality of
salmon and subsistence in these cultures. This assessment benefits from the time-depth of
relationships developed by Boraas and Knott. Overall, this section of the report is based on
standard ethnographic methods, although the research design and analysis could be explained in
more detail (and described in a separate methods section). The "voices of the people" sections
are helpful to present directly the perspectives given by local people. These quotes reveal the
complexity of subsistence and contemporary village concerns in this region. At times, the
cultural assessment can minimize this complexity.
RESPONSE: Additional detail was added to the methodology section of Appendix D.
As detailed in the specific comments below, potential risks and impacts to subsistence are
underestimated and at times framed in the report as primarily ones of physical health and
economic factors. As described in Appendix D, harvesting, processing, sharing, and consuming
wild foods are central to social, cultural, spiritual, psychological, and emotional well-being in
Yup'ik and Dena'ina cultures. The subsistence lifestyle is considered central to the health of the
people and communities of this region. This is particularly important to note for indigenous
communities who continue to cope with the legacies of colonialism. This point is made in
Appendix D (but at times could also be strengthened there, as suggested below), and is
articulated in some of the quoted interview material.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 41
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The assessment text regarding the importance of the subsistence way of life has
been expanded to recognize the centrality of subsistence to the social, cultural, and spiritual
well-being of the indigenous cultures.
Recent data on subsistence harvests, use areas, and local context collected for the PLP
Environmental Baseline Document (as well as evaluation and discussion of such data, e.g.,
Langdon et al. 2006) and by the Alaska Department of Fish and Game (e.g., Fall et al. 2012)
would be a useful addition to the cultural characterization. Other studies of local traditional
ecological knowledge (e.g., Kenner 2005) may help to supplement the assessment of the
abundance and distribution of fish species in this region, or to supply information on other less-
studied freshwater fishes. Recent research on the contemporary salmon-based livelihoods of the
region (e.g., Holen 2011, 2009a, and 2009b; Hebert 2008: Donkersloot 2005) would also be
helpful to include. An inclusion of case studies of salmon-hased cultures that have suffered
depletions of their resource base would add to the presentation of likely fish-mediated impacts to
culture (e.g., Colombi and Brooks 2012).
RESPONSE: The suggested references were consulted during the revision of the report and
the discussion of subsistence has been expanded. In addition, case studies have been cited
where applicable in the discussion of potential effects to indigenous cultures in Chapter 12.
Appendix E also characterized the economic baseline of the region. Why is this dimension not
asked about here?
RESPONSE: The focus of the assessment is potential effects on salmon from large-scale
mining. There are two secondary endpoints: salmon-mediated effects on wildlife and Alaska
Native culture. The economics related to potential salmon-mediated effects are not evaluated
because they are outside the scope of the ecological risk assessment. Appendix E presents
information regarding the economic value of salmon is presented as background for the
descriptive material in C liapter 5. and could be used as a basis for future analyses. However,
this assessment does not include an economic endpoint.
Dennis I). Dauble, Ph.D.
As noted in the iippnmch, characterization of and risk to ecological resources emphasized salmon
and other important sport and commercial fish species. Consequently, the description of non-
salmonid species generally lacked estimates of population size, except for sport and subsistence
catch statistics. There was a long list of other resident fish in Appendix A, but their role in the
Bristol Bay watershed (including the Nushagak River and Kvichak River watersheds) is not
described in any detail there or in the main report. Available data on known or perceived
ecological interactions among salmonid and resident fish should be included in the assessment.
RESPONSE: The assessment endpoints—salmonid fishes and their effects on wildlife and
Alaska Native cultures—have been clarified in Chapters 2 and 5; other fish species are thus
outside the scope of the assessment. However, we recognize in the text that other fishes (as
well as other biota) are important components of the ecosystem, and have included a table of
all documentedfish species in the region in Chapter 5 to better reflect the fish fauna in the
region.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 42
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Another limitation to the salmon-centric assessment is that risk assessment endpoints, described
in Chapter 3 of the main report, do not address other aquatic ecological resources. Consequently,
while there was acknowledgment of ecological dependencies among salmon, other fishes, and
land mammals, very little information was provided on primary and secondary production
processes of aquatic communities. For example, the relative importance of marine-derived
nutrients (MDN) in the form of salmon eggs and carcasses is discussed, but there is only brief
mention of aquatic insects in the diet salmonid species. What nutrient levels occur in these
stream systems with and without MDN?
RESPONSE: We recognize that nutrient status, and more important prey availability, is a
critical component of habitat capacity for fish in these systems, and may be strongly driven by
salmon derived nutrients. We concur that more information is needed regarding potential
limiting factors for salmon productivity and capacity. and that food availability may be one
such factor. The role of aquatic invertebrates in the diet of sahnonids receives more attention
in the revised draft, and is an essential part of the risk assessment for miter treatment and
discharge, given the relatively high sensitivities of aquatic invertebrate taxa to metals.
However, because water chemistry data may not provide a complete picture of trophic status,
particularly where direct consumption of salmon flesh, eggs, andfry is of such high
importance as it is in many of the area streams, we determined that nutrient status of area
streams is outside the scope of this assessment.
A description of major groups of aquatic in\ erlchralcs in terms of hiomass and seasonal
abundance should be included in the main report I'ni lher. aquatic and terrestrial food webs and
linkages need more embellishment One approach might he to add narrative text with the
conceptual model discussion, including descriptions of community structure, function, and
biomass.
RESPONSE: Additional detail on food webs is beyond the scope of this assessment (as
detailed in Chapters 2 and 5). Further, available data are inadequate to assess risks at that
level of specificity. For example, there are no acute copper toxicity data for any aquatic insects
and only one old chronic value for a caddisfly.
More detail on ri\ er and lake limnology would be helpful. For example, the hydrology of the
watershed is mainly limited to a brief discussion of salmonid habitats. The geology of the basin
emphasizes geology of mining areas and mineral processes. A more landscape-based description
is warranted given the importance of geology to surface water processes and groundwater
movement. The report u on Id benefit from having a summary table listing lake size/volume and
river length/discharge for watersheds potentially affected (and not affected) by mining activities.
RESPONSE: We now include maps of geology and estimated mean annual flow for the study
region (Chapter 3).
Also missing were specific habitat requirements for rearing of juvenile salmon. A brief
description of where pink and chum salmon spawn and rear in the Bristol Bay watershed relative
to other salmon species should be included in the main report. There was nothing in Appendix A
on where coho, pink, and chum salmon reside within the Bristol Bay watershed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 43
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Identified spawning and rearing habitats for the five Pacific salmon species are
reflected in Figures 5-3 through 5-8, and additional text on salmon life histories has been
included in Chapter 5.
Each appendix has a wealth of supporting information and could serve as a stand-alone
document. However, having to work back-and-forth between the main report and appendices to
interpret critical aspects of the assessment presents a challenge. Don't assume the average reader
will read (and interpret) these appendices. To help remedy, the authors of the main report should
strive to directly cite relevant information (and/or a specific appendix) that supports their
conclusions.
RESPONSE: Additional information from Appendices . I and li lias been pulled into Chapter
5 of the main assessment. In addition, the purpose of the appendices has been clarified in
Chapter 2.
Gordon H. Reeves. Ph.D.
The assessment, which included the report and appendices, was comprehensi\ e and thorough
regarding the ecological resources of ilie Vushagak and k\ icliak watersheds. The best available
data on fish numbers and distribution ( Alaska Dept of I'isli and Game's aerial escapement
counts, records from the Anadromous \Vtilers Catalog and Alaska Freshwater Fish Inventory,
and the Environmental Baseline Document of the Pebble Limited Partnership (2011)) were used
for the assessment. These data formed the foundation for much of the assessment on potential
impacts to anadromous salmonids and their freshwater habitat in these watersheds and their
characterization appeared to be accurate The authors also appeared to have thoroughly
identified and considered all of the appropriate literature
RESPOXSE: Xo change suggested or required.
Charles H 'eslev Slaughter, Ph.l).
If only Volume I (the Main Report) is considered, the characterization of some aspects of the
Nushagak and K\ icliak watersheds would have to be termed cursory. Chapter 2, Volume 1
(Characterization of Current Condition) provides only a superficial overview of the landscape of
the Bristol Bay watersheds, a reader would preferably have access to Wahrhaftig (1965) or
Selkregg (1976), as only two (relatively dated) suggestions, to gain a more comprehensive
understanding of the region. Similarly, Volume 1 provides a relatively superficial discussion of
non-fish wildlife concerns, or human/cultural concerns
RESPONSE: Additional information on the region's physical environment from Selkregg
(1974) has been included in Chapter 3. We have also clarified that our discussion of biological
communities focuses on the assessment endpoints, as defined in Chapters 2 and 5.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 44
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
By contrast, the information provided in Appendices A-H appears to be comprehensive and
complete for each subject field. (Appendix I appears to be a general "template" summary, not
tailored to the Bristol Bay watershed environment).
RESPONSE: The purpose of the appendices vs. the main assessment document has been
clarified in Chapter 2. Appendix I is not meant to be specific to any given region, but discusses
options that are possible and notes that their applicability is dependent on site-specific
constraints. What would be chosen for the Bristol Bay watershed environment, given a mining
plan and permit application, also would be dependent on regulatory decisions.
As noted in the Executive Summary, the Assessment docs NOT address several major
components of the (hypothetical) Pebble project, including electrical generation and
transmission, a deep-water port, or "secondary development" and associated infrastructure,
which would follow an initial mining project. A truly comprchcnsi\ c analysis should
incorporate a full analysis of these aspects.
RESPONSE: The scope of this assessment was tailored to its purpose, as clarified in the first
two chapters. We would expect that a full evaluation of any future mining permit applications
and subsequent National Environmental Policy . Ict Environmental Impact Statements would
consider these components.
JohnD. Stednick, Ph.D.
The site characterization needs to be expanded. The report needs to better characterize the
physical setting. There are a variety of data sources that can be used to better describe the
physical setting. It would be useful to see geology, ueomorphology, soils, vegetation, digital
elevation maps, hypsometric cur\ es of the watersheds in question, streamflow data, and
precipitation data -especially storm e\ cuts and water quality data for surface and groundwater
over time and space. Various geographical information system maps would be useful here.
RESPO N SE: Maps displaying information about the physical setting have been added in
Chapter 3.
The salmon populations and habitat linkage needs to be better documented since many of the
mine impacts are resulted from hydrologic modification. Figures 3-2A to 3-2E represent good
thinking and an understanding of the linkages and potential effects of mining on these resources.
The linkages to indigenous peoples is illustrated in Figure 3-2E, but little text is presented,
referring the reader to the Appendix. The other conceptual models are not adequately addressed
in the text. These flow charts provide an opportunity to present processes and linkages as related
to potential effects of mine development activity and need to be developed within the text.
Indeed, they seem to stand alone with little discussion of potential effects. Additionally, not all
charts have adequate materials in the appendix for coverage, thus the variability in resource
coverage is inconsistent and infers either a writing bias or data (lack of) bias.
RESPONSE: Conceptual models are now linked with relevant text, and are included in each
of the risk analysis and characterization chapters.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 45
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The assessment concludes that a hydrologic modification will have detrimental salmon habitat
consequences. The groundwater contributions to streamflows are important, both hydrologically
and ecologically. Additional streamflow and groundwater data are needed to represent this
linkage. Similarly, additional water quality data over time and space are needed and should
include water hardness for metal standards. Depth to groundwater as related to streamflow, age
dating of waters, and streamflow modeling would all be useful to illustrate the groundwater
upwelling and hyporheic exchanges.
RESPONSE: We have incorporated a figure illustrating modeled and observed groundwater
upwelling zones (Chapter 7).
Site disturbance will be significant, yet there is no discussion of soil erosion. Soil erosion and
subsequent suspended sediment transport would have the potential to have significant effects on
water quality, channel delivery efficiency, salmon, salmon habitat. and metal transport. There is
a generic discussion of road construction related lo erosion. In.il road standards, road location,
road usage, road maintenance (salting, grading, or watering), and length of roads would help in
the risk assessment.
RESPONSE: Soil erosion on the mine site is not assessed because the scenario prescribes that
runoff will be directed to retention basins. Salts used to reduce dust and improve winter
traction on roads are discussed in Section 10.3.3 (Chemical Contaminants in Stormwater
Runoff). Road usage and length are also factored into the risk assessment (e.g., in the
assessment of chemical spills (Section 10.3.3) and potential impacts from dust (Section
10.3.5)). Potential mitigation measures for stormwater runoff, erosion, and sedimentation are
discussed in Box 10-3.
Are any endangered or threatened species present, either state or federally listed?
RESPONSE: Text has been added to C liapter 5 stating that there are no state or federal
endangered or threatened species in the region.
Roy A. Stein, Ph.D.
Overall Characterization. The characterization of the resources of the Nushagak and Kvichak
watersheds was appropriate and accurate in the ecological arena save for the issues discussed
below. Geological and cultural resources seemed adequately characterized, but they are not
within my expertise. I'inalK. given the emphasis on these two watersheds (not the entire Bristol
Bay watershed), might there he some consideration of a more circumscribed document title?
RESPONSE: The assessment deals with multiple spatial scales (now clarified in Chapter 2).
The Bristol Bay watershed is the largest spatial scale considered in the assessment, and it is
the only one that encompasses all of the issues discussed in the document.
Broad Scale Comments:
Global Climate Change I. Risks to salmonids seem far greater than what is reviewed
throughout this portion of the document. Missing, in my view, is any consideration of Global
Climate Change, especially in light of the expected life of the mine (25-78 years), applied
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 46
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
directly to the Bristol Bay Watershed (save for a brief mention on page 5-28, 2nd full paragraph).
Given our current understanding, general changes likely include more intense precipitation
events and increased temperature (and then of course, all that follows from these two changes
and as models become more sophisticated, more specific geographically localized impacts could
be assessed). With more intense storms come a greater likelihood of a failure of Tailings Storage
Facilities (i.e., commensurate with more frequent and more intense flooding), more acidity from
Pre-Tertiary waste rock (which will enter quite vulnerable, poorly buffered streams), and greater
sediment influx into streams (and increasing fines in the gravel by as little as 5%, quite a small
proportion, "...causes unacceptable effects on salmonid reproduction" (page 8-6; also see
Chapter 7), which could occur during "routine operations", especially in light of the fact that
sediment influx into streams is a cumulative process). Increased stream temperatures, depending
on the absolute increase over a period of 78 years (and beyond, see "in perpetuity" comments
below), could lead to reductions in salmon spawning success, as extant populations are
specifically adapted to the current temperature regime. As is apparent, both increasing intensity
of storms and increasing temperature will likely compromise salmon spawning success, and
growth and survival of their offspring in the freshwater environment of Nushagak and Kvichak
rivers.
RESPONSE: Climate change projections and potential impacts are now included and
discussed in Chapter 3 and are included as important external factors in the risk analyses
presented in Chapters 7, 9,10 and 14.
What this would entail, at the very least, is a discussion of a monitoring system to quantify the
impacts of Global Climate Change whose impacts on the ecosystem can then be differentiated
from mine impacts. My concern is that if the mine is built, all neuati\ e impacts of the mine on
salmonids, etc., could be attributed to Global Climate Change rather than the true culprit which
would be the mining activities.
RESPO N SE: Climate change projections and potential impacts are now included in Chapter
3, and Box 14-2 includes a discussion of the need for future monitoring to differentiate
climate change effects from large-scale mining effects.
Global Climate Change II. Indeed, climate change is affecting Alaskan salmon as
demonstrated (in a paper that just appeared online July 11, 2012) by a loss of a late-migrating
population of pink salmon in a small stream near Juneau, in favor of an early-migrating one.
Genetic evidence supports this explanation for Kovach et al. (2012) had 17 generations of data
(since 1979) showing the reduction of the September spawners in favor of the late-August ones
in response to increasing stream temperatures. As Kovach et al. (2012) write in their concluding
paragraph:
"We no longer observe the clear phenotypic distinction between early- and late-
migrating individuals that was once present in the system. Apparently, the very-
late-migrating phenotype has been greatly reduced or potentially lost. Although
microevolution may have allowed this population to successfully track
environmental change, it may have come at the cost of a decrease of within-
population biocomplexity - the loss of the late run. This is not a surprising result;
by definition, directional selection will decrease genetic variation. However, it
does highlight the importance of maintaining sufficient genetic and phenotypic
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 47
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
variation within populations in order for them to have the ability to respond to
environmental change."
The ramifications of this work are obvious. As pointed out in the report (pages ES-8, 2-22, 5-28
as just a few examples), the exceptional quality of the Bristol Bay salmon stocks depend on the
pristine quality of a set of quite diverse aquatic habitats, which has led to the development of
genetically diverse stocks of salmon within species, each uniquely adapted to particular habitats.
Reducing this variability by mining on top of the rivers that produce >50% of the wild sockeye
salmon in Bristol Bay serves to reduce the flexibility with which these stocks respond to any
environmental change (most notably Global Climate Change), and most notably during the time
course of the Pebble Mine.
RESPONSE: Climate change projections and potential impacts are now discussed in Chapter
3 and include these points.
Groundwater Exchange. One of the key aspects of this system is the importance of
groundwater exchange with surface streams and this groundwater contributes mightily to
salmonid egg incubation success and survival (page 2-21). Simultaneous with this is the fact that
the water demands of the proposed mine will require more than just surface waters available to
it, but rather the mine will have to exploit groundwater resources to support its operations. This
is yet another risk to salmonid success lor reduction in the a\ ai lability of groundwater will lead
to increased temperatures in summer (see pages 3-7. 5-2S. 5-21)) and less inviting overwinter
habitats (pages 5-20, 5-29), further exacerbating both mining and climate change effects.
RESPONSE: We have updated our hydndogic analyses to represent the probable influence of
mine scenarios on surface water, groundwater interaction. Climate change projections and
potential impacts are now discussed in C liapter 3.
Exploration Effects. During the public testimony segment, several Alaskan Natives argued that
impacts owing to exploration ha\ e already occurred A series of points were made: 1)
exploration equipment was left behind, despoiling the landscape, 2) noise from helicopters
frightened moose making them less \ ulnerable to exploitation, and 3) habitat change has already
begun just due to exploration actix ities
RESPONSE: EP. I acknowledges this testimony, but potential or actual impacts of exploration
activities are outside the scope of the assessment as defined in Chapter 2.
"In Perpetuity." I'oil owing up on the idea of increased risk (see previous points) to salmon, I
struggled with the idea of this mine being monitored and maintained "in perpetuity" (e.g., pages
ES-2, 4-32, 4-34). First, this relates directly to the Global Climate Change issues, in that these
changes likely will continue to build through time, further exacerbating negative impacts on
salmon. Even without climate change, salmon are in peril from mining operations in the
Nushagak and Kvichak rivers; with climate change, the cards are stacked against them.
RESPONSE: The post-closure phase of mining begins when reclamation is completed and
monitoring and maintenance commences using the controls put into place during closure;
exactly how long the site would require monitoring and maintenance is unknown, and thus
may be 'in perpetuity'. There are no existing examples from which to evaluate success of
treatment in perpetuity. No mine in Alaska has maintained a tailings pond into post-closure,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 48
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
although one small mine did maintain a pond during a many year hiatus from operations.
Under AS 72.90.040, financial assurance is required to be sufficient to cover expenses for as
long as treatment need is predicted, even into perpetuity (e.g., Red Dog Mine). Maintaining a
water cover over the tailings is a part of the reclamation and closure plan for the Red Dog
Mine. The comment is noted and understood. No changes suggested or required.
Second, what regulatory or institutional mechanisms currently available place the responsibility
of these efforts on the corporation "in perpetuity"? Because mining companies come and go,
might there be mechanisms that come into play if this particular company goes bankrupt? Might
there be some sort of bonding process that protects the en\ iron men 1; from the mine's remains into
the long-term future? If not, should new legislation be piiisiied'' Acknowledgement of this
important issue should be front and center in the document, in my \ iew.
RESPONSE: There are many requirements that have to be met including compliance with the
CWA § 404(b)(1) Guidelines and adequate financial assurance under AS 72.90.040. The
former would lead to the least environmentally damaging, preferred alternative and the latter
to having adequate financial resources to cover the cost ofperpetual treatment. These issues
would be addressed in a permit process. Our purpose in the assessment is to evaluate the
potential effects of the primary features of a mine, assuming conventional modern mitigation
measures. Additional information on the regulations and financial assurance issues
associated with mining has been added in C hapter 4 (Boxes 4-2 and 4-3). The comment is
noted and understood.
Third, I began the re\ iew process with idea that the mine would be built, would capture its
resources, and then would end by restoring the site The scenario that includes monitoring and
maintenance 1,000 years into the future continues to bother me. One solution that comes to mind
is that Federal or state government would be charged with these monitoring and long-term
maintenance activities, paid for by a hefty tax on the minerals removed from this site.
RESPONSE: C urrently, the solution is the requirement and provision of adequate financial
assurance (under. I.V 72.90.040, when speaking specifically about Alaska) by the company.
The comment is noted and understood. No change suggested or required.
Finally, 1 am not encouraged In any of the text surrounding this issue, the two most relevant
quotes (pages 4-3 I and 5-45. respectively) being:
"There are no examples of such successful, long-term collection and treatment
systems for mines, because these time periods (100's to 1000's of year) exceed
the lifespan of most past large-scale mining activities, as well as most human
institutions."
"We know of no precedent for the long-term management of water quality and
quantity on this scale at an inactive mine."
RESPONSE: The post-closure phase of mining begins when reclamation is completed and
monitoring and maintenance commences using the controls put into place during closure;
exactly how long the site would require monitoring and maintenance is unknown, and thus
may be 'in perpetuity'. There are no existing examples from which to evaluate success of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 49
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
treatment in perpetuity. No mine in Alaska has maintained a tailings pond into post-closure
although one small mine did maintain a pond during a many year hiatus from operations.
Under AS 72.90.040, financial assurance is required to be sufficient to cover expense for as
long as treatment need is predicted, even into perpetuity (e.g., Red Dog Mine). Maintaining a
water cover over the tailings is a part of the reclamation and closure plan for the Red Dog
Mine. The comment is noted and understood No changes suggested or required
And, finally, a quote from Chapter 8 on page 8-13:
"The promises of today's mine developers may not be carried through by future
generations of operators whose sole obligation is lo the shareholders of their time
(Blight 2010)."
RESPONSE: The comment is noted and understood. So changes suggested or required.
William A. Stubblefield. Ph.D.
The EPA's assessment document presents a seemingly comprehensive compilation of the data
associated with the ecological, geological, economic, and cultural resources of the Bristol Bay
area. The characterization as presented seems to pro\ ide appropriate background information for
the assessment considering the hypothetical nature of the e\ alualion. Without having specific
knowledge of the area in question, it is not possible to provide an assessment as to whether the
characterization was accurate. I'm unaware of significant literature, reports, or data that were
specific to the site and u ould be useful for consideration. The assessment should be expanded to
include greater detail regarding the environmental aspects of the site.
RESPONSE: Additional information on the physical environment of the region and
assessment endpoints lias been incorporated into Chapters 3 and 5.
Dirk van Zr/, Ph.D., P.li.
The geological information was taken from documents prepared to conform to and in compliance
with the standards set by ISational Instrument 43-101 (NI 43-101) (Ghaffari et al., 2011). This
regulatory instrument emphasizes resource information for projects. While I cannot comment on
the accuracy of the regional geological information, the document should reflect accurate
geological information of the Pebble District as known at the time when the report was prepared.
RESPONSE: The assessment uses geological information available for the Pebble site area.
Geological information from Selkregg (1974) has been incorporated into Chapter 3.
My review did not include the Environmental Baseline Document (EBD) of the PLP. However,
in scanning that document, it seems that more site-specific information on site hydrogeology
may be available than was described in the EPA Assessment. While the latter refers to the EBD
extensively in terms of fish populations, etc., it does not refer to it for much of the site physical
characterization. EPA should address this in edits to the Draft Assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 50
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Additional site-specific hydrogeology information has been incorporated in
Chapters 3and 7 and in the calculation of water quality values in Chapter 8. EBD data were
used along with USGS data for hydrologic analysis in both drafts of the assessment, but the
sources of data were not discussed as extensively in the previous draft.
Phyllis K. Weber Scannell, Ph.D.
The Environmental Assessment presents a well-documented discussion of the fish and wildlife
resources of the Nushagak River and Kvichak River Watersheds, with more limited discussions
of the remainder of the Bristol Bay Watershed. The document discusses interactions among
species, including nutrient flows and the importance of groundwater systems; however,
information on contributions of marine-derived nutrienls and existing pressures on the
environment are not as complete, or lacking. The information is general in nature. Should mine
development go forward, it will be necessary to obtain ecological information specific to the
potentially affected areas. The information should include timing offish spawning, egg hatch,
in-migration and out-migration, and similar specific life-history information for important
wildlife species.
RESPONSE: We have clarified the use of information at different scales (Bristol Bay
watershed and Nushagak and Kvichak River watersheds in the problem formulation chapters,
smaller spatial scales in the risk analysis and characterization chapters). General information
on assessment endpoints is included in C liapter 5. with more detailed information included in
the appendices.
Paul Whitney. Ph.l).
Fish Population Lsti males There are several places in the text where impacts of the loss and
degradation of habitat on lisli populations was not quantified because of the lack of demographic
data lor salmonids (e.g., page ES-26, third bullet). These statements are only partially accurate. It
is true that population models such as life tables or Leslie matrices require population age class
data to estimate population numbers. However, even if demographic data are available, these
population models do not relate population estimates to habitat quality. Incomplete data and
relating fish population estimates to habitat quality are not an uncommon problem in ecology and
there are many approaches lor dealing with this issue. Approaches such as Ecosystem Diagnosis
and Treatment (McElhany et al 2<)10), Expert Panels (Marcot et al. 2012), Bayesian nets (Lee
and Reiman 1997), Discussion with experts (Appendix G), or Weighing Lines of Evidence
(Section 6.1.5) are just some of the methods for relating habitat quality to fish abundance.
Models and expert opinions, of course, bring their own uncertainties but it seems better to have
quantitative estimates (and discussion of the estimates) of all the potential fish losses due to
habitat loss than no estimate at all.
RESPONSE: Approaches such as EDT, mentioned above, were considered, but rejected due to
lack of stream-reach specific information needed to provide the sort of quantitative estimates
desired. Expert panels and Bayesian Belief Networks are recognized as potentially providing
useful guidance for identifying key uncertainties and directing future research and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 51
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
monitoring efforts. However, this was deemed outside the scope of this assessment, and we
insteadfocus on the risks associated with the types of habitat change that would be expected
under the mining scenarios outlined. We restrict quantitative estimates of population level
effects to the most severe cases where total losses of runs could be reasonably assumed
Even though the Executive Summary indicates that the impacts of loss and degradation of habitat
on fish populations could not be quantified, the text does provide some estimates. For example,
the assessment (page 6-11, first full para) estimates "that the combined effects of direct losses of
habitat in the North Fork Koktuli, down stream in the mainstem Koktuli and beyond, and
impacts on macroinvertebrate prey for salmon could adversely alTcct 30 to 50% of Chinook
salmon returning to spawn in the Nushagak River watershed." This type of statement, and the
basis for the statement followed by a discussion of uncertainty, is a good example of the
estimates that would better describe possible impacts of the example mine on salmonids. Another
example estimate appears on page 6-39 for four species of salmon
RESPONSE: We restrict quantitative estimates of population level effects to the most severe
cases where total losses of runs could be reasonably assumed, such as the example given
above. The text of the revised assessment has been clarified for consistency regarding
feasibility of estimates.
Question 2. A formal mine plan or application is not available for the porphyry copper
deposits in the Bristol Bay watershed. HP. I developed a hypothetical mine scenario for
its risk assessment, based largely on a plan published by Northern Dynasty Minerals.
Given the type and location of copper deposits in the watershed, was this hypothetical
mine scenario realistic and sufficient for the assessment? Has EPA appropriately
bounded the magnitude of potential mine activities with the minimum and maximum
mine sizes used in the scenario? . I re there significant literature, reports, or data not
referenced that would be useful to refine the mine scenario, and if so what are they?
David A.. if kins. M.S.
The hypothetical mining scenario presented in the Assessment is based on a "Preliminary
Assessment Technical Report" ol'ihe Pebble deposit prepared for Northern Dynasty Minerals by
Wardrop (referred to as GhalTai i el al. 2011), in conformance with Canadian National Instrument
43-101 (NI43-101) which is used to set standards for public disclosure of scientific and
technical information about mineral projects of companies on bourses supervised by the
Canadian Securities Administrators. By most accounts, the Pebble deposit is a world-class
deposit and the Wardrop report counts nearly 11 billion tonnes of total resource. It is unlikely
that all the ore currently identified would be mined, so 11 billion tonnes would be an upper
bound for this particular deposit. It is also certain that exploiting the Pebble deposit would have
to be at a scale large enough to justify the capital investment to build an infrastructure in such a
remote area. Although the Assessment is ostensibly about any mining development in the Bristol
Bay watershed, the use of the Wardrop scenario for Pebble effectively makes the report an
assessment of mining the Pebble deposit.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 52
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The purpose of the assessment is to estimate potential impacts of large-scale
surface porphyry copper mining on salmon ecosystems in the Bristol Bay watershed. The
preliminary plan for mining the Pebble deposit was used as the basis for the assessment
because that deposit is the most likely to advance in the near term. Also, the Agency believes
that mining of other porphyry copper deposits in the watershed would proceed with a similar
approach, since the scenarios used are similar to what has been done at other porphyry copper
deposits. Therefore, it is appropriate to use Northern Dynasty Mineral's 2011 plan for the
Pebble deposit (Ghaffari et al. 2011) as the basis for the scenarios; however, a final mining
plan may differ from what is presented in Ghaffari et al. (2011). Chapter 13 of the revised
assessment also considers the potential cumulative effects of additional smaller copper
porphyry mines in the watershed No change suggested or required
The question then becomes what size mine is feasible IV0111 a technical and economic point of
view. The Pebble deposit mine plan, as presented in the Wardrop report, outlines three scenarios:
• An "investment decision case" for a 25-year mine life that would mine 2 billion tonnes of
ore;
• A "reference case" for a 45-year mine life that would mine 3.8 billion tonnes of ore; and
• A "resource case" for a 78-year mine life that would mine 6.5 billion tonnes of ore, or
55% of the total measured, indicated and inferred resource.
The Assessment chose minimum and maximum mine sizes of 2 billion and 6.5 billion tonnes of
ore, respectively. Thus, the resource estimate used for the Assessment is the same as that for the
two end members presented by Wardrop. This would make the mine one of the largest in the
world, exceeding the size of the 10th percentile of global porphyry copper deposits by an order of
magnitude (see Appendix H of the Assessment) Mines that ultimately become this size usually
expand by increments, as exploration discovers new ore zones and expansion permits are
granted.
RESPO N SE: Yes. the scenarios represen t large-scale mines. The purpose of the assessment is
to estimate potential impacts of large-scale surface porphyry copper mining on salmon
ecosystems in the Bristol Hay watershed, so large mine sizes are appropriate. It is quite likely
that large mines would be created in increments, but this would not influence our assessment,
as we have evaluated impacts based on volumes of material released in the event offailures or
accidents and on material processed as proposed in Ghaffari et al. (2011) as reasonable for a
deposit of this size, regardless of the time period for mine operation. However, we have
included a third, smaller mine in our revision to represent the median-sized porphyry copper
mine on a worldwide basis (250 million tons).
The Wardrop report further delineates Pebble West as a low-grade deposit near the surface that
would most efficiently be mined using open-pit methods, with Pebble East as a deeper, higher-
grade deposit that would most efficiently be mined using underground methods (specifically
block-caving). Mine facilities, as outlined in the Wardrop report, would include:
• Open-pit mining utilizing conventional drill, blast and truck-haul methods for near-
surface deposits.
• Underground, block-cave methods for deeper deposits.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 53
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
• A process plant with throughput of 200,000 tonnes/day that utilizes conventional crush-
grid-float technology with secondary gold recovery.
• Other mine-site facilities, including:
o Tailings storage.
o Waste rock storage (the estimated waste/ore strip ratio is 2:1).
o A natural-gas fired power plant,
o Shop, office, and camp buildings.
o Pipelines to ship ore concentrate slurry to the port facility; return water from the
tailings slurry after separation at the port facility; and fuel.
RESPONSE: No change suggested or required.
This mining and ore processing approach is convent ion a I. and the Assessment includes these
elements. A mine developer may present alternative plans thai could \ ary or alter how the mine
is developed, but the fundamental components would most likely remain the same.
RESPONSE: The EPA agrees with this comment. So change suggested or required.
Because the Assessment is presented as a general assessment of mining risks and impacts in
Bristol Bay and not a specific analysis of the Pebble Project, reliance on the scenario presented
in Wardrop makes the assessment o\ ei ly specific. Further. Chapter 7 provides more specific
information on "Cumulative and Watershed-Scale Ffleets of Multiple Mines," which presents
analysis of potential impacts from mining five additional deposits in various stages of
development (presumably from early exploration to pie-feasibility) The information presented
in Chapter 7 seems more like another mining scenario than a cumulative impacts assessment.
Therefore, I would suggest a broader range of potential mining scenarios be organized as
follows, with the detail of assessment necessarily becoming more speculative with each
subsequent scenario in the list (due to the lack of geologic and engineering information on the
other deposits):
• l)e\ elopment of one. a\ erage-sized porphyry copper deposit (50th percentile or 250
million tonnes of ore as described in Appendix Fl) in the location of the Pebble deposit.
• l)e\ elopment of a mega-mine in the location of the Pebble deposit (of the range between
2 and (•> 5 billion tons of ore) that may develop after multiple expansion and permitting
cycles.
• Development of a mining district consisting of an average-sized Pebble mine and other
potential mines (i e . those presented in Chapter 7).
• Maximum development of all identified potential resources to their most likely ultimate
extent.
Considering this broader range of scenarios would help the reader to better understand the range
of potential risks and impacts.
RESPONSE: The Pebble deposit is located in the watershed of interest, the deposit is similar
to other copper porphyry deposits in the world, and components of the scenarios are common
and anticipatedfor any such deposit of this type; thus, we feel that use of the Pebble deposit
characteristics and location is appropriate. The revised assessment includes an additional
mine size scenario (Pebble 0.25), representing the worldwide median size porphyry copper
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 54
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
mine (Singer et al. 2008). The revised assessment expands the cumulative impacts discussion
(Chapter 13) further by including transportation corridors and secondary impacts.
Steve Buckley, M.S., CPG
Additional mine scenarios are necessary to appropriately bound the magnitude of potential mine
activities. The maximum mine size in the mine scenario seems appropriate given the existing
public information on the Pebble deposit. The minimum mine size of 2 billion tons exceeds the
90th percentile of global porphyry copper deposits. Using a minimum mine scenario in the range
of 250 million tons or in the 50th percentile range of global porphyry copper deposits would be
more appropriate to bound the lower end of the magnitude of potential mine activities. It would
also be useful to include some variation in mining methods. This could include incremental
development of a smaller open pit in the lower grade zones of a deposit, along with a portion of
the higher grade deposit being mined by underground block caving methods to further assess the
minimum potential impact of the mine scenario.
RESPONSE: The revised assessment includes a mine size scenario (Pebble 0.25) representing
the worldwide median-sized porphyry copper mine as presented in Singer et al (2008). The
revision does not evaluate risks from hazards for underground mining, but a brief discussion
of underground mining is included in Chapter 4. The failures assessed would apply whether
the mining technique were underground or surface.
The revised assessment is a major improvement with the addition of mine scenario 0.25 and the
brief discussion of underground mining.
Courtney Car others, Ph.D.
The hypothetical mine scenario was closely based on a probable mine prospect under
development. As such, it appears to be realistic and sufficient, if challenging to conceptualize as
fully hypothetical given this association.
RESPONSE: No change suggested or required.
The report notes that the Pebble deposit may exceed 11 billion metric tons (4-17). The rationale
for choosing 6.5 billion metric tons as a maximum size is based "most likely mine to be
developed (4-19)." The rationale for not choosing a higher potential maximum could be
explained.
RESPONSE: Both the 2 and the 6.5 billion ton scenarios were presented in Ghaffari et al.
(2011) as economically viable, technically feasible, and permittable. The purpose of the
assessment was to estimate potential impacts of large-scale surface porphyry copper mining on
salmon ecosystems in the Bristol Bay watershed; the 6.5 billion ton mine is a large mine.
Because this size mine is on the lower bound of a maximum size, it is a conservative
assumption for the risk assessment. Thus, for the purposes of the assessment, it is not
necessary to hypothesize an even larger mine.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 55
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Dennis D. Dauble, Ph.D.
The hypothetical mine scenario initially appeared realistic and useful in terms of potential project
scope. However, it was apparent during the public hearing, and upon further discussion between
members of the panel, that assumptions on mine size should be revisited based on deposit
characteristics and extraction potential. Also, assumed practices and operations should be
verified against current best-practice and State of Alaska permitting guidelines.
RESPONSE: The revised assessment includes a smaller sized mine that is based on the
median-sized porphyry copper mine on a worldwide basis. The State of Alaska does not have
permitting guidelines that address the size of a mining operation. Land use activities were
previously subject to stipulations meant to minimize surface damage or disturbance under 11
Alaska Administrative Code (AAC) 96.140, but this regulation was repealed in December
2002. The State does have statutory and/or regulatory requirements for an approved Plan of
Operations (11 AAC 86.800), a Reclamation Plan (Alaska Statute (. I.V) 2ZI9.30) and
appropriate Financial Assurance (AS 27.19.040).
Referenced literature provides appropriate context, however. I cannot help belie\ e that
information on environmental impacts from past mining acli\ ilies conducted in the Rocky
Mountain metal belt would be relevant to this assessment in some cases. It is also possible that
recent published information from Hoklen Mine in northern W ashington State would help
establish context for effects of leachates and model results thai predict downstream transport of
tailing material in a wilderness selling, lor example
RESPONSE: Environmental impacts from historic mining are the basis for understanding
that risks from hazards of mining need evaluation. Modeling of tailings transport was based
on the expected characteristics of tailings for the Pebble deposit. There is an expanse of
literature on Superfund sites and interactions of metals associated with sediments and their
leaching. II e included a number of selected sites in our background information, but to
include all possible sites would get further away from the scope of the assessment, which was
to evaluate potential effects within the Bristol Bay watershed
Gordon H. Reeves, Ph.D.
No comments on this question
Charles Wesley Slaughter, Ph.D.
Given the available information base for the ore deposits of the Bristol Bay watershed, and the
publicity which has attended the Pebble planned development over the past several years, the
Assessment's hypothetical mine scenario seems fairly realistic. Further, it is appropriate that the
Assessment consider the probable impacts of other future mineral development projects once an
initial entry (presumably Pebble-Northern Dynasty Minerals) has been accomplished. Such
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 56
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
subsequent development - "cumulative effects over a long time period" - could (and should)
receive more emphasis than is accorded in the Assessment.
RESPONSE: The assessment of cumulative effects of multiple mines is given more emphasis
in the new Chapter 13.
John D. Stednick, Ph.D.
The document does not adequately bound the range of mine scenarios. The minimum mine
development scenario is not adequately addressed. A frequent criticism during the public
comment session was that mine plans presented in the assessment are not representative of
current standards. A compilation of existing world porphyry mine complexes as well as other
types of mines specific to Alaska would better inform the reader of mining processes and
potential risks. The physical setting in Southwest Alaska is not the same as the Bingham Mine in
Salt Lake City. Currently, the document refers lo a particular mine in a particular risk
assessment (stressor), e.g., the Fraser River for salmon, Aitika for chemisliy. and Altiplano for
pipeline failures.
RESPONSE: The revision includes a smaller mine size that represents the worldwide median
size for a porphyry copper mine (Singer et al. 200,S). to help with the issue of the range of
scenarios. EPA disagrees that the mine scenarios evaluated are not representative of current
standards. This view apparently stems from use of the term "good" rather than "best"
practices in the draft assessment. The reason for using that term is that the term "best
management practices " is a term generally applied to specific measures for managing non-
point source runoff from stormwater (40 CFR Part I30.2(m)). Measures for minimizing and
controlling sources of pollution in other situations are often referred to as best practices, state
of the practice, good practice, conventional practice, or simply mitigation measures. We
assume that these types of measures would be applied throughout a mine as it is constructed,
operated, closed, and maintained post-closure, regardless of the qualifier that one wishes to
place with it. A text box was added to the revised Chapter 4 that discusses terms to help clarify
our intention for descriptors used. The Fraser River example was considered because it had
been used as an analogue by others, but was dismissed as not representative of Bristol Bay.
Other mines that are noted in the assessment are illustrative of specific issues only and not
used for risk evaluation in the Bristol Bay watershed. While physical settings are not the same,
the components and the impacts are similar and thus included to help a reader understand
where and how these things occur.
The Bureau of Land Management has identified certain lands that will be excluded from
development. This reference needs to be followed up.
RESPONSE: The purpose of the assessment was to estimate potential impacts of large-scale
surface porphyry copper mining on salmon ecosystems in the Bristol Bay watershed This
presupposes that our mine scenarios are located in areas that are not excluded from
development. The majority of land in the two watersheds is state land that is available for mine
development, and Figures 2-3 and 2-4 now indicate protected areas within the Bristol Bay
watershed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 57
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Roy A. Stein. Ph. IX
Hypothetical Mine Scenario. Though mining does not lie within my area of expertise, I
thought that this scenario helped me understand the potential impact of a mine of this magnitude
in a wilderness, pristine watershed. I find it difficult to comment as to whether this scenario is
realistic and sufficient, though I did use this scenario to guide my comments below. From the
text, it is apparent that this is a realistic scenario, based on documents filed by the company with
the Canadian government. This makes this scenario the most realistic one could expect.
a. Minimum and Maximum Mine Size. For 111c. as an ccologist, this bounding helped me
to understand the potential impacts of the Pebble Mine, though I did not understand what
the probability of either mine size happening in the near terra. Understanding these
probabilities would be helpful to the readers
b. Mine-Size Continuum. Is it more likely that the initial Pebble Mine will be maximum
or minimum in size? Wouldn't it be lar better to review a continuum of mine sizes from
the smallest that is economically feasible to one that is intermediate in size and then to
one (or two) that would take to the largest realistic mine size? Willi this continuum, the
reader begins to understand the overall impact of \ arious mine sizes 011 the Bristol Bay
ecosystem. Some reflection 011 these mines sizes and their impacts would have helped
me interpret the Environmental Risk Assessment with some additional insight.
RESPONSE: The State of Alaska does not have permitting guidelines that address the size of
a mining operation. Many identified deposits never become developed mines for various
reasons, and it is unknown how many deposits exist and are economically viable for
exploration of mining feasibility. Once it is decided to develop a site, there are a number of
things that must occur before a mine begins operation. Thus, it is not possible to predict the
probability of either mine size happening in a specific period, at least with any certainty. All
we can say is that there are deposits that have the potential to be mined in the future.
It is more likely that an initial mine would begin at a smaller scale and become larger and
perhaps be permitted in stages of increasing size. However, there are different approaches in
how plans are presented and these depend on multiple factors, including economics and
projected costs/gains in prices of the metal being mined. For example, if it were not
economically viable to mine only a small part of a known large deposit, a larger mine would be
proposed and planned for. The revised assessment includes a size scenario that represents the
worldwide median-sized mine to provide more of a continuum of sizes.
One Watershed. Gi\ en the productivity of salmon from these two river systems (50% of the
sockeye salmon in Bristol Bay are produced from these rivers), might there be some thought
given to limiting the mining operations to a single watershed, either the Nushagak or the Kvichak
(page ES-2)? In so doing, in a single stroke, the impact of this mine on salmon is reduced by
50% or more. Could the Pebble Mine be confined to one watershed, such as where the majority
now falls - in the Nushagak River (both the north and south forks of the Koktull River)
watershed? Even so, this suggestion becomes especially pertinent to Chinook salmon spawning
in the Nushagak River, for this run is "near the world's largest" (page ES-5), but yet the
Nushagak watershed is small relative to other watersheds (such as the Kuskokwim and the
Yukon) where Chinook salmon are abundant. As a result, any impacts to the watershed by a
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 58
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
mine of this size are magnified, another concern when considering this location. Without mining
expertise, I cannot judge whether it would be possible to mine in only one of the watersheds,
rather than both. Even so, some consideration should be given to this suggestion.
RESPONSE: Restricting impacts to one watershed would change the risks, and could be a
part offuture mine plans. We chose to represent a suite of realistic mine scenarios based upon
preliminary mining plans and the location of the ore deposit which lies in both watersheds.
William A. Stubblefield, Ph.D.
No comments on this question.
Dirk van Zvl, Ph.D., P.E.
The hypothetical mine scenario adopted by the I-PA relied almost exclusi\ civ on the document
prepared for Northern Dynasty Minerals (NDM). one of the partners of the Pchhlc Limited
Partnership. Developing a mine plan for a specilic ore body is a large task and is undertaken by a
large team of engineers and scientists In the process of developing a mine plan many options are
considered for each facility and its components, including mining methods, process design
options, waste rock management options, tai lings management options, shipment of product, etc.
The hypothetical mine scenario was prepared by an independent consulting company for one of
the partners and this plan does not necessarily represent the design and management options that
will be selected for developing this ore body. Because of ore grades and the deposit style, it is
most likely that an open pit mine will be developed as assumed in the report for the western
lower grade ore body and that underground mining will be used for the eastern higher grade ore
body. The size of the ore body and the strip ratio for an open pit mine are completely dependent
on metal prices and production costs at the time of mine development. Metal prices and
production costs will also be a major factor in deciding whether to first develop an underground
mine instead of an open pit mine. While some of the components of the final mine may contain
elements of the conceptual mine, it is impossible to know whether the hypothetical mine scenario
is realistic, as will be further discussed in the comments below.
RESPONSE: It is acknowledged in the assessment that the mine scenarios might not look
exactly like a mine presented in a mining plan. The assessment is not a mining plan and is not
an evaluation of a mining plan; it simply uses current information for the Pebble deposit
because it is a large ore deposit which has had extensive exploration with potential for
development in the near future. An additional scenario has been included in the revision to
match the worldwide median mine size and show a better continuum of scenarios possible. We
consider our scenarios to be realistic, as they were presented as possible layouts for the Pebble
deposit and stated in Ghaffari et al. (2011) as being "economically viable, technically feasible,
and permittable".
To address the issue of sufficiency it is necessary to understand the range of potential outcomes
related to the various options. For the most part, the EPA study used the information from the
NDM document for evaluating impacts to salmonids. Using different options, both technological
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 59
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
as well as site selection, for some or many of the facilities could result in impacts that are
different from those described in the report. I would therefore suggest that using only the present
hypothetical mine scenarios is insufficient. There could be a range of impacts, such as the
surface areas of facilities, which in some cases could be smaller than what was chosen and in
other cases larger. However, this does not mean that the hypothetical mine represents "average
conditions." I therefore consider the mine scenario not sufficient for the assessment.
RESPONSE: It is acknowledged in the assessment that the components in the mine scenarios
might not be exactly what would be proposed in a mining permit application for this location
or for other locations within the Bristol Bay watershed. The purpose of the assessment was to
evaluate potential impacts of large-scale surface porphyry copper mining in the Bristol Bay
watershed, so the assessment was not meant to represent "average conditions". However, an
additional mine size scenario has been included in the revision to match the worldwide median
porphyry copper mine size and show a better continuum of scenarios possible.
The minimum and maximum mine sizes selected In I -PA are 2 billion tonnes mined over 25
years and 6.5 billion tonnes mined over 78 years, in both cases, the daily ore processing rate is
200,000 tonnes. As indicated above, the final economic mine size at the time of development
will be determined by metal prices and production costs Note that production costs, as used
here, include all the considerations related to regulatory. en\ ironmental and social aspects of the
mine and its environs. Mining companies typically make in\ estment decisions for periods of 20
to 30 years. It is seldom, if ever, that a new investment will be made based on a 78 year mine
life; however, the upside potential will be taken into account u hen an investment for a shorter
mine life is made. Tt is also unlikely that environmental regulatory agencies will consider issuing
a permit, including closure plans, etc. for a 78-year project. Furthermore, even if the mine
ultimately continues lor 7S years, it is certain that the operating and environmental control
technologies and societal expectations will change in that period and therefore the elements used
by EPA for the maximum size hypothetical mine will certainly not be valid for such a long mine
life. It is therefore my conclusion that assuming the development of a 2 billion tonne ore body is
realistic, but that assuming development of a 6.8 billion tonne ore body, using static technology
assumptions, is not.
RESPONSE: II hile it is true that an actual mine likely would be permitted in increments, the
assessment never stated that the ~H-yr scenario would not be done this way. In fact, the
assessment does not discuss how such scenarios would be permitted at all, and to do so is
outside the scope of the assessment. The purpose of the assessment is to estimate potential
impacts of large-scale surface mining of copper porphyry on salmon ecosystems in the Bristol
Bay watershed, and that necessarily assumes that the mine scenarios are permitted. It is true
that some technologies would have advanced over time from 'day one' of mine scenario
development, and we acknowledge this in our assessment. It is impossible to predict, however,
how impacts from use offuture technologies would differ from those in use today, or how they
would change conditions existing at the time they began being used We can only present and
predict potential impacts based on use of the most appropriate technologies available at the
current time. No change suggested or required.
The EPA assessment report includes a range of the literature and reports in evaluating the
selected mine scenario. However, I have a number of specific comments about various aspects of
the report as well as the references.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 60
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Good practice vs. best practice. On p. 4-1 of the report, the EPA states: "Described mining
practices and our mine scenarios reflect the current practice for porphyry copper mining around
the world, and represent current good, but not necessarily best, mining practices". EPA does not
clarify this decision, nor does the report clarify the distinction between "good" and "best"
practices. It can only be concluded that "best" will be better than "good". On the basis of this, it
is inconceivable to me that the Bristol Bay communities, the Alaska regulatory authorities as
well as Federal Regulatory Authorities will not demand that the company follow "best mining
practices", however that is defined at the time. It is also inconceivable to me that the company
will not follow "best mining practices" in the design and development of such a mine. During the
engagement processes, the stakeholders will have to agree what represents "best" practice in the
design of the mining project. It is important to note that most of the failure statistics used as a
basis for the evaluations in the report are derived from data gathered over the last 50 years or so
(e.g. refer to p. 4-45 of report). It may be argued that this i nlbimalion is mostly for mines
following "good" practices and, in many cases, for projects that had a lower standard of care. To
my knowledge, there are no statistics available thai compare failure rales of facilities designed
and operated under "good" practice to those designed and operated under "best" practices,
whatever definitions are used for "good" and "hesf "
RESPONSE: The term "best management practices " is a term generally applied to specific
measures for managing non-point source runoff from stonnwater (40 CFR Part 130.2(m)).
Measures for minimizing and controlling sources of pollution in other situations often are
referred to as best practices, state of the practice, good practice, conventional, or simply
mitigation measures. We assume that these types of measures would be applied throughout a
mine as it is constructed, operated, closed, and post-closure, regardless of the qualifier that
one wishes to place with it. To remove any ambiguity and subjectiveness of terms "good" or
"best", we have removed them in the revision and have added Box 4-1, which includes
definitions for several terms used.
The EPA also is not aware of any statistics available that compare dams designed (and/or
operated) under different standards; however, the probabilities for dam failure used in the
assessment were not derived solely from the historical record. Historical failures were
discussed as supporting background information and present a defensible upper bound on the
failure probabilities. The failure probabilities used in the assessment are based on Alaska's
dam classification and required safety factors applied to the method of Silva et al. (2008). The
data presented by Silva et al. (2008) consider only the annual probability offailure from slope
instability, but the methodology is equally applicable to other failure modes. The discussion of
failure probabilities in the revision (Chapter 9) is expanded to clarify this issue.
Mine scenarios. The executive summary indicates (p. ES-11): "The mine scenario includes
minimum and maximum mine sizes, based on the amount of ore processed (2 billion metric tons
vs. 6.5 billion metric tons), and approximately corresponding mine life spans of 25 to 78 years,
respectively". This seems to indicate that the mine life cycle in the first case consists of 25 years
of operational life followed by closure and, similarly for the second case, 78 years of operational
life followed by closure. However, a careful review of the water management section (section
4.3.7) indicates that this is not the case. The EPA water balance calculations are simplified to a
set of deterministic values in Table 4-5 for four water management stages during the overall
mine life cycle: start-up, operations minimum mine (25 years), operations maximum mine (78
years), and post-closure. For post-closure, only the 78-year mine life numbers are used. It
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 61
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
therefore seems that EPA is not considering that the 25-year mine will close, but that its life will
automatically be extended to 78 years. Does this mean that the EPA really does not evaluate the
minimum mine size completely, i.e. the 25-year mine life followed by closure? It is important
that this be clarified as it would be inconsistent not to evaluate closure of the 25-year mine. It is
possible that additional evaluations, or at least additional explanations, will be required to clarify
this.
RESPONSE: The reviewer is correct. Our water balance calculations only explicitly present
the closure of the Pebble 6.5 scenario. The water balance for the Pebble 0.25 and the Pebble
2.0 scenarios would be similar, but in those scenarios there would be no water captured in
TSF 2 or TSF 3 and the amounts captured in the pit would be proportionally smaller, as they
are in the operating scenarios. While the pit is filling in each of the three scenarios, the total
amount of water captured is slightly less than the amount captured during operations, and
about 35% to 45% of the water captured is available for reintroduction to the streams. Once
the pit is full, the amount captured at the mine pit area drops substantially and 100% of the
captured water is reintroduced. Post closure flows for the three mine scenarios are presented
in Table 6-8.
Tailings management technologies. Ongoing technology dc\ dopment has resulted in a broader
range of tailings management options than only slurry tailings disposal. Filtered dry stack
tailings can be considered as a realistic option, even for mines with higher production rates.
Flotation of remaining sulfides in the tailings before deposition is also a realistic option for
mines; it has been done successfully at the Thompson Creek Mine in Idaho for the last 18 plus
years. While these technologies are mentioned, they are not selected for reasons such as
technology not being appropriate for the climatic conditions and concerns with disposal of pyrite
waste. Both of these are not insurmountable technical issues and adopting such management
options will reduce failure probabilities and potential impacts following a failure. The failure
mode of a filtered dry stack facility not containing sulfides will be completely different from a
slurry impoundment and the potential environmental impacts of these other tailings management
options will definitely be far smaller than those for the selected mine scenario using slurry
tailings disposal.
RESPONSE: Selective flotation to lower the pyrite content has been added to the discussion of
processing in C liapter 4 and referenced in Chapter 6 as being done in the assessment
scenarios. How tailings are managed within the impoundment can affect water chemistry and
a dry stack with sulfides removed may produce the best water quality results after reclamation
if the fate of the sulfide tailings is never considered. According to Ghaffari et al. (2011), 14%
of the tailings produced would be pyritic (with the selective flotation method used), which
equates to an average of28,000 tons/day in a 200,000 tons/day mining operation, or over 255
million tons during a 25-year estimated lifetime for that scenario. These tailings need to be
managed in such a manner that oxidation does not lead to acidic drainage and the most
effective way is to deposit them subaqueously. Additionally, a "rule of thumb" for design of
dry stack tailings is to allot 25 acres for every thousand dry tons of tailings per day over the
life of a 20-year operation (SME Mining Engineering Handbook 1973). Assuming our
scenario of200,000 tons per day processed and that 99% of this material was waste tailings,
this would amount to 4900 acres (25 *200,000 *.99/1000) or 19.8 km2 over 20 years. This
exceeds the internal surface area taken up by the TSF for tailings deposited over 25 years for
the same mass of material processed per day by 5.6 km2; thus, even if there were not a risk of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 62
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the PAG tailings acidic leaching potential, dry stack tailings disposal at this site would create
additional surface area loss versus the scenario's traditional dam.
Waste rock management. The waste rock management plan on p. 4-13 calls for the potentially
acid generating (PAG) waste rock to be separated from the rest of the waste rock and states that
the "PAG waste rock might be placed in the open pit at closure to minimize oxidation of sulfide
minerals and generation of acid drainage". However, on p. 4-33 it is stated that: "PAG waste
rock will be processed through the flotation mill prior to mine closure, with tailings placed into
the TSF (tailings storage facility) or the mine pit." These two alternatives represent completely
different management, economic and environmental conditions and are not consistent. Milling
the PAG waste rock represents a higher cost than placing the P.\( i rock in the pit and placing the
PAG waste rock tailings in the TSF will increase the size of the TSF. Placing the PAG tailings in
the pit will set up a completely different management scenario than placing the PAG waste rock
in the pit. The EPA should clarify which option or range of options they select for evaluation and
use that consistently in the assessment.
RESPONSE: These statements have been made consistent. The revised assessment scenarios
include processing the PAG waste rock over the course of operations to minimize the length of
time a PAG waste rock pile would be on the surface, and thus minimize its potential for
oxidation and subsequent release of acidic leachate. There is no longer mention of PAG rock
being disposed in the pit at closure (Section 6.3.3).
Water balance and management - waste rock. Mine site water balance and management is a
very complex issue as recognized by the N\\ on p 4-27' " water balance development is
challenging and requires a number of assumptions" Because of these uncertainties, complex
probabilistic dynamic models are employed at mines where the site details are better defined than
that of the EPA hypothetical mine scenario. The information in Box 4-2 indicates that the
"captured flows include water captured at the mine site and the TSFs (Table 4-5). The total
amount of water captured at the mine site includes net precipitation (precipitation minus
evapotranspi ration | footnote (.luring operations most of these areas will not be covered with vegetation
and the correct terminology here is ""evaporation" |) over the areas of the mine pit, the waste rock
piles, and the cone of depression (without double-counting any areas of overlap)". On p. 4-23 it
is stated that. ""Monitoring and recovery wells and seepage cut-off walls would be placed
downstream of the piles to manage seepage, with seepage directed either into the mine pit or
collection ponds" I 'iuure 4-1) show s this schematically where leachate from the waste rock enters
the groundwater that then llow s to the mine pit or to the monitoring and collection well.
However, if net precipitation only includes the components above (precipitation minus
evapotranspiration), effect i\ ely excluding infiltration, and if this net precipitation is captured
from that waste rock pile (as stated in Box 4-2), then there should not be any water available to
infiltrate into the waste rock pile, i.e. there should not be any leachate. All references to seepage
from the waste rock piles are incorrect following the EPA's assumptions of total capture of net
precipitation. In addition, the approach that is used in the water balance is inconsistent with
observed field performance and descriptions in the literature, as is it difficult to imagine a case
where there is zero infiltration into a porous waste rock pile (e.g. Nichol et al., 2005 and Fretz et
al., 2011). The EPA must clarify the whole water balance model and the evaluations. For the
assessment to have any credibility, the water balance and management evaluations should reflect
realistic conditions.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 63
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The term "evapotranspiration" has been corrected to 'evaporation' in the
revised assessment when discussing the operational phase. Total precipitation equals the sum
of evaporation, transpiration (where applicable), runoff, and infiltration. Net precipitation
included in the water balance includes all water falling onto the site components minus water
leaving only via evaporation (i.e. it includes both runoff and infiltration). Leachate/seepage
water originates from precipitation which has infiltrated the waste rock piles or tailings onto
which it fell. Therefore, the discussion of seepage is not incorrect, and there is no indication
that the scenarios represent "zero infiltration", as we discuss (as noted in the comment made)
how seepage and leachate is managed. However, the water balance section has been revised
for clarity (new Section 6.2.2). As the commenter notes, the assessment discusses seepage
collection systems. However, the assessment does not assume total capture. Our water balance
assumes that 50% of the leachate that is lost from the TSFs and from the portion of the waste
rock piles outside the drawdown zone of the pit escapes into the groundwater and eventually
into the streams.
Dam failure - tailings storage facilities. During operations, "water falling within the perimeter
of a TSF would be captured directly in the TSI . hul runoff from catchment areas up-gradient of
the TSF would be diverted downstream" (p. 4-27) At closure, water would he removed from the
TSF providing more storage, but also maintaining a small pool lo "keep the core of the tailings
hydrated and isolated from oxidation" (p. 4-32). This seems lo assume that the diversion systems
will be kept in place and most likely will be upgraded to di\ eri up-gradient surface water around
the tailings impoundment. It is likely that the design criterion for the upgraded diversion system
during the post-closure period will be the probable maximum llood (PMF) as is done at a number
of mines. Dam failure analyses were done assuming thai the llood leaving the TSF includes the
PMF inflow from the np-gradienl catchment, excess water 011 top of the tailings and 20% of the
tailings volume (Box 4-8). While one can argue that a failure including all these materials may
be a plausible, although a very low likelihood e\ ent during operations, it seems less probable that
such a failure will take place lor the mine closure period when an upgraded diversion system is
in place Also, during the closure phase, the tailings will consolidate and be less mobile. Note
that the densilication heha\ ior of oil sand tailings referred to on p. 4-32 (i.e. the Wells, 2011
reference) does not apply to copper tailings. The presence of clay minerals and bitumen in the
mature fine tailings portion of the oil sand tailings is the source of the different behavior
(Znidarcic et al . 2'H I).
RESPONSE: It is assumed that post closure scenarios will have drainage facilities in place
that could safely pass storm events. However, the decision was made to assume a failure that
was consistent for both the operational and post operation scenarios. The PMF would overtop
the main dam of the TSI' and drain into the North Fork Koktuli as a representative failure,
assuming that drainage facilities were either not yet operational or had failed. It is important
to note that the PMF peak flow generated only 291 cms, which is small when compared to the
peak flow release at the failing dam (149,263 cms and 11,637 cms for the large and small
failures, respectively). 20% was selected to represent the volume releasedfrom the TSF
because it fell within a reasonable range when compared to release volumes of historic
failures. The 20% volume included both solids and pore water.
Reclamation slope of waste rock. On p. 4-32 it is stated that: "We assume that NAG waste rock
would be sloped to a stable angle (less than 15%) (Blight and Fourie, 2003)". I contacted Profs.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 64
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Geoff Blight and Andy Fourie about this statement and received the following response from
Prof. Blight: "The only reference to 15 degrees (not 15 %) slopes is the following, talking about
the outer tailings, not waste rock covered, slopes of decommissioned TSFs: "it must be
remembered that the outer slopes will need to be rehabilitated, and that for vegetation to be
stable, and surface erosion minimal, the maximum outer slope should not exceed 15 degrees."
This error in reference must be corrected; it is recommended that more typical closure slopes of
about 30% (or 3H: IV, about 18 degrees) for waste rock should be used in the evaluations.
RESPONSE: The text has been changed to correctly read 15 degrees in Chapter 4 and
Appendix I.
Phyllis K. Weber Scannell, Ph.D.
The Environmental Assessment discusses a liypt>lhelical mine (gi\ en that mine plans have not
been developed). Page 4-5 of the document stales that "rocks associated with porphyry copper
deposits tend to straddle the boundary between net acidic and net alkaline . "" The Pebble
Project Environmental Baseline Report (SRK 201 I. Chapter 11) summarizes testing on the
samples from the pre-Tertiary porphyry minerali/ed rock in Pebble East Zone (PEZ) and Pebble
West Zone (PWZ). The metals leaching acid rock drainage study showed acidic conditions
occurring immediately in core with low \P. but the a\ eraue delay to onset of acidic conditions
was estimated to be about 20 years. Copper w as leached in the highest concentrations, but Co,
Cd, Ni, and Zn also leached from samples from PI-/ Wacke (sedimentary rock) samples from
PEZ and PWZ leached As. Sb. and Mo, in addition to Cn (SRK. page 58). The available
information on acid generation and metals leaching appears to be preliminary. Development and
permitting of a viable mine plan will require e\tensi\ e sampling and data analysis of ore
samples, plans for classifying waste rock (as PAG and NAG), and, possibly, plans for collecting
and treating runoff and seepage waters
RESPON SE: EP. I agrees that developing and permitting a viable mine would require
extensive in formation. The assessment is not a mining plan. The scenario presents a suggested
treatment option for mining influenced water and settling ponds for water that is simply
stormwater runoff. No change required.
The Environmental Assessment seems a bit premature in making an assessment of the potential
for acid rock drainage (ARD) or metals leaching (ML). Data on metals other than Cu are
insufficient and possible toxicities to fish are not addressed. Further, the description of the
potential mine may no1 relied a likely mine scenario. It is difficult to calculate potential risks to
the environment without a specific mine plan. The section of the Environmental Assessment
should be revised as more data on ARD and ML become available.
RESPONSE: The assessment uses the geochemistry data that are available from the Pebble
Limited Partnership. Copper was emphasized in the review draft because the EPA believed,
and still believes, that it is the contaminant of greatest concern. Toxicities to fish of the other
metals were not discussed because they had been screened out. However, the revised
assessment explains the screening process and the selection of copper in more detail in a new
section on the identification of stressors of concern (Section 6.4.2) and more metals have been
added to the screening assessment. The toxicities of all metals reported in the leachate are now
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 65
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
addressed either as individual elements or, in the case of major ions, as contributors to total
dissolved solids. The mine scenario is based on the most recent preliminary plan released by
Northern Dynasty Minerals (Ghaffari et al. 2011).
Paul Whitney. Ph.D.
Reclamation Plan. I am not familiar with the Northern Dynasty Minerals mine plan. I wonder if
their mine plan includes a Reclamation Plan. If not, why not? Tf their mine plan includes a
Reclamation Plan, why isn't it presented as part of the Bristol Bay Assessment? The feasibility
of reclaiming the waste rock and tailings areas and possibility ilie pit (page 4-23, last para, last
sentence) seems important for evaluating the acceptability of the example mine. I am not aware
of any mine regulating agency that does not require a Reclamation Plan as part of a mine
application. I wonder if a Reclamation Plan that in\ olved placing waste rock and tailings back in
the pit and reducing surface infiltration would greatly reduce the need lor water treatment.
RESPONSE: The assessment is not a mine plan. The NDM document relied upon for
information in the assessment (Ghaffari et al. 20 J I) is a preliminary mine plan and does
include conceptual reclamation measures which we have used in our scenarios. The State of
Alaska requires a Reclamation Plan of all mining facilities unless they are very small (AS
27.19.010). EPA's revised scenarios utilize blending throughout the mine life to eliminate the
need for long-term storage of PAG waste rock. Reclamation activities for the scenarios are
discussed more clearly in the revision and are activities that are considered feasible and
common at other similar existing mining sites. The scenarios in the assessment assume that
reclamation is properly completed, but concentrates the discussion on impacts that are
expected even with such activities.
Best Mini iv-* Practices. The assessment refers to the example mine plan as having both the "best"
mining practices (eg. page ES-10, five lines from the bottom) and "not necessarily best" mining
practices (e g , page 4-17. four lines from the top). Both of these statements can't be accurate.
RESPONSE: II ith regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed is that we have assumed modern mining
technology and practices. The terms are qualitative when generally interpreted, or have a
regulatory meaning (best management practices), and thus we have eliminated their use in the
revised assessment to avoid confusion.
Noise Levels. The mi lie plan should provide information on the location, frequency, and size of
blasting, sound level isopleths around the mine, and efforts to minimize sound levels as the mine
develops. I wonder if a majority of the sound levels will attenuate as mining activities move
deeper into the ground or if will there be a hundred years of blasting at the surface level. The
interviews with the villagers indicate that blasting and helicopter noise is a concern (Appendix
D, Cultural Characterization, page 94). A characterization of current noise levels in relation to
the area and timing of current and past wildlife use would help to determine if the whole or parts
of the watersheds are less than pristine.
RESPONSE: Noise from large-scale mines is outside of the scope of this assessment because it
is not known to affect salmonid populations. EPA agrees that a full evaluation of any future
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 66
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
mining permit applications and subsequent National Environmental Policy Act
Environmental Impact Statements would consider these direct effects on wildlife.
Water treatment during the winter. I wonder if it will be possible to treat water during the winter.
Will such treatment have to occur in a warm building? If so, what are the temperature
consequences of releasing warm treated water into streams?
RESPONSE: Although water quality standards for temperature are directed toward summer
maximum temperatures, we emphasize the importance of the year-round thermal regime to
which salmon are adapted Abnormally warm temperatures could accelerate egg maturation
and reduce survival of incubating salmon. We state that a protective approach would
discharge water that met baseline thermal regime conditions throughout the year, and
describe the risks associated with failure to meet this regime.
Yes, it is possible to treat water during the winter, and treatment facilities would be contained
in structures to prevent freezing of the treatment plant. While the Alaska water quality
standards (18 AAC 70) do contain maximums for water temperature, even the most stringent
of these (13°C) may be too high for winter-time discharge. Very likely, following a permit
application, ADF&G would determine an appropriate temperature for winter discharge and
the facility would have to meet this value, especially for discharges to anadromous streams.
Temperature can be easily lowered during the winter by exposure to ambient air.
Cone of Depression. I have worked on pit mines u here hyd ideologists model the lateral extent
of the cone of depression and have mapped the lateral extent as an area around the pit. The
lateral extent of the cone of depression, illustrated in I'igure 4-9, appears to be underestimated
and has no effect 011 si renins or uelhinds The figure has no scale. Is the lateral extent of the cone
of depression in Figure 4-1) based on modeling (see Box 4-2, para 3, last sentence)? If so, how
manyNWI wetlands and meters of si renin are in 1 he area used for the model? If there are
wetlands or streams in the modeled area, how far down stream will the cone of depression
influence stream flow and welland hydrology '
The information in lio\ 4-2 doesn't clearly (at least to me) deal with the proportions of run-on
and run-off water. If the di\ cried run-on water is supposed to mitigate the cone of depression,
will it be a\ ailahle for down stream resources? Why won't diverted water seep back into the
near-by pit \ ersus mitigating the cone of depression? The answer to these questions is on page 5-
72, but merely indicating there will be a reduction is not very informative.
RESPONSE: The cone of depression is now projected to dewater streams and wetlands with
which it intersects. These losses are incorporated into the water balance calculations used to
estimate changes in downstream streamflow presented in Chapter 7. The lateral extent of the
cone of depression was estimated using the Dupuit-Forcheimer discharge formula for steady-
state radial flow into a fully penetrating well in a phreatic aquifer with a diameter equal to the
average mine pit diameter. The radius of influence was determined by balancing the net
precipitation falling within the cone of depression with the calculatedflow into the mine pit.
Section 6.2.2.1 and Box 6-2 in the revised draft (Section 4.3.7 and Box 4-2 in the original draft
assessment) contain additional details on the methodology.
Our estimate of the mine pit inflow agrees closely with the estimate provided in Ghaffari et al.
(2011). The estimated cone of depression would extend about 1.2 km beyond the pit rim The
water within the cone of depression that wouldflow to the mine pit is included in the water
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 67
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
balance. The revised assessment presents estimates of the changes in stream/low at individual
downstream gages.
Figure 4-9 (now Figure 6-5) is merely a schematic of the flows and is not to be interpreted as
an exact representation of the expected flow regime. The figure is not to scale and is not based
on modeling.
All of the precipitation falling within the cone of depression is considered to flow into the mine
pit. These losses are incorporated into the water balance calculations used to estimate changes
in downstream streamflow presented in Chapter 7. Streams within the cone of depression
would dry up. The assessment assumes that water from streams upstream of the cone of
depression would be diverted through pipes or channels to locations on streams downstream of
the cone of depression. The diverted water would not seep or flow into the mine pit.
Run-on and run-off water terminology. I am used lo referring to up gradient or adjacent water
that runs onto the pit or tailings facilities as run-on water and to water from the mine or storage
facilities as run-off water. The assessment doesn't always distinguish these two types of water.
For example, on page 4-13, line 6 refers to precipitation run-off water as up gradient water. On
page 4-26, the first bullet refers to run-off water as water running off mine facilities. The
terminology overlap makes it difficult (at least for me) to understand how the run-on and run-off
water will be captured and diverted around the mine facilities or used for other purposes. In
addition to calculations, diagrams of the di\ ersions would he helpful. Will there be parallel
diversion ditches around the facililies. one lor run-on and one for run-off water? Will one or both
of these ditches be lined? How will the water in these ditches he inlluenced by the cone of
depression? These questions are alluded lo in the discussion on page 4-27(second para), but are
not explicitly addressed. I am sure engineers can and ha\ e answered these questions for other
mines with water balance analyses It would he interesting to see an explicit summary of the
water balance for the various facililies Such analyses would be good for the example mine plan
during operation and once the mine is no longer a net consumer of water (page 5-44, para 2).
Without the water balance analyses, potential impacts are not easily understood or quantifiable.
RESPOSSE: A single diversion ditch intercepts up-gradient surface water and routes it
around the areas disturbed by mining, preventing it from mixing with waters that have
encountered site materials. The revised assessment includes a more detailed water balance
(Chapter 6).
Some ideas for how to manage and separate run-on and run-off water might help determine
which streams might dry up and what type of mitigation measures (i.e., lining ditches) could
minimize the impact. In addition, if run-on water can be maintained in a diversion ditch, what is
the opportunity for developing a reclamation plan for the ditches? Such plans might be able to
minimize and partially compensate for lost reaches of headwater streams.
RESPONSE: The revised assessment does include some possible reclamation activities for
closure, although others could be proposed for an actual mining plan. Water would continue
to be diverted around areas where the water could encounter contaminants. Mitigation to
compensate for lost streams would be addressed through a regulatory process that is beyond
the scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 68
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Protective approach. A "protective approach" is mentioned on page 5-30 (para 3, last sentence).
This has something to do with water management and would be good to explain.
RESPONSE: This has been clarified in Chapter 7.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 69
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 3. EPA assumed two potential modes for mining operations: a no-failure
mode of operation and a mode involving one or more types offailures. Is the no-failure
mode of operation adequately described? Are engineering and mitigation practices
sufficiently detailed, reasonable, and consistent? Are significant literature, reports, or
data not referenced that would be useful to refine these scenarios, and if so what are
they?
David A. Atkins. M.S.
The no-failure scenario attempts to quantify the impacts from de\ eloping the footprint of the
project alone. In reality, various failures and accidents ine\ iliihl\ occur, and they may have a
range of impacts from inconsequential to large. So this scenario is presented to describe the
minimum impact that could be expected from project de\ elopmenl assuming everything works
as planned.
RESPONSE: This comment is a correct interpretation of our "no-failure" scenario. Because
this distinction was not clear to many other readers, the revision no longer uses the term "no
failure". The no-failure scenario from the draft assessment has been changed to a chapter on
the effects of the footprint of a mining operation, without regard for operational problems
(Chapter 7).
The mine will, by necessity, remove those streams and wetlands lhat are beneath the pit, waste
rock, tailings and processing plant development areas There should be some flexibility in siting
facilities other than the pit or underground workings I or the "110-lailure' scenario, the
Assessment presents lengths of stream and areas of wetlands lhat would be lost due to physical
displacement of the aquatic resources from mine de\ elopment and reduction in flows from mine
water management. The assessment presents the lb I lowing resources that would be lost and that
have been shown to be spawning or rearing habitat for coho, Chinook, and sockeye salmon, or
have resident populations of rainbow trout and Dolly Varden:
25-year scenario
78-year scenario
Eliminated or blocked streams (km)
87.5
141.4
Reduced flow (>20%; km)
2
10
Eliminated wetlands (km')
10.2
17.3
Given the range of uncertainty with the proposed mine plan, presenting stream lengths and
wetland areas to the tenth place implies unrealistic accuracy. Significant figures should be
checked and consistent throughout the document, and ranges should be presented if known (e.g.,
results for the pits could be presented with more accuracy since we know where they will be,
whereas other facilities that could be located in different areas should be presented with an
appropriate range of uncertainty).
RESPONSE: The authors have carefully addressed this issue. Numbers from the literature or
from the PLPEBD retain the number of significant figures in the original. Numbers derived
for this assessment have the appropriate number of significant figures given the precision of
the input data and uncertainties due to modeling and extrapolation.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 70
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The impacts as presented appear substantial, mainly because of the very large nature of the
project. However, it would be helpful to describe the significance of this loss, specifically with
regard to the following questions:
• What impact would the loss to streams and wetlands have on the fishery within the
Nushagak and Kvichak basins?
RESPONSE: Impacts of habitat loss and alteration are very difficult to quantify given the
lack of information on limiting factors, production and capacity estimates. We were unable
to comprehensively evaluate impacts at the population level. except for the most severe
cases where total losses of runs could be reasonably assumed.
• Is this loss significant in comparison to the fishery as a u hole?
RESPONSE: Losses of streams and wetlands under the mine footprint could not be related
to the fishery due to reasons listed above. For the TSF failure scenario that completely
eliminates or blocks access to suitable habitat in the North Fork Koktuli River, we estimate
that the entire Koktuli portion of the run (~28% of Nushagak escapement) could be lost.
Higher proportional losses would occur if significant downstream effects occurred due to
transport of toxic tailings fines beyond the Koktuli as modeled under the Pebble 2.0 TSF
failure.
• Are there local communities thai could lx- al'leclcd hy this specific loss?
RESPONSE: M ildlife and resident fish communities would be affected by reductions in
spawning salmon. Local communities would also be affected by the reduction, which is
now discussed in Chapters 12 and 13 (e.g.. Chapter 13 now contains tables which refer to
specific subsistence resources used by individual communities).
• Is fragmentation of the resource from this loss a significant impact (i.e., are there stocks
that are unique to the project area)?
RESPONSE: Stock structure and genetic diversity are not well known at the project scale,
but based on evidence from other parts of Bristol Bay watersheds, local adaptation is
highly likely. Discussion of fragmentation effects is now included in Chapters 7 and 10.
There is no discussion of engineering and mitigation practices in this section. The responsible
regulatory authority would require the project proponent to present a mitigation plan to
compensate for these impacts before permitting. Measures would include minimization of impact
through facility siting, reclamation if possible, and compensation if reclamation were not
feasible. A thorough analysis of possible mitigation approaches and the likelihood of their
success are necessary to fully evaluate impacts from the 'no-failure' scenario.
RESPONSE: Mitigation measures for design and operation are more clearly called out in the
revised assessment. While measures chosen here may differ from what is required during the
regulatory process, the assessment is not a mining plan and not an evaluation of a mining
plan. The assessment assumes that measures chosen for the scenarios would be as effective as
possible and examines only accidental failures rather than a failure to choose a proper
mitigation measure. Mitigation to compensate for effects on aquatic resources that cannot be
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 71
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
avoided or minimized by mine design and operation would be addressed through a regulatory
process that is beyond the scope of this assessment. Nevertheless, in response to public and
peer comments we have included a discussion of compensatory mitigation in Appendix J of
the revised assessment.
Steve Buckley, M.S., CPG
The engineering and mitigation designs associated with the no-failure mode of operation are
inadequate. There is no detailed discussion of engineering practices. There is insufficient
discussion of any potential mitigation measures and there is a lack of any detailed research into
applicable engineering and mitigation methods. Appendix I provides some engineering and
mitigation practices along with water quality mitigation and monitoring during closure; however,
these are not discussed or accounted for in the main assessment document.
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included as part of the mine scenarios in the draft
assessment, and this discussion has been expanded in the revised assessment. The mitigation
measures proposed within the mine scenarios are those that could reasonably be expected to
be proposedfor a real mine (they are a subset of options presented in Appendix I), all of which
were presented as appropriate for the Pebble deposit in Ghaffari et al. (2011). Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation would be addressed through a regulatory process that is beyond the
scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment.
The reorganization and additional discussions presented in Appendix J clarify when and where
mitigation would become part of the regulatory process. The chapter organization is much better
than the first draft with respect to Failure/No Failure modes.
Courtney Car others, Ph.D.
The no-failure mode of operation appears to be described adequately. The engineering and
mitigation practices appear to be sufficiently detailed, reasonable, and consistent, although I have
no particular expertise with which to evaluate this part of the assessment.
RESPONSE: No change suggested or required.
It would be helpful to have a clear statement about how well the local (geotechnical, hydrologic,
and environmental) conditions in this region have been studied and characterized. How much is
understood about the seasonal variation in these conditions and how those variations would
affect these scenarios? How well are statistics from mines and TSFs constructed in very different
environments likely to apply here?
RESPONSE: TSF failure probabilities are based on Alaska's dam classification and required
safety factors applied to the method of Silva et al. (2008). The discussion of failure
probabilities is expanded to clarify this issue. Therefore, the failure is not a consequence of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 72
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
any specific site conditions or seasonal phenomena. However, the discussion of local and
regional conditions has been expanded in the revised assessment. It is recognized that some
issues such as hydrology are very complex and additional information will be useful in future
analysis of any mine plans.
Dennis D. Dauble, Ph.D.
The description of the no-failure mode for mine operation appears adequate in terms of potential
mitigation measures that might be employed. I have limited knowledge of current engineering
practices and subsequent risks to the environment from best practices of modern mines,
including those operating under optimal conditions. However, it would be helpful to include a
short discussion on which mitigation measures would be mosi applicable to mining activities in
the Bristol Bay watershed.
RESPONSE: Standard design mitigation measures consideredfeasible, appropriate, and
'permittable' (as per Ghaffari et al. 2011) were considered and are discussed in Chapter 6 and
Appendix I of the revised assessment; these are standard measures common to other copper
porphyry mines. Evaluation of measures that would be proposed for an actual mine would
occur through the regulatory process. II h ether these same measures would be appropriate for
all locations within the Bristol Bay watershed would depend on the given site's specific
characteristics.
Gordon H. Reeves. Ph. I).
No comments on this question
Charles Wesley Slaughter, Ph.D.
Based on the actual history of other major resource extraction projects in Alaska and throughout
the world, a "no failure" assumption seems unrealistic. Rather, the assumption should be that
there will be failures, of \ aryinu modes and magnitudes, over the life of the project. This reality
is recognized in se\ eral sections of text.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. The "no
failure" chapter has been eliminated. The revised assessment differentiates between potential
effects from the footprint of a mine (Chapter 7), water treatment (Chapter 8), TSF failures
(Chapter 9), the transportation corridor (Chapter 10), and pipeline failures (Chapter 11).
In some sections in the Assessment, presumed "mitigation practices" are either cursory,
optimistic, or so general as to be un-supported. Examples include Section 4.3.7's cursory,
generalized statements about handling water: "Uncontrolled runoff would be eliminated.. .The
mine operator would capture and collect surface runoff and either direct it to a storage
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 73
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
location... or reuse or release it after testing and any necessary treatment"; "...water from these
upstream reaches would be diverted around and downstream of the mine where practicable";
"precipitation would be collected and stored..."; and "Assuming no water collection and
treatment failures, this excess captured water would be treated to meet existing water quality
standards and discharged to nearby streams, partially mitigating flow lost from eliminated or
blocked upstream reaches." Other examples from Chapter 6: ".. .assuming no water collection
and treatment failures" and "excess captured water would be treated.. .and discharged to nearby
streams..."
RESPONSE: Water management (mitigation) measures are more clearly described and
discussed in the revised Section 6.1.2.5, and in sub-sections for the mine components in the
scenarios. However, the intent of the assessment is not to specify technologies, beyond those
already identified by the existing preliminary mining plans. Rather, the assessment focuses on
the environmental outcomes of conventional modern mining practices and effluents.
John D. Stednick, Ph.D.
The no-failure mode is not adequately described Assessment of the effects of the mine is based
on large risk failures of low probability and did not include low risk failures of higher
probability. The report concludes (and emphasizes) that the mine footprint will disrupt/disturb
contributing watershed and wetland areas and result in hydiolouic modification. The hydrologic
modification affects salmonid habitats, particularly in low How conditions. Regulatory oversight
will include the State of Alaska, and certainly mitigation measures would be required. The task
is to address the adequacy of these mitigation measures
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. The revised
assessment no longer uses the term "no-failure", and does address effects from scenarios
having higher probability and lesser magnitude (e.g., failure to collect or treat leachate water)
in addition to those with lower probability and higher magnitude (e.g., TSF failure).
Mitigation to compensate for effects on aquatic resources that cannot be avoided or minimized
by mine design and operation would be addressed through a regulatory process that is beyond
the scope of this assessment. Severtheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment.
Pollutant/toxicity assessment focused on copper. Other metals can be presented to show the
range of metal concentrations for chronic and acute toxicity. Suitability of treatment processes
for all wastewaters can be included to address potential effects on receiving waters.
RESPONSE: Although copper is emphasized due to its dominance of the toxicity of leachates,
other metals were and are presented (Chapter 8).
The discussion of roads is mostly related to fish blockage and some soil erosion. Information on
current design standards was not included and tended to relay on dated references from logging
roads
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 74
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The discussion of roads covers risks related to filling and alteration of wetlands,
stream crossings, fine sediments, dust deposition, runoff contaminants, and invasive species.
Information on current design standards is now included within text boxes throughout
Chapter 10, and relies on recent literature. The failure frequencies cited in the assessment are
from modern roads and not restricted to forest roads. One of the papers used for general
information (Furniss et al. 1991) focuses on forest and rangeland roads, but it is a seminal
publication on the potential effects of roads, particularly as they relate to salmon.
There were no engineering or mitigation practices described in this section or in the document.
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included in discussion of the mine scenarios in the
draft assessment as part of the design. This discussion has been expanded in the revised
assessment.
Roy A. Stein. Ph. IX
No Failure Operations and Their Impact. What about the failure of continued monitoring, of
continual inspection, of continual, rigorous oversight ' This is more insidious than a catastrophic
failure of some sort, but perhaps just as dangerous (in lad. one research geochemist testified
during public testimony that of 150 hard-rock mines, none operated without leakage of leachate).
How can we be sure that mine operators will be held strictly accountable for their actions with
regard to best mining practices (a point emphasized by those who testified in favor of the Pebble
Mine that indeed best management practices would be used), meeting all the various and sundry
regulations, and communicating all of these activities back to the regulatory organization? Will
there be a force of will 011 the part of EPA or other regulatory body to be sure that all activities of
the operator are appropriate and within regulatory limits? The down-side of poor monitoring and
lack of rigorous o\ ersight is the loss of salmonid populations. These losses are, in my view, less
important than compromising human health and life. Yet, at the Upper Big Branch Mine in West
Virginia, dust standards have been exceeded for years, leading to a dust explosion that killed 29
miners 011 April 5, 2010. In turn, even surviving miners were not immune to these dust impacts,
for they suffer from "black lung", a condition that literally shortens their life by decades. In turn,
much of the monitoring of these conditions has historically been the responsibility of the owner
corporation, rather than an independent regulatory body, much like "the fox guarding the
chickens". Here at the I'ehMe Mine site, where only fish (but, of course, Native Alaskan
subsistence users, plus other human users as well) are at stake, would one expect rigorous
oversight by appropriate regulatory bodies? Skepticism leads to cynicism when contemplating
the Upper Big Branch Mine case history in the context of the Pebble Mine proposal.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. The "no
failure" chapter has been eliminated The revised assessment differentiates between potential
effects from the footprint of a mine (Chapter 7) water treatment (Chapter 8), TSF failures
(Chapter 9), the transportation corridor (Chapter 10), and pipeline failures (Chapter 11).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 75
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Holding the mining company accountable is done through the regulatory process and is
outside scope of this assessment.
Engineering Practices and Mitigation. I did not think that mitigation was well described in
text, but Appendix I is quite well developed and was instructive to me as I moved through the
documents. I would suggest including the ideas in Appendix I in the mitigation section of the
main report. Other comments on mitigation issues can be found below associated with Question
12.
RESPONSE: Mitigation measures for design and operation are included in the assessment
(Chapter 6), and are those that reasonably could be expected to be proposed for a real copper
porphyry mine (they are a subset of options presented in . \ppendix I). Many, if not all, of these
measures were presented as appropriate for the Pebble deposit in Ghaffari et al. (2011).
Mitigation to compensate for effects on aquatic resources that cannot be avoided or minimized
by mine design and operation would be addressed through a regulatory process that is beyond
the scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment.
Mitigation measures have been explained more dearly throughout the revised assessment.
William A. Stubblefield. Ph.D.
It is interesting and appropriate that the N\\ has included holli modes of operation in conducting
this assessment. This iippnmch provides some degree "hi>iniclinti"' for the assessment;
however, the degree of accuracy (i e . precliclahi lil\) lor either scenario cannot be known at this
time. The document appropriately acknow ledges that there are a variety of potential mitigating
factors (e.g., acts of Ciixl. accidents, market changes) that may render the assumptions used in
this assessment incorrect.
RESPO N SE: Our intention for the "no-failure " scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. There were a
number of comments on this approach so the revised assessment no longer uses the term "no-
failure" and addresses effects from the foot print (i.e., the no failure scenario), failure
scenarios having higher probability and lesser magnitude (e.g., failure to collect or treat
leachate water), and failure scenarios having lower probability and higher magnitude (e.g.,
TSF failure).
No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
The no-failure mode of operation failures is based on surface disturbances and potential
blockages caused by the various facilities. For example, for the mine pit, TSF and waste rock
facility, the surface areas of these facilities are used as a basis for calculating the streams and
wetlands affected by the mining activities. While the failure mode is adequately described,
engineering and mitigation practices are not adequately described by EPA.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 76
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Mitigation measures associated with the mine components (e.g., waste rock pile,
TSF, etc.) were discussed in the sections presenting those components. They were not repeated
in the section on failures. These measures are included again in the revised assessment
(Chapters 4 and 6), but are described in greater detail. However, the emphasis of the
assessment is on the consequences of failures rather than on the details of how inadequate
mitigation may cause the failures.
The EPA Assessment states on p. 8-1 "Routine operations are defined as mine operations
conducted according to conventional practices, including common mitigation measures, and that
meet applicable criteria and standards". The adverse effects listed are: direct impacts as a result
of removal of streams in footprint of mine pit and waste storage areas; reduced streamflow
resulting from water retention; removal of wetlands in the footprint of the mine; indirect impacts
of stream and wetland removal; diminished habitat quality in streams below road crossings; and
inhibition of salmonid movement from culverts that may block or diminish use of full stream
length.
RESPONSE: No change suggested or required.
Any mine in Bristol Bay will have to undergo a rigorous and lengthy regulatory review and
permitting process. I do not know of a process that will exclude consideration of the impact of all
mine facilities on the streams and wetlands in the region. Therefore, I would suggest that the full
implications of "mine operations conducted according to coin enlional practices, including
common mitigation measures, and that meet applicable criteria and standard" should have been
addressed in the report The EPA (2003) document on Generic Ecological Assessment Endpoints
for Ecological Risk Assessment specifically details the applicability of Section 404 of the CWA
in addressing community and ecosystem-level endpoints. "The CWA provides authority for the
Corps to require permit application to avoid and minimize wetlands impacts and requires EPA to
develop, in coordination with the Corps. the criteria used for Section 404 decisions. When
damages to wetlands are una\ oidaMe. the Corps can require permitees to provide compensatory
mitigation" It is unclear why this was not included in the evaluations.
RESPO N SE: This is not a permitting document. The purpose of the assessment is to evaluate
the effects of the operation of a mine on salmon ecosystems in the Bristol Bay watershed,
while following conventional practices, including common mitigation measures. Once those
effects are described. then it is appropriate to determine 1) if unacceptable environmental
effects are likely to occur and 2) whether those effects can be offset (made acceptable from a
regulatory standpoint) with compensatory mitigation. In other words, compensatory mitigation
is a next step and not within the scope of this assessment. Nevertheless, we have included a
discussion of compensatory mitigation in Appendix J in the revised assessment.
Similarly, one would expect that the regulatory reviews will require that the impacts resulting
from loss of streams, streamflow and road crossings will be addressed through engineering
designs, proposed mitigation measures, as well as regulatory and community engagement best
mining practices (see discussion above on "good" vs. "best" practices).
RESPONSE: Our scenarios included mitigation measures through engineering design and
operations to reflect standard industry practices. The purpose of the assessment was to
evaluate risks in the presence of these measures. The commenter is correct that alternatives
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 77
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
for such measures would be evaluated during the permitting (regulatory) process. No change
suggested or required.
On p. 4-33, it is stated that "Environmental impacts associated with premature closure may be
more significant than those associated with planned closure, as mine facilities may not be at the
end condition anticipated in the closure plan and there may be uncertainty about future reopening
of the mine". Further text describes potential negative impacts from such a premature closure.
One of the outcomes of the regulatory review and permitting will be the establishment of
financial assurance that will provide State and Federal Regulatory Agencies with the financial
resources to accommodate a closure. These obligations are typically reviewed on a 3 or 5-year
interval to also make sure that they are adequate to cover premature closures. If the mining
company is still managing the site, then they will have responsibilities under all Federal and
State Regulations and the dire picture painted by the Y. l\\ Assessment should not come to pass.
RESPONSE: The revision includes language that addresses financial assurance (Box 4-3).
Because of this major oversight of the realities when permitting and operating a mine it is
essential that the scenarios be reviewed by e\ alualinu the effects that regulatory requirements
and resulting mitigation methods would have on the no-failure conditions before completely
reworking the no-failure mode of operations and their impacts Other significant reports and data
that should be reviewed include typical permitting documents and resulting requirements for
similar mines in the US and Canada to obtain a range of potential outcomes. The results from
such an evaluation will also contribute significantly to the discussions in Alaska when the Pebble
Mine and other mines in Bristol Bay are brought forward to permitting.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. The revised
assessment no longer uses the term "no-failure", but simply presents effects from scenarios
having higher probability and lesser magnitude (e.g., failure to collect or treat leachate water)
as well as those having lower probability and higher magnitude (e.g., TSF failure).
Phyllis A". Weber Scannell, Ph.I).
Chapter 4 provides a detailed description of a hypothetical mine design for a porphyry copper
deposit in the Bristol liny watershed. Some of the assumptions appear to be somewhat
inconsistent with mines in Alaska. In particular, the descriptions of effects on stream flows from
dewatering and water use do not account for recycling process water, bypassing clean water
around the project, or treating and discharging collected water.
RESPONSE: The issues mentioned were discussed in Sections 4.2.3 (tailings storage) and
4.3.7 (water management) in the original draft document. They are now addressed in Chapter
6 of the revised document, which describes the mine scenarios, and in Chapter 4, which
provides generic background on porphyry copper deposits and mining. Streamflow effects
presented in Chapter 7 now reflect a complete water balance, including water capture and re-
use, bypass, and discharge from the wastewater treatment facility, as suggested by the
commenter.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 78
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Section 4.3.8, Post-closure Site Management, raises critically important issues - can a mine in
this area be designed for closure? Is it acceptable to develop and operate a mine that will require
essentially perpetual treatment? It is my belief that these are the essential questions that should
be addressed during any mine permitting process.
RESPONSE: EPA agrees that these are important questions to be addressed, but they are risk
management, not risk assessment, questions. The purpose of the assessment is to evaluate
risks to the salmon fishery from large-scale mining. Risk management decisions will be made
during the permitting process. Thus, no changes to the assessment were made in response to
this comment.
Section 4.3.8.1 raises concerns about long term water quality and quantity from the mine pit.
These concerns need to be addressed during a mine permitting process. Pit water quality
depends on how the pit is developed, what reclamation will occur, if reclamation will be
concurrent with mining, and what kinds of water treatment will be used. Tailings storage facility
(TSF) water quality depends on how the mine tailings are managed; it may lx- possible to use dry
stack tailings with sulfide removal rather than submerged tailings.
RESPONSE: EPA agrees that water quality can be influenced by design and reclamation, but
when the latter entails creating a pit lake there is little flexibility for reclamation concurrent
with mining. How tailings are managed within the impoundment can affect water chemistry,
and a dry stack with sulfides removed may produce the best water quality results after
reclamation if the fate of the sulfide tailings is never considered. According to Ghaffari et al.
(2011), 14% of the tailings produced will be pyritic. which equates to an average of28,000 tpd
in a 200,000 tpd mining operation (over 255 million tons during a 25 year mine life). These
tailings need to be managed in such a manner that oxidation does not lead to acidic drainage,
so the most effective wtiy is to deposit them subaqueously. Some suggested common mitigation
measures for management of the pit at and post closure are included in the revised assessment
(Chapter (>).
Paul Whitney. Ph.D.
Mitigation Plan Most mine perm it applications I have worked on include both mitigation to
minimize environmental impact and mitigation to compensate for environmental impact. The
assessment outlines a \ ariely of mi tigation measures to minimize impact, but no compensatory
mitigation. This is a concern, for 1 wonder if compensatory mitigation for the example mine is
even possible in the watersheds.
RESPONSE: The purpose of the assessment is to evaluate the effects of the footprint and
operation of a mine that follows conventional practices, including common mitigation
measures. Once those effects are described, then it is appropriate to determine 1) if
unacceptable environmental effects are likely to occur and 2) whether those effects can be
offset (made acceptable) with compensatory mitigation. Determining compensatory mitigation
is a next step and outside the scope of this assessment; however, we have included a discussion
of compensatory mitigation in Appendix J in the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 79
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The watersheds are characterized with descriptors such as "pristine" (e.g., page 6-29, last para,
second line), "nearly pristine" (e.g., pages 2-25 and 7-2) and "exceptional quality" (page 2-20). It
is also stated that the return of the salmon "fuel" (i.e., provide energy to) the terrestrial food web.
If in fact the watersheds are pristine or nearly pristine, the habitat is high quality and there is
little, if any, opportunity for compensatory mitigation (i.e., improving low quality habitat) in the
terrestrial and fresh water environments. For example, if 55 miles of streams and streamside
wetlands are lost to the mine footprint (page ES 15, first bullet), is it possible to find miles of
very degraded stream to plan for and implement compensatory mitigation? If one assumes a
mitigation ratio of 3:1 for enhancement, one might have to find 165 miles of degraded stream for
compensation. I suspect (but don't know) that there are \ cry few (i f any) miles of degraded
stream where compensatory mitigation could occur in the Bristol Bay watershed(s). If this is the
case, it might not be possible to demonstrate no net loss lor waters of the US, and this is
something EPA should be interested in.
RESPONSE: The comment is correct in stating that the exceptional quality of the Bristol Bay
environment leaves little opportunity for compensatory mitigation. Determining compensatory
mitigation is outside the scope of this assessment; however, we have included a discussion of
compensatory mitigation in Appendix J in the revised assessment.
I agree that the ecological resources can be ranked as ha\ ing high quality because the human
footprint on the habitat is small (i.e., few roads and \ i I limes), hut from an energetics (i.e., fuel)
and food web perspective, the pristine characterization may not he accurate. The commercial
catch of approximately 27.5 million fish each year ('up to 7<>"„ of the total number of sockeye
produced) is a lot of calories that are not flowing through the ecological foodwebs of the
watersheds. Granted, some of the commercial catch (if not caught) might not enter the
watersheds, but some and perhaps a lot would, especially in good run years. While the harvest
level might be sustainable, the loss of energy to commercial fishing causes pause to characterize
the watersheds as pristine or nearly pristine The potential impact of fisheries on energy flow has
been addressed by Piiuly et ill (2<)<)<)) and I .ihraliito etal. (2008). I wonder if it is technically
possible that a reduction in the commercial fishery is a compensatory mitigation measure.
RESPO N SE: Compensatory mitigation requirements address the need for project proponents
to replace aquatic resources and ecosystem functions that their project has impacted. Reduced
fishing harvests would not replace lost spawning and rearing habitat. Further, it would
remove the burden of compensation from the party that caused the damage. Determining
compensatory mitigation is outside the scope of this assessment; however, we have included a
discussion of compensatory mitigation in Appendix J in the revised assessment.
Effluent treatment. Water quality information in the assessment for benchmarks, background,
and leachate is extensive. A thorough review of the water quality and toxicity information is
beyond the scope of work of this review. After several reads of this information, it appears that
the work is good for copper. For example, work on salmonid olfaction and copper conducted by
McCarthy et al. (2007) is potentially important and is cited. The inhibiting effects of copper on
olfactory receptor neurons cited by McCarthy et al. (2007) at or above 2 |ig/L are lower than the
Alaska hardness-based standards and the biotic ligand model (BLM) standard in Table 5-14, but
are above the biotic ligand model standard in Tables 5-15 and 5-16.1 assume this is due to
differences in binding of copper by dissolved organics but I am not sure. Whether one decides to
use the 2 |ig/L benchmark, or the even lower BLM benchmarks that are in some cases below
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 80
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
background values in Table 5-19,1 think the key question is whether proposed leachate
processing can cost-effectively achieve benchmarks that hover around background
concentrations. The answer is beyond my level of expertise.
RESPONSE: BLM-derived copper criteria are derivedfor the different leachates. The values
depend on the co-occurring ions, which differ considerably among leachates and, in the case
of ambient waters, on dissolved organic matter. The low copper benchmarks would be
achievable with treatment by reverse osmosis, which has been used at other mines. Whether it
is economically feasible to achieve benchmarks close to background concentrations is outside
the scope of the assessment. No change suggested or required.
I do not agree with the assessment's critical question u hellier or not effects are observed at
these low levels (page 5-57, Exposure-Response Data from Analogous Sites, second sentence). If
effects are observed at background concentrations, it seems unreasonable to ask for an even
lower benchmark than background concentrations The uncertai lilies assessment at the bottom of
page 5-57 also seems unreasonable. The possibility that background concentrations are not
protective in particular cases seems highly unlikely lor one of the most productive salmon
communities in the world.
RESPONSE: The passages cited by the reviewer refer to the possibility of effects at copper
concentrations below criteria, not below background. Similarly. the studies mentioned found
effects below criteria levels but above background concentrations.
I can think of many questions that are more critical than looking lor effects on salmonids at
background or near background concentrations of copper I-'or example, it might be more
important to ask what concentrations of copper will result in a significant impact on the salmonid
populations and to ask what impact a mixing /.one would have on salmonid populations. Last but
not least, what are 1 lie potential impacts of all toxics on the many other non-salmonid species?
RESPO NSE: See response to previous comment regarding background concentrations versus
water quality criteria. Because water quality is so high at the site, the threshold for copper
toxicity is low based on the biotic ligand model. Effects on salmonid populations from
exceeding toxic thresholds cannot be estimated because the available monitoring data do not
characterize salmonid demographics or productivity in the streams draining the site. However,
the analysis has been expanded to include estimates of kilometers of stream habitat that would
be exposed to copper levels sufficient to cause aversion, sensory deprivation, decreased
reproduction, or kills. Mixing zones are not allowed by the State of Alaska for water quality
compliance in anadromous streams, and the available stream data are not sufficient for
mixing zone modeling. However, a discussion of mixing zones, including the amount of
mixing that would be required to reach nontoxic levels, has been added Non-salmonid fish
are not included as endpoint species. However, because copper and most other metals are most
toxic to arthropods, the assessment implicitly addresses non-salmonidfish (which depend on
arthropods in the food web) as well.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 81
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 4. Are the potential risks to salmonid fish due to habitat loss and
modification and changes in hydrology and water quality appropriately characterized
and described for the no-failure mode of operation? Does the assessment appropriately
describe the scale and extent of risks to salmonid fish due to operation of a
transportation corridor under the no-failure mode of operation?
David A. Atkins. M.S.
For the no-failure mode of mine operation, the risks to salmonid fish due to habitat loss and
modification in the vicinity of the project are described in terms of loss of lengths of stream or
areas of wetlands. Project proponents state that the mine will only impact a very small fraction of
the watershed (under a no-failure scenario). It is important lo establish whether the modeled
impact (e.g., the loss of 87.5 km of streams) is significant, both in terms of the absolute impact,
as well as the effect on ecosystem fragmentation
RESPONSE: Footprint effects on habitat loss are now characterized in relation to the
distribution of habitat conditions throughout the larger watersheds. Fragmentation effects are
not anticipated at the mine site, apart from blockage of headwater streams as described, but
are anticipated in the case of the transportation corridor (C liapter 10) and TSF failure
(Chapter 9).
In addition, project proponents often stale they will |">iesei\ e and e\ en improve the fishery. As
mentioned in the answer lo the previous <.|iiestion. il would be helpful lo know what kinds of
mitigation efforts could he employed - minimization, reclamation and compensation - and have
some assessment of the potential effectiveness
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included in discussion of the mine scenarios in the
draft assessment, and this discussion has been expanded in the revised assessment. The
mitigation measures proposed within the mine scenarios are those that could reasonably be
expected to be proposed for a real mine (they are a subset of options presented in Appendix I),
all of which were presented as appropriate for the Pebble deposit in Ghaffari et al. (2011).
Reclamation is not mitigation. but the revised assessment includes also some suggested
measures to be used in closure post-closure to reclaim the disturbed areas. Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation would be addressed through a regulatory process that is beyond the
scope of this assessment. Severtheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment.
The Assessment determines that construction of the transportation corridor could alter the
habitat, chemistry, and the migration path across the corridor for the over 30 streams that the
corridor will cross or come near. The report further states that the corridor could affect 270 km of
streams below the corridor and 240 km of streams above, but that there is no way to assess the
magnitude. Therefore, the impacts of the corridor on fish populations are unknown, and this
impact is not described in a way that can allow a reviewer to draw any conclusion.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 82
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The revised assessment states that "the exact magnitudes of changes in fish
productivity, abundance and diversity cannot be estimated at this time," but summarizes the
species, abundances, and distributions that would potentially be affected Also, the assessment
concludes that, assuming typical maintenance practices after mine operations, approximately
15 of 32 culverted streams with restricted upstream habitat would be entirely or in part blocked
at any time. "As a result, salmonidpassage—and ultimately production—would be reduced in
these streams, and they would likely not be able to support long-term populations of resident
species such as rainbow trout or Dolly Varden."
Further, the references for road design and construction practices seem to be more representative
of forest and rangeland roads than the type of road that would likely be constructed for this type
of project. It would be helpful to cite experience from other transportation corridors constructed
for mining and oil and gas projects and developed recently in Alaska.
RESPONSE: Because the proposed mining would take place in an undeveloped area, the
literature is necessarily from areas outside of Bristol Hay. Further, we found no data
concerning the performance of culverts for mining or oil and gas projects in the region.
However, to the extent possible we used examples from representative environments. The
failure frequencies cited in the assessment are not restricted to forest roads. One of the papers
used for general information (Furniss et al. 1991) focuses on forest and rangeland roads, but
it is a seminal publication on the potential effects of roads, particularly as they relate to
salmon. The general conclusions of that paper should be applicable to the transportation
corridor considered in the assessment. Information on current design standards is now
included within throughout Chapter 10. and relies on recent literature.
Steve Buckley, M.S., CPCi
Risks lo fish due lo habitat loss and modification and changes in hydrology and water quality are
overl) simplified ui\ en the broad parameters used to model these potential risks. More specific
details 011 the water balance would help define potential risks to fish from dewatering and habitat
loss. For example, there is 110 attempt to identify groundwater flow paths or the specific response
of various landlbims to seasonal changes in precipitation and runoff, yet 34 pages are dedicated
to an attempt to quantify these impacts. More detailed information is needed to accurately
quantify the changes in anticipated runoff and infiltration in the proposed area to determine
potential impacts to hydrology and water quality.
RESPONSE: A revised and much more detailed water balance and streamflow analysis is now
incorporated in the assessment (Chapters 6 and 7). The revised draft also includes estimates of
the specific changes in flow at individual gages. The commenter is correct that the assessment
does not specifically address the potential changes in landforms or, for that matter, to
vegetation in the land areas such as the drawdown zone which would be among the areas most
affected by the mine development. We consider that any such changes would have only a
secondary impact on salmon relative to the impacts that the assessment does address.
The additional information presented in Chapter 3 on stream characterization along with the
information in Figure 7.14 on groundwater and fish habitat are major improvements to the
document.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 83
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Additional ecological information on the contributing watershed area for each fish bearing
stream crossing would help identify the potential impacts to fish due to the construction and
operation of a transportation corridor.
RESPONSE: Additional information on watershed attributes (discharge, channel gradient,
floodplain potential) of streams crossed by the transportation corridor, and their importance to
salmonids, is now included in the analysis presented in Chapter 10.
This information is a critical addition to the revised assessment.
Courtney Car others, Ph.D.
Six key direct and indirect mechanisms are identified to pose potential risk to salmonid fish
species: eliminated or blocked streams (87.5-141.4 km), reduced stream flow, removal of
wetlands (10.2-17.3 km), indirect effects of stream and wetlands removal (downstream effects
likely diminishing fish production), diminished habitat quality downstream of road crossings,
and blocked movement of salmonids at road crossings. These mechanisms are described clearly.
The report appears to appropriately describe the scale and extent of risks under a no-failure mode
of operation, although I have no particular expertise with w liich to evaluate this assessment.
RESPONSE: No change suggested or required.
DennisD. Dauble, Ph.D.
The assessment describes the number of stream miles impacted under each mode of operation,
including miles blocked and eliminated I .ess specific were descriptions of impacts due to
sedimentation and leachates W hat is lacking is quantitative estimates of spawning and rearing
habitat that would be lost relative to the total habitat available. Having this information would
help |iro\ ide perspective of overall risk to individual watersheds and the Bristol Bay watershed
as a whole. Risks to salmonid fish due to changes in water quality (i.e., toxic materials) need to
consider differences in sensitivity and behavioral response according to salmonid life stage.
RESPONSE: Stream habitat losses are now characterized in relation to the distribution of
habitat conditions throughout the larger watersheds (Chapters 3 and 7). The assessment of
risks from aqueous toxicity distinguishes overt toxic effects on early life stages from
behavioral effects on adults.
Surface water characteristics of site watersheds within the area of probable impact are detailed in
Table 5-17, but not so for other streams and lakes in the broader watershed. More information
should be presented where available. It is not clear whether potentially affected streams and
lakes might be nutrient limited (seems that they might be given their dependence on MDN). For
example, include N or P concentrations and some discussion about primary and secondary
productivity.
RESPONSE: We recognize that nutrient status, and more importantly prey availability, is a
critical component of habitat capacity for fish in these systems, and may be strongly driven by
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 84
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
salmon derived nutrients. We concur that more information is needed regarding potential
limiting factors for salmon productivity and capacity, and that food availability may be one
such factor. However, because water chemistry data may not provide a complete picture of
trophic status, particularly where direct consumption of salmon flesh, eggs, and fry is of such
high importance, and because nutrient status is a water quality or habitat parameter not
directly influenced by mining operations as outlined in our conceptual models (e.g., Figure 7-
1), we determined that nutrient status of area streams is outside the scope of this assessment.
I found risks to salmonid fish due to operation of the transportation corridor well-described with
respect to spatial distribution of fish and their habitats.
RESPONSE: No changes suggested or required.
Gordon H. Reeves. Ph.D.
The potential risks to the freshwater habitat of anadromous salmonids arc appropriately
characterized and described for the no-failure mode of operation. The report considered the
primary potential impacts of mine development and operation that could impact habitat and
quantified the impacts where possible The analyses seemed sound and logical, given the
acknowledged limitations about the actual mine location and operation.
RESPONSE: No change suggested or required.
One possible factor that could inlliience the results was the use of the USGS 1:63,360 maps for
developing the stream network These maps generally undcrrcprcsent the amount of small
streams, which can he ecologically important contributors to the overall productivity of the
freshwater habitat of anadromous salmonids This is acknowledged in the limitations (p. 5-46).
Thus, the potential loss and modification of habitat that the report describes could be considered
minimal at this time It would he prudent to confirm the accuracy of the stream layer developed
from the I (•>3,3(',l) maps in any future analysis
RESPONSE: The EPA agrees with this comment. No change to the assessment suggested or
required.
The potential impact (if the mine development and operation on the productive capacity of the
various river systems could he de\ eloped more fully to gain better insights into potential impacts
of the mine. The authors considered the amount of habitat that could potentially be impacted by
mine development and operation by estimating the stream length that would be impacted and by
considering the percent of spawners of the various species (from ADF&G surveys) observed in
potentially impacted areas. However, the productive capacity of given stream reaches for a
given fish species can vary widely. Any additional analysis could consider using Intrinsic
Potential (IP) (Burnett et al. 2007. Ecological Applications 17:66—80), which considers local
geomorphic features to estimate the potential of a given stream reach to provide high quality
habitat for a given species. The concept, developed for use in the Pacific Northwest (PNW), has
been applied successfully for Chinook salmon in the upper Copper River (A. Bidlack, EcoTrust,
Cordova, AK, unpublished). The IP model for Chinook salmon from the PNW that was used in
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 85
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the Copper River was modified after discussion with local biologists. Similar modification may
be needed for the PNW IP model for coho salmon to be used in Bristol Bay.
RESPONSE: We now include a characterization of stream channel gradient, watershed
terrain (% flatland), and mean annual flow for all streams in the two watersheds. We are
unable to build a complete IP model, as this would require validation and more elaborate
construction of metrics appropriate to this region, but our preliminary characterization
provides the building blocks for assessing the distribution of key habitat-forming and
constraining features across these watersheds.
Another factor that I believe merits further consideration is the potential impact of altered
thermal regimes of discharge water from treatment facilities (p 5-28). Warmer water could have
potential ecological impacts, particular during the time u hen eggs are in the gravel. Eggs could
develop more quickly and fry could emerge earlier as a result of e\ en minor changes in water
temperatures (see: McCullough, D.A. 1999. A re\ iew and synthesis of effects of alternations to
water temperature regime on freshwater life stages of salmonids, with special reference to
Chinook salmon. US Environmental Protection Agency, Seattle, EPA ^ I <)-R-i)9-010. 279 p.; and
McCullough, D.A., J.M. Bartholow, H.I. Jager, and I I co-authors. 20<)1) Research in thermal
biology: burning questions for coldwater stream fishes. Reviews in Fisheries Science 17: 90-
113.). These changes could be significant ecologically.
RESPONSE: This is now addressed in Chapter 8.
The report noted in several places that the potential impact on groundwater flows was not
understood at this time hut that disruptions of flow paths could have critical impacts on aquatic
resources. One impact that was not mentioned is the loss of over-wintering habitat. K.M.
Burnett (U.S.D.A. forest Scr\icc. PNW Research Station, Corvallis, OR., draft report) found
that the major overwintering areas lor coho salmon in the Nome River, AK were at points of
groundwater inputs The groundwater influx created areas that were less likely to freeze during
winter
RESPONSE: Overwintering effects from thermal changes are now described in Chapter 7.
Charles H eslev Slaughter, Ph.D.
Yes, the risks to salmonids are well characterized with regard to the hypothetical mine operation
itself. However, I suggest that the concept of "no failure," if taken as applying to the entire
operation from inception through operation, is not realistic.
RESPONSE: The "no failure" scenario was not meant to represent a realistic scenario.
Rather, it was meant to illuminate the effects that would occur solely from a mine footprint,
even in the absence of accidents or failure. The revised assessment no longer uses the term
"no-failure", but simply presents effects from scenarios having higher probability and lesser
magnitude (e.g., failure to collect or treat leachate water) and those having lower probability
and higher magnitude (e.g., TSFfailure).
The Assessment makes a fair start toward considering the risks to salmonids from the potential
transportation corridor. However, the many issues regarding stream and wetlands directly or
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 86
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
indirectly affected by roads and pipelines are not fully explored. The extent (length, area) of
streams and wetlands affected, as outlined in the text, should be considered a very optimistic
lower estimate. The specific issues mentioned, such as bridge or road maintenance, culvert
blockage or failure, erosion from cuts, fills, and the roadway itself, are all significant. I simply
suggest that the potential consequences of imposition of the (hypothetical) transportation
corridor, and future expansions consequent to ancillary infrastructure development and further
additional resource extraction projects, would be broader, more severe and of more consequence
(and thus should receive more emphasis) than the Assessment indicates. I suggest more fully
incorporating Frissell and Shaftel's Appendix G into the body of the Assessment.
RESPONSE: The revised assessment notes that the characterization of both stream length and
wetland area affected likely represents a conservative estimate of the potential effects of the
transportation corridor on hydrologic features of this area. The cumulative risk section
(Chapter 13) has been expanded to include the transportation corridor, ancillary mining
development and secondary development. Additional information from Appendix G is
incorporated into the main text, and the appendix is referenced in a number of places.
John D. Stednick, Ph.D.
To address this question, a water balance needs to be de\ eloped for the study area watersheds.
Develop a water balance that includes all the principal components and how they may vary in
time and space. The site characterization needs significant impro\ ement, particularly as related
to hydrologic inventories and processes. Little to 110 data are presented on temperature,
precipitation, evaporation, frozen soils, soil moisture storage, and groundwater storage and
movements. The data that are presented often ha\ e unreasonable significant figures. The linkage
between surface and groundwater needs to be better demonstrated. Hyporheic exchanges are
recognized as being important, but the assessment does not demonstrate this linkage.
RESPONSE:. I complete annual water balance with patterns of temporal variability has been
developed and incorporated into the revised assessment. Detailed temporal variability is
beyond the scope of the assessment. The significant figures of reported data and analysis
results have been reduced to a reasonable number. Detailed data presentations on
temperature, precipitation, evaporation, frozen soils, soil moisture storage, and groundwater
storage and movements are generally not reported. We cite sources of information, and data
are cited and reported as needed for assumptions of analyses conducted for the assessment.
Additional information on surface water- groundwater interactions has been added.
Iliamna Lake hydrology needs to be characterized. What are the inflows, outflows, and turnover
rates? What is the existing water quality in the lake? Aquatic life should be characterized as well.
What is the risk of pollutants entering the lake from the road corridor or upstream mine
development operations?
RESPONSE: The overall hydrology of the lake was not included because none of the
scenarios would result in a change in the hydrology or water quality of the lake as a whole.
Rather, any effects in the lake would be limited to the vicinity of outflows from the affected
streams. Risks from contamination of tributaries to Iliamna Lake are discussed in Chapters 8,
10 and 11 of the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 87
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Climate variability is recognized as a game changer. What are the potential future scenarios for
temperature and precipitation changes in southwest Alaska, and how will these scenarios affect
the water balance? How will climate change affect the availability of water for mine operations,
including processing and potable uses?
RESPONSE: Climate change projections and potential impacts are now discussed in Chapter
3.
Similarly, a complete water quality characterization is lacking. What is the water quality in
surface waters, groundwaters, in time, and in space? What is the definition of background water
quality? Numerous exploratory activities have taken place in the watershed and have the
potential to affect water resources. How were these separated or addressed? Given the geologic
and geomorphologic settings for the study area, are we comfortable that the watershed ridges
delineate the watershed area? Groundwater mo\ enients may ignore the physical watershed area
boundary and follow groundwater gradients. Stream llow measurements from the gauged
watersheds could be useful in answering this question. Similarly, the linkage of groundwater and
hyporheic exchange needs to be better demonstrated Do these exchanges occur in all stream
segments and gradients? What effect does the groundwater ha\ e on stream temperatures? Are
depth to groundwater readings available ' Is a groundwater monitoring program in place?
RESPONSE: The water quality described as background by the PLP in their Environmental
Baseline Document was accepted as such in the assessment. II ater quality in the three streams
that drain the site was presented in Table 5-1 ~ (now Table 3-4). The assessment now includes
information on estimated groundwater interaction strength across the project area streams
(Figure 7-14).
The tables and hydrographs (pages 5-32 to 5->M are unclear. What streamflow changes are
associated with what salmon species and life stage ' A boundary condition for adults is different
than for fry
RESPONSE: Maps of species distributions in relation to affected stream segments are now
included in Chapter 7. The environmental flow analyses are not species or life stage specific,
but assume an overall risk associated with proportional deviations from the baseline flow
regime.
The proposed mine will use large quantities of water in ore processing and transport. How much
is required and how ill this affect water resources; both surface and groundwater?
RESPONSE: The water balance has been extensively revised (Chapters 6 and 7), and updated
estimates are incorporated into the streamflow computations provided in Chapter 7. Effects on
groundwater resources are explicitly incorporated in the analysis of the pit dewatering,
associated cone of depression, leachate leakage from the TSF and the waste rock piles, and
interbasin groundwater transfers.
The no-failure mode of operation is predicted to change the watershed contributing area and
hence streamflow, and uses the boundary condition of a 20% change in streamflow as significant
salmonid habitat loss. The assessment assumes a liner response between watershed area and
streamflow contribution, and a linear response between habitat productivity and watershed area.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 88
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: A more comprehensive water balance is now used to estimate stream/lows,
which incorporates losses and additions due to pit dewatering, and wastewater treatment plant
processing and distribution, such that the relationship of watershed area and streamflow is not
linear. We do not assume a linear response between habitat productivity and watershed area.
Upland settings are probably more productive in terms of productivity and should be addressed
as such.
RESPONSE: Relative productivity of aquatic habitats has not been extensively documented
across the region.
Toxin assessment focused on copper, and other metals can he presented to show the range of
metal concentrations for chronic and acute toxicity, i.e. arsenic, molybdenum, silver, barium,
and lead. Given the very clean waters (low hardness and organic carbon), the chronic toxicity of
various metals should be evaluated. Water quality \ aries in time and space in the study area, and
a better characterization of water quality could he developed. Metal loads could be calculated
with streamflow records. What is the proportioning of dissolved versus lolal metals? Are metals
transported with sediments? Do organic carbon lluxes change in space or lime0
RESPONSE: The influence of receiving water chemistry was incorporated to the extent that
current science allows. The toxicity of copper was corrected for water chemistry using the
biotic ligand model, and the toxicity of other metals was corrected for hardness, when models
were available. In each case, water chemistry of the individual receiving stream was used for
the correction, Also, instream concentrations are based on streamflows, including changes in
streamflows due to mine operations. Both sediment and organic matter concentrations are
quite low in all three streams at the Pebble site, so copper remains dissolved in the model and
other metals in leachates are likely to remain dissolved as well.
Salmonid risk from travel corridor The proposed road location has the potential to affect 270 km
of stream between stream crossings and Lake lliamna. The expected road erosion and sediment
production has known effects 011 salmonid resources. The discussion of the travel corridor does
not include the potential for road failures, landslides, blocked culverts, or ditch failure. The
discussion does not talk about traffic volume or the potential of hazardous material transport on
the travel corridor Need to address road maintenance, fugitive road dust, and road chemicals
either dust or ice control.
RESPONSE: The original draft assessment included discussion of the potential for road and
slope failures, blocked culverts, and soil erosion from road cuts, borrow areas, road surfaces,
shoulders, cut-and-fill surfaces, and drainage ditches. The revised assessment factors traffic
estimates into assessments of chemical spills from transport truck accidents (Section 10.3.3)
and impacts from dust (Section 10.3.5). Salts used for to reduce dust and improve winter
traction are discussed in Section 10.3.3. Potential mitigation measures for stormwater runoff,
erosion, and sedimentation are discussed in Box 10-3.
There is no discussion of water processing after delivery of the slurry to the sea port and return
of waters back to the mine site.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 89
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Section 6.1.2.5 of the revised assessment discusses that the water would be
returned to the process water ponds. The scenarios indicate that slurry water would be
returned to the site without treatment other than removal of solids.
Roy A. Stein. Ph.D.
No-Failure Mode of Operation. My comments regarding the no-failure mode of operations and
their impact on salmon can be found under Question 3.
RESPONSE: See response to Question 3.
Road Use I: Page 5-60. Beyond calcium chloride, how can we lie confident that the typical
chemicals that derive from highway use will not occur 011 this mine road (as noted on page 5-
60)? Is it because the low volume of traffic? If so. would not we expect accumulation through
time... over the 78 years of the mine operation (see Appendix G for some detailed analysis:
should some of this material be added to the main report?)? What about the impact of road dust
on nearby aquatic systems (wetlands, streams, ri\ ei s. etc )°
RESPONSE: The text relating to traffic and contaminated runoff has been modified in the
revised assessment to read: "It is unlikely that the potential transportation corridor would
have sufficient traffic to significantly contaminate runoff with metals or oil, but stormwater
runofffrom roads at the mine site itself might contain sufficient metal concentrations to affect
stream water quality. " Though traffic-associated contaminants may be expected to increase
over time, they would probably not be as significant as stormwater runoff-associated metals
from the mine site. . I.v noted in the revised assessment, the main impact of dust from the
transportation corridor on salmonids would likely be a reduction in riparian vegetation and
subsequent increase in fine bed sediment. The main impact of dust at the mine site would be a
direct increase in fine bed sediment due to mine construction and operation (the effects of
increased sediment loading are discussed in Section 10.3.4).
Road I'se II: Page 5-62 to 5-63 (plus Appendix G: again, as with other appendices, include
more of this information in the main report). Will there be frost heave of the road bed such that
specific structures will have to be installed to prevent this movement of the road bed? These
roads will be treated with chemicals, such as calcium chloride, to keep the dust down and
contribute to an ice-liee condition, but no data are available for the impacts of these chemicals
on nearby streams. How then do we deal with this issue (page 5-62 and 5-63)? The suggestion
is that one needs to ha\e roads built at least 8 meters from streams, but this cannot be the case in
this situation, simply because of the large number of streams, rivers, and wetlands along the road
corridor? More detail as to the impact of the transportation corridor should be added, including
issues, such as truck accidents, fuel spills, other chemical spills, etc.
RESPONSE: The assessment does not address potential frost heave of the road bed, although
this factor will need to be considered during design of a road. Additional information has been
added to the revised assessment on the potential impact of calcium chloride on nearby
vegetation, surface water, groundwater and aquatic species. According to the USDA Forest
Service (1999), application of chloride salts should be avoided within 8 m of water bodies. We
agree that the 8 m buffer zone for salts would be difficult to maintain, but it could be achieved
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 90
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
at some cost in dust suppression and winter road salting. It would not require keeping roads
out of that zone.
Additional information from Appendix G is incorporated into the main text, and the appendix
is referenced in a number of places. The revised assessment contains greater detail on the
potential impact of the transportation corridor. For example, the assessment now factors
traffic estimates into assessments of chemical spills from transport truck accidents (Section
10.3.3) and potential impacts from dust (Section 10.3.5). Fuel spills are covered in Chapter 11.
Road Use III: page 5-71 (plus Appendix G). The road will intersect multiple streams and rivers
along the northern end of Iliamna Lake, where as many as one third of the sockeye salmon in this
lake spawn. And this is where the causeway across Iliamna Lake will be built as well. From my
perspective, it seems that impacts on spawning sockeye will be large in this area (without saying
anything about causeway: will there be culverts or bridges to allow water and fish to
communicate with the rest of the lake)? I would argue this is important, given salmon are
attracted to certain odors and water-flows and these odors and water-flows are coming from
inlets streams into Iliamna Lake. Preventing any soil of blockage of water llow or salmon
migration would be the goal. Are there other issues that should be considered when building this
causeway?
RESPONSE: The assessment makes no mention of a causeway across Iliamna Lake. We
believe the commenter is referring to a proposed causeway over the upper end of Iliamna Bay,
which is part of Cook Inlet. Culverts or bridges would be built to allow fish access between
streams crossed by the proposed road and Iliamna Lake. The issue of culvert blockage is
discussed in the assessment.
Road Use IV. Points made In public testimony reinforces the idea that as this area is opened to
the public, the opportunity for new. invasive species to colonize this pristine ecosystem increases
dramatically, likely to 100% Simply put. in\asi\e species will now be carried by humans via
the road. inad\eitenlK. into this pie\ iously inaccessible watershed.
RESPO N SE: EPA assumes that the proposed road would be closed to the public during
mining operations but potentially could become a public road after mining operations cease.
Even when not open to the public, construction and operation of the proposed transportation
corridor increase the probability that new terrestrial and aquatic species will be transported to
and potentially establish themselves in the Bristol Bay region. If the road were opened to the
public, the probability of colonization by invasive species may increase further, but rates of
introduction by industrial and public vehicles cannot be distinguished given available
information. Invasive species are addressed in Section 10.3.6 of the revised assessment.
William A. Stubblefield. Ph.D.
The document appears to adequately address potential questions associated with habitat loss due
to hydrologic changes, especially considering the hypothetical nature of the mine and the lack of
specific detailed information regarding an actual proposed facility and all of the associated
operational details of the facility. The assessment of potential impacts and ecosystem protection
parameters is predominately based upon the publication of Richter et al. (2011). Additional
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 91
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
support and evaluation of these recommendations for fisheries populations in the Bristol Bay
area should be closely evaluated.
RESPONSE: Prompted by this comment, we consulted with regional biologists and
hydrologists to evaluate the suitability of the sustainability boundary approach for flows. We
asked them if there was any reason that fish populations in these streams, or the specific
hydrology of the area, made it exceptional with regard to this approach (e.g., was there any
reason to think that the Richter approach was not applicable here). We received uniform
support for applying this approach to Bristol Bay streams. We strengthen our emphasis that
this is a precautionary approach, and that the detailed hydrologic and habitat modeling work
that PLP contractors have begun will help provide a useful basis for more sophisticated flow-
habitat modeling.
Dirk van Zvl, Ph.D., P.E.
Chapter 5 of the EPA Assessment is entitled: "Risk Assessment: No Failure" Chapter 5 presents
an evaluation of habitat loss and modification resulting IV0111 the hypothetical mine. A summary
of the "risks" associate with the "no failure" case is pro\ ided in Chapter 8. There is specific
focus on evaluating the magnitudes of llic losses and modifications to the environment.
RESPONSE: No change requested or required.
A risk assessment addresses three questions (Kaplan and (iarrick. N81):
• What can happen' (i e . What can go u long')
• How likely is it that that will happen '
• If it does happen, what are the consequences''
There are a large number of risk assessment methods and it is common to express the magnitude
of risk as a combination of likelihood of occurrence and consequences (TEC, 2009). This is the
typical outcome for engineering assessments of systems For example, in the case of a Failure
Mode and Effects Analysis (FMEA), it would be typical to develop a risk matrix to combine
likelihood of occurrence and consequences to express the level or magnitude of risk in
qualitative terms (Robertson and Shaw, 2012).
The EPA Assessment describes the two components of risk but does not provide any information
on the magnitude of the risk I'or example, for the no-failure condition it describes the length of
streams, areas of wetlands, etc that will be impacted by developing the mine, i.e. the
consequences. One may argue that the likelihood of occurrence of these consequences is unity
(or certainty) if the mine is developed, as this is not specifically addressed by the report.
RESPONSE: The risk assessment does address engineering risks in the manner specified,
including magnitudes of spills, leakage, etc., and the consequences for water and habitat
quality. We now explicitly clarify that losses of stream length are unavoidable for a project of
this magnitude (Chapter 7). The magnitude of the risk or the likelihood that the stream
lengths will be lost if the mine is constructed is 100%.
One would next expect an expression of the magnitude of this risk based on some comparison of
the consequences to a set of outcomes that could result in acceptable or unacceptable risks. The
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 92
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
EPA suggests this as an approach in its 1998 Guidelines for Ecological Risk Assessment (EPA,
1998): "In some cases, professional judgment or other qualitative evaluation techniques may be
used to rank risks using categories, such as low, medium, and high, or yes and no". Quantitative
approaches such as fuzzy logic has also been used to develop expressions of magnitude of risk as
described by EPA (1998): "For example, Harris et al. (1994) evaluated risk reduction
opportunities in Green Bay (Lake Michigan), Wisconsin, employing an expert panel to compare
the relative risk of several stressors against their potential effects. Mathematical analysis based
on fuzzy set theory was used to rank the risk from each stressor from a number of perspectives,
including degree of immediate risk, duration of impacts, and prevention and remediation
management. The results served to rank potential environmental risks from stressors based on
best professional judgment".
RESPONSE: Although the 1998 ERA Guidelines describe professional judgments as an
acceptable method for ranking risks in appropriate circumstances, it is not appropriate for this
assessment. The purpose of this watershed assessmen t is not to rank risks. It is to estimate, as
far as existing data and knowledge allow, the risks associated with proposed and potential
mining activities in the Bristol Bay watershed. EPA decided during the problem formulation
that this assessment would be based on published science. . 1// assessment based on elicitation
of expert judgment could be performed in the future, if desired.
It is unclear to the reader how significant a loss of 87 5 km of streams in the Nushagak River and
Kvichak River watersheds is to the o\ em 11 ecosystem. Are there any criteria that can be used to
develop such an expression? Can a multi-stakeholder workshop (as is often done) be used to
develop such criteria and expressions of risk magnitude ' Without ha\ ing such expressions of
risk magnitude it is impossible lor those without specific expertise in salmonids to evaluate
whether this is a significant risk. Price et al. (201 states that: "Between 1999 and 2008, 3,500
fish passage barrier culverts were replaced with fish-passable structures, reportedly opening
nearly 5,955 km offish habitat in Washington streams (Governor's Salmon Recovery Office
2008)" Comparing the loss of S5 km to this gain of 5,955 km seems to imply that 85 km loss
may represent a relatix ely small risk, which may not be the case at all. However, the EPA
Assessment does not pro\ ide any insight in the magnitude of risk except to provide a value for
the consequences.
RESPONSE: The purpose of the assessment is not to assign significance to the risks, but to
provide information for decision-makers on the consequences of mining.
The EPA did conduct a multi-staUeholder conference to determine the significant endpoints
and exposure pathways for the assessment. However, the EPA decided during the problem
formulation that this assessment would be based on published science. Therefore, a multi-
stakeholder workshop would not be an appropriate mechanism to estimate risks or their
significance.
Comparing the potential loss of salmon-supporting streams in the Nushagak and Kvichak
River watersheds to restoration of streams in Washington for salmon recovery may not be very
useful. The comment does illustrate that seemingly inconsequential or insignificant losses
have frequently led to diminished or even lost salmon stocks. In the example cited, apparently
3,500 seemingly inconsequential actions had to be remedied at public expense because of their
cumulative impacts.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 93
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Similar comments can be made with respect to the relative risks associated with the other losses
of ecological functions for other failure modes.
RESPONSE: See response to previous comment.
It is recognized that it is important to maintain separation between the risk assessment and risk
management functions. As expressed by the National Research Council Panel in their report on
Science and Decisions (NRC, 2009): "The committee is mindful of concerns about political
interference in the process, and the framework maintains the conceptual distinction between risk
assessment and risk management articulated in the Red Book. It is imperative that risk
assessments used to evaluate risk-management options not be inappropriately influenced by the
preferences of risk managers".
RESPONSE: The EPA agrees with this comment and its implication that assessors should not
judge significance. No changes suggested or required.
Providing an expression of risk magnitude should not interfere at all in the separation of risk
assessment and risk management, but should pix>\ ide the risk manager with one extra level of
analysis and insight from the expert assessor of i lie problem ill hand. Multi-stakeholder
interaction will only serve to enhance the value of the risk milking.
RESPONSE: See response to previous comment on this topic.
On p. 4-33, it is stated that after closure "No l\\(i waste rock would remain on the surface". It is
also stated in Chapter 4 that PAG and NAG waste will be segregated. On p. 5-48, it is stated that:
"However, the primary concern during routine operation would be waste rock leachate. That
leachate would become more voluminous as the waste rock piles and uses of waste rock for
construction increased during operation. After mine closure, it would be a major source of
routinely generated wastewater along with water pumped from the TSF and pit. Leachate
composition from tests of the three waste rock types (Tertiary, East Pre-Tertiary, West Pre-
Tertiary) is presented in Tables 5-.1.4 through 5-16". There is no specific indication which of
these waste rock types could be described as PAG or NAG and Chapter 5 seems to assume that
these 3 samples are representati\ e of the total amount of waste rock, about 4 billion tonnes for
one mine scenario If all the PAG material will be removed from the surface, as stated in the
scenario in Chapter 4. and the Y\(i will not generate acid drainage, then it is difficult to
understand why the waste rock piles and waste rock used for construction (supposedly all NAG
at this stage) would be the major source of "routinely generated wastewater."
Note that it is further unclear w liy there would be water pumped from the tailings and the pit if
the TSF were closed, as discussed above, and if it will take the mine pit 100 to 300 years to fill.
Some clarification is in order.
RESPONSE: This issue has been clarified in the revised document. PAG and NAG waste
rocks would be identified during the course of mining, but the available test results indicate
thatpre-Tertiary rock is PAG. The PAG waste rock would be segregated and none would
remain at mine closure. However, some will be on the surface during operation. In addition,
the NAG waste rock produces potentially toxic leachates that must be collected and treated
Water could be pumpedfrom the TSFs and the pit for treatment before discharge. Treatment
of pit water would occur once it is a source rather than a sink.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 94
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
A further reference to the fate of waste rock after closure is found on p. 5-77 of the EPA
Assessment: "Under the mine scenario, the mine pit, waste rock piles, and TSF would remain on
the landscape in perpetuity and thus represents permanent habitat loss." It should be noted that
the scenario states that PAG will not remain on the surface, whatever volume and area of land
surface that represents.
RESPONSE: Section 6.3.3 of the revised assessment clarifies that no PAG waste would
remain on the surface.
The descriptions of exposure and exposure-response resulting from the transportation corridor in
Section 5-4 of the EPA Assessment focus on potential impacts and make use of references that
are clearly not representative of the expected road construction. A number of these references
date from 1975 and 1976 (p. 5-59) and are not necessarily representative of road design and
construction practices in 2012. On p. 5-62, the following statement and reference is given:
"Sediment loading from roads can severely affect si reams below the right-of-way (Furniss et al.,
1991 and references therein)". This reference is specifically focused on forest and rangeland
roads, clearly not representative of a major trail spoliation road between a mine and the port
facilities from where its products are shipped This publication contains main recommendations
specifically for forest and rangeland roads and some of them are indicative that it is not
applicable to the transportation corridor for a major mine access road: "Design cut slopes to be as
steep as practical. Some sloughing and bank failure is usual ly an acceptable trade-off for the
reduced initial excavation required" (p. 306); and "stream crossings can be considered dams that
are designed to fail. The risk of failure is substantial for most crossings, so how they fail is of
critical importance" (p 3 I'M. The reference also refers to the application of oil as a dust
abatement additive on p. 3 12, w liich is hardly acceptable practice. In my review, I did not find
that any of the references used in the EPA Assessment refer specifically to mine roads such as
those considered for the transportation corridor at the Pebble Mine scenario.
RESPONSE: The information cited from the two publications noted by the commenter is still
true today. The first use of Darnell (1976) was incorrect, and has been changed to Furniss et
al. (1991). Although Furniss et al. (1991) focuses on forest and rangeland roads, it is a
seminal publication on the potential effects of roads, particularly as they relate to salmon. The
general conclusions of that paper should be applicable to the transportation corridor described
in the assessment. Furniss et a I. (1991) lists a number of guidelines for road design and
construction that will help minimize adverse effects on salmonid habitats. It does not
specifically advocate the application of oil as a dust abatement additive. It merely states that
whatever chemicals are used, they should be applied so as not to enter streams, and that
subsequent transport of these substances into water courses should be evaluated.
The failure frequencies cited in the revised assessment are from modern roads and not
restricted to forest roads. Because the proposed mining would take place in an undeveloped
area, the literature used in the assessment is necessarily from areas outside of Bristol Bay.
However, we used recent literature from representative environments to the extent possible.
Lastly, information on current design standards that would be used along the proposed
transportation corridor is now included within text boxes throughout Chapter 10.
It is further interesting that it is stated on p. 5-60 that there will be 20 bridges and 14 culverts
along the road without referring to this as an assumption, and no reference is cited for this
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 95
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
information. Will there be a change in impact if the decision is made to build 30 bridges and 4
culverts or 34 bridges and no culverts?
RESPONSE: The estimate of 20 bridges came from Ghaffari et al. (2011). In the revised
assessment, crossings that would be bridged (now 18) are based on mean annual stream flows,
as explained in the text. If a decision was made to build more bridges and fewer culverts, there
would be a change in impact, but scenarios with 30 or more bridges are probably not realistic.
The discussion on the potential impacts of the transportation corridor on salmonids serves the
purpose of highlighting some aspects that engineers and fish biologists must take into account
when designing and maintaining the final transportation corridor lor the Pebble Mine and other
mines in the Bristol Bay area. However, this assessment docs not appropriately describe the scale
and extent of the risks to salmonid fish due to operation of a transportation corridor under the no-
failure mode of operation.
RESPONSE: The "no failure" mode of operation was meant to illuminate the effects that
would occur solely from a mine footprint, even in the absence of accidents or failure. This
term has been eliminated in the revised assessment. The no-failure scenario from the draft
assessment has been changed to a section on the effects of the footprint of a mining operation,
without regardfor operational problems (Chapter ~). The revised assessment places the
streams along the transportation corridor into the context of the entire Nushagak and Kvichak
River watersheds with respect to important watershed attributes such as discharge, channel
gradient, andfloodplain potential. Potential risks to fish habitats and populations associated
with the proposed corridor are then evaluated in some detail.
Phyllis K. Weber Scannell, Ph.D.
The no-failurc model makes a number of assumptions about how the mine will be developed -
some may be accurate, some may be considerably different. It is important to take under
consideration that Pebble is currently a prospect, not a mine. Should this project proceed to mine
development, it will be incumbent on the mining company to develop a rigorous mine plan that
includes detailed information on all aspects of a future project. This mine plan will be reviewed
by state and led era 1 staff with experience in large project development.
RESPONSE: The HP. I agrees with this comment. No changes suggested or required.
The no-failure model discusses the amount of riverine habitat that will be lost to mining by the
mine pit, tailing storage facility, and waste rock dumps. Anadromous fish habitat is protected
under Alaska Statute 16.05.840-870. The statute requires review of a project potentially
affecting fish habitat and, where necessary, avoidance, mitigation, or compensation. A project
must provide free passage of fish; the project cannot be placed in such a way that fish are
prohibited from moving into the upstream reaches. Estimates of habitat loss from the mine
footprint are not possible without a more detailed plan of operations for the mine.
RESPONSE: The scenarios presented are meant to represent those expected as typical for
mining of porphyry copper deposits of this type, and are based on preliminary mine plans from
NDM (Ghaffari et al. 2011). Although layout of mining components at a site may differ
somewhat from what we present in the scenarios, the main components of mining will remain
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 96
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the same for open-pit mining. Given stream density in the area, direct losses of stream and
wetland habitat of a similar magnitude would be inevitable with projects of the specified
magnitudes.
There are many aspects of the development of a large mine project that need thorough review to
ensure that habitats are protected. These include, but are not limited to: classification and
storage of waste rock, lower grade ore, overburden, and high grade ore; development and
maintenance of tailings storage facilities; development and concurrent reclamation of disturbed
areas, including stripped areas and mine pits; collection and treatment of point and non-point
source water; quantity and timing of discharges of treated water, monitoring of ground water,
seepage water and surface water; and biomonitoring. The transportation corridor will require
review and permitting of every stream crossing of fish-hearing waters. In addition, plans should
be developed for truck wheel-washing to minimize transport of contaminated materials.
RESPONSE: The EPA agrees that these aspects would need to be subject to a thorough review
during the development and approval of a detailed mining plan. So changes suggested or
required.
Paul Whitney. Ph.D.
Material Resource Areas. Material resource areas, mentioned 011 page 4-34, for the road and
pipelines should be discussed in more detail W ill aggregate lx- required? If so, where are the
aggregate resources in relation to floodplains ' I spent a summer surveying material resource
areas for a proposed arctic and subarctic pipeline and access road Suitable material resource
areas are sizeable and are often important (e.g., aggregate) for wildlife (such as bears that
hibernate or survive the winter in dens) and fishery resources. Sometimes dens can only be
excavated in non-permafrost (i e . aggregate) soils. It appears the project area is in a zone of
discontinuous permafrost, but permafrost could be more continuous in the higher elevations
along the road through the Kenai Mountains An accurate assessment should determine the
permafrost location(s), as well as the area and importance of material resources for fish and
wildlife. I11 addition, Reclamation Plans for the material resource areas should be briefly
discussed to ensure that areas mined for aggregate will not avulse and capture streams.
RESPONSE: EPA agrees with this commenter that the impacts of material resource areas to
wildlife, fishery and other subsistence resources could be significant and must be addressed in
an environmental impact statement and as part of the 404 permit review. Review of potential
material resource areas was not included in the scope of this assessment.
Water for Dust Control. Dust control for the 86-mile proposed haul road will likely require a lot
of water. Where will this water come from? Withdrawal from streams crossed by the haul road
could have impingement and flow reduction consequences. Adequate screening could solve the
impingement issue. Some back-of-the-envelope calculations could determine if water
withdrawals for dust control could alter the projected hydrographs when salmonids are present in
the streams.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 97
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: We expect water for dust control to be a small amount from any one source.
Permits would be required from the State of Alaska that would address impingement issues.
We do not expect this to be a major issue.
Question 5. Do the failures outlined in the assessment reasonably represent potential
system failures that could occur at a mine of the type and size outlined in the mine
scenario? Is there a significant type of failure that is not described? Are the
probabilities and risks of failures estimated appropriately? Is appropriate information
from existing mines used to identify and estimate types and specific failure risks? If
not, which existing mines might be relevant for estimating potential mining activities in
the Bristol Bay watershed?
David A. Atkins. M.S.
The Assessment focuses on some low probability, hiuh impact failures (e g . TSI failure), and
presents summaries of failures at existing mines. The majority of the focus is on catastrophic
failures, such as TSF, pipeline, water collection and treatment. and road and culvert. Anecdotal
information regarding mine failures is numerous, but often not well documented, so it is difficult
to get information on the details of failures of other projects It is also difficult to extrapolate the
probability of failure from one site to the next, and the report stresses the wide range of
uncertainty, depending 011 design and environment Without a more detailed understanding of the
mine plan and associated engineering, as well as additional detailed analysis, it is difficult to
determine if the failure probability estimates presented in the Assessment are reasonable.
RESPONSE: The authors concur with the commenter that it can be "difficult to get
information on the details of failures of other projects". The statistics for historic tailings dam
failures are derived from the largest available database and include many tens of thousands of
dam-years. The pipeline failure data cover millions of kilometer-years ofpipeline experience.
The data on failures of water collection and treatment systems and of culverts are less
extensive. M e also recognize that even with detailed engineering and design information, the
prediction of failure probabilities is extremely difficult. Finally, since all of these low-
probability failures are statistical phenomena, the actual experience at any one site could be
vastly different than another similar site, even when the failure probabilities have the same
distribution.
The focus on catastrophic failures also takes attention away from what is probably a more likely
scenario. Every project is subject to accidents and smaller, non-catastrophic failures that have
varying degrees of consequence. Sometimes these failures are easily identified and fixed and
other times they can go un-noticed for periods of time.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative wastewater treatment plant failure, truck
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 98
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
accidents and spills, and refined leachate seepage scenarios) and explains why these
particular failure scenarios were chosen.
It would be helpful to describe some smaller-scale failures that have occurred at mine sites. A
partial list includes: accidents and spills along the transportation corridor or within the mine site;
unanticipated seepage of contaminated water that may be difficult to detect, collect and treat;
movement of water along preferential flow pathways that are difficult to characterize; temporary
failure of water collection and treatment systems; mistakes in engineering analysis that
underestimate the volume of water that must be collected and treated or overestimate the volume
of water available for use; and designing based on incomplete data and understanding of climate
conditions.
RESPONSE: There is a wide variety of failures that could occur, including those provided by
the commenter. Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative wastewater treatment plant failure, truck
accidents and spills, and refined seepage scenarios) and explains why these particular failure
scenarios were chosen.
Steve Buckley, M.S., CPG
The engineering failures reasonably represent potential system failures outlined in the mine
scenario based on historic porphyry copper deposits of this type It is less clear if the failures
reasonably represent a mine scenario hased on state of the art engineering and mitigation
practices. Appendix I pro\ ides some information related to potential system failures and possible
mitigation measures designed to minimize these risks but these are not treated in any detail in the
assessment. It would lx- difficult to pull together the most modern engineering and mitigation
practices from around the world hut it could help hound the risks associated with modern mine
development
RESPONSE: Our purpose in the assessment is to evaluate the risks from hazards resulting
from a mine operated with appropriate mitigation measures for design, operation, monitoring
and maintenance, and closure. Accidents and failures happen regardless of mitigation
measures; thus, effects of several failures are evaluated Mitigation measures related to our
mine scenarios are now clearly discussed in Chapter 6.
The revisions clarify when and w here potential mitigation measures would be treated in the
regulatory process.
The Red Dog mine in northwest Alaska might be relevant for estimating potential mining
activities in the watershed. Although the characteristics of the deposit differ significantly, at
roughly 150 million tons it is half the size of a reasonable minimum mine scenario and would be
helpful to characterize some minimum mine development scenario.
RESPONSE: A third mine size scenario, representing the worldwide median size porphyry
copper mine (Singer et al. 2008), is included in the revised assessment.
This improves the assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 99
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Courtney Car others, Ph.D.
The potential failures outlined in this assessment include: tailings dam failures, pipeline failures,
water collection and treatment failures, and road and culvert failures. These failures appear to
represent the key potential failures for this mining scenario, their risks appear to be estimated
reasonably, and statistics from existing mines appear to be used appropriately, although I have no
particular expertise with which to evaluate this assessment. As we discussed in our peer review
panel, the focus here is on catastrophic failure. More detail should be provided on likely non-
catastrophic failures, ones that would be more difficult to detect
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
leachate seepage scenarios) and explains why these particular failure scenarios were chosen.
We did include leachate seepage scenarios (C haptev 8), which would be more difficult to detect
than catastrophic failures.
Dennis D. Dauble, Ph.D.
My experience in system failure of mines of the size and type outlined in the scenario is limited.
However, what does seem to be missing is the long-term effects of leachates to receiving water
bodies in any type of risk scenario, including both 11011-lailiirc and lailure modes. That is,
assuming no catastrophic failure, how might tailings constituents interact with aquatic habitats
seasonally, such as during periods of snowmelt and severe rainfall events?
RESPOXSE: The original draft assessment contained a scenario in which tailings leachate
was not fully contained and reached a stream (Section 6.3 in the May 2012 draft). The revised
assessment includes estimates of leachate escaping from the TSFs and from the waste rock
piles, bypassing the collection systems, and entering the streams (Chapter 8). The estimated
loadings of copper and other elements from these leachate flows are included in stream
concentration estimates. The assessment discusses the impacts of these concentrations on the
aquatic habitat and biota. The commenter is correct that we did not include a scenario in
which the dam does not fail, but snowmelt and severe rainfall would result in overtopping and
release of untreated water. That is very plausible, but there are just too many possible failure
scenarios to include more than a few of them.
Gordon H. Reeves, Ph.D.
No comments on this question
Charles Wesley Slaughter, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 100
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Potential failures seem reasonable, based on history of other mining operations. However, the
consequences of hydrologic extremes during winter (frozen soil) conditions are not adequately
addressed. The possibility of the mining operation and the transportation network encountering
discontinuous permafrost is not mentioned, although at least some soils maps indicate permafrost
presence.
RESPONSE: Warhaftig (1965) reports that permafrost is sporadic or absent in the Nushagak-
Bristol Bay Lowland (Table 3-1). If permafrost were detected during project development or
construction, the designs would need to address any potential impacts on the infrastructure
and potential impacts of the infrastructure on the permafrost. Frozen soil could improve
vehicular access to parts of the project site and minimize the disturbance from such access.
The occurrence of an extreme hydrologic event, such as heavy rainfall on frozen soil or heavy
rain on an existing snowpack, could produce unusually heavy runoff and higher than normal
stream flows. We analyzed the impact of a tailings dam failure during the Probable Maximum
Precipitation (PMP) event, thereby generating the Probable Maximum Flood (PMF). The
increased flow due to precipitation was, in most of the scenarios, small compared to the flow
released from the TSF. Any increases in the peak runoff due to frozen soil or melting
snowpack would also be small relative to the TSF release, so the modeled scenario can be
considered a reasonable bounding estimate.
Designs for the components of the mine infrastructure would need to consider the natural cold
region conditions and incorporate appropriate design features and safety factors to achieve an
acceptable level of performance, (ihaffari et al. (2011) says "The Pebble deposit is located
under rolling, permafrost-free terrain in the Iliamna region of southwest Alaska..." and "The
deposit is situated approximately 1,000ft amsl, in an area characterized by tundra, gently
rolling hills and the absence of permafrost." (ihaffari et al. (2011) describes the transportation
corridor thusly: "The road route traverses terrain generally amenable to road development.
... There are no significant occurrences of permafrost or areas of extensive wetlands."
Nevertheless, if sporadic areas of permafrost are discovered, the designs will need to address
the interactions between the infrastructure and the permafrost.
The probability approach outlined for potential TSF dam failure is unpersuasive. It is difficult to
relate to a number like "0.00050 failures per dam year," or to the implication on p. 4-47 that one
can expect a tailings dam failure only once in 10,000 to one million "dam years." This could
suggest to the casual reader thai failure of the hypothesized TSF1 dam (for which one "dam
year" is one year) should nol be anticipated in either the time of human occupation of North
America, or the span of human e\olution.
RESPONSE: The comnienter is correct. The proposed dams, if designed, built, and
maintained to current engineering best practices, would be anticipated to have a low annual
probability offailure. However, the failure probability would not be zero. The writers concur
with the commenter that these low probability numbers may be difficult for the casual reader
to grasp, so we now also present estimates of probability in terms of probability failure over
different time periods. The discussion of this issue has been expanded to clarify that the
failure rate is a design goal and is not based on empirical evidence.
Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-magnitude event at the
Pebble locale has an estimated return period of 200 years. Such a return interval probability is
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 101
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
difficult to interpret, given the lack of historical seismic records for the region; in any event, such
a return period estimate is in no way predictive of future seismic activity, in year 2012 or year
2212. (The suggested 200-year return period should also be viewed in light of the 79-year
suggested operating life of the hypothetical Pebble operation, probable longer-time operations at
other mineral extraction sites which would be developed following implementation of Pebble
and building from the infrastructure associated with Pebble, and also the projected very long
persistence of the TSFs following cessation of active mining).
RESPONSE: Box 9-2 in the revised assessment (Box 4-6 in the May 2012 draft of the
assessment) provided the OBE and return period determined by NDM in the Preliminary
Assessment. A detailed engineering design and safety evaluation is outside the scope of this
assessment. The discussion of seismicity (Section 3.6) addresses the uncertainty in interpreting
and predicting earthquake magnitude and recurrence in the Pebble area.
Box 4-6 does note that "The return periods stated in Alaska dam safely guidance are inconsistent
with the expected conditions for a large porphyry copper mine de\eloped in the Bristol Bay
watersheds, and represent a minimal margin of safely
RESPONSE: The return periods used are consistent with the Alaska Dam Safety Guidance,
however the operator could include additional margin of safety in the design for critical
structures. The return period and seismic safety factors do not inform the failure analysis in
this assessment, but would be important considerations during the review process for any
future mine plan.
JohnD. Stednick, Ph.D.
The assessment reasonably addresses potential large system failures, but should include a variety
of smaller and perhaps more frequent failures (see Question 4). A large tailings storage facility
failure compared lo a blocked road drainage culvert. The level of detail in the assessment of the
potential system failures \ aries considerably and baits the question—why? Does this demonstrate
lack of understanding (if failure prediction, lack of failure prediction, or writing team expertise?
RESPONSE: The EPA does not believe that a blocked culvert requires or deserves the same
level of analysis as a tailings dam failure and spill. The latter is a much more complex
phenomenon with multiple consequences that require evidence and analysis such as the
potential toxicity of the spilled tailings. In contrast, blockage of a culvert is a relatively
straightforward phenomenon. Further, a TSF failure poses a much greater concern for
stakeholders and local communities due to the large magnitude of potential effects.
Tailings storage facility: The liquefaction phenomenon, internal and external erosion, seepage,
and overtopping are some of the main failure modes of tailings storage facilities. A large quantity
of stored water is the primary factor contributing to most tailings storage failures. The risk of
physical instability for a conventional tailings facility can be reduced by having good drainage
and little (if any) ponded water. Some suggest that the tailing pond freeboard should be able to
accommodate the 100-year, 72-hour storm/streamflow event. What are the State of Alaska
standards? Discuss the probability of failure of a TSF from other than overtopping by a
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 102
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
precipitation/streamflow event. The potential of seismic activity and its effect on tailings storage
and other earthworks needs to be addressed.
RESPONSE: The historic failure probabilities for tailings dams presented in the assessment
include all modes of failure, including slope instability, liquefaction, overtopping, erosion, and
seismic activity. Historical failures were discussed as supporting background information and
present a defensible upper bound on the failure probabilities. The failure probabilities used in
the assessment are based on Alaska's dam classification and required safety factors applied to
the method of Silva et al. (2008). The data presented by Silva et al. (2008) consider only the
annual probability offailure from slope instability, but the methodology is equally applicable
to other failure modes. The discussion offailure probabilities in the revision (Chapter 9) is
expanded to clarify this issue.
Chemical transport spill: Mine development and ore processing will require significant loads of
petroleum and chemical products. Although the exact processing formulations are not given,
most copper porphyry mines use similar formulation in ore flotation and processing. How will
chemicals be stored, transported, and recycled ' What are the opportunities for accidents to
occur?
RESPONSE: Storage of chemicals is addressed in the assessment. Petroleum transport is by
pipeline and an assessment of the risk of spills has been added in Chapter 11 of the revised
assessment. Process chemicals would be transported by truck. Discussion of truck transport
and potential for accidents has been included in Chapter 10. including a quantitative analysis
of the risk of wrecks and of wrecks that cause spills into streams or wetlands.
Roy A. Stein. Ph. I).
Failures Appropriate for Mines of this Size and Type. Given my background, I can't answer
with any authority, though the comparisons seemed appropriate, though clearly no extant mines
are as large as the one proposed herein. Some of the public testimony spoke directly to
comparisons with existing mines in dry areas would be completely inappropriate because it is the
hydrology of the Bristol Bay watershed that would make it so very vulnerable to mining impacts.
RESPONSE: No change suggested or required.
Failures Not Described? I speak to failures associated with routine operations previously in
this review, as well as chemical spills along the transportation corridor. Also included herein
should be impacts of the Cook Inlet Port and potential spills, accidents, etc., on the marine
ecosystem.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
seepage scenarios) and explains why these particular failure scenarios were chosen.
Discussion of transport and potential for accidents has been included in Chapter 10, but risks
are not evaluated. Potential impacts in Cook Inlet are important and should be considered in
the regulatory process should a company submit a permit. However, because this is an
assessment of the Bristol Bay watershed, evaluation of impacts from the port (and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 103
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
transportation corridor outside the Nushagak and Kvichak River watersheds) is outside the
scope of this assessment.
Probabilities and Risks of Failures. These seemed reasonably well documented, though again
this falls outside of my expertise. Even though out of my realm, I still would have liked a more
quantitative assessment of these risks, developed in a rigorous, defensible way. I am
discouraged when I understand that history (in the eyes of the mining company) is not a good
predictor of the future because technology has taken us so much farther along, reducing risks of
whatever failure significantly. In my view, this is a specious argument and one that should be
roundly put to bed by the authors of this report. History is indeed Ihe absolute best predictor of
the future and technological changes that have occurred since past mines must be absolutely and
critically evaluated to determine if indeed risks do go dou n. This is a serious issue and one that
should be addressed with some rigor by the authors.
RESPONSE: Historical failures were discussed as supporting background information. The
failure probabilities used in the assessment are based on Alaska's dam classification and
required safety factors applied to the method nfSilva et al. (2008). The EP. I has strengthened
the discussion of failure probabilities (Chapter ')) to the extent that available evidence allows.
Existing Mines as Comparisons. Given my background, I can't answer with any authority,
though the comparisons seemed appropriate, though clearly 110 extant mines are as large as the
one proposed herein (nor does it incorporate all we have learned in the mining business since the
last mine for which we have data). Even so, 1 again caution the authors that these existing,
historical mines provide the best data we have to estimate risks of failure and that, under no
circumstances, should one accept the indefensible argument that progress with mine safety
(speaking broadly, not just human health) has progressed to the point that these risks, previously
quantified in these older mines, are now small. As suggested above, we have no other
quantitative values for risk, except for existing mines and we cannot simply erroneously lower
the risk based 011 new, untested technologies
RESPOSSE: II e agree that it is appropriate to include in our scenarios technologies that are
in current use and not those that are untested, and thus have done so. No change suggested or
required.
Proportional Losses of Salmon, [s it possible to estimate the proportion of the salmon runs
compromised in the lace of major failures in tailing storage facilities or other failures? In other
words, I would recommend adding a chapter that uses best estimates of salmon produced within
the Bristol Bay watershed and then assess the maximum impact of, let's say, a Tailing Storage
Facility failure—with this failure, might we lose 10%, 40%, or 75% of our salmon productivity?
In addition to this estimate, one might estimate the number of stocks or unique genetic units lost
with a major failure? I know these numbers are difficult to get, but if one begins with
escapement from these systems as well as insights from harvest, we may be able to bound these
impacts. Only in this fashion can we put these data into context. This exercise also will serve to
counter the argument by the mining company that they are only destroying some small
percentage of salmon habitat and hence (assuming a linear relationship between habitat lost and
salmon eliminated) only some very small percentage of salmon. Because losing 2% of critical
headwaters habitat may translate to huge losses of salmon (say 20%), one simply cannot assume
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 104
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
a linear relationship between habitat and salmon. Explicitly making this argument improves the
rigor of the main report.
RESPONSE: The fact that salmon productivity cannot be assumed to be linearly related to
habitat has been more explicitly elaborated in the revised assessment. Due to lack of
comprehensive estimates of limiting factors across the impacted watersheds, population level
effects could not be quantitatively estimated, except in the most severe cases where total losses
of runs could be reasonably assumed Our ability to estimate population level effects was
limited to situations that were assumed to completely eliminate habitat productivity and
capacity in an entire watershedfor which estimates of escapement could be inferred. For this
assessment, these conditions are only met in the TSE failure scenario that completely
eliminates or blocks access to suitable habitat in the Sorth Eorli Koktuli River. In that case,
we estimate that the entire Koktuli portion of the run of Sushagak escapement) could
be lost. Higher proportional losses would occur if significant downstream effects occurred due
to transport of toxic tailings fines beyond the Koktuli as modeled under the Pebble 2.0 TSF
failure.
Correlations between Ocean and Terrestrial Conditions From the literature (see Irwin and
Fukuwaka. 2011. ICES J. Mar. Sci. 68: 1122 as just oik- example), we know what climatic
conditions lead to poor rearing conditions in Ihe ocean, thus com promising growth and
ultimately survival of salmon. Gi\ en these ocean conditions, might we have correlative effects
in the terrestrial environment, thus leading to a cumulative effect of ocean and terrestrial
impacts? For example, if particular oceanic conditions underlie poor survival, might this
correlate with an increased probability of flooding, leading to a higher probability of a Tailings
Storage Facility failure, leading to a cumulative synergistic effect that could be multiplicative in
its negative impact on salmon populations? W as there any attempt to correlate these impacts?
Just how realistic is it to reflect on these sorts of multiplicative effects? Might these effects have
a catastrophic effect on salmon populations '
RESPO NSE: Pursuing this hypothesis would require a research effort that is beyond the
scope of this assessment.
William A. Stuhhlefield, Ph.D.
The scope of failures described in the assessment seems to be sufficiently comprehensive and all
likely failure-types are considered The probabilities and risks for failure seem to be adequately
estimated, given the state-of-tlie-science; however, these estimates are likely to be very sensitive
to site-specific concerns and operational considerations. Once site-specific information is
available, it is likely that much better estimates of failure potential at a site can be developed.
RESPONSE: We agree that more site-specific information could support more site-specific
estimates. No change required.
Dirk van Zvl, Ph.D., P.E.
Failure modes outlined in the EPA Assessment do not reasonably represent the potential failures
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 105
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
that could occur at a mine of the type and size outlined in the mine scenario.
The EPA Assessment considers a series of large, or catastrophic, type failures. The failure modes
considered in the mine scenario include four major items: tailings management facility;
pipelines; water collection and treatment; and, road and culvert. These failure modes are
included in the risk assessments for a mine of this type and size, either qualitatively as is
typically done in FMEA, or quantitatively. However, the range of failure modes will also
consider many other types of failures that can also occur during regular mine operations. These
other types of failures may result in the full spectrum of risks (from insignificant to very high)
depending on site and mine life-cycle conditions. Performing extensive analyses for the four
major items implies that they could occur and that they are the only failure modes that will be
significant.
RESPONSE: It is true that any of the failures analyzed could occur, as based on historic and
current knowledge of mining; however, a focus on these failures does not imply that they are
the only failures that will be significant. Because the number of potential failures is extremely
large, it is necessary to choose a representative set of failure scenarios. The revised assessment
includes more failure scenarios (e.g., diesel pipeline failure, quantitative water treatment
failure, and refined seepage scenarios) and explains why these particular failure scenarios
were chosen.
Responses to the other parts of the current question (Are the probabilities and risks of failures
estimated appropriately? Is appropriate information from existing mines used to identify and
estimate types and specific failure risks ' II" not. u hich existing mines might be relevant for
estimating potential mining activities in the Bristol liny watershed ') are further discussed in my
responses to other questions below.
RESPONSE: No change suggested or required.
Phyllis A". Weber Scanne//, Ph.I).
This section focuses on catastrophic failures; however, there are a number of non-catastrophic
failures that can occur at a mine site. Non-catastrophic failures include leakage of contaminated
water to ground or surface waters from PAG waste rock, the tailing storage facility, and exposed
ore surfaces, and from emergency discharge of untreated water from the TSF and ore spills from
trucking accidents Such failures can be minimized or prevented with good site planning and
monitoring. An additional "failure" has been experienced at a mine in Alaska when the water
elevation of the tailing pond was sufficiently high to cause groundwater flow across a natural
divide into an opposite drainage.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The review draft contained a scenario in
which tailings leachate was not fully contained and reached a stream (Section 6.3 in the May
2012 draft). The revised assessment document includes more failure scenarios (e.g., diesel
pipeline failure, truck accidents, quantitative water treatment failure, and refined seepage
scenarios) in Chapters 8,10 and 11 and explains why these particular failure scenarios were
chosen.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 106
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Paul Whitney. Ph. IX
Sediment Transport. The failure analysis indicates a sediment transport study was beyond the
scope of the assessment. Not only is such a study important for fish resources, it is important for
all ecological resources, especially plant community succession along the stream and a delta into
Bristol Bay.
RESPONSE: EPA agrees that a sediment transport study is important to fully characterize the
watershed and potential impacts from large-scale mining. The scenarios indicate that runoff
eroding sediment from the site would be directed to sediment retention basins. Sediment
runoff from the road is considered in Chapter 10. The development and implementation of a
model of sediment (i. e., tailings) transport after the initial tailings outflow would require a
research and development effort that is beyond the scope of this assessment.
This might seem extreme, but the failure analysis indicates the Koklnli. Mulchatna, and
Nushagak Rivers would stabilize into a new channel after a failure and nol continue to work their
way across the floodplain and eventually transport materials hundreds of miles down river. The
Mount St. Helens eruption, given as an analogy in the assessment (page 6-3), certainly moved
sediment into the Columbia River Channel, the Columbia Ri\ er Kstuary, and Pacific Ocean over
a hundred miles away. Copper concentrations in the Columbia River estuary as a result of the
eruption ranged from 1 to 43 |ig/L (Lee 1996). The upper limit of the range is approximately 20
times greater than the no-effect benchmark listed in the assessment The down stream
consequences of changes in sediment transport and water sediment chemistry for fish and
wildlife are sometimes \ei y large, not anticipated (see Peace Athabasca Delta response below),
and costly to remediate Solutions for the Peace Athabasca Delta involving check dam
construction may not be directly applicable to a tailings dam failure analysis in the assessment
but may have some \ alue short of dredging I'm not sure how it would work, but a mitigation
effort (page 4-32. para 3. last sentence) using bulk tailings would apparently be placed down
gradient to catch tailings in the event of a failure. Is this a safety check dam? It is possible that
the assessment could be impim ed if this and other redundant efforts to minimize risk could be
discussed in more detail and considered in the failure analysis.
RESPONSE: II e have clarified this in Chapter 9 of the revised assessment. Our intent was to
describe a channel becoming reestablished in the deposited tailings and slowly reworking the
entire valley bottom, moving the fine material downstream. We agree that tailings would be
carried downstream to Bristol Hay over the long term with adverse impact to waters. The
mitigation addressed in the comment, where a levee was created downstream to capture the
flow of tailings, was not our intent. We described a reclamation scenario where the tailings are
dewatered and sloped to eliminate the chances of mass failure of the impoundment. Mitigating
impoundment failure by creating a second dam downstream would greatly increase the mine
footprint on salmon habitat.
I agree with the assessment's statement on page 6-11 (6 lines down) that impacts of a tailings
dam failure to fish would extend down the mainstem Koktuli River and possibly further. If the
Mount St. Helens analogy is a good one, the impacts could reach Bristol Bay. Even if the Mount
St Helens analogy is not a good one, I suspect sediments would continue to move down river as
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 107
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the river(s) moved across their floodplains through time. I also agree with the assessment's
statement that the time to reach dilution approaching background would be very long, as the
sediments in tributary rivers to Bristol Bay will be continually reworked (page 6-25) and
resuspended.
RESPONSE: EPA agrees with the commenter. No change suggested or required.
As a terrestrial ecologist, I have always been impressed by the impact of Bennett Dam on
sediment transport in the Peace Athabasca Delta ecosystem over 600 miles away (Cordes 1975).
The assessment does mention the potential impact of toxics on the likelihood of plant community
succession on deposited tailings (page 6-10), but this causal pathway was not assessed. Given the
increases in metal concentrations in the Columbia River Estuary cited by Lee (1996) for the
Mount St. Helens eruption, an assessment of down stream sedimentation and changes in
sediment chemistry should be addressed. The likelihood that far reaching impacts of a failure
could influence plant succession and wildlife habitat quality is probably not anticipated but given
the Mount St. Helens analogy and the lessons learned from Bennett Dam. the likelihood of such
impacts deserve more attention.
RESPONSE: This comment recommends expanding the scope of the assessment to include
direct effects on terrestrial plants and midlife habitat. The scope was set during the planning
and problem formulation processes to encompass the expressed concerns of the Alaska Native
organizations that asked the Agency to address potential mining in the Bristol Bay watershed,
as well as the needs of the Agency's decision makers. This scoping process is in keeping with
the Guidelines for Ecological Risk Assessment (I SEP. I 1998). The commenter may be more
familiar with Environmental Impact Statements, which are full disclosure documents and
therefore are more broadly focused.
Sediment Benchmarks. Once again, as a terrestrial eeolouist with ecological risk assessment
experience. T know that sediment chemistry and determining toxic benchmarks for sediments is
very complex and subject to \ aried opinions. 1 admire the MacDonald et al. (2000) effort to
reach consensus on sediment benchmarks but 1 have three concerns. First, the consensus values
listed in MacDonald et al. (2<)<)<)) are geometric means of values from several sources. The mean
consensus \ allies likely do not equate to No Observed Effect Concentrations, which would
probably be lower than the mean \ alues. Second, the lack of observed effects was sometimes for
a "majority'" of sediment dwelling organisms, but not all (MacDonald et al. 2000, Table 1).
Third, some of the sources used lor the mean values included interstitial water, but apparently
not all (Table 1). I will always remember Dr. John Stein (currently the acting director of
NOAA's Northwest Fisheries Science Center in Seattle) standing up at a workshop for the
Columbia River Channel Deepening Project. He had a small bottle of sediment and water in his
hand and, while shaking it, he said something to this effect: It's the pore water we are interested
in. Considering that a proposed mine, at some point, will be reviewed by NOAA, it seems
appropriate to consult with NOAA regarding benchmarks for all the species of sediment/pore
water-dwelling organisms likely to occur in the potentially effected watersheds addressed in the
assessment.
RESPONSE: The EPA agrees that pore water concentrations are generally useful predictors
of toxic effects of sediment, and sediment pore water toxicity is addressed in the assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 108
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
However, the MacDonald et al. (2000) benchmarks are also a useful line of evidence. NOAA
has been consulted during the development of this assessment. No change required.
Question 6. Does the assessment appropriately characterize risks to salmonid fish due
to a potential failure of water and leachate collection and treatment from the mine site?
If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
David A. Atkins. M.S.
Water treatment failures of varying scale occur ill \ irtually e\ cry site thai treats water, and mine
sites are no exception. The risk of failure of water treatment described in the assessment is useful
as background, but as the report states, the risk is highly uncertain. A non-catastrophic water
treatment system failure is fairly likely to occur al some point during the mine life, and, hence,
requires a detailed assessment. The treatment in the Assessment is cursory (less than one page).
This type of failure is much more likely than a TSI fail Lire (w liich receives more than 20 pages
of analysis), and therefore requires a much more thorough treatment given the probability of
occurrence and likelihood of impact to salmon species
RESPONSE: The wastewater treatment failure scenario has been expanded and is now
detailed and quantified in Chapter
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
failure could go unnoticed for some time or the resources may not be available to correct it
quickly, depending on how long after closure the failure occurs and the stewardship of the
treatment system.
RESPONSE: A failure during operation would be corrected quickly only if it did not require
extensive repairs or the manufacture and import of replacement equipment. The EPA agrees
that water collection and treatment failures after mine closure would be less likely to be
corrected in a timely manner. In addition, events after closure, such as filling of the pit, would
affect water quantities and qualities in ways that would affect treatment success. A discussion
of this issue has been added to Chapter 8 of the revised assessment. A discussion of financial
assurance has been added in Box 4-3.
Steve Buckley, M.S., CPG
Less than a page (4-39) is devoted to the failure of water and leach tile collection and treatment.
This seems inadequate given it would be one of I lie main systems thai could impact fish at the
potential mine site. In contrast, 20 pages are de\ oled lo tailings dam failure (p 4-39 to 4-60).
RESPONSE: The wastewater treatment failure scenario lias been expanded and is now
detailed and quantified in Chapter S of the revised assessment.
This reorganization is helpful along with llie additional information on wastewater treatment
failure scenarios.
Courtney Carothers, Ph. I).
The report concludes that wastewater and leachate treatment and collection failures could expose
local streams to mildly to highly toxic water harmful lo invertebrates and fish species.
Depending on the type of failures, these exposures could last from a period of hours to years.
The report notes that in the case of Red Dog Mine, Alaska, the water treatment system was
inadequately designed, but does not discuss why such a design was approved and allowed to be
implemented, nor does it discuss the likelihood of replicating such a design flaw in future mining
scenarios.
RESPONSE: The Red Dog Mine's treatment system was inadequately designed because the
amount of water to be treated was underestimated. Mine site hydrology is often difficult to
predict and treatment systems may fail despite the intent of mining companies and regulators.
Dennis D. Dauble, Ph.D.
More information on local hydrology, including seasonal runoff patterns (e.g., peak flows) and
groundwater movement would be useful. I found no description of existing water quality
characteristics of potential receiving waters, except what is included in Table 5-17 of the main
report. Are these values (such as hardness, which moderates metal toxicity) consistent
throughout the watersheds, including downstream lakes? Other questions include: What
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 110
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
volumes of leachates might be collected and treated versus volumes not captured and
subsequently released downstream? Is copper the only constituent of concern to aquatic animals?
Are there processing chemicals that would also be toxic?
RESPONSE: Monthly flow patterns for area streams are now presented in Chapter 7. The
three streams described in Table 7-17 are the three potential receiving waters for any site
effluents. The water balance, including leachate volumes, is now described in the assessment
(Chapter 6). Copper is the primary contaminant of concern. Others are described in the new
Section 6.4.2.3 and discussed in Chapters 8 and 11.
The assessment should also consider and discuss relative risk to aquatic ecosystems from
downstream transport of sediment-bound metals to Iliamna I .ake, if deemed probable.
RESPONSE: Although metals in aqueous emissions would partition to sediment, and the
sediment would mobilize during high flows and eventually reach the lake, this route is not
judged to be significant. Toxicity is caused by dissolved metals, and concentrations from
release of metals from transported sediment to lake water are likely to be minor.
Gordon H. Reeves. Ph.D.
The report focused primarily on the lethal effects of the contamination from leachate and water
treatment and collection failures. 1 lo\\e\ er. could there be ecological consequences to fish and
invertebrates that are not directly lethal hut that could ha\e ecological consequences over the
long term? I suggest that this needs to be considered more fully in this assessment.
RESPONSE: The report addresses both acute lethal effects and chronic lethal and nonlethal
toxic effects to both fish and the invertebrates on which they feed. It is highlighted in the text
when estimated concentrations are sufficient to cause acute lethality. Such effects are
important not only because they are severe but also because they could occur during episodic
exposures.
Sockeye salmon are most abundant salmon in Bristol Bay and a primary species of focus in this
analysis. The direct impacts of mine and mine-related activities have been considered but there
appears to be a lack of consideration of the impact on zooplankton, the food source for sockeye.
If this were a deliberate omission, then a statement about why it was omitted is required. The
revision should include this if it was an oversight.
RESPONSE: Any effluents would be released to the streams draining the site. In those
ecosystems, zooplankton are rare and the primary food organisms for fish are benthic
invertebrates. However, the toxicological data used to assess effects on invertebrates are
dominated by planktonic crustaceans. Therefore, if toxic concentrations of metals reach
Iliamna or other lakes, the toxicity assessment would actually be more relevant to the plankton
that occur there than to the insects in the receiving streams. Hence, the numerous references
to effects on aquatic invertebrates are relevant to zooplankton and no new analyses are
needed
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 111
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Charles Wesley Slaughter, Ph.D.
No. Text suggests that a monitoring well field downslope from the TSF (and presumably from
all hypothetical TSFs) would detect seepage; such seepage would then be intercepted and either
returned to the TSF or "treated and released to the stream channel." Either action presupposes
adequacy of monitoring seepage and subsurface flow (both spatially and temporally); returning
such water to the stream further presupposes fully adequate treatment to meet both regulatory
and aquatic biota requirements for water quality and flow regime.
RESPONSE: The water treatment and leachate capture discussions have been expanded and
are now detailed and quantified in Chapter 8 of the revised assessment.
Assumptions are very generalized and optimistic: "assuming 110 water collection and treatment
failures" and "excess captured water would be treated . and discharged to nearby streams..." -
this assumes both "no failures" over the life of the operation, and that such treated "excess
captured water" could be successfully treated before release to fully meet both regulatory water
quality criteria and the possibly more sensith e biological requirements of individual
invertebrates and fish stocks (Appendices A & li)
RESPONSE: Water management (mitigation) measures are more clearly described and
discussed in Section 6.1.2.5 of the revised assessment, and in sub-sections for the mine
components in the scenarios. Our intention for the "no-failure'' scenario was to identify and
evaluate the unavoidable environmental effects if all systems and mitigation measures
operated perfectly, and to separate those effects from a scenario where systems periodically
failed. However, the "no failure" scenario is no longer included. The purpose of the
assessment is to describe the potential adverse environmental effects that could exist even with
appropriate and effective site mitigation measures. The assessment is not intended to duplicate
or replace a regulatory process, which is where required permit discharge limits for water
quality would be determined.
John I). Stednicli, Ph.D.
The TSF is designed to hold the tailings under water to minimize the oxidation of pyritic
materials and limit ARl) or AMI) production. The TSF will be underlain by hypalon to capture
leakage waters. There is the possibility of failure to collect waters from the TSF—either surface
runoff or leakage with or without storm (precipitation) events. There is also the possibility of
failure of the treatment plant to treat the wastewater. Such treatment systems in Colorado usually
have a bypass pond to temporarily hold waters for later pump back and treatment as a result of
power failure, plant going off-line, storm events, or plant maintenance.
RESPONSE: The original draft assessment contained a scenario in which tailings leachate
was not fully contained and reached a stream (Section 6.3 in the May 2012 draft); however,
this has been refined with new data in the revision for seepage from the TSF and waste rock
piles that escapes capture from the mitigation measures (Chapter 8). Additionally, the scenario
presents a suggested treatment option for mining influenced water and settling ponds for
stormwater runoff. A wastewater treatment plant failure scenario in the revised assessment
assumes emergency storage capacity has been exceeded or the bypass system fails.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 112
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The waters in the study area have very low buffering capacity; metal toxicity would occur at low
concentrations and dilution of metals would require time and space. The maximum index counts
on page 6-39 are confusing and not well related to the risk characterization. Copper was used as
an example metal, but other metals are also toxic and further characterization of the waste rock
can be presented. Further analysis of a water and leachate collection failure can be made over
time: the effects of dilution flows over the various months with low flows, or when adult salmon
are present in the stream as opposed to juveniles - or, when juveniles are emerging. The toxicity
quantification is difficult and appears more of an academic exercise here, rather than site
specific.
RESPONSE: Undiluted leachate concentrations are used to calculate the hazard quotients (we
assume that is what the reviewer means by "maximum index counts"), to screen the
constituents for contaminants of concern, and because the State of Alaska does not allow
mixing zones in anadromous streams. However, dilution was considered in the failure
scenarios for waste rock leachates and tailings leachates. Copper was emphasized because it is
by far the most toxic metal relative to concentrations in rock, tailings and concentrate
leachates. Site specific water chemistry was used to estimate the toxicological benchmarks for
copper and for metals with hardness-dependent toxicity. More detailed analyses of dilution are
included in Chapter 8 of the revised assessment.
Leachate collection from the tailings area is only briefly described. What are the State of Alaska
standards for collection and treatment' What are the potential effects of not collecting or treating
the tailings leachate waters0 Compare the detail and length of leachate discussion to the TSF
failure discussion (see earlier comments).
RESPONSE: There are no specific AK State standards for collection and treatment of tailings
leachate, but any discharge to waters of the United States requires a Clean Water Act Section
402 (CMI $ 402) permit. In . Maska. the Department of Environmental Conservation issues
these permits under the Alaska Pollutant Discharge Elimination System program Such
permits would contain effluent limitations that are protective of the State WQS. The original
draft assessment contained a scenario in which tailings leachate was not fully contained and
reached a stream (Section 6.3 of the May 2012 draft); however, this has been refined with new
data in the revision (Chapter 8) for seepage from the TSF and waste rock piles that escapes
capture from the mitigation measures.
Given the hydrologie connection between surface and groundwaters, what effect will
interception of all waters on the TSF do to the surrounding wetlands and groundwater levels?
Again the lack of a water balance does not let the reader determine if this water interception is
significant or will have significant resource effects.
RESPONSE: The water balance presented in the revised assessment quantifies the amount of
water captured at each TSF under each scenario. The assessment also presents the percentage
change in streamflow at existing gages in the North Fork Koktuli, the South Fork Koktuli,
and the Upper Talarik.
The water balance presented in the assessment looks at the overall amount of water dedicated
to consumptive uses, which therefore would no longer be available to contribute to
groundwater recharge or stream flow. Furthermore, the assessment attempts to quantify the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 113
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
annual amount of and percentage increase or decrease in stream/low for individual stream
reaches. These stream flow results are presented in Chapter 7 of the revised assessment.
Our analysis does not attempt to explicitly calculate groundwater levels or drawdown except
within the cone of depression around the mine pit. Our analysis assumes that all precipitation
that falls outside the actual mine footprint (i.e., the physically disturbed areas) would continue
to contribute to groundwater recharge and stream flow, although the exact flow patterns
would change due to collection of surface runoff and stream diversions. Areas within the
footprint would most likely see some decrease in the groundwater levels due to the collection,
management, and possible treatment of surface runoff, with some potential increases
downstream of the points of water release from the wastewater treatment plant. Overall, the
cumulative reduction in streamflow would be approximately equal to the amount of water
retained on the site as tailings pore water.
Most, if not all of these failures are the result of human error \\ hal safeguards will be in place?
What are the best mining practices to minimize human error?
RESPONSE: The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine (they are a subset of options
presented in Appendix I), all of which were presented as appropriate for the Pebble deposit in
Ghaffari et al. (2011). We assume that these types of measures would be applied throughout a
mine as it is constructed, operated, dosed, and maintained through post-closure. Because the
possibility for human error is inherent in any activity in which a human is involved, accidents
and failures happen. Human error-caused accidents and failures are minimized when the
humans involvedfollow careful and responsible management of the mining site. No change
suggested or required.
Roy A. Stein, Ph.D.
GroundwsHcr C'onncclivily mid TSI-" ( onslruolion I. Extensive connectivity between surface
water and groundwater means thai any failure will allow contaminated waters to flow quickly to
areas of importance for salmon And. indeed how does one build a tailings pond (coarse textured
glacial drift in llie Pebble Mine area) with this much permeability? Why would one only line the
tailings dam; shouldn't the entire Tailings Storage Facility be lined? Would not this be "Best
Practice"?
RESPONSE: Our estimates of the expected leakage from the full TSFs range from about 2 to
6 m /min. If a mine at the Pebble deposit goes forward, the design of the TSFs should include
a more thorough flow analysis that would calculate the expected rate of flow and associated
flow paths from the TSFs. If the calculated leakage rates were unsatisfactory from an
environmental, operational, or economic perspective, the designer could incorporate other
design elements (e.g., a liner) to reduce the expected leakage rate. Full liners beneath TSFs
are not always used; however, there is a growing requirement to use liners to minimize risks of
groundwater contamination, with new mines in Australia being required to justify why one
wouldn't be required (Commonwealth of Australia 2007). Liners are not required in the US.
Whether something is "best" depends on the specifics of the site. The mitigation measures
proposed within the mine scenarios are those that could reasonably be expected to be proposed
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 114
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
for a real mine (they are a subset of options presented in Appendix I), all of which were
presented as appropriate for the Pebble deposit in Ghaffari et al. (2011). Evaluation of
different mitigation measures to determine if lining the facility would be "best" for this site
would be part of the regulatory process and thus is outside the scope of this assessment.
Groundwater Connectivity and TSF Construction II: A water-impermeable barrier will be
installed only on the interior dam face and nowhere else. To prevent communication between
these facilities and the groundwater, is it feasible to map groundwater inputs before the facility is
filled, place barriers over these areas, and thereby reduce influx of groundwater into the facility
and perhaps prevent movement of toxic water into the groundw atei? I make this point with some
hesitation, given the point made on page 5-29:
"Projecting specific mining-associated changes lo ground water and surface water
interactions in the mine area is not feasible al this lime ""
RESPONSE: See response to previous comment.
Failure of Leachate and Water Collections See my comments under Question 12
below. In addition, I see this as a huge undertaking for which monitoring and response
(mitigation) are clearly as important as the actual plan to capture these wastes
RESPONSE: EPA agrees with this comment. No change suggested or required.
William A. Stubble field. Ph.D.
The risk assessment attempts to consider the effects of metal discharges for water and leachate
from the mine site. This assessment is based on metals concentrations measured in potentially
"similar" mine waters from other sites; concentrations of metals are likely to differ based on
source material and operational differences The effects concentrations used in the evaluation
are based on I S N\\ ambient water quality criteria (AWQC) for metals, and this approach is
appropriate lor "screening le\ el" e\ aluations It should be noted that exceedence of an AWQC
does not portend the occurrence of adverse effects. Ambient water quality criteria are derived in
such a way that they are intended to represent "safe concentrations." In other words, if
environmental concentrations remain below the AWQC, it is assumed that unacceptable adverse
effects will not occur, evceedence of an AWQC suggests that adverse effects may occur to some
species, but thai this must be e\ aluated more closely. Salmonid species are not the most
sensitive organisms in the copper AWQC species sensitivity distribution (SSD); therefore, direct
effects on salmon are e\ en less likely at concentrations in the range of the AWQC.
RESPONSE: Effluents or ambient waters from mines at other sites were not used The
leachate concentrations used (except for the product concentrate leachate) are from available
results of material leaching tests from the Pebble deposit. Otherwise, the Agency agrees with
these comments. The assessment used criteria for screening, but then examines the toxicity
data more closely, including field data to determine potential effects (e.g., aversion, sensory
inhibition, mortality and reduced reproductive success of salmonid fish). The greater
sensitivity of aquatic insects was described in the May 2012 draft of the assessment and is
further highlighted in the revised assessment. The protectiveness of the copper criterion is
considered in both the original and revised assessments.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 115
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
It is interesting to note that the risk assessment document states that copper is one of the "best-
supported criteria. However, it is always possible that it would not be protective in particular
cases due to unstudied conditions or responses." Further, the document goes on to suggest that
organisms such as mayflies etc. are important to the aquatic ecosystem but are not considered in
the copper AWQC and therefore may not be sufficiently protective. It also suggests that because
an acute-chronic ratio approach is employed to correct the final acute value to obtain a final
chronic value, there may be increased uncertainty associated with the protectiveness of the
chronic criterion. This appears to be an area where EPA might benefit by conducting research
(either alone or in concert with industry) to reduce uncertainty in the criteria to an acceptable
level. In addition, additional chronic toxicity data may be available from research conducted in
response to the European REACH regulations and consideration of this research may reduce the
level of uncertainty in the criteria. Bioavailability correction \ ia the BLM approach is only
considered for copper in the risk assessment; biotic ligand models ha\ e been developed for a
number of metals (e.g., zinc, nickel) and these should be considered in the assessment as well.
Finally, the assessment approach seems to use a sum TU-based approach for assessing "metals
mixture" impacts. This is based on an assumption of additive interactions among the metals.
Although this is probably the best assumption in going forward, limited data are available to
support this approach.
RESPONSE: The EPA has examined the El "s 200,S I ohuttury Risk Assessment of Copper
(the relevant REACH document). . Mthough the authors could derive a chronic species
sensitivity distribution, that is because the wtiy that they include and aggregate data differs
from the EPA '.v method. They apparently did not generate lie it' test data for the assessment. In
particular, they have no data for sensitive aquatic insects, so the EU assessment does not
resolve that problem. The BLM was used for copper because it is the contaminant of greatest
concern and because the copper BLM has been approved by the EPA Office of Water. Other
metals with BLMs, such as zinc and nickel, occur at lower levels in leachates relative to their
toxicities, so BLM modeling was not justified.
Areas where additional research would be beneficial include:
• Mixtures Information regarding the potential interactive effects of multiple metal
exposures would be useful and would reduce assessment uncertainty.
• Species sensitivity concerns: there is extremely limited data (esp. chronic datat) on all of
the salmon species of concern in Bristol Bay
• Additional data, especially chronic toxicity data and data for additional metals for which
no water quality criteria exists, would be extremely helpful.
RESPONSE: The EPA agrees that these are good research topics. No change is suggested or
required.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to a potential failure of water and leachate collection and treatment from the mine site. It
only estimates the likelihoods of occurrence and the consequences. See discussion under
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 116
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 4 above regarding suggestions for improving estimation and expression of the
magnitude of risks to salmonid fish due to potential failure of water and leachate collection and
treatment from the mine site.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows. We have added to the discussion of leachate
contamination, including estimation of the magnitude of leachate escaping the collection
system and entering surface waters (Chapter 8).
Water collection and treatment failure likelihood. An eslimale is presented for the amount of
seepage that may flow from the TSF. Similar estimates are not presented for the waste rock piles.
The effects of the effective exclusion of oxygen from the saturated or partially saturated tailings
should be considered in developing an estimate of the water quality of the resulting seepage. It
may be an important factor in reducing the oxidation of sulfide minerals remaining in the interior
of the TSF. This same effect could also mitigate the release of poor water quality in the long-
term following closure. The precipitation on the site may be sufficient to effectively retain a
suction saturated profile in parts of the TSF.
RESPONSE: The revised assessment presents an estimate of the amount of seepage from the
waste rock piles (Table 6-3). The water quality of the leachate from the TSF was modeled with
the concentrations reported from the tailings humidity cell tests. The reported copper
concentration of 0.00533 mg/l is lower than some of the reported groundwater concentrations,
and is the same order of magnitude as the highest reported stream concentration of 0.0013
mg/l. The TSF at closure was modeled with free standing water at the surface and with
downwardflow with an estimated gradient between 0.07 and 0.12.
Water collection and treatment failure consequences. Water collection and treatment is being
done at a number of mines in North America. Past experience at the Red Dog mine is quoted;
however, there are many other examples that could have been examined. An important example
is that of the Equity Silver Mine in British Columbia (Aziz and Meints, 2012): "Acid Rock
Drainage (ARD) was discovered at the Equity Silver Mine in the interior of British Columbia in
1981. The latest water treatment plant was installed in 2003, 9 years after the mine closed in
1994, and is the fourth successi\ e treatment plant for the site that has treated ARD for a period of
over 30 years. Discharge water quality was maintained since 1991 except during two high flows
associated with freshet conditions in 1997 and 2002. ARD collection and treatment system
upgrades were installed after 2<>o2 and these have performed well through three large freshet
conditions in 2007, 201 I and 2012. The timeframe for treatment is perpetuity and financial
assurance is in place for a total amount of $56,291 million through a long-term security bond
(letter of credit) with the BC Provincial Regulatory Authority. The security bond is reviewed by
stakeholders every 5 year". Collection and treatment at Equity Silver indicates that companies
are committing to long-term water treatment of ARD and that regulatory frameworks are in place
to protect water quality in downstream streams and rivers. It is recommended that EPA perform a
more thorough review of other sites where water treatment occurs to better characterize this
failure mode.
RESPONSE: We have been unable to obtain a copy of the cited report. The 2010
Sustainability Report, which is available, does not contain these specifics. This mine
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 117
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
apparently has begun perpetual water treatment. A comprehensive review of mine sites with
water treatment systems is beyond the scope of this assessment.
Phyllis K. Weber Scannell, Ph.D.
This section of the report provides an in-depth discussion of possible sources and fates of
contaminated water. Chapter 6.3 discusses possible adverse effects from early mine closure or
prematurely shutting down a water treatment system. These issues highlight the need for a mine
plan that includes concurrent reclamation, sufficient bonding to conduct reclamation in the event
of an early shut down, and plans and specifications for collection and bypass of clean water and
collection and diversion to a water treatment system of contaminated water.
RESPONSE: The EPA agrees with this comment. \o change suggested or required.
The Risk Characterization (Section 6.3.3) discusses possible contaminant loads to downstream
waters. As stated in this section, it "serves to indicate the large potential risk from improperly
managed waste rock leachate." This statement highlights the need for an in-depth mine plan
with sufficient monitoring and fail-safe provisions An emergency discharge of untreated waters
from a tailings storage facility could he made to a collection pond for later treatment or the
tailings pond could be engineered to accommodate a higher flood event so the likelihood of
overtopping is minimized.
RESPONSE: The waste rock leachate scenario referenced by the commenter has been
eliminated and replaced with a more detailed and realistic scenario for waste rock leachate
collection and treatment (C liapter S).
Section 6.3.3 (Risk Characterization) states " \ltei nati\el\. water collection and treatment failure
could be a result of an inadequately designed water treatment system which could result in the
release of inadequately treated water as at the Red Dog Mine, Alaska (Ott and Scannell 1994,
USEPA I ). In that case, the failure could continue for years until a new or upgraded
treatment system is designed and constructed." This statement is misleading and overly
simplistic; the water treatment system at the Red Dog Mine was designed to treat the predicted
flows. Howc\ er. the stream bypass and collection systems were constructed in 1991 to intercept
seepage waters. The additional water that was collected and treated dictated construction of a
second water treatment system in 1992. Sand filters were added in 1993 to remove fine
particulate Zn. The issue was not that the water treatment was inadequate, but that the pre-
mining hydrologic data was insufficient and that state, federal, and mine officials lacked
experience in mine construction on permafrost soils.
RESPONSE: The water treatment failure at Red Dog was, as the commenter describes, the
result of an unintendedfailure to design a plant that was adequate for the mine and site. The
passage has been expanded to clarify the nature of the failure.
Overall, the discussions of risks to salmonid fish due to a potential failure of water and leachate
collection and treatment from the mine site highlight the need for more comprehensive
information on groundwater, including delineating flow pathways, depth to surface, and water
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 118
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
volumes. Additional information is needed on water collection, storage, and treatment at future
mine facilities.
RESPONSE: Discussions of wastewater collection and treatment have been considerably
expanded in Chapter 8.
Paul Whitney, Ph.D.
No comment on this question.
Question 7. Does the assessment appropriately characterize risks to salmonid fish due
to culvert failures along the transportation corridor? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so what are
they?
David A. Atkins, M.S.
The description of culvert failure is necessarily general because there are currently no designs.
The general data on ai1\ ei l failures presented lor the types of cul\ erts described in the references
cited (principally for forest and range land) indicate a high prohiihi I ity of failure (30-66% failure
rate). It is probable. ho\\e\ er, that the transportation corridor for the project would be
constructed to a higher standard than most of the roads included in these papers. It would be
helpful to know if similar data are a\ ai lahle for highly engineered roads of the type likely to be
built for the project
RESPO N SE: The assessment assumes modern mining technology and operations. We did not
find information explicitly from highly engineered roads, but to the extent possible we used
recent literature from representative environments. The failure frequencies cited in the revised
assessment are from modern roads and are not restricted to forest or rangeland roads.
Information on current design standards is now included in text boxes throughout Chapter 10.
Steve Buckley, M.S., CPG
The references provided in this section emphasize culvert failures in the Pacific Northwest and
Tongass National Forest. The streams and culverts in these regions are heavily influenced by
large woody debris loading. It would be more appropriate to classify the various potential stream
crossings by watershed and the amount of large woody debris available to be recruited to the
stream and influence culvert blockage.
RESPONSE: Many of the references in the assessment relate to the Pacific Northwest,
because much work on culverts and potential impacts on salmon have been performed there.
Flanders et al. (2000) has been deleted in the revised assessment. The failure frequencies cited
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 119
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
in the revised assessment are from the best available literature concerning modern roads, and
are not restricted to forest roads.
The bulk of the information presented on culvert failure research relates to roads in forested
watersheds and these may not be as applicable to the proposed transportation corridor. The
additional information provided on the character of each potential stream crossing is more useful
for the assessment of the potential impacts of the mine road to fish passage.
Courtney Car others, Ph.D.
Culvert failures due to blockage and erosion are noted to be common and are likely to occur in
this scenario. Culvert failures would prevent the movement of fish, which could eliminate a year
class from blocked stream systems and fragment upstream and downstream populations,
increasing likelihood of localized population depletions and extinctions. Monitoring and
maintenance of culverts can be expected to decrease after mine operation, increasing the risks of
these failures. The report appears to appropriately characterize risks to salmonid fish due to
culvert failures along the transportation corridor, although I have no particular expertise with
which to evaluate this assessment.
RESPONSE: No change suggested or required.
Dennis D. Dauble, Ph.D.
Mitigation practices, such as new culvert design, was well described, as was bridging of
roadways and porous fills to mitigate risks due to culvert failure along the transportation
corridor. This assessment also included appropriate risk characterization for both the no-failure
and failure scenarios. There should be literature available from the Washington State Department
of Transportation on fish passage relative to culvert placement and design. Otherwise, I have no
suggestions for improvement.
RESPONSE: We have examined literature on fish passage relative to culverts from the
Washington State Department of Transportation. Culvert failure frequencies from their 2012
paper, in which approximately 62%of culverts were identified as total or partial barriers, were
not used in the assessment because we could not determine the age of examined roads.
Gordon H. Reeves. Ph.D.
The literature review of culverts and their potential impacts on fish and fish habitat is very
thorough and the presentation of results is accurate. However, most of the cited material is from
studies done in areas outside of Bristol Bay and the direct applicability of results is problematic.
This should be done in the revision.
RESPONSE: Because the proposed mining would take place in an unde\>e!oped area, much of
the literature is necessarily from areas outside of Bristol Bay. However, to the extent possible
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 120
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
we used examples from representative environments, and applied the results to the proposed
mine.
I think that there are potential mitigation measures that were not presented. The primary one,
besides the use of bridges, as suggested in the report, is making all culverts be arch culverts.
These culverts make use of the stream bottom, which reduces the potential for the culverts to
become perched and impede upstream movement, and are less likely to change the gradient than
other culvert types. All culverts could, as recommended in the report, be at least one bank width,
which is larger than required by the state of Alaska. This would minimize the possibility of
plugging with debris.
RESPONSE: A number of culvert types (e.g., arch culverts) may he used on the proposed
transportation corridor; it is not in EPA '.s purview to suggest which ones should be used.
However, culvert design approaches specified in the Memorandum of Agreement between the
Alaska Department of Fish and Game and the. Maska Department of Transportation and
Public Facilities (ADF&G and ADOT 2001) are described in Box 10-2 of the revised
assessment.
The review of potential road impacts lacked two possible consequences. One is lliat roads could
be corridors for the introduction of in\asi\e species, planls and animals. The consequences of
the successful establishment of non-naiive species could ha\ e critical ecological impacts on
native species and the ecosystem. The second consequence is that a road will allow greater
access to streams where access was previously limited. Fish populations could be more easily
and intensively harvested in sport and subsistence fisheries, which adds additional stresses to the
populations. Lee et al. (1997. Assessment of the condition of aquatic ecosystems in the Interior
Columbia River basin. Chapter 4. Eastside Ecosystem Management Project. PNW-GTR-405.
Portland, OR: U.S. Department of Agriculture. I oresl Service, Pacific Northwest Research
Station.) found a direct relation between road access and the status of salmonid populations in
the Columbia liasin
RESPONSE: II e agree with the commenter that roads could be corridors for the introduction
of invasive species. The potential impacts of invasive species as a result of construction and
operation of the proposed transportation corridor are discussed in Section 10.3.6 in the revised
assessment. II itli respect to stream access, EPA assumes that the proposed road would be
closed to the public during mining operations but potentially could become a public road after
mining operations cease. If that were the case, there would be greater access to streams and
fish populations. However, the potentially important impact to fish would likely occur from
secondary (induced) development, or development resulting from the introduction of industry,
roads and infrastructure associated with mining. This is briefly discussed in Section 13.3 in
the revised assessment. Improved accessibility would increase hunting and fishing pressure, as
well as competition with existing subsistence users.
Charles Wesley Slaughter, Ph.D.
No. The Assessment does not adequately address the road/stream crossing/culvert issue. Given
the projected transportation corridor, Pebble locale to Cook Inlet, and the inevitability of a
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 121
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
further network of "minor" roads in the mine and TSF locale, plus additional infrastructure
linkages, road/culvert/stream crossings are a major concern for aquatic habitat and fisheries.
Readers of the Assessment should be directed to Frissell and Shaftel's Appendix G for a more
comprehensive discussion of this important topic.
RESPONSE: The revised assessment addresses the road/stream crossing/culvert issue in
detail. Secondary development is now described qualitatively in Chapter 13 of the revised
assessment. The reader is repeatedly directed to Appendix G for more details.
The specific consequences of a failure on salmonid habitat and biology are portrayed well.
RESPONSE: No changes suggested or required.
John D. Stednick, Ph.D.
No. It is unclear how the estimate that 50% of i lie cul\ erts would I ail was obtained, given that
the literature shows a range of 30 to 60% (Section 4 4 4). What literature was used? Are road
BMPs satisfactory in this environment? Have the Alaska BMPs been audited ' Culvert repair
taking a week to several repairs in a month seems high If the road crosses a critical salmon
rearing stream, conservative pipe sizing or hiidgework could he considered. The direct loss or
inaccessibility of upstream salmon hahitat does not necessarily translate to salmon loss. Timing
of culvert blocking event with salmon migration and duration of blockage should be considered.
Need to include references to Alaska Department of Natural Resources and Alaska Department
of Transportation.
RESPONSE: The literature showed a range of 30 to 66%, with an average of roughly 50%.
The literature used was noted in Section 10.3.2. I. However, Flanders et al. 2000 was deleted in
the revised assessment, bringing the average culvert failure frequency to 47%.
Best management practices (BMPs) or mitigation measures would be used to minimize
potential impacts to salmon ecosystems from construction and operation of the proposed
transportation corridor. These BMPs, and their likely effectiveness, are now discussed in text
boxes throughout Chapter 10 (e.g., liox 10-2). Environmental characteristics along the
transportation corridor would likely render the effectiveness of standard or even "state of the
art" mitigation measures highly uncertain. Further discussion on this is contained in Box 10-
5 of the revised Assessment.
We are not aware of any audits of Alaska BMPs.
The text relating to culvert repair considers that the proposed mining would occur in an often
harsh, remote environment.
Design considerations (including sizing) for culverts are now discussed in Box 10-2. Culverts
would be designed in accordance with guidance in Alaska Highway Drainage Manual (ADOT
1995) and the Memorandum of Agreement between the Alaska Department of Fish and Game
and the Alaska Department of Transportation and Public Facilities (ADF&G and ADOT
2001). Both of these documents are cited within Box 10-2.
Blockage of a culvert by debris or downstream erosion would prevent the in-and-out migration
of salmon and the movement of other fish among seasonal habitats. The direct loss or
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 122
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
inaccessibility of upstream salmon habitat does not necessarily translate to loss of a
population of salmon, but production would ultimately be reduced If blockage of a culvert by
debris or downstream erosion occurred during in-migration of salmon and persisted for
several days, it would result in the loss of spawning and rearing habitat. If it occurred during
out-migration and persistedfor several days, it could cause the loss of a year class of salmon
from a stream
What are the design considerations for the culverts? What precipitation/streamflow relationships
will be used for sizing purposes? What are the usual casual mechanisms for culvert failure? How
much woody material do these streams carry? Do culverts fail from debris plugging, road
slumps, or overtopping by storm events? What road BMPs will be implemented?
RESPONSE: Design considerations (including sizing) for culverts are now discussed in Box
10-2. Culverts would be designed in accordance with guidance in Alaska Highway Drainage
Manual (ADOT1995) and the Memorandum of. \greement between the Alaska Department of
Fish and Game and the Alaska Department of Transportation and Public Facilities (ADF&G
and ADOT 2001).
As noted in Section 10.3.2, culverts are deemed to have failed if culverts (and thus fish
passage) are blocked (e.g., by debris. ice. or beaver activity) or if stream flow exceeds culvert
capacity, resulting in overtopping and potential road washout. The causal mechanisms for
culvert failure are briefly discussed in Section 10.3.2.
We do not have an estimate of the amount of woody material carried by streams in the
assessment area,
BMPs (e.g., stormwater runoff and fine sediment mitigation) that would be implemented are
now discussed in text boxes throughout Chapter 10 of the revised assessment.
Roy. 1. Stein, Ph.D.
Sizing Culverts: Page 5-61 I lav the suggestion is that "Culverts must be 0.9 times the ordinary
high-water width..and w here the channel slope is less than 5%, the "the culvert is allowed to
be 0.75 times" this same marie Does this take into account global climate changes, which
would mean higher flow rales lhaii historically has been the case? Shouldn't culverts be sized
larger than what historical flow rales would suggest, given that Climate Change will likely result
in more intense storms and therefore greater stream flows than has historically been the case?
RESPONSE: The commenter makes a good point. We do not believe that the sizing
suggestions noted by the commenter (found in the Memorandum of Agreement between the
Alaska Department of Fish and Game and the Alaska Department of Transportation and
Public Facilities (ADF&G and ADOT 2001) take into account global climate change. Climate
change projections and potential impacts are now included in Chapter 3. We note in Chapter
10 of the revised assessment that climate-related changes, such as increased flood frequency
and shorter return interval for major flood events, would likely undermine the structure of the
proposed transportation corridor and stream crossings. The variability and magnitude of
stream flows could also enhance other impacts described in the chapter. Interactions between
climate change and mining risks are discussed in Box 14-2.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 123
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Culvert Failures: Page 6-42. Culvert failure rates of 30-66% suggest we are doing something
wrong with establishing these culverts to maintain stream flow under a road. How might we
reduce this rate of failure (larger culverts? placement issues? solutions of any sort?)? In fact, if
indeed 50% of the culverts will be blocked (see bottom of page 6-43), are we not dealing with an
unacceptable solution of running streams under roads? Might there be some replacement of
these culverts with bridges; certainly, bridges are more expensive to build, but they simply do
not have the failure (i.e., blockage) rate that culverts do. Might there be a trade-off here between
initial investment costs (high for bridges) and salmon protection (fewer blockage events)? What
would Best Management Practices tell us in this context0
RESPONSE: In the revised assessment, culvert failure frequencies are reported as 30 to 58%,
with an average failure estimate of 47%. The likelihood of extended blockages would be low
during mine operations because the roadway would be monitored daily to ensure that failures
could be rapidly identified and repaired. However, the likelihood would increase after mine
operations cease, if inspection and maintenance frequencies declined to those of typical roads.
Best management practices (BMPs) or mitigation measures would be used to minimize
potential impacts to salmon ecosystems from construction and operation of the proposed
transportation corridor. These BMPs. and their likely effectiveness, are now discussed in text
boxes throughout Chapter 10.
Bridges would generally have less impact on salmon than culverts, but can result in the loss of
long riparian side channels if they do not span the entire floodplain. The actual decision as to
what type of structure (bridge versus culvert) would be constructed at each crossing would be
made by industry engineers in consultation with state permitting staff.
William. I. Stubblefield, Ph.I).
Potential effects on salmonid populations were e\aliialed due to culvert blockage and failures.
Culvert Mockages will pie\ eni salmon passage leading to possible effects on reproductive
success. I .i lei aliire data for the incidence of culvert failures were used in assessing failure
probability This seems to be an appropriate approach given the hypothetical nature of the mine
used in the assessment; howe\ er. this is not my area of expertise and I am not aware of
additional data that should be considered.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
Road and culvert failure likelihood. The likelihood of road and culvert failures is discussed in
Section 4.4.4 (p. 4-62). This section relies on the paper by Furniss et al. (1991) for a number of
aspects. As was pointed out above, this paper is focused on forest and rangeland roads and is not
applicable to the access road for the Pebble Mine. It is recommended that further evaluations be
done of similar roads at mines constructed between mines and port facilities to update this
section.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 124
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Furniss et al. (1991) does focus on forest and rangeland roads, but it is a
seminal publication on the potential effects of roads, particularly as they relate to salmon. The
general conclusions of that paper should be applicable to the transportation corridor proposed
in the assessment. The failure frequencies cited in the revised assessment are from modern
roads and not restricted to forest roads. Because the proposed mining would take place in an
undeveloped area, much of the literature used in the assessment is necessarily from areas
outside of Bristol Bay. However, to the extent possible we used recent literature from
representative environments. We found no literature concerning the operational success of
culverts on roads between mines or connecting mines and ports.
Road and culvert failure consequences. The failure consequences discussed in Section 6.4
seem to be based on almost total regulatory failure during and after operations. The information
also serves to highlight the aspects that should be considered u lien designing, operating and
maintaining the access road during operations and subsequently during closure.
RESPONSE: In the revised assessment, consequences of failures and routine operations are
covered in Chapter 10. The failure consequences are not based on almost total regulatory
failure. Rather, they take into account the use of best management practices (BMPs) or
mitigation measures that are discussed in text boxes throughout Chapter 10. \onetheless,
environmental characteristics along the transportation corridor would likely render the
effectiveness of standard or even "state of the art" mitigation measures highly uncertain.
Further discussion on this is contained in Box 10-5 of the revised assessment.
Phyllis K. Weber Saumell, Ph. I).
The risks to salmonid fish due to culvert failures would be minimized by implementation of
permits by Alaska Department of l-'ish and (ianie (ADF&G), Habitat Division. Under A.S.
16.05.840-870, Alaska has some ol'lhe most proleelix e laws for fish and fish habitat in the
United States. Further, given the lack of specific information on road alignments, construction
methods and stream crossings, it is not possible to calculate lengths of affected streams, quantify
loss of fish habitats, or predict failures of culverts, side slopes, etc. The document would be
strengthened if it included speeilie information on locations of spawning and rearing habitats and
estimated the contribution of fish habitats in the Nushagak River and Kvichak River Watersheds
to the Bristol Bay fishery
RESPONSE: Best management practices (BMPs) or mitigation measures that would be used
to minimize potential impacts to salmon ecosystems from construction and operation of the
proposed transportation corridor are now discussed in text boxes throughout Chapter 10. Box
10-2 specifically refers to fish habitat regulations under Title 16.
As noted in the revised assessment, uncertainty exists in the characterization of streams and
wetlands affected by the proposed transportation corridor. Based on the chosen road
alignment scenario (which agrees with that proposed in Ghaffari et al. 2011) we feel that we
are justified in estimating the potential footprint of the proposed corridor and its potential
impact on fish habitats and populations. We note in the revised assessment that "Although this
route (the one proposed in the EPA scenario) is not necessarily the only option for corridor
placement, the assessment of potential environmental risks would not be expected to change
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 125
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
substantially with minor shifts in road alignment. Along any feasible route, the proposed
transportation corridor would cross many streams, rivers, wetlands, and extensive areas with
shallow groundwater, including numerous mapped (and likely more unmapped) tributary
streams to Iliamna Lake (Figures 10-1 and 10-2)."
Specific information on locations of spawning and rearing habitats along the proposed
transportation corridor is difficult to obtain; as noted in the revised assessment, the Alaska
Anadromous Waters Catalog and Alaska Freshwater Fish Inventory do not necessarily
characterize all potential fish-bearing streams because of limited sampling along the corridor.
Nonetheless, the revised assessment summarizes the species, abundances, and distributions
that would potentially be affected, and places the streams along the transportation corridor
into the context of the entire Nushagak andKvichak River watersheds with respect to
important watershed attributes such as discharge, channel gradient, andfloodplain potential.
As far as placing potential mining impacts in the context of the entire Bristol Bay watershed,
we are unable to build a complete IP model, as this would require validation and more
elaborate construction of metrics appropriate to this region. However, our preliminary
characterization provides the building blocks for assessing the distribution of key habitat-
forming and constraining features across these watersheds.
Paul Whitney. Ph.D.
Criteria for bridge versus culvert installations alonu the proposed haul road. The dynamic
process of beaver dams causing streams lo mo\ e across the floodplain should also be a criterion
for determining if and where culverts are installed lor a potential road (pages 4-36 and 4-63).
Even if salmonids are not present at a stream crossing, the mosaic of active and decayed beaver
ponds in the floodplain can be important rearing areas for forage fish and benthic drift that are
utilized by salmonids (Snodgrass and Mel'le. 1l^7. Schlosser and Kallemeyn, 2000). If beaver
dams (but not salmonids) are present abo\ e proposed stream crossings, bridges or causeways that
allow the streams to move across the floodplain should be recommended versus a culvert.
RESPONSE: Our revised assessment is based on the assumption that crossings over streams
with mean annual flows greater than 0.15m3/s would be bridged. However, the actual decision
as to what type of structure (bridge versus culvert) would be constructed at each crossing
would be made by industry engineers in consultation with state permitting staff. We agree with
the reviewer that beavers can have an important influence of channel location and
morphology. However, beavers move frequently, so over the life of the road the locations of
beaver dams would change and therefore seem unlikely to provide a good criterion for
crossing designs.
Beaver are known to block culverts at the upstream ends. Beaver-proof culverts are an option,
but all the designs I am aware of would certainly hinder, if not block, movement of forage fish
and benthic drift. Causeways or bridges are the best way to encourage beaver activity (i.e.,
functions) and all the benefits that accrue.
RESPONSE: The actual decision as to what type of structure (bridge versus culvert) would be
constructed at each crossing would be made by industry engineers in consultation with state
permitting staff.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 126
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Toxic plume. Spills of transported chemicals are not quantified in the culvert failure section of
the assessment. I have participated in a mine risk assessment and a landfill risk assessment where
spills of cyanide and landfill leachate have been modeled. While I did not conduct the plume
movement analyses, stream hydrologists readily calculated how far spilled materials would move
down stream until the concentrations of chemicals in streams reached acceptable benchmarks.
The longevity of a spill of chemicals for copper processing should be calculated. It appears that
the Water Treatment failure assessment on page 6-39 conducted some sort of plume analysis to
determine the potential for an impact on Iliamna Lake. Perhaps it is possible to use this analysis,
or at least the model, to address the consequence of a spill of i ran sported chemicals.
RESPONSE: We have modeled the transport of spills from pipeline failures and emissions
from the wastewater treatment plant and leakage of leac hates. In Section 10.3.3.1 of the
revised assessment we estimate the number of reagent spills that would occur over the roughly
25-year life of Pebble 2.0 scenario. We did not conduct a plume analysis for a spill of
transported chemicals, but given the toxicity of sodium ethyl xanthate (Section 6.4.2.3) we
expect that a spill of this compound into a stream along the transportation corridor would
cause a fish kill. Given the uncertainty concerning the nature and magnitude of a truck
accident and spill, we decided that a quantitative analysis of transport and fate would not
materially contribute to the value of the assessment.
Question 8. Does the assessment appropriately characterize risks to salmonid fish due
to pipeline failures? If not, what suggestions do you have for improving this part of the
assessment? Are significant literature, reports, or data not referenced that would be
useful to characterize these risks, and if so what are they?
David. I.. Uhins, M.S.
The discussion of pipeline failures is based on published failure rates, principally for oil and gas
pipelines This analysis results in a pipeline failure rate of one per 1,000 km per annum. This is a
pretty generic iiLimhcr that docs not consider actual pipeline design. Rather it indicates that
pipelines designed using standard practices do fail with a fairly high frequency. The Assessment
does not apply this failure rale to the gas and diesel pipelines because "they are not particularly
associated with mining "" Wiihoui the mine, there would be no pipeline. So given that this rate of
failure is quantifiable based on good data and that the pipeline would be built to serve the
project, this risk should be considered.
RESPONSE: The assessment assumes that pipeline design follows standard ASME practices.
A diesel pipeline failure and resultant spill into two creeks has been added in Chapter 11 in
the revised assessment. A gas leak is considered but is not analyzed because of the lack of
significant causal linkage to fish production.
A concentrate pipeline spill would have differing impacts depending on when and where the spill
occurred, with deposition in Lake Iliamna likely being the worst outcome. As noted in the report,
it is likely that a pipeline spill would be detected rapidly and that the volume of the spill would
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 127
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
be limited and amenable to remediation. A better description of how concentrate pipeline failures
have occurred would be helpful to better understand the risk for this project (e.g., the July 2012
Antamina concentrate pipeline failure, although this pipeline would operate under a much
different pressure regime due to extreme altitude change).
RESPONSE: Consolidated, statistically representative data on concentrate pipeline failures
are not readily available, although anecdotal evidence from some case studies can be found.
The pipeline failure statistics reviewedfor the assessment come primarily from oil and gas
pipelines, but also include some water and hazardous liquid pipelines. The performance of
mining concentrate pipelines is not expected to be better than the performance of oil pipelines,
because concentrate pipelines would be expected to be more susceptible to internal corrosion
and abrasion. The 2012 Antamina concentrate pipeline failure in Peru was reportedly caused
by the rupture of a pipe elbow in a valve station. The regulatory, geographic, and operating
conditions of the Antamina pipeline may differ greatly from those of the concentrate pipeline
in the assessment scenarios. A discussion of causes and probabilities of pipeline failures is
included in Section 11.1 of the revised assessment. The revised assessment also includes
discussions of concentrate spills at the Bingham Canyon, Utah, and Alumbrera, Argentina,
copper mines (Section 11.3.4.2).
Steve Buckley, M.S., CPG
The assessment does generally describe the potential risks lo iisli iVoin hypothetical pipeline
failures.
RESPONSE: No changes suggested or required.
Courtney Cur others, Ph.l).
A pipeline failure would be expected lo release toxic leachale into stream systems in the
transportation corridor, none of which would dilute the leachate enough to prevent severe toxic
effects (both immediate and long-term) The report discusses three pipeline failures in the Bajo
de la Alumbrera mine in Argentina. The largest pipeline failure lasted two hours (compared to
only two minutes of exposure hypothesized in the current mine scenario). The report could more
clearly describe this case and its likely effects. The report appears to appropriately characterize
risks to salmonid fish due to pipeline failures, although I have no particular expertise with which
to evaluate this assessment.
RESPONSE: The EPA agrees that it would be desirable to have more information on the
effects of the largest Bajo de la Alumbrera spill, but we have included all information on that
failure that is available.
Dennis D. Dauble, Ph.D.
The risks to salmonid fish due to release of pipeline concentrate/slurry and leachates (as return
water) are well described. However, risks of a diesel fuel spill are not. More detail could be
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 128
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
provided on reclaimed water. For example, what toxic constituents (and at what volumes) would
be released to the environment if these pipelines failed?
RESPONSE: A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11 in the revised assessment. New data on concentrate leachate in the slurry have
been added, and it is assumed to also describe the return water.
Gordon H. Reeves. Ph.D.
Assuming that characterizations of the pipeline failure arc accn rale, the potential impacts on fish
and fish habitat are appropriate and reasonable. It was clear that the effects of a pipeline failure
could be major, depending on the duration and timing of i lie spill, because of the concentration
of metals in the slurry, the particular life-stage present, flow conditions, and the reduced
potential to fully remove the material from a stream or wetland afterwards
RESPONSE: No change suggested or required.
The one question that I had about this section was the potential impact on |ih\ toplankton and
zooplankton in Lake Iliamna, particularly at the local scale I assume that any spill from pipeline
failure would have potential impacts on the lake and 011 ph\toplankton and zooplankton, the
major food for juvenile sockeye salmon. The ecological consequences would depend on the
extent and intensity of any spill and 011 how ju\ enile sockeye use areas near tributary streams. I
would expect that a spill could be particularly detrimental if ju\ enile sockeye use the area near or
adjacent to natal streams when they enter the lake I think this should be considered in more
detail in any additional analysis.
RESPONSE: The toxicity data for species sensitive to copper and most other potentially toxic
metals are derived from planktonic crustaceans, so it is already addressing planktonic lake
species more directly than stream bentliic species. However, the more detailed analyses of
aqueous releases in the revised assessment (Chapter 11) provide a better basis for addressing
the risk of exposure in Iliamna Lake. Because the aqueous phase of the product concentrate
slurry would enter the stream briefly and would be rapidly diluted in the lake, it is not judged
to be a major risk to plankton relative to the potentially sustained direct effects on salmon eggs
and larvae of the deposited solid phase or relative to stream invertebrates, which would have
much less benefit of dilution.
Charles Wesley Slaughter. Ph.D.
No. Concerns with pipelines crossing streams, watercourses and wetlands are similar to those
earlier expressed for the road corridor. On-site investigation may well reveal many more
"watercourses," including intermittent and ephemeral streams, than the 70 crossings cited;
possible pipeline failures thus may have much wider potential for impacting salmonids than is
indicated in the Assessment.
RESPONSE: The document has been edited to indicate that 70 is a minimum value.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 129
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and recently
in Montana (?), suggesting the probability (with what confidence limits?) that there would be
only 1.5 stream-contaminating spills or two wetland-contaminating spills over 78 years of
operation seems wildly optimistic (and what is half a spill?).
RESPONSE: If you think of the mine as a repeated experiment (i.e., many mines with
concentrate pipelines of that length and duration), we would expect a mean frequency of 1.5
spills. A more straightforward explanation is that we expect 1 or 2 spills, given the scenario.
These frequencies are based on a large data set, not just the TAPS experience (which is an
atypical pipeline design and much larger than the diesel line) or the spill in the Yellowstone
River (which is a single event). An explanation of the frequency has been added in Section
11.1 of the revised assessment.
Assuming that any spill (over the 78-year project span) would last only two minutes (p. 6-32, p.
6-34), with a consequent minimal volume of spilled material, also seems highly optimistic. Even
highly-automated systems, with redundant sensors and automatic responses, are susceptible to
error or failure, and the Bristol Bay watershed environment is not benign with regard to
mechanical apparatus. The authors appear to recognize this u illi their discussion of the
Alumbrera incident.
RESPONSE: The scenario has been modified to a 5-minute response. Also, more information
has been added about the possibility of system failure or human error in response to this and
similar comments. The uncertainty discussion in Chapter 11 indicates that the exposure
scenario is predicated on the successful operation of a remote shutoff. There may be extreme
weather or geological events that render the remote shutoff system inoperable. We did not
evaluate those events. so the assessment may underestimate these risks.
The specific consequences of a failure on salmonid haMtat and biology are portrayed well.
RESPO.WSE: \o changes suggested or required.
JohnD. Stednick. Ph.l).
The pipeline corridor consists of lour pipelines over a distance of 86 miles. No information was
provided on pipeline structure or placement, other than mentioning of stream crossings. The
pipeline failure of concentrate slurry was modeled using chemistry from the Aitik (Sweden)
mine. Is this best approximation? That mine is about 80 years old and is processing ore from the
edge of the pit, with much lower sulfur content than Pebble.
RESPONSE: More information on the pipeline structure and placement was provided in
Section 6.1.3.2 of the revised assessment. We have replaced the USGS leachate data from the
Aitik product with analyses from actual concentrate slurry provided by Rio Tinto to describe
the aqueous phase of the slurry and the return water, which would be alkaline (Section
11.3.2.1). However, the Aitik data are used to estimate the risks from deposition of the product
in a stream or wetland. The environmental leaching of the concentrate would resemble the
USGS's leaching test of the concentrate, not the alkaline solution in the pipeline slurry. To the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 130
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
best of our knowledge, no other aqueous leach test of a porphyry copper product concentrate
is available.
Pipeline failures can be significant in any environment and spill or pipe break prevention
requires significant monitoring. Will automatic shutoff controls be included? Are workers
stationed 24 hours/day every day? Some of the past Alaska failures were in winter conditions,
when things were not easily visible—under ice or snow cover. How will this be addressed?
RESPONSE: Our scenarios include remotely-operated valves tied to a supervisory control and
data acquisition (SCADA) continuous monitoring system tied to a continuously staffed control
room The remotely-operated valves could be triggered either manually or through software,
but we expect that except in the event of a clear rupture of the pipeline, the SCADA signals
would trigger an alarm that would initiate manual intervention by the operator. Consistent
with best management practices, we have assumed that the pipeline would be visually
inspected along its full length at least daily, and that the inspection protocols would consider
the difficulties of detecting a spill under snow or ice. Potential remedial actions for cleanup up
of diesel, product concentrate and return water spills are discussed in C hapter 11.
The toxicity approach seems reasonable. What is the anticipated chemistry of the return waters?
Diesel spill monitoring? The geometric mean of three \ allies (u hich references) indicates that
there is a 14% probability of failure in each pipeline in each year. This is not acceptable at any
level.
RESPONSE: The return water is assumed to be the same as the aqueous phase of the slurry
water. A diesel spill scenario has been added in C hapter / /. The pipeline failure rate reported
in the assessment is similar to the rates calculated in studies by others and represents several
large datasets. Some of the reported failures are due to corrosion, which tends to result in
small releases; some of the failures are due to mechanical impacts, which tend to produce
larger releases: some are due to other causes. The buried pipelines in our scenarios would be
expected to have a lower incidence of rupture from third party activities, but may have a
higher incidence of rupture due to landslides or earthquakes than the overall dataset.
Corrosion or corrosion leaks could go undetected for longer periods of time compared to an
aboveground pipeline.
Roy A. Stein. Ph. I).
Pipeline Failures I. Whal dictates 14 km between automatic shut-off valves; shouldn't this
distance be shorter as the pipe becomes larger, i.e., related to the amount of liquid/slurry that
would be spilled upon pipe failure? Shouldn't all of these pipelines be double-walled? What
would Best Management Practices tell us in this context?
RESPONSE: The distance between valves appears to be consistent with current practices. The
pipelines are described as double-walled in above-ground reaches. A double-walled pipeline
along the entire length of a pipeline might be desiredfrom a purely environmental protection
standpoint; however, it may not be feasible or cost effective to do this. Therefore, we have
proposed the most commonly used (and accepted method) of double-walled construction over
any water bodies. There are a large number offactors that go into standard practices for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 131
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
design, construction, and testing of pipelines. In the draft assessment, we included a selected
number of these factors as examples of our design mitigation measures. In the revised
assessment (Section 6.1.3.2), we have included a statement that the design would follow the
standards of the American Society of Mechanical Engineers, to indicate that modern
mitigation measures would be included.
Pipeline Failures II. Like all other failures, it seems to me that "Standard Operating Procedures
(SOP)" for mitigation should be in place in anticipation of any future spill or contamination of
the environment. I do not think that these procedures need to be in this report, but an
acknowledgement of their presence and that mining companies will follow these SOPs in
response to any spills that occur, be it pipeline, TSF, 1 ruck, leachale bed, etc.
RESPONSE: The EPA agrees that a mining company should have an SOPfor remediating
spills, but we have not been able to find such an SOP for remediating a product concentrate
pipeline spill.
William A. Stubblefield. Ph.D.
Potential effects on salmonid populations were evaluated due lo potential pipeline failures as part
of the risk assessment. This evaluation focused on potential failures associated with the pipelines
for the product concentrate slurry and return water. No consideration of the natural gas or diesel
pipelines was presented, stating that such pipelines "are common and the risks are well-known."
Although I would acknowledge the failures in natural gas and petroleum pipelines are common, I
would not discount the potential effects to salmon populations associated with such spills.
RESPONSE: A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11. Natural gas is not a contaminant of concern because it would vaporize and, at
worst, burn, which would not pose a significant risk to salmonid fish.
Evaluation of potential impacts due to a spill of product concentrate slurry or return water was
based 011 extant data from an existing copper mine in Sweden; to the extent that this slurry and
return water is representath e of similar materials coming from the Pebble mine, this approach is
appropriate. The assumptions used in the amount of material that might possibly be spilled seems
appropriate and hasecl 011 past experience and realistic assumptions; however, these assumptions
need to be reconsidered if and w hen a real mine plan is prepared.
RESPONSE: TheAitiU leacliate is no longer used to estimate the aqueous phase of the slurry,
in response to other comments. Because the slurry would be alkaline, appropriate analytical
data were obtained from Rio Tinto (Section 11.3.2.1). However, the Aitik data are still used for
the leaching of the concentrate in a stream or wetland where neutral water would be the
leaching agent. No other relevant data are available.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to pipeline failures. It only estimates the likelihoods of occurrence and the consequences.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 132
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
See discussion under Question 4 above regarding suggestions for improving estimation and
expression of the magnitude of risks to salmonid fish due to pipeline failures.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows.
Pipelines failure likelihood. The EPA Assessment focuses on the failure of the concentrate
pipeline because "We do not assess failures of the natural gas or diesel pipelines here because
such pipelines are common, their risks are well known and they are not particularly associated
with mining". I find this statement puzzling because all pipeline failures should be of concern. It
is further puzzling because the likelihood of pipeline failures for the concentrate pipeline is
derived from the failure statistics for pipelines in the oil and gas industry (p. 4-60). Failure of the
Baja de la Alumbrera concentrate pipeline in Argentina is suggested as an analog, indicating that
such failures can occur; however, I disagree thai "it suggests that concentrate pipeline failures
are common at a modern copper mine." This last statement is not supported by any further
analysis of concentrate pipeline failures at oilier modern copper mines. It is recommended that
such analyses be performed or that the text be edited to indicate this shortcoming
RESPONSE: The revised assessment contains an analysis of the risks of a diesel pipeline spill
in Section 11.5. Natural gas is not a contaminant of concern because it would vaporize and, at
worst, burn off, which would not pose a significant risk to salmonidfish (Chapter 11). The
quoted statement is literally correct, liaja de la Alumbrera is a modern copper mine and
pipeline failures are common there. This suggests that failures are common. Further,
concentrate pipeline failures at the Antamina and Hingham Canyon mines are now discussed.
However, the statement in the revised assessment has been weakened by changing "common "
to "not uncommon." To support that statement, a review of pipeline failures at U.S. porphyry
copper mines has been added (Section 11.1). However, the EPA has not found a sufficient
record of product concentrate pipeline operations to develop a probability offailure that is
specific to that pipeline type.
Pipelines failure consequences. Failure consequences are focused on the release of concentrate
into water As indicated in Appendix H of the report, the analog concentrate from the Aitik mine
is dominated In chalcopyrite, a sulfide mineral which contains the copper. If the concentrate is
submerged under water in relati\ ely slow flowing streams then very little long-term release of
the copper will occur, as the water does not contain sufficient oxygen to allow for sulfide
oxidation. It is only w hen the concentrate is transported to locations above the water level that
oxidation and release of metals will occur.
RESPONSE: Subaqueous oxidation of sulfides does occur in well oxygenated waters as this
comment recognizes by stipulating "relatively slow moving streams". However, as the flow
data in the spill scenario indicate, these are not slow moving streams. Subaqueous leachate
column tests are conducted to assess the potential for subaqueous oxidation. Tests conducted
by the Pebble partnership indicate the potential for subaqueous oxidation. Salmonids require
high dissolved oxygen levels, so salmonid streams are necessarily well oxygenated. For
example, the EBD states that mean dissolved oxygen levels for the North Fork Koktuli River
was 10.2 ppm. This oxygen would oxidize the sulfides.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 133
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Phyllis K. Weber Scannell, Ph.D.
This section of the document focuses on effects of pipeline failures; however, without a viable
mine plan, descriptions of pipelines and estimates of possible effects are speculative. The
resource developer may opt to build a pipeline to transport fuel from the coast to the mine site or
slurry concentrate to the port. Construction of any pipelines would require review and approval
by state and federal agencies, such approvals would likely contain monitoring plans to ensure
pipeline integrity. However, the risks of pipeline failures should not be minimized; the Fort
Knox Mine near Fairbanks recently experienced a 45/100 u;i11on spill of cyanide solution after a
bulldozer struck a supply line (Fairbanks Daily News Miner. August 24, 2012).
RESPONSE: The EPA agrees that the specific locations of pipelines and requirements for
monitoring may differ from our scenario, which is based on preliminary mine plans (Ghaffari
et al. 2011). However, we believe our scenario is plausible and allows an evaluation of
potential impacts from pipeline failures. The I'ort Knox spill has been added to the assessment
as an example of spills due to human error in Section 11.1.
The risks of a pipeline failure to salmonid fish depend on the duration of the spill, the type of
material spilled (return water or concentrate). the location of ilie spill (in the uplands or in a
waterway), and the timing. The effects of a pipeline failure in a waterway when juvenile salmon
are present would be far more severe than a pipeline failure in an upland area.
RESPONSE: EPA agrees with this comment. Discussion of spill location and life stage
exposed has been expanded in Section /1.3. and the other issues have been carried over from
the May 2012 draft.
Given that there currently is no information on road alignments or locations of future pipelines, it
is not possible to estimate the number of stream crossings (70, page 6-30) or an exact length (269
km, page 6-30) of potentially affected waterways. The risks from pipeline failures outlined in
the draft document should be revised when more specific information on the mine plan of
operations becomes available
RESPONSE: The road alignment and length in Northern Dynasty Mineral's preliminary
mine plan (Ghaffari et al. 2011) were used in this assessment. The number of stream crossings
was also detailed in that plan, and was checked by the EPA using USGS data. We would
expect that risks would be re-evaluated as part of a future specific transportation corridor
plan. No change required.
Paul Whitney. Ph.D.
Refer to comments/responses to Questions 2 and 7.
RESPONSE: See responses to those questions.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 134
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 9. Does the assessment appropriately characterize risks to salmonid fish due
to a potential tailings dam failure? If not, what suggestions do you have for improving
this part of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
David A. Atkins, M.S.
The Assessment generically describes tailings dam failures and the potential impact in detail. It
also uses some site-specific information on tailings supernatant and humidity cell leachate. There
is no question that a tailings dam failure would be catastrophic lor lhe fishery and the project,
and although low probability, is the single largest risk lo the fishery. A tailings dam failure could
harm a very large area of the watershed for a very long period of li me and could require a
massive and expensive remediation effort.
RESPONSE: Agreed. No change suggested or required.
The tailings deposition and storage methods outlined in the Wardrop \I 4.i-l
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
salmonid fish due to a potential tailings dam failure, although I have no particular expertise with
which to evaluate this assessment.
RESPONSE: No change suggested or required.
Dennis D. Dauble, Ph.D.
Tailings deposition is described in Chapter 4 of the main report, but I could not find anything
that described potential risks to fishes, including effects to aquatic food webs and loss of fish
spawning and rearing habitat.
RESPONSE: The TSF failure description and the assessment of risks were presented in
separate chapters in the original draft assessment, but have now been moved into one location
(Chapter 9) for clarity
As noted in the text, the sediment transport model used could only simulate sediment transport
and deposition -30 km downstream of the mine site Thus, potential effects to fish habitats were
not well quantified for the mainstem Koktuli Rh er (and beyond), in addition lo the Mulchatna
and Nushagak rivers. Is there a likelihood that any tailings material might reach Lake Illiamna?
If not, say so in the document. It is equally useful to say u here impacts will not occur (as it
relates to sensitive habitat) as it is to describe where impacts are likely and reasonable.
RESPONSE: None of the 3 TSFs are in the watershed of Ilia nma Lake. The hydrology of the
site has been clarified and better maps added to clarify this issue. Risks to Iliamna Lake from
water treatment failures, even if they occur in the Sushagak drainage, are now noted (e.g.,
transport of toxic leachate from the South I'orli Koktuli to Upper Talarik Creek via
groundwater exchange between theses basins: Chapter 8). However, sediment (i.e., tailings)
would not follow that route.
The assessment deemed that it was "not possible" to determine how far the initial slurry
deposition would extend, how far re-suspended sediments would travel, and how long erosion
processes would continue. It seems that information from other mine closure sites could be used
by assessment authors to infer effect by analogy. The statement alluding to potential sediment
run out distance at the bottom of page 4-56 of the main report should be included in the summary
of effects. This is an important point.
RESPONSE: The revised assessment now includes a clearer description of the magnitude and
duration of effects. We apply the runout distance equations of Rico et al. (2008) to conclude
that under Pebble 2.0 scenario dam failure conditions, runout distance exceeds 307 km (190
miles), reaching the marine waters of Bristol Bay (Section 9.3.2).
Gordon H. Reeves. Ph.D.
Assuming that characterizations of the dam failure are accurate, the potential impacts on fish and
fish habitat are appropriate and reasonable. Impacts, like those of a pipeline failure, are likely to
be widespread in the watershed and to be long lasting, resulting from inundation of areas by
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 136
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
sediments and contaminants in the water. I think that potential impacts across the broader scale
could be developed and highlighted more fully. Also, consideration of Intrinsic Potential (see
response to Question 4) could provide additional insights into potential impacts of a tailings dam
failure.
RESPONSE: While we were unable to conduct a full Intrinsic Potential modeling exercise, we
have quantified the distribution of classes of gradient, mean annual flow, and % flatlandfor
reaches of the site watersheds where TSFs would be placed. As highlighted in the report,
streams — especially those downstream of the TSFs in the mainstem North Fork Koktuli and
South Fork Koktuli - are low-gradient, floodplain prone channels that currently support
spawning populations of several salmon species, rainbow trout, and Dolly Varden.
I think that potential consequences of climate change 011 h\ diouraphs should have been
considered in this section. More precipitation is projected to occur as rain in the winter rather
than snow for many parts of AK. How would this potentially impact the tailings dam facilities?
This seems to be a key piece of information lhal is needed to better understand the risk of dam
failure and the potential for impacts on aquatic resources.
RESPONSE: Climate change projections and potential impacts are described in Chapter 3,
and referenced here as a key uncertainty for in perpetuity management of the TSFs and other
mine infrastructure. The probabilities for dam failure discussed in this assessment involve
dams constructed in a variety of climates. . I ny potential increase in precipitation due to
climate change is but one of the conditions for which the dams would need to be designed. A
design developed to handle a higher level of precipitation would be expected to have a similar
failure probability to one designed for an area of lower precipitation as long as the same
design standards, e.g. the safety factor against overtopping, were used in both designs. Of
course, if the design did not consider the possibility of increased precipitation as a design
factor, and climate change did cause such an increase, then the probability of failure would be
higher.
I thought lhal results to date of the i 111 pacts of the \ olcanic eruption at Mt. St. Helens, while not
exaclly the same as a mine operation, were not useful in considering long-term impacts and the
response of aquatic ecosystems to such major disturbances. The impacts on streams are still
more prevalent and extensi\ e than what is described in the report. Most stream systems are
transporting large amounts of line sediment and areas of exposed gravels are rare.
RESPONSE: Extensive discussion of the Mount St. Helens analogy has been removed at the
prompting of several reviewer comments including this one. We have retained key references
from the region that are illustrative for considerations of likelihood offine sediment transport
and recovery (or lack thereof).
Charles Wesley Slaughter, Ph.D.
Yes. Physical consequences of TSF dam failure are fairly portrayed. I would only suggest that
effects of initial sediment deposition and long-term remobilization and redeposition would
extend beyond the spatial and temporal limits of the modeling used in the Assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 137
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Agreed. We now more clearly state that remobilization and deposition could be
extensive; potentially reaching Bristol Bay.
Employing advanced eco-hydraulic modeling tools such as MIKE-11, MIKE-SHE (DHI,
Copenhagen), and consultation with state-of-art practitioners (IAHR-International Association
for Hydraulics Research, UI Center for Ecohydraulics Research, and others), along with
improved high-resolution input data such as LIDAR survey of the complete Kvichak and
Koktuli/Nushagak systems, would allow a more complete estimate of potential hydrologic and
sedimentation (and consequently biotic) consequences of TSF dam failure for the entire river
system, headwaters to Bristol Bay.
RESPONSE: We agree that LIDAR survey data would greatly improve the understanding of
the project area topography. Alternative modeling platforms coupled with LIDAR have the
potential to improve the estimates provided in this assessment, but these data collection and
modeling efforts were not within the scope of this project.
John D. Stednick, Ph.D.
The tailings dam failure was modeled and the distance of sediment transport was estimated. The
modeled tailings dam failure used an estimate of 20% mobilization from the tailings ponds. How
was this value determined? The model was run to a stream length of 30 km (the rivers
confluence), yet the report acknowledges that a sediment pulse could run for hundreds of
kilometers. The moisture content of the tailings is estimated to lx- 45% by volume (page 4-50);
the 20% volume of sediment may he underestimated. This initial risk to salmonid fish is clear,
but the persistence of the sediment affect could be discussed.
RESPONSE: 20% was selected as a reasonable estimate, falling within the range of historic
failure release volumes (e.g... I zam and IJ 2010 state 1/5 (i.e., 20%) and Dalpatram 2011
states 20-40",,). II e agree that the total volume released during a failure would vary depending
on water content, consolidation of tailings, and meteorological conditions in the valley during
the time of the failure. We also agree that the sediment initially deposited on the valley
floodplains and the sediment that remains downstream of the failed dam would become a
continuous source of tailings, with the potential to re-suspend during each subsequent
rainfall-runoff event that occurs before any sort of mitigation/clean-up efforts could be
implemented to control this process. The 20% used was based on recent literature for the
amount that is generally released from these types of dam failures. The tailings dam failure
was intended to be a conservative analysis to shed light on whether a failure is a significant
concern in the Nushagak River watershed
The Mount St. Helens analogy is inappropriate for a variety of reasons and such comparisons
should be removed from the assessment.
RESPONSE: Use of the Mount St. Helens analogy has been removed.
The probable maximum precipitation (PMP) value was extrapolated from Miller (1963) and the
assessment commented how this value might be reduced upon further analysis. Conversely,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 138
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
additional data could increase this value. There was no discussion of the recurrence interval of
this 24-hour storm.
RESPONSE: The assessment used the published PMPfrom Technical Paper No. 47 (the
current guidance available from NOAA, at www.nws.noaa.gov/oh/hdsc/studies/pmp.html).
Within TP 47 the 24- hour PMP values presented exceed those provided for the 100-year 24-
hour event. While actual planning, design, and construction of a TSF dam will require
additional study, for the purposes of this assessment we relied on readily available information
and current published guidance for the determination of a design storm event with the
magnitude that would likely be considered during design of the dam The published PMP was
applied to a Type-1 storm within HEC-HMS to determine a flood hydrograph that would
result in the watershed located above the TSF as described for the purposes of this assessment.
This hydrograph is expected to peak at 291 cms and was applied to the hydraulic model to
cause an overtopping and subsequent failure of the TSF dam. The flood wave generated by the
dam failure had a peak flow of149,263 cms and 11.637 cms (large and small dam failures
respectively). The hydrology discussed and applied related to the dam failure represents a
precipitation and runoff event of the magnitude that the dam would be expected to
accommodate, but in this assessment was used as a mode of failure.
No hydrologic data were provided. The streamflow gauging stations operated l\\ the US
Geological Survey near the study area suggest peak si ream flow rates from snow melt and from
rain events. The hydrograph shape and magnitude help detenuine if rain or snowmelt
dominated. In the assessment, the peak flow estimate from the Natural Resources Conservation
Service runoff method used a Type la storm distribution, the least intense precipitation
distribution, but the literature would suggest that a Type I distribution would be more
appropriate for Alaska I low does this storm event compare to the measured flood at Ekwok
(page 4-50)? The curve number (CN) was not identified, nor the methods used to calculate that
value. Similarly, the watershed slope, time to peak, hydraulic length, channel routing functions,
and channel resistance methods or results were not presented. What precipitation data are
available ' The design of culverts, bridges, and storm water ponds all require good precipitation
records and the confidence in that estimate is based on record length.
RESPONSE: We agree that the Type-1 storm is appropriate for Alaska. This has been
corrected and updated. Additional hydrologic parameters were also included in the update to
better describe the development of the HEC-HMS model. It is important to note that this
hydrologic event was used to provide a mode offailure for the dam. The flood wave generated
by the dam dwarfs the PMP hydrograph and the runoff from the storm event is not relevant.
3 2
The comparison is unclear for a 3,313 m /s flow in a 2,551 km watershed area to the TSF flow
3 2
of 1,862 m /s and an area of 1.4 km . What was the precipitation and recurrence interval for the
Ekwok storm? The relation of groundwater flows to streamflow during storm events needs to be
evaluated. The flood producing precipitation events in this area no doubt add to groundwater
flows.
RESPONSE: This discussion was provided to help the reader understand the magnitude of the
dam failure flood wave relative to an event that was experienced by a local community. We
also wanted to draw attention to the fact that such a failure would occur in the upper end of
the watershed where typical storm events do not generate floods of such magnitude. We
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 139
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
acknowledge that the two watersheds are not similar and the intent was not to draw
comparisons of runoffpotential from one watershed to another, just to help the reader gain
context.
With a new estimate of precipitation depth of known recurrence interval, the design storm could
result in a higher flood event with greater velocities and greater sediment transport ability, along
with a greater sediment volume released from the TSF, resulting in a greater risk to salmonid fish
and habitats.
RESPONSE: We agree. The tailings dam failure was intended to be a conservative analysis to
shed light on whether a failure is a significant concern in the Sushagak River watershed. No
change suggested or required.
Roy A. Stein. Ph. IX
TSF Failure and Remediation. The text on pages (->-1 to 6-2 states.
"Remediation may occur following a tailings spill. In.il il is uncertain A spill
would flow into a roadless area and into streams and ii\ u s that are loo small to
float a dredge, so the proper course of remediation is not obvious."
At this juncture in time, this statement points to the fact thai we do not have the technology, or
the appropriate operating procedures, in place to remediate a TSF spill. Does this essentially let
the mine operator "olT the hook"? Should we be promulgating mining activities in locations
where we cannot remediate spills, given our current state of knowledge or ability to apply current
techniques? What guidance would "Best Management Practices" provide for this situation?
RESPONSE: The intent of the statement is to point out the challenges of remediation of a
large-scale tailings spill in a remote area. Large-scale tailings cleanup is challenging even in
urban and more developed rural areas where there is road/rail access for equipment and
transport of contaminated sediment. In the assessment watersheds, the only current access to
downstream areas is provided by the rivers themselves, which are generally too small for
large-scale dredging equipment.
William A. Stubhlefiehi Ph.D.
Potential effects on salmonid populations were evaluated due to tailings dam failures. Tailings
dam failure would potentially result in the release of large volumes of mine tailings and
associated contaminated waters, leading to possible acute and long-term effects on salmon
populations. It is also important to note that direct effects on salmon may be very species
dependent, due to life-cycle differences, and the time at which the dam failure occurs. Potential
effects due to sediment inundation/impaction can adversely affect habitat, leading to decreased
spawning. Evaluation of the potential for tailings dam failure effects considered acute and
chronic risks due to aqueous exposures, chronic risks due to sediment exposures, and risks due to
dietary exposures. All of these seem to be appropriate exposure pathways and all were
adequately considered, although site-specific information will improve risk predictions.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 140
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Agreed. No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to a potential tailings dam failure. It only estimates the likelihoods of occurrence and the
consequences. See discussion under Question 4 above regarding suggestions for improving
estimation and expression of the magnitude of risks to salmonid fish due to potential tailings dam
failure.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. In this case we have
estimated those for a tailings dam failure to the extent that existing information allows.
Although fish abundance data are limited for this region, we did identify the potential
magnitude of a tailings dam failure to the C hi nook run of the Nushagali River.
TSF failure likelihood. The failure statistics gi\en 011 p 4-45 arc based 011 tailings failure
statistics over the last 50 years or so. Was there also a re\ ieu of Ihe operational histories, and
therefore failures, of tailings impoundments designed and constructed in the last 10 to 15 years?
It is recognized that one of the failures identified in Box 4-4 (Auiul S.A. Mine, Baie Mare,
Romania) falls in this category. However, many of the failures included in the analyses are
associated with older tailings facilities, especially those associated with large releases of tailings
solids. A significanl impro\ emenl in tailings management is the implementation of an
Independent Tailings Dam Re\ ieu Board (ITRB) for large mining projects (Morgenstern, 2010).
An example of the acli\ ities of an ITRB is ui\ en in Minera Panama (2012). Morgenstern (2010)
provides a listing of tailings failures from 2<)o| and 2010 and comments that "in no case, to the
knowledge of the Writer, was there systematic third party review" of the failed facilities as
would he the case u hen an ITRB is active I expect that a tailings review board will also be used
for the IVhble Mine and the heha\ ior of a tailings management facility designed and operated
under these conditions will be more representative of the potential failure likelihoods expected
for such a facility. It is expected that this likelihood will be much lower than those used in the
evaluations of the scenario in the I-PA Assessment.
RESPONSE: The probabilities for dam failure used in the assessment were not derived solely
from the historical record. Historical failures were discussed as supporting background
information and present a defensible upper bound on the failure probabilities. The failure
probabilities used in the assessment are based on Alaska's dam classification and required
safety factors applied to the method of Silva et al. (2008) which compares dams designed,
constructed, and/or operated under different standards. The discussion of failure probabilities
in the revision (Chapter 9) is expanded to try to clarify this issue.
TSF failure consequences. It is difficult to estimate the volume of tailings that will be released
when a tailings impoundment fails. The release of 20 percent of tailings from a slurry deposited
TSF may be realistic when it contains a large pool and is subjected to a large flood, but it is
unrealistically high for a TSF containing a small or no pool (such as in the case of a filtered dry
stack). I would consider the assumption that a release of 20% of the tailings material for the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 141
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Pebble mine scenario is on the high side, even during operations. When the mine is closed and
the tailings reclaimed I would consider the 20% release assumption as unrealistic, especially if
the closure implementation included a diversion system designed for the PMF. It is further
unrealistic to assume that the released tailings will remain in the downstream channels and flood
plains following the failure. In the case of the Aznalcollar Tailings Dam failure in Spain, all the
released tailings downstream of the mine were removed. While such a removal action will
impact parts of the watershed, it will help to recover the area faster than leaving all the tailings in
place and will also reduce the longer-term impacts on downstream water quality. I therefore
disagree with the assumption on p. 6-2 that "the assessment assumes that significant amounts of
tailings would remain in the receiving watershed for some time and remediation may not occur at
all." Box 6-1 provides "background on relevant analogous tailings spill sites" and three historic
sites are used as analogs. These are not realistic analogs, as they all relate to historic mining
under completely different scenarios. While the material historically released in these streams
were from base metal mines, the circumstances of their release, especially in the case of the
Clark Fork and the Coeur D'Alene Rivers, were \ cry different I .ong-lci ni uncontrolled releases
occurred in these river systems due to regulatory ciivuinstances or historically acceptable
practices that differ significantly from those in the 21Century.
RESPONSE: We concur with the commenter that the estimation of the potential release
volume is difficult and inexact. The release of 20% of impounded tailings used in our analysis
is well below the reports of 30% to (>(>"» in the historic record. Dry stacking was not proposed
in the Ghaffari et al. (2011) report and lias not been proposed in our scenarios. Ghaffari et al.
(2011) proposed maintaining a pond on the top of the TSi' to keep the pyritic tailings
submerged, implying that the bulk of the tailings would remain saturated. We acknowledge
that there are other tailing management strategies that could reduce the potential risk of
failure.
The remediation of the I99S. iznalcdllar Tailings Dam failure in Spain was completed within
the 6 months between the April failure and the onset of heavy rains in October. The
Aznalcollar area, near Seville, has a much drier and warmer climate than the Bristol Bay
area. It is also a heavily farmed area with flatter topography andfar better access. The success
at removing the tailings from the Aznalcollar failure would be difficult to replicate in the
Bristol Bay area, and significant amounts of tailings would remain in the receiving watershed
for some time.
Box 6-1 in the original draft assessment (now Box 9-1) clearly states that the analogs
presented "provide evidence concerning the nature of exposures to aquatic biota". They are
not used to address the probability of a tailings dam failure or other aspects of the release of
tailings.
Phyllis K. Weber Scannell, Ph.D.
The assessment considers two possible failures of the tailings dam: a partial-volume failure
occurring during mine operations and a catastrophic failure occurring during or after mine
operations. The partial-volume failure (as modeled in the assessment) would result in a greater
than 1,000-fold increase in discharge and the catastrophic failure in a greater than 6,500-fold
discharge.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 142
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: No change suggested or required.
The discussion of tailings dam failures describes possible changes in channel and floodplain
morphology and briefly mentions that the tailings deposition would be a source of easily
transportable, potentially toxic material.
RESPONSE: No change suggested or required.
The potential for increased metals loadings to river and lake systems is understated. Although
there are no current predictions of tailings water quality, the water quality of tailings water from
similar mines could be used to model increases in metals loading from dam failures.
RESPONSE: The original draft and the revised assessment use tailings leachate data from the
PLPEBD (see Chapter 8). We believe that is more defensible than use of tailings leachates
from other mines.
In addition to the partial-volume failure and the catastrophic failure, there are other possible
sources of metals loadings from the tailings pond I samples are emergency releases of untreated
tailings water, seepage of tailings water into the groundwater, and flow from the tailings pond to
groundwater in an adjacent drainage as the head (i e hydrostatic pressure) is increased as the
tailings pond is filled. The last example was experienced at the Red Dog Mine when the
increased elevation of the tailings pond caused water to flow underground into the Bonns Creek
drainage instead of the Red Dog Creek drainage Interception ditches were installed after the
increases in metals loading to Bonns Creek were detected
RESPONSE: Because the number of potential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The original and revised drafts of the
assessment include "seepage of tailings miter into the groundwater, andflow from the tailings
pond to groundwater in an adjacent drainage" but not emergency releases of tailings water.
Paul Whitney. Ph.I).
Duration I agree with the assessment that it would take a "very long time" (page 6-25, first full
para, last line) to reach concentrations that would not exceed threshold exposure levels. A "very
long time" could mean hundreds of years to one person or geological time (i.e., millions of
years) to another person The assessment could be improved if some sidebars are put on the time
likely required for no risk dilution or "more normal channel and floodplain." One suggestion
would be to estimate the amount of time it would take the river/stream to move across the
floodplain in the "relatively undisturbed" Bristol Bay watershed. I would also like to know
whether reclamation or rip rap or rock weirs in areas with spilled tailings would reduce or extend
the time to reach "more normal" conditions.
RESPONSE: We agree that geomorphic analyses of channel/floodplain recovery following a
TSF failure would help improve the estimation of recovery time, but such an analysis would
be a research and development effort beyond the scope of this assessment. Analogous
examples of recovery following massive contributions of sediment to stream systems were
explored (e.g., Mount St. Helens) but were not uniformly valued by reviewers, and were
droppedfrom the assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 143
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 10. Does the assessment appropriately characterize risks to wildlife and
human cultures due to risks to fish? If not, what suggestions do you have for
improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
David A. Atkins, M.S
The assessment does a good job analyzing the importance of lisli resources to other wildlife and
to Alaska Native communities. The lack of site-speciiic information in the report results in only
very general conclusions that there 'would be some elVecl " Of course, wildlife in the project area
and any traditional use of these lands would be affected by project construction under the no-
failure scenario. However, due to the lack of information, it is unclear if this is an area rich in
other wildlife or if there are traditional native land users that rely on the area Although the
conclusion of this section is necessarily general, it would be helpful to have more detailed
characterization of wildlife and native use in the project area
RESPONSE: More information about subsistence harvest use in the area of the mine
scenarios and other areas of the watershed that may be a ffected by mining activities has been
added to the report. Because the scope of the assessment includes potential effects to Alaska
Native communities related to salmon'uls, this has been the focus of the evaluation ofpotential
effects on indigenous cultures. However, EPA recogniz.es that there are potential effects due to
loss of subsistence harvest areas for other species, as well as potential effects on non-Alaska
Natives who practice a subsistence way of life. This is clarified in the revised assessment
(particularly in Chapters 12 and 13).
As the commenter notes, there is limited information about the use of specific mining claim
areas by wildlife. The Pebble Limited Partnership Environmental Baseline Data Report
includes data on the presence/absence of wildlife species around the Pebble claim, but we are
not aware of any data on abundance of wildlife, so a more detailed characterization of a
specific mine claim area may not be possible. There is some information about subsistence use
of the mining claim areas which has been published by Alaska Department of Fish and Game.
General areas of wildlife harvests have been added as a figure to the revised assessment.
Under the failure scenario, a tailings dam failure, in particular, would be catastrophic for wildlife
and Alaska Native communities that use the area.
RESPONSE: A new Figure 5-2 illustrates that subsistence use of fish is extensive in areas
downstream of the TSF in the Mulchatna and Nushagak Rivers, and this fact is now
referenced. Effects of a TSF failure on wildlife are now discussed in Section 12.1.
Steve Buckley, M.S., CPG
No comments on this section.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 144
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Courtney Carothers, Ph.D.
Wildlife: The sections discussing risk to wildlife resulting from effects on salmonids are fairly
short. Those animals that directly feed on these fish are likely to be impacted, as well as those
that depend on other resources enhanced by the marine-derived nutrients supplied by salmon
carcasses. The report concludes that the primary aquatic contaminant is copper (5-75), but notes
that the ore processing chemicals are unknown, as are their toxicities (5-59). These unknowns
could be noted as potential contaminants.
RESPONSE: Section 8.2.2.5 now includes a discussion of ore processing chemicals and their
toxicity.
Human cultures: Overall, the main report (and Appendix I)) describes the central role that
salmon play in both Yup'ik and Dena'ina culture, both traditionally and in contemporary
communities. As noted above, the scope of the assessment focusing on these two cultural groups
should be made more clearly. Appendix E, for example, focuses on other human groups local to
this region, and those who migrate to the region lor commercial fishing and recreation, who may
also be affected by risk to fish in this region The vulnerabilities listed in Appendix D (p. 4-5)
could be listed in the main report more clearly as risks
RESPONSE: The assessment text has been expanded to identify these vulnerabilities.
Literature on the effects of contaminated or declining resources on subsistence communities
could be utilized to describe likely impacts in more detail I-'or example, the report notes: "the
actual responses of Alaska Nativ e cultures to any impacts of the mine scenario is uncertain" (ES-
26). While the sped lie responses are uncertain, likely responses can be predicted (and many are
articulated in Appendix I)) There are data on the psychological, social, cultural, and economic
disruptions caused In the l-xxon Vaklex oil spill (e.g., Braund and Kruse 2009; Palinkas et al.
1993), the cumulativ e effects of oil and gas development in the North Slope region (e.g., Braund
and Associates 2001). NRC 2<)i).i). and social impacts related to mining development in Alaska
(e.g., TetraTcch 2009; Storey and Hamilton 2004). Drawing on some of this literature could help
provide likely scenarios for impacts to Alaska Native subsistence-based communities from
decreased quality, quantity: or cli v ersity of salmonids. Current and recent responses to salmon
shortages in the Yukon-kuskou kirn region may also be helpful to include.
RESPONSE: The references provided have been reviewed and information from case studies
used where appropriate to provide data on responses of Alaska Native communities to
disruptions in subsistence resources (see Chapter 12).
Clearly the impacts to subsistence are not just lost food sources, but loss of healthy subsistence
lifeways, loss of practices, loss of cultural connections to the past, loss of connection to specific
places, loss of teaching and learning, loss of sharing networks, loss of individual, community,
and cultural identity, among others as detailed in Appendix D. This point could be made more
forcefully. As noted above and detailed in the specific comments below, subsistence is framed at
times in the report as primarily important for physical health and economic necessity. The
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 145
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
cultural, social, psychological, and spiritual aspects of subsistence livelihoods should also be
consistently highlighted.
RESPONSE: The assessment has been expanded to bring forward more information on the
psychological, cultural, social, and spiritual connections between Alaska Native cultures and
fish.
As discussed Appendix D, Alaska Native cultures in this region and other regions in the state are
also dependent upon the cash economy, both for subsistence production and for other needs. The
role of commercial salmon fishing or other wage engagements related to salmon in the study
communities, while discussed in Appendix E, is not gi\ en much discussion in the main report.
How dependent is the subsistence economy upon commercial and recreational fisheries and in
this region?
RESPONSE: Additional information about the connections between the subsistence way of
life and the commercial and recreational fisheries has been added to the text of the revised
assessment.
There is a brief mention of non-fish related impacts lo Alaska Xative communities in the main
report (5-77). Unless a full treatment of these impacts (posiii\ e and negative) is included, these
paragraphs should be removed. While in general. I am sn pptxli \ e of an increased scope (i.e., it is
incredibly difficult to isolate only salmon-mediated impacts to Alaska Native communities),
these other economic, social, and cultural impacts are not presented fully in the analysis, nor was
the ethnographic research designed to in\ estiuate these impacts, so passing mention of them here
does not seem appropriate
RESPONSE: The text of the assessment has been clarified to indicate the scope of the
assessment does not include an evaluation of direct effects on indigenous culture or human
welfare, because it is limited to aquatic pathways and impacts. However, the fact that there
would be direct effects is acknowledged in Chapter 12. consistent with comments from peer
reviewers and the public.
Dennis I). Dauble, Ph.D.
There is considerable detailed information in Appendices D and E relating to impacts of the
project to the economy This information includes how salmon affect all segments of the
population, such as cultural resources of Native Peoples. However, not addressed in detail were
long-term impacts to Nali\ e Peoples that might occur after losing a way of life that includes
salmon. The description of potential impacts to their health and welfare should be expanded.
There are numerous examples of how Columbia River tribes have been negatively impacted due
to loss of fish resources (and fishing as a lifestyle) as a result of dam construction. These
impacts go beyond simple economics.
RESPONSE: EPA recognizes that there is a great deal of information about cultural and
health effects on indigenous populations from loss offish resources. We have referenced some
case studies in the assessment (with a primary focus on Alaska), and have expanded the
discussion ofpotential effects from a loss of salmon resources in Chapter 12.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 146
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The report should include a discussion of effects specific to unique user groups. That is, some
communities rely almost solely on sockeye; kings are more important to others. These impacts
could be segregated by watershed, for example. Also, some groups have more option for
subsistence gathering if sockeye and Chinook salmon resources are impacted. Potential impacts
of a declining salmon population due to mining operations would be less for them than groups
"on the edge" who currently rely mainly on salmon.
RESPONSE: The text of the revised assessment (Section 5.4.2.2) has been expanded to
acknowledge the differences between communities with regard to use and reliance on salmon
and other fish resources.
Disturbance of wildlife from noise and roadways should be included with respect to migration
corridors and critical habitat. Highlight species most likely ill risk Irom human disturbance,
habitat loss/displacement (from the project footprint), and loss of salmon. For instance, do some
piscivorous species have the ability to shift their diet to include another source of protein? If so,
how would this shift affect the human culture with reference to a subsistence lifestyle?
RESPONSE: EPA recognizes that there are potential direct effects on wildlife from large-
scale mining operations, including disturbances from noise and roadways; however direct
effects on wildlife are outside of the scope of the assessment, as explained in Chapter 2. We
would expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
direct effects on wildlife.
Gordon H. Reeves. Ph. I).
These topics are outside my area of expertise so I cannot comment on if the report adequately
characterizes the risk to wildlife and human cultures 1 am familiar with some of the literature on
the importance of salmon to wildlife and the report represents the finding fairly and accurately. I
cannot pro\ ide any additional literature, reports, or data.
RESPONSE: No change suggested or required
Charles Wesley Slaughter, Ph.D.
No. The Assessment clearly qualified that its objective was to consider risks to salmonids, and
only inferentially consider "salmon-mediated" effects.
RESPONSE: No change suggested or required
Appendix C provides a comprehensive discussion of non-fish wildlife and the relation of those
populations to salmon. However, the Assessment itself (Volume 1) provides only a brief
summary in Chapter 2.2.3, which could allow a cursory reader to perhaps conclude that wildlife
populations have little risk of impact from the hypothetical Pebble project. Is this the intent of
the Assessment authors? A more in-depth reading of Appendix C allows inferring potential
consequences to wildlife and birds of "salmon-mediated" impacts of mining development.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 147
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: We have clarified assessment endpoints and provided additional background
information from Appendices A and B in Chapter 5. Direct effects on wildlife are outside the
scope of the assessment (as stated in Chapter 2), but their potential importance is
acknowledged in Chapter 12.
The assessment has been expanded to include more information from Appendix C of the draft
report (now an independent U. S. Fish and Wildlife Report).
The scope of the assessment is focused on potential risks to salmon from large-scale mining
and salmon-mediated effects to indigenous culture and wildlife. EPA recognizes the
complexity of potential direct, secondary, and cumulative effects on wildlife. We would expect
that a full evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
complexities and utilize the information in U.S. Fish and Wildlife report (Appendix C of the
draft report.)
USFWS RESPONSE: We acknowledge the comment about the comprehensive discussion of
wildlife and the relation of selected species to salmon in Appendix C. Appendix C could be
used as the basis of an assessment ofpotential impacts of mining on wildlife, but decisions
related to assessment scope with regard to potential impacts of mining on wildlife is the
responsibility of EPA.
Appendix D provides a comprehensi\ e and useful discussion of the indigenous people of the
Bristol Bay region and of their traditional ecological knowledge and cultures. Appendix D
clearly lays out the vulnerabilities and risks (summarized on p. 4-5 and pp. 150-153) associated
with the (hypothetical) major resource extraction projects. However, the Assessment (Volume 1)
provides only cursory consideration of these human aspects of the potential project - on p. ES-
23, and in Section 2.2.5. Presumably, this is because the EPA mandate is to conduct an
ecological risk assessment, rather than assessment of consequences for human populations,
whether indigenous. nati\e. resident, non-native, non-resident, or the larger cash economy world
as represented In the State of Alaska, Northern Dynasty Minerals, or Pebble Limited
Partnership
RESPONSE: The scope of the assessment related to Alaska Native cultures is limited to
salmon-mediated effects—that is. potential effects on indigenous culture if there are negative
effects on salmon. The assessment draws on the more comprehensive information in Appendix
D to evaluate these risks. The text of the revised assessment has been expanded to include and
discuss the vulnerabilities and risks to Alaska Native culture from potential large-scale
mining.
John D. Stednick, Ph.D.
The effect on wildlife section largely focuses on the return of nutrients to the land in various
shapes and forms and adds little to the risk discussion. Other than marine-derived nutrients,
other stressors exist. What are the consequences of the mine operation on other wildlife habitats?
Habitat fragmentation? Noise and light disruptions, etc.?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 148
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The scope of the assessment is focused on potential risks to salmon from large-
scale mining and salmon-mediated effects to indigenous culture and wildlife. The EPA
recognizes the complexity of potential direct, secondary, and cumulative effects on wildlife and
would expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
complexities. The final assessment acknowledges the potential for direct effects on wildlife
(Chapter 12).
Mine development will require the use of explosives. What are the effects of the nitrate and
ammonia in the air following each detonation? The National Trends Network data suggest that
the area receives about 1 kg/ha/yr of nitrogen in precipitation Thus, the increase in atmospheric
inputs from the explosions may exceed the marine-deri\ ed nutrients. What are the
consequences?
RESPONSE: See response to previous comment.. Mso. stressors evaluated were based on how
they would significantly affect our primary endpoint of interest (the region's salmon
resources) and their relevance to EPA's regulatory authority and decision-making context.
Although atmospheric inputs from mining operations are a potential source of contamination
to streams, they are not considered as significant as other stressors and are not regulated
under the Clean Water Act. Finally, because of terrestrial nitrogen fixation, nitrogen is not
believed to be the most important marine derived nutrient.
The potential loss or change in lifestyles of indigenous peoples is important, but this information
seems relegated to Appendix D. Include more of this information in the main body of the text.
Actually, there is a significant amount of information in the appendices that should be brought
forward.
It is unclear why there is such variability in the detail or depth of assessment of each of the
stressors Why does the TSF failure section ha\ e 34 pages while the section on potential effects
on nati \ e peoples only has a few pages with a reference to an appendix? The unevenness of the
coverage needs to be addressed.
RESPONSE: The text of the revised assessment has been expanded to incorporate more of the
background information from . \ppendix D and a more comprehensive evaluation ofpotential
salmon-mediated risks to indigenous culture. The assessment's primary endpoint is the
abundance and production of salmonid fish, so effects on those receptors are treated in more
depth than the secondary endpoints, fish-mediated effects on wildlife and Alaska Native
cultures.
Roy A. Stein. Ph. I).
Risks to Wildlife. The importance of salmon in bringing Marine Derived Nutrients (MDN) into
these freshwater ecosystems and watersheds and their role in influencing wildlife and associated
interactions between wildlife and human cultures was well described. Loss of these nutrients
would severely compromise wildlife, and thereby human (through reductions of subsistence
harvest), populations. The appendix dealing with wildlife was quite detailed and well done and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 149
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
it would serve the main report well for the authors to include critical information from this
appendix.
RESPONSE: No change suggested or required.
USFWS RESPONSE: We acknowledge the comment regarding the quality of Appendix C.
Comments related to the scope of the assessment are not specific to Appendix C and are the
responsibility of EPA.
Risks to Native Alaskan Culture I. I thought that the assessment could be approved if the same
approach used for the mine, i.e., a case history approach, was used for human cultures. Surely,
there exist situations where salmon have declined or have been reduced by
development/exploitation (the Fraser River, perhaps?) where subsistence by Native Alaskans
was historically paramount. Once the salmon were reduced, w luil was the impact on the Native
Alaskans subsistence culture? How did the Nati\ e Peoples respond ' from whence did they get
sustenance, cash, etc.? What sort of displacement occurred?
RESPONSE: The EPA recognizes that there is a great deal of information about cultural and
health effects on indigenous populations from loss offish resources. If e ha ye referenced some
case studies in the assessment (with a primary focus on Alaska), and have expanded the
discussion ofpotential effects from a loss of salmon resources.
Risks to Native Alaskan Culture II. The only risks reviewed here came via salmon-mediated
impacts on the culture of Native Alaskans. Direct impacts of llie mine through jobs (potentially
positive), wildlife (likek negative), etc. all could be discussed briefly, serving to broaden the
overall impact of the mine and its associated activities on Native Alaskans.
RESPONSE: The HP. I acknowledges that the scope of the assessment is limited to salmon-
mediated risks to indigenous culture and that there would be many other potential effects
(positive and negative) from development of large-scale mining in the region. Direct impacts to
Alaska \ative culture have been acknowledged in the revised report, but are outside of the
scope of the assessment.
Risks to Other Cultures. I was a little surprised that little text was spent on recreational
anglers, commercial fishers, subsistence users (other than Native Alaskans), etc. Appendices
provide some guidance here and this text need not be voluminous, but mentioning these impacts
of the mine on Ihese groups would improve the main report.
RESPONSE: The discussion in the revised assessment has been expanded to acknowledge
potential effects on non-. Maska Native subsistence users. The EPA acknowledges that there
are potential impacts on recreational anglers and commercial fishers. However, because the
assessment focused on fish-related effects to Alaska Native culture, an evaluation ofpotential
impacts to the recreational and commercial sectors is outside of the scope of this assessment.
William A. Stubblefield. Ph.D.
Potential effects to wildlife and human cultures are briefly addressed in the risk assessment. No
"quantitative" assessment of potential effects is provided. For the most part, it appears that
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 150
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
potential affects to both wildlife and human cultural endpoints are directly proportional to the
injury suffered by salmon populations as a result of any spills or failures. Given the level of
detail available at this point in time regarding mine operations and closure, that is probably about
as far as any assessment could go. I'm not aware of any literature reports or data that would
assist in further characterization of these potential injuries.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to wildlife and
human cultures due to risks to fish. It only estimates the likelihoods of occurrence and the
consequences. See discussion under Question 4 abo\ e regarding suggestions for improving
estimation and expression of the magnitude of risks to wildlife and human cultures due to risks to
fish.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information alhnvs.
Phyllis K. Weber Scannell, Ph.D.
The document focuses 011 effects to wildlife that would occur from failures - tailings dam failure,
pipeline failure, etc There are other sources of disturbance to wildlife that should be addressed
in a future mine plan and agency re\ ieu and permilli ng. Other mines in Alaska limit truck
traffic 011 the haul road during caribou migrations, incinerate all kitchen waste, educate workers
on bear safety, and prohibit inappropriate disposals of food containers or other wildlife
attractants Other factors that might need to be addressed to protect wildlife are limiting air
traffic and noise during certain limes of the year. Unless addressed, these issues are more likely
to cause detrimental effects to wildlife than dam or pipeline failures.
RESPONSE: So change suggested or required. The EPA agrees with the reviewer that any
future mine plan would require an evaluation of and mitigation for, direct effects on wildlife.
Paul Whitney. Ph.D.
Problem Formulation for Wildlife. If one addresses the problem formulation for wildlife at face
value, the answer is pretty straight forward. The assessment tells us that the consequences of loss
and degradation of habitat on fish populations could not be quantified because of the lack of
quantitative information concerning salmon, char, and trout populations (page ES-26, third
bullet). Furthermore, we learn that indirect effects, such as risks to wildlife, cannot be quantified
(page 5-75, para 1, last line). Stating that reduced salmon production would reduce the
abundance and production of wildlife (page 5-75, para 1, last sentence) is accurate but not
appropriate for a document that is intended to provide a scientific and technical foundation for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 151
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
future decision making (page ES-1, para 1, last sentence). It is certainly possible to provide a
more complete scientific and technical response to a question regarding impact to wildlife. I
respectfully suggest that the first step in developing a more complete scientific and technical
response is to modify Question 10 by deleting the words "due to risks to fish" and to separate the
risk analysis for wildlife from the risk analysis for human culture. A revised wildlife question
should address both indirect and direct risks. A revised human culture question should address
(among many other risks) both direct and indirect risks to fish and wildlife.
RESPONSE: As the reviewer notes, the scope of the assessment is focused on potential risks to
salmon from large-scale mining and salmon-mediated effects to indigenous culture and
wildlife. The EPA agrees with the reviewer that direct effects on wildlife are likely to be
important. We would expect that a full evaluation of any future mining permit applications
and subsequent National Environmental Policy Act Environmental Impact Statements would
consider these direct effects. The revised assessment acknowledges the potential for direct
effects on wildlife as well as risks due to fish.
Mixed Messages. The assessment and questions ui\ en to the review committee give mixed
messages regarding the scope of work. A variety of statements and conclusions in the assessment
are inter-related, as one would expect for a document that addresses ecosystem issues. While
acknowledging the overlapping issues and responses. I offer the following discrete points that
give me mixed messages regarding the scope of work and an appropriate characterization of risks
to wildlife. Since these points are related, there is some repetition of information but each of the
following items address a discrete issue
1. The PREFACI- (para 2) clearly states "Our goals in conducting this assessment are to
complete an olijecti\ e assessment of the potential impacts of large scale mining on
aquatic resources in the Bristol Bay watershed." This statement is subject to wide
interpretation In this re\ iewer and by the document itself. For example, aquatic resources
include many species of wildlife and the Cederholm papers indicate that there are 137
species of wildlife that are associated or closely associated with fish. Wildlife species and
their associations are aquatic resources but have been glossed over in the assessment. The
lack of follow-through in implementing the above goal statement (i.e., not addressing
aquatic wildlife and wildlife associated with fish) is problematic. One can readily
understand that timing and budget constraints result in some statements indicating that
data are just not available or that certain types of modeling, while possible, were not
conducted. I lowe\ er. it is difficult to understand why potential impacts to wildlife are
limited to indirect effects due to loss of fish if the assessment was truly conducted as an
assessment of aquatic resources. The message is mixed because the goal is to assess
aquatic resources, but many aquatic resources are not assessed.
RESPONSE: The assessment has been revised and reorganized to better explain the
goals and objectives of the assessment. As stated in previous responses to comments,
the focus of the wildlife assessment is limited to salmon-mediated effects.
2. The assessment was supposedly conducted as an ecological risk assessment (Abstract,
line 5). One of the first steps of an ecological risk assessment is to state well defined
endpoints. Suter (1993, page 22) presents 5 criteria that any endpoint should satisfy. My
notes are in brackets.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 152
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
a. Societal relevance (Assessment appears to do a good job of identifying
concerns of people living near the area of the proposed mine. Yet, many of the
potential endpoints listed in site models (e.g., Figure 3-2E) are not addressed.
The reason for not assessing endpoints in the site models (such as wildlife
quality and quantity) should be provided. Better yet, clearly state both direct
and indirect endpoints for wildlife.)
b. Biological relevance (The biological - especially ecological - relevance is
incomplete. Wildlife and the functions they provide are relevant in that
wildlife comprise many of the secondary and tertiary consumer species in the
ecosystem. The upland and aquatic habitats utilized by wildlife are the indirect
and direct sources of many nutrient and energy inputs to fish in the Bristol
Bay ecosystem. The upland components of the Bristol Bay ecosystem should
be assessed in more detail to provide a biologically relevant assessment. If
not, the assessment should directly indicate why not).
c. Unambiguous operational definition (Overall a good job for a hypothetical
mine. See response to Question 2 for suggestions)
d. Accessibility to prediction and measurement (The inability to estimate the
impact on salmon numbers is problematic Is the inability due to a lack of
population data, the unwillingness to utilize peer-reviewed methods other than
demographic population modeling, or some oilier reason? Furthermore, the
potential reduction in marine-derived nutrients and the potential impact on
wildlife can be assessed and predicted - see below).
e. Susceptibility to the hazardous agent (The assessment does a good job on the
hazards lor aquatic resources but not susceptibility to wildlife biotic (e.g., loss
of habitat) and abiotic stressors (e.g., ice dams and scouring, as well as
hydrographs and sedimentation influence on ecological succession).
Furthermore, an adequate ecological risk assessment should address ALL
ecological stressors (e.g., loss of habitat and disturbance) not just toxics).
Suter's fi\ e criteria should be addressed when selecting endpoints. All the endpoints
selected and illustrated in figures should be assessed. The incomplete assessment of
quantitative estimates of all the impacts on salmon, non-salmonid fish, wildlife, as
well as community and ecosystem parameters should be remedied.
RESPO.WSE: The criteria above apply to the specific assessment endpoints considered
in the assessment, as described in Chapter 5. Each of these endpoints meets the above
criteria, and additional information describing endpoints within the ecological risk
assessment process has been added to the beginning of Chapter 5.
3. A complete assessment of aquatic resources as stated in the PREFACE should address
both direct and indirect risks to the many species of wildlife and wildlife habitat that are
aquatic resources and closely associated to aquatic resources. The list of all the aquatic
species and habitat is lengthy. A partial list of categories of wildlife and wildlife habitats
that are aquatic resources includes grebes, ducks, shorebirds, beaver, muskrat, otter,
mink, and riparian, emergent, and aquatic vegetation. The impact of mining on all of the
aquatic resources is given scant coverage in the assessment, with the possible exception
of impact to wetlands. The PREFACE provides wide ranging goals but the assessment
has many gaps, and I wonder why.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 153
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The latest draft of the assessment has been revised and reorganized to
better explain the goals and objectives of the assessment. Those goals and objectives
focus the assessment on salmonid fish and a limited qualitative discussion of the
salmon-mediated effects on wildlife and Alaska Native cultures. The EPA
acknowledges and agrees with the reviewer that there will be direct and indirect effects
of mining on wildlife, but the detailed evaluation of those effects will be left to the
NEPA and permitting processes should a mine be proposed.
4. The Executive Summary (para 1) indicates . USEPA launched this assessment to
determine the significance of Bristol Bay's ecological resources and evaluate the
potential impacts of large-scale mining on these resources "" The emphasis on salmonids
does not fairly represent the entirety of ecological resources To me, ecological resources
should include ecological parameters such as plant community succession, species
diversity, energy flow, and structure and function, in addition to the information on
salmonids.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and the
specific endpoints considered in the assessment are described in Chapter 5. We
recognize throughout the text that other ecosystem components are important and
potentially will be affected by large-scale mining, but fall outside this assessment's
scope. Ecological risk assessments focus on endpoints that are important to decision
makers and stakeholders, are susceptible to the stressors, and meet other criteria; they
do not attempt to cover all ecosystem components (Sitter 1993, EPA 1998).
5. The Scope of the assessment (ES-, 1 last para) indicates that "wildlife ... as affected by
changes in the fisheries are additional endpoints of the assessment." This informs the
reader that only wildlife species that are affected by changes in the fisheries will be an
endpoint Yet many species of wildlife that are affected by changes in the fisheries are
not addressed In addition, such an endpoint glosses over the importance of assessing fish
that are affected In changes in wildlife. I or example, beaver modify aquatic habitat and
terns prey on fish The scope of the assessment seems to brush over many of the
important fish and wildlife interactions discussed by Cederholm et al. (2000) and
Cederholm et: al. (2<)i)|) and to focus mainly on marine-derived nutrients. I am pleased
with the discussion of the marine-derived nutrients but am not clear how the emphasis on
this one interaction, which is not quantified in the assessment, furthers the understanding
of potential mining impact on the Bristol Bay ecosystem.
RESPONSE: The HP. I agrees with the reviewer that there are many complex fish and
wildlife interactions in the study area and the revised assessment acknowledges this
fact. The assessment focused on loss of MDN to the system and on potential loss offish
as a food source for wildlife as important potential salmon-mediated effects on wildlife.
We would expect that a full evaluation of any future mining permit applications and
subsequent National Environmental Policy Act Environmental Impact Statements
would include analysis of other potential causes of effects on wildlife.
6. Gende et al. (2004) use a variety of methods to quantify both the marine-derived nutrients
and energy that transfer from salmon to terrestrial wildlife and habitats. For example,
they determined that bears moved nearly 50% of the salmon-derived nutrients and energy
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 154
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
from streams by capturing salmon and dragging the carcasses from the stream. Gende has
also worked with University of Alaska Associate Professor Mark Wipfli (Gende et al.
2002) on the relationship of salmon to terrestrial habitats. Alaskan and national experts
are likely available to assist EPA in quantifying the movement of marine-derived
nutrients to the terrestrial ecosystem.
RESPONSE: The references are appreciated. The EPA acknowledges that a more
quantitative evaluation of salmon-derived nutrients to the terrestrial ecosystem may be
possible. However, this type of evaluation would require further data collection and
significant analysis, which is beyond the scope of this assessment. A detailed
evaluation of the movement of marine-derived nutrients to the terrestrial ecosystem is
not needed to assess the impacts of mining on sahnonids, and is more detailed than is
needed for a qualitative discussion of the salmon-mediated effects on wildlife and
Alaska Native culture. We would expect that a full evaluation of any future mining
permit applications and subsequent National Environmental Policy Act Environmental
Impact Statements would address potential effects to salmon-derived nutrients.
7. As mentioned above, the assessment docs a line job of emphasizing the importance of
marine-derived nutrients transported 1o the terrestrial environment hul lias relatively little
information on the importance of terrestrial-derived allochthonous nutrients transported
to the aquatic environment. Doucett et al. (2007) discuss methods for measuring
terrestrial subsidies to aquatic food webs using stable isotopes of hydrogen. The potential
loss of terrestrial subsidies due to mining might be as great as a potential reduction of
marine-deri\ cd nutrients It would be informative to discuss the relative importance of
marine-deri\ ed. autochthonous, and allochthonous nutrients. Such information might
influence "best mining practices" and reclamation that can partially compensate for lost
allochthonous inputs
RESPONSE: See response to previous comment.
8. I'igurc .1-21- indicates wildlife quality, quantity or genetic diversity, as well as wildlife
prcdation are important to Alaska Native Cultures. Tribal elders are said to have concern
for "potential direct effects on other subsistent resources" (which includes wildlife and
vegetation) In addition, Appendix D comments on concerns regarding wildlife (e.g.,
caribou) and vegetation (e.g., berry gathering) that are not directly linked to the fisheries.
An assessment of the direct impact of potential mining and the ongoing exploration on
wildlife and vegetation would address the concerns mentioned above. Once again, I
wonder why the site model figures outline so much detail that is not addressed in the
assessment text.
RESPONSE: The conceptual models were drafted to gain an understanding of all
potential sources, stressors, and effects from large-scale mining and have been revised
to clarify the scope of the assessment. The EPA focused the assessment on potential
risks to salmonids and subsequent salmon-mediated effects on indigenous culture and
wildlife. There are many other potential risk pathways which are outside the scope of
this assessment, including potential loss of plant and wildlife subsistence and other
cultural resources (see Figure 2-1).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 155
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The conceptual models have been modified to more accurately reflect issues within the
assessment's scope, and to more clearly identify those pathways considered outside of
scope when they are included.
9. Comments in the main assessment Volume 1 (page 5-76, first lines) indicate that
information on wildlife and potential direct impacts are being collected but are not
included. Including measures of direct wildlife impact in addition to indirect impact as a
result of changes in fisheries would greatly improve the assessment. There are many
ways to estimate the impact of habitat loss on wildlife (Morrison et al. 1998). A better
understanding of impacts on wildlife would presumably provide a better technical basis
for designing a reclamation plan.
RESPONSE: The scope of the assessment is focused on potential risks to salmon from
large-scale mining and salmon-mediated effects to indigenous culture and wildlife.
EPA agrees with the reviewer that direct effects on wildlife are important and that
there are ways to estimate effects related to habitat loss. II e would expect that a full
evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
direct effects. The revised assessment acknowledges the potential for direct effects on
wildlife. The assessment does not include a reclamation plan and is not intended to
support the development of a reclamation plan.
10. The assessment (page 5-77, last sentence) stales "Although this assessment is focused on
salmon, the non-salmon-related impacts on nali\ e ail lures from routine mine operation
are likely to lx- more significant..If this is in lad the case, what better reason could
there be for increasing the scope to include wildlife, vegetation, community, and
ecosystem structure and function (i.e., non-salmon)? Methods and mine-related examples
for assessing such an increased scope are available from the Northwest Habitat Institute
(nulii org) in Corvallis Discussions with Tom O'Neil, Executive Director, indicate that
a\ ailable data from the Alaska GAP Analysis Project and the Alaska Natural Heritage
Program could likely be used lo make such assessments.
RESPONSE: See response to previous comment. This type of evaluation would be
valuable but would require further data collection and significant analysis, which is
beyond the scope of this assessment.
11. The risk assessment's locus on "indirect effects on wildlife" (page 5-1, para 2) is
consistent with some of the several goal statements that lean toward fish influence on
wildlife. I have difficulty reconciling this emphasis because further on in the text (page 5-
75, para 1, last line) it is stated that the "indirect effects cannot be quantified." If the
indirect effects cannot be quantified, one should think it is even more important to get a
handle on the direct effects, which can be assessed by any one of several wildlife
methodologies (Morrison et al. 1998).
RESPONSE: Wildlife is not a main focus of this assessment. We agree that a more
detailed assessment of direct and indirect impacts of mining to wildlife will have to be
done as part of the NEPA and permitting processes. This assessment is focused on
salmon because of the world class, outstanding salmon fisheries in Bristol Bay.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 156
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Further, as a keystone species, salmon serve as an excellent indicator of overall
impacts to the ecosystem
12. The assessment indicates the exceptional quality of the fish populations and their
importance to the region's wildlife is due to five key characteristics. The fourth
characteristic is "the increased ecosystem productivity associated with anadromous
salmon runs" (page 2-20, Section 2.3). Similar statements are made by Woolington
(2009). How is this increase measured, and what is the baseline for the increase? I am
sure the biologists that wrote about the increase are trying to convey the concept that the
energy and marine-derived nutrients provided by salmonids is incorporated into the
primary and secondary production in the terrestrial ecosystems in the Bristol Bay
watersheds. I am not sure that such a statement considers the loss of energy and marine-
derived nutrients as a result of commercial fishing. Is il possible that commercial harvest
may decrease ecosystem productivity compared to producti\ ily prior to pre-European
(e.g., 200 hundred years ago) harvests ' Perhaps it would be more accurate to delete the
word "increase." On the surface, this might seem like a small edit but this edit is one of
many edits needed to address, not only the existing conditions in the watersheds, but also
the existing conditions relative to "pre-Buropean conditions " My understanding of a
Cumulative Impact Analysis is that it includes past assessments of ecosystem resources
and functions as well as current and future conditions (see discussion of Cumulative
Impacts in response to Question I I)
RESPONSE: The purpose of the assessment is to evaluate potential effects of large-
scale mining on the fishery, not the effects of commercial fishing. Therefore, the
potential effects on wildlife associated with loss of marine-derived nutrients considers
the present condition of the watershed and fishery as a baseline for an evaluation of
potential future effects of large-scale mining.
13 Pauly¦ et al (2<><><>)and I.ibralato el al (2<)i)S) address the energy impacts of fisheries. I
would be interested to know if the impact of fisheries on marine-derived nutrients and
energy available to the wildlife and terrestrial ecosystems in the Bristol Bay watersheds
could be compared to the potential impact of the example mine on marine-derived
nutrients and energy a\ ailable to the watersheds. Is the potential impact of the mine small
in relation to the impact of commercial fishing or very large compared to commercial
fishing? Such information would be informative for determining acceptability of risks. It
is possible that such a comparison could stimulate discussion of the possibility of
compensatory mitigation for losses due to mining.
RESPONSE: Compensatory mitigation for losses of aquatic resources from mining is
governed by a regulatory process outside the scope of this project. Nevertheless, in
response to public and peer comments we have included a discussion of compensatory
mitigation in Appendix J of the revised assessment.
14. The assessment could be greatly improved if more of the linkages and pathways
illustrated in the various Conceptual Models were addressed and if impacts on ecological
parameters, such as community succession (down gradient, in the lake and on tailings and
waste rock) and aquatic and upland structure and function, were addressed. For example,
Site Reclamation is illustrated and highlighted as orange boxes in three places on Figure
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 157
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
3-2C (I'm not sure what the orange highlighting signifies, for there is no legend on this
figure). Information on possible wildlife limiting factors (e.g., calving and nesting
habitat) and plant communities in the watershed could improve the quality of Site
Reclamation. A Reclamation Plan needs to address the likelihood that reclamation can be
implemented and if so what benefits it might accrue through time. This type of
information would likely result in a better risk assessment.
RESPONSE: EPA agrees that more detailed information on wildlife limiting factors
would need to be collected and evaluated as a part of a site reclamation plan. We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would include
this information as a basis for a site reclamation plan.
15. The discussion of wildlife species in Appendix (' is \ cry good but little of the insight
provided in these descriptions of hunted or Happed species is reflected in the assessment
of impact of potential mining practices 011 wildlife. I cannot help but wonder (i.e., mixed
message) why Appendix C is so thick and so little of the good information in Appendix C
is reflected in the assessment document I or example, the impact of noise and human
presence related to mining and roads is addressed 011 page 54 (Appendix C). Such an
impact on certain sensitive species could be equal to or greater than the loss of wildlife
habitat in the mine footprint. Woolington (20<)l>) indicates that the Red Dog Mine in
Alaska has implemented certain measures that ha\ c reduced the impact of the haul road
on wildlife. It would be good to know what these measures are and if such measures
would be equally effective for a road relati\ ely close to Anchorage.
RESPONSE: The purpose of the appendices r\. the main assessment document has
been clarified in Chapter 1. The former \ppendix C has now been published as a
separate US Fish and Wildlife Service document because it serves purposes for the
I'SFIVS beyond the scope of the EP. I \ assessment.
I SI'IIS RESPONSE: We acknowledge the comment about the quality of Appendix C.
C 'onunents related to the scope of the watershed assessment are not specific to
Appendix C and are the responsibility of EPA.
Will the proposed haul road to the mine be closed to the public and hunting? If so, the
impact of the road may not be high. I can remember Val Geist saying that wildlife behave
the way one leaches them to behave. Examples are deer in the streets of Banff Park
during the day time and big horn sheep and elk foraging along the Trans Canada Highway
in the Park. The big horn sheep often stop traffic as people stop to observe them. Outside
the Park, where deer and elk are hunted, one is less likely to see deer and elk along the
highway and deer take on a more nocturnal existence.
RESPONSE: For purposes of the assessment, EPA assumes that the road would be
closed to the public during mining operations but potentially could become a public
road after mining operations cease. About 80% of the transportation corridor is on
private land owned by various Alaska Native Village Corporations, with which the
Pebble Limited Partnership has existing commercial partnerships. Alaska Native
Regional Corporations are charged with managing land and resources within their
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 158
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
control in the best interests of their shareholders. Chapter 13 includes a discussion on
the potential impacts of increased access due to additional roads.
The scope of the assessment is focused on potential risks to salmon from large-scale
mining and salmon-mediated effects to indigenous culture and wildlife. EPA has not
evaluated direct effects from a transportation corridor on wildlife. We would expect
that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider
the direct effects of the transportation corridor on wildlife.
16. As an arctic and sub arctic small-mammal wonk. I anticipate that any discussion of
Alaskan ecosystems will include some discussion of the less charismatic fauna (small
mammals and songbirds) and the functions they |no\ ide Such discussion is given little
attention in the assessment.
RESPONSE: These issues are outside the scope of the assessment, as specified in
Chapter 2. Appendix C (now an independent US Fish and If ildlife report) selected key
wildlife species for characterization, based on their direct dependence on salmon or on
their roles in distributing marine derived nutrients through the ecosystem It was not
possible to fully characterize every species. The EP. I recognizes that there are multiple
species that were not fully characterized and that have important functions in the
terrestrial and aquatic ecosystem. II e would expect that a full evaluation of any future
mining permit applications and subsequent National Environmental Policy Act
Environmental Impact Statements would consider all species present in these
watersheds.
USFWS RESPONSE: The I SI II S selected key species to include in Appendix C and
the selection method is described in the report. We certainly concur that small
mammals are important from an ecosystem perspective, but they are less linked to fish
or to public concerns than the selected species. Songbirds were included in the wildlife
report under landbirds.
Question 11. Does the assessment appropriately describe the potential for cumulative
risks from multiple mines? If not, what suggestions do you have for improving this part
of the assessment?
David A. Atkins. M.S
According to the Assessment, cumulative risks result from the potential development of at least
five additional prospects: Humble, Big Chunk, Groundhog, Sill, and 38 Zone. Exploiting these
prospects would amount to development of a mining district (see discussion for Question 2 in
regards to appropriateness of the mining scenario).
RESPONSE: No change suggested or required.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 159
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The Assessment quantifies the loss of stream lengths and wetland areas that potentially support
salmon and resident fish from the development of these projects under a 'no-failure' scenario.
The assessment is highly speculative given that mine development plans are not available for
these prospects.
RESPONSE: Cumulative impacts assessments evaluate past, present and reasonably
foreseeable future actions that are temporally and spatially linked to the project under
consideration. The potential mines in the revised assessment are reasonably foreseeable based
on State of Alaska planning documents and industry exploration activities. We have expanded
the discussion of cumulative impacts in Chapter 13.
As with the Pebble scenario, it would be helpful to put this loss of resource in perspective in
terms of the fish resources as a whole. It would also be helpful lo describe any mitigation
measures that are feasible to offset the impact of loss of streams and wetlands. Furthermore, it
would be helpful to better understand the role these developments could have in further
fragmenting salmon populations.
RESPONSE: Due to lack of comprehensive estimates of limiting factors across the impacted
watersheds, population level effects could not be quantitatively estimated except for the most
severe cases where total losses of runs could be reasonably assumed Our ability to estimate
population level effects was limited to situations that were assumed to completely eliminate
habitat productivity and capacity in an entire watershed for which estimates of escapement
could be inferred. For this assessment, these conditions are only met in the TSF failure
scenario that completely eliminates and blocks access to suitable habitat in the North Fork
Koktuli River. In that case. we estimate that the entire Koktuli portion of the run (-28% of
Nushagak escapement) could be lost. Higher proportional losses would occur if significant
downstream effects occurred due to transport of toxic tailings fines beyond the Koktuli as
modeled under the full TSF failure.
Compensatory mitigation is governed by a regulatory process outside the scope of this
assessment but will be an important part of any permitting process. Nevertheless, in response
to public and peer comments we have included a discussion of compensatory mitigation in
Appendix J of the revised assessment.
We have expanded the cumulative effects discussion (Chapter 13) to give a broader description
of streams, wetlands and fish at risk and have incorporated a discussion of the possible effects
on biological complexity and fragmenting salmon populations.
The following potential subsidiary impacts from development of a mining district of this scale
should also be described in more detail or at least mentioned:
• The extensive road network required to support mines in the area and the attendant
development associated with this network.
• The camps associated with the project, in migration of workers to the project areas, and
the demand for resources to be imported from outside the area.
• Invasive species that may follow this scale of development.
RESPONSE: The discussion about the cumulative effects of roads (including invasive
species), secondary development, and mining camps has been expanded in Chapter 13.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 160
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Steve Buckley, M.S., CPG
The assessment does describe the potential for cumulative risks from the development of
multiple mines in the area; however, the section is misleading in that it describes specific mine
footprints and tailings disposal sites at these prospects where there is no information on the size
or character of the potential future mine sites. The assessment does not describe the cumulative
effects of mine development.
RESPONSE: We have modified the descriptions ofpotential future mines to make them less
specific (e.g., recognizing that mining infrastructure could go anywhere in the subject claim
blocks). We have expanded the discussion of potential cumulative effects to include the
cumulative effects of multiple mines and development induced by the advent of large scale
mining in the watershed (Chapter 13).
These revisions are a major improvement to Ihe document.
Courtney Car others. Ph. I).
In general, the report suggests lhal efleels IV0111 multiple mines would increase the prevalence
and cumulative impacts of the risks described lor the one-mine scenario. Again, for the cultural
assessment, the conclusion is made that effects 011 humans would be primarily "direct and
indirect loss of food sources" (7-15) As the number of large-scale mines increases in this region,
the entire subsistence way of life could come under threat This would be a much larger impact
than lost food sources
RESPOXSE: The text of the revised report (Chapter 12) has been expanded to clarify that a
loss of the subsistence uv/r of life goes well beyond the loss of food resources.
Dennis I). Duuble, Ph.D.
Individual risk is described in \ arying levels of detail with overall risk or effects considered to be
largely additive. I low ever, the relative magnitude of the effects of mining each ore deposit is
difficult to discern. It is possible that one of the smaller ore sites could be developed within an
acceptable risk scenario, but it is difficult to determine given that the assessment is largely built
on potential impacts of the Pebble Mine. To put things in perspective (individually and
cumulatively), there should be a discussion of habitat lost given each individual mine footprint,
during normal operation (includes water treatment and withdrawal) and as a result of pollutant
exposure. Also, Section 7.4.1 of the main report provides estimates of stream miles affected due
to blockage and elimination, but provides nothing quantitative for other direct and indirect
impacts of mine operation. The cumulative risk discussion in Chapter 7 could be expanded to
link up with the conceptual model described in Chapter 3.
RESPONSE: We cannot reliably predict the habitat loss due to additional mines. We have
tried to describe plausible examples of where additional mines could be developed on the basis
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 161
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
of active exploration on existing claim blocks. In the revised assessment (Chapter 13) we
predicted aquatic resource impacts based on a typical mine footprint being constructed
anywhere in the block and an average stream and wetlands density. We also expanded
discussion of the impacts of ancillary mine infrastructure and induced development. This
provides a conservative estimate of the cumulative effects of multiple mines and sheds light on
whether cumulative effects are a significant concern. We have developed a specific conceptual
model for cumulative impacts and used it to enhance the discussion.
Gordon H. Reeves. Ph. I).
I found this chapter well done, even though the analysis was less extensive than what was done
for the Pebble Mine. It is clear that multiple populations u on Id be put at varying degrees of risk
simultaneously if mine development occurs ns is portrayed in this report. This certainly could
compromise the "portfolio effect" (Schindler el ill 2d 10 Nature -105,3 June
2010|doi: 10.1038/nature09060), which has maintained the long-term productivity of sockeye
salmon in Bristol Bay.
RESPONSE: We have added the portfolio effect to the discussion of the potential effect of
multiple mines.
Charles Wesley Slaughter, Ph.D.
Yes - but a qualified "yes "" The Assessment appropriately outlines the probability of additional
resource extraction projects beyond Pebble itself, and recognizes that additional resource
opportunities (beyond the claims depicted in I 'iunre 4-6), currently unknown or unverified, could
become viable or desirable to some interests in the future. Section 7.4 summarizes many of the
risks. Ho\\e\ er. the brief co\ eraue (Jo pages) accorded the entire subject of "cumulative risks"
is not consonant with the \er\ long-term, spatially dispersed (and presumably linked by
transportation and communication corridors) impacts and risks of multiple mines (and associated
infrastructure) in many different sectors of the Bristol Bay watershed.
RESPONSE: II e have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
John D. Stednick. Ph.D.
Cumulative risks can result from multiple risks (effects) from a single mine or individual risks
from multiple mines. This chapter identified potential risks from proposed mine activities;
without consideration of design standards or performance criteria, which is difficult without
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 162
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
specific mine designs/plans. Environmental risks were weighted equally - a TSF failure as
compared to a blocked culvert. The simple addition of stream length, as affected by various mine
footprints, does not represent a cumulative risk.
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
Roy A. Stein. Ph.D.
Cumulative Risks from Multiple Mines. (" I earl \. as amply demonstrated in the Environmental
Risk Assessment, cumulative risks would be greater than those from just the Pebble Mine, even
though these risks are difficult to quantify. Important points made in this section deal with the
economies of scale that would benefit additional mines coming to the Bristol liny watershed.
After the first mine, new mines become more profitable sini pi \ because some of the
infrastructure (roads, power, fuel pipelines, etc.) has already been provided, thus reducing cost
outlays for the establishment of new mines. To me, this seems as quite an insidious process, for
once the door is swung open for the first mine, then many more will follow owing to
infrastructure considerations; with these additional mines come far greater cumulative
environmental risks Ouanlifying these risks would help the reader and the public understand
what the ramifications of allowing one mine to begin operations might be. More text attempting
to quantify these cumulative impacts would be useful and instructive.
RESPOXSE: fi e have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
William A. Stubblefiehl. Ph.D.
The potential for cumulative risks associated with the development of multiple mines in the
Bristol Bay watershed is not treated with a great degree of detail. Although each of the potential
stressors (e.g., water withdrawal, habitat illumination, road and stream crossings) are
acknowledged and addressed, little quantitative consideration is given to the potential effects
associated with development of multiple mines. This, however, is probably appropriate given the
hypothetical nature of the single mine scenario and the potential for greater impacts associated
with the development of multiple lines. Short of concluding that "failures at one mine could be
bad and failures at multiple mines could be worse," little else could be concluded. It is noted that
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 163
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the multiple mines scenario leads to multiple tailings impoundments, more roads and culverts,
increased discharge potential of contaminated waters and increased habitat loss and reduction of
water resources and all of these lead to potentially greater environmental injury as a result of
failures.
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not appropriately describe the potential for cumulative risks from
multiple mines. In fact, the assessment does not identify the risks, only the likelihood of
occurrence and the consequences. See discussion under Question 4 above about estimating and
expressing the magnitude of risks and u luit is ret|Lii reel to appropriately describe the potential for
cumulative risks from multiple mines
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows.
The cumulative assessment is \ cry conceptual at best, as there are no specific proposals from any
of the other potential resource areas Cumulati\ e impacts can only be evaluated once further
details about other potential mines and their plans are a\ ailable. At this time, this section can at
best be seen as speculation
RESPO NSE: The cumulative effects discussion is not meant to be a definitive, quantitative
evaluation. II e have presented a plausible example of how a mining district could develop and
a simple estimate of the impacts to aquatic resources andfish. It is intended to shed light on
whether cumulative effects are a significant concern.
It is impossible to impro\ e this part of the assessment with the information on mine development
currently available; it can only be done when further information is published by the various
mining companies.
RESPONSE: Definitive quantitative risk assessments offuture mines will need to wait for
more information on those mines and on the resources they could potentially impact. This
assessment is meant to identify potential issues of concern with development of large scale
mining in the two watersheds. Based on the information available now, is the possibility of
cumulative impacts from multiple mines and induced development a significant concern? We
have utilized the existing information on the number of mining claims, the extent of recent
exploration activities, local and state land use plans, and the ubiquitous nature of fish habitat
in the watersheds to shed light on that question. Future environmental impact statements will
provide more definitive analyses. We have tried to improve the cumulative analysis by
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 164
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
discussing the potential impacts from entire mine footprints, transportation corridors, induced
development and increased access.
Phyllis K. Weber Scannell, Ph.D.
There are two issues that should be considered: cumulative effects from a single mine and
cumulative effects from multiple mines. Cumulative effects from a single mine might include
aquatic habitat degradation from non-point sources, including run-off from exposed mineralized
rock, seepage from the tailings impoundment, contaminated dusl. noise, and other forms of
disturbance.
RESPONSE: We added a discussion on the cumulative effects from a single mine (Box 6-1),
as well as discussion of multiple transportation corridors, induced development and increased
access (Chapter 13).
Cumulative effects from multiple mines are difficult to predict because there are too many
unknowns. It is frequently to the advantage of a mining company to take ad\ antage of existing
infrastructure, without building new camps, new mills, etc It is also possible to use an old mine
pit for tailings or waste rock disposal from a new site. ho\\e\ er. none of these features can be
determined until there is sufficient exploration to determine if mining is feasible, to characterize
the deposit, and to develop a detailed mine plan. To date, there is not sufficient information to
predict cumulative effects from multiple mines
RESPONSE: We have expanded the discussion of cumulative impacts from multiple mines in
Chapter 13. The analysis of cumulative effects is largely qualitative rather than predictive. It is
a plausible example of future cumulative effects that can shed light on whether cumulative
effects is an important topic for consideration as plans for mining in the Nushagak and
Kvichak River watersheds move forward.
Paul Whitney. Ph.l).
Need to Address Past Impacts. l-IWs guidance for reviewing cumulative impact analyses (EPA
1999 - most recent guidance on EPA's website) asks that past present and reasonably
foreseeable actions lx- considered The assessment's coverage of cumulative risks from multiple
mines certainly addresses "reasonably foreseeable" mines. Not addressed in the analysis is past
impact(s) and how such impact(s) might be additive to current and foreseeable impacts. As
commented above, commercial fisheries remove a lot of the salmon (up to 70% - over 10 million
fish) annually from the drainages associated with the mine. The continual annual reduction, or
loss of energy and nutrients that might otherwise return to the ecosystems, should be considered
a past impact as part of the cumulative impact analysis.
RESPONSE: The purpose of the assessment is to evaluate potential effects of large-scale
mining on the fishery, not the effects of commercial fishing. Therefore, the assessment
considers the present condition of the watershed and fishery as a baseline for an evaluation of
potential future effects of large-scale mining. We have expanded the discussion on past,
present and reasonably foreseeable actions in Chapter 13. Since the fishery has sustained
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 165
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
itselffor over 100 years despite commercial fish harvesting, it is unclear to what degree
reduced escapement to the watersheds has had negative impacts. As Hilborn (2006) states,
"Recent paleoecological analysis of returns to Bristol Bay show no indication of decreased
production since commercial exploitation began." Furthermore, the fishery is tightly managed
by the Alaska Department of Fish and Game, which reduces the harvest to increase
escapement when necessary.
I have prepared/managed cumulative impact analyses of fish and wildlife for a relatively small
(e.g., a 200-acre aggregate mine) and a relatively large 7,100-acre coal mine in Washington. A
cumulative impact analysis of past, present, and future fish and wildlife losses for the coal mine
expansion in a large watershed indicated cumulative impacts of the expansion were less than one
percent of past mining, agriculture, urban and forest activities in the watershed. In addition, the
cumulative loss could be fully mitigated by compensatory restoration. I would be interested to
know what the estimated fish and wildlife loss is due to the example Clipper mine in comparison
to the loss related to the commercial fishery. In addition, I would he interested to know if
potential fish and wildlife losses due to mining could be fully mitigated. If the watershed is
pristine or nearly pristine, the opportunity for compensatory mitigation may he low. If the past
impact of the commercial fishery is large and the watersheds are not pristine, there may be
opportunities. There is not much degraded habitat that could he improved by a mitigation plan. If
such a cumulative impact analysis were conducted, it may stimulate conversation about reducing
commercial fishing to compensate lor impact losses due to mining.
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment. Appendix J includes a discussion of the challenges of creating equivalent
habitat for anadromous and resident fish and why this would be particularly challenging in
the context of the Bristol Hay watershed. C ompensatory mitigation requirements address the
need for project proponents to replace aquatic resources and ecosystem functions that their
project has impacted. Reduced fishing harvests would not replace lost spawning and rearing
habitat. Further, it would remove the burden of compensation from the party that caused the
damage.
Question 12. Are there reasonable mitigation measures that would reduce or minimize
the mining risks and impacts beyond those already described in the assessment? What
are those measures and how should they be integrated into the assessment? Realizing
that there are practical issues associated with implementation, what is the likelihood of
success of those measures?
David A. Atkins. M.S.
The Assessment describes what is considered to be conventional 'good' mining practice, but
does not adequately describe and assess mitigation measures that could be required by the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 166
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
permitting and regulatory process. A thorough analysis of possible mitigation measures as
employed for other mining projects and the likelihood that they could be successful in this
environment would be necessary.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed modern
mining technology and operations. The terms are qualitative when generally interpreted, or
have a regulatory meaning. The term "best management practices" is a term generally applied
to specific measures for managing non-point source runofffrom stormwater (40 CFR Part
130.2(m)). Measures for minimizing and controlling sources ofpollution in other situations
are referred to as best practices, state of the practice, good practice, conventional, or simply
mitigation measures. We have added a text box in the revision Chapter 4 to discuss terms.
Mitigation measures considered feasible, appropriate. and 'perniittable' (as per Ghaffari et al.
2011), were considered in the assessment, and these are measures common to other copper
porphyry mines. Evaluation of alternative strategies (e.g., other options presented in Appendix
Ifor the mitigation of the same issue) should be part of the permitting process for a specific
mining plan. The State of Alaska does have statutory and/or regulatory requirements for an
approved Plan of Operations (11AAC 86.800), a Reclamation Plan (Alaska Statute (AS)
27.19.30) and appropriate Financial Assurance (AS 2719.040) and the revised Chapter 4
notes these requirements.
It is highly likely that for mines located in the lirislol Bay watershed, conventional engineering
practices would not be sufficient. Therefore, it is important to consider mitigation on numerous
fronts when determining the \ i ability of the project A section on innovative and state-of-the-art
approaches for both mitigation and construction of mine facilities would be helpful to better
understand if risks can be minimized or eliminated given sufficient funds.
RESPONSE: The HP. I agrees generally that conventional engineering practices may not be
sufficient, but it may not be appropriate to test innovative approaches in a watershed such as
the Bristol Hay watershed. The EPA is not aware of innovative mitigation measures that have
sufficient history to be applicable to this location.
Under the no-failure scenario, the footprint of the mine (open pit, block-cave subsidence zone,
waste rock and tailings areas) will by necessity destroy habitat. There may be ways to create
equivalent habitat to compensate for lost habitat in areas within the watershed that are currently
not producti\ e lor lisli This form of mitigation may work for resident fish, but it is unclear if it
would work for anadromoiis fish that return to very specific locations to spawn.
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment. Appendix J includes a discussion of the challenges of creating equivalent
habitat for anadromous and resident fish and why this would be particularly challenging in
the context of the Bristol Bay watershed.
It is also becoming common practice to offset impacts from project development with
preservation of equivalent habitat areas that are also at risk from development (http://bbop.forest-
trends.org/). It is unclear if this is a feasible consideration for this project as this could involve
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 167
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
allowing one development (e.g., Pebble), while potentially taking away the development rights
of others (presumably for proper compensation).
RESPONSE: See response to above comment. The potential efficacy of using habitat
preservation as a form of compensatory mitigation for impacts in the Bristol Bay watershed is
discussed in Appendix J.
Steve Buckley, M.S., CPG
There are many reasonable mitigation measures that could reduce the risks and impacts beyond
those described in the assessment. Some of these are contained in the Appendices and referenced
therein but not discussed in detail or described in the assessment ll is beyond the time
constraints provided in this review to develop an exhausli\ e research list of these potential
mitigation measures; however, EPA could include measures designed to: reduce the mine
footprint and limit the number of potentially affected watersheds; reduce, isolate, or eliminate the
amount of potentially acid generating waste; pix>\ ide secondary containment measures for all
pipeline corridors; and use natural streambed arch cul\ erts and bridges at lisli bearing stream
crossings.
RESPONSE: The mitigation measures proposed within the mine scenarios are a subset of
options presented in Appendix I, all of which were presented as appropriate for the Pebble
deposit in Ghaffari et al. (2011)., 1 ppendix I includes multiple options for each source of
contamination, with some having specific applicability (e.g.. site and material characteristics)
and others being more broadly applicable. Some additional mitigation measures have been
added to the scenarios in Chapter 6 of the revised assessment (e.g., processing PAG waste rock
over the life of the mine and selective flotation to minimize pyrite in tailings). Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation (e.g., by reducing the footprint or limiting the number of affected
wetlands) would be addressed through a regulatory process that is beyond the scope of this
assessment. Nevertheless, in response to public and peer comments we have included a
discussion of compensatory mitigation in Appendix J of the revised assessment.
This seems to he a reasonable approach
Courtney Car others. Ph. I).
While I do not have knowledge of mitigation measures, a more thorough discussion of mitigation
measures could be included. Even if mitigation measures are largely deemed to be ineffective in
this case, they should be presented and evaluated as such.
RESPONSE: The EPA agrees with this comment and Appendix I is intended to provide that
discussion. The mitigation measures proposed within the mine scenarios are those that could
reasonably be expected to be proposedfor a real mine (they are a subset of options presented
in Appendix I), all of which were presented as appropriate for the Pebble deposit in Ghaffari
et al. (2011). Evaluation of alternative strategies (e.g., other options presented in Appendix I
for the mitigation of the same issue) should be done during a permitting process for a specific
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 168
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
mining plan. The assessment assumes that measures chosen for the scenarios would be as
effective as possible.
Dennis D. Dauble, Ph.D.
Potential mitigation measures are well described in Appendix I. I have no suggestions for
additional measures. Implementation of mitigation measures is entirely dependent on the
regulatory framework for operations and the oversight and monitoring practices that would be
mandated as a condition of the mining activity. Thus, sonic discussion of how/which mitigation
practices would be most applicable in the Bristol Bay watershed (and limitations thereof), given
constraints and characteristics of local hydrology and ucolouy. is warranted.
RESPONSE: The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine, are a subset of options presented
in Appendix I, and were presented as appropriate for the Pebble deposit in Ghaffari et al.
(2011). Evaluation of alternative strategies (e.g.. other options presented in . ippendix Ifor the
mitigation of the same issue) should be done during a permitting process for a specific mining
plan. Permit applicant mitigation measures that reduce the risks identified in the assessment
would be welcome during the application process.
Gordon H. Reeves. Ph. I).
I identified one polcnlial mitigation lor ail\ oris the use of open arch types that are at least one
bankfull width i 11 size As described in my response lo Question 7, this could reduce many of the
potential impacts raised by the authors
RESPO N SE: .Suggestion noted. In our scenarios (Chapter (>). we assumed state-of-the-art
practices for design, construction, and operation of the road infrastructure, including design
of bridges and culverts for fish passage.
Charles H eslev Slaughter, Ph.D.
If it is assumed thai the PI V project, or some similar development, were to go forward, I cannot
suggest mitigation measures beyond those discussed above. Since a major concern for salmonids
- perhaps THE major concern is with consequences of the transportation corridor, simply
having the mine without the roads/pipelines would alleviate much potential risk. However, there
is presumably no practical, economically feasible way to not have the transportation corridor; air
transport of all materials to and from the site might technically be possible, but would not be
economically feasible.
RESPONSE: We agree that it would be infeasible to have a mine in this location without also
having road access. No change suggested or required.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 169
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
John D. Stednick, Ph.D.
The purpose of this assessment is not to identify mitigation measures. This suggests that things
can be fixed by mitigation. Risks were identified for a variety of situations, and the preventative
measures would better address the mining impacts. Mitigation measures are also a mining cost
that needs to be determined by the mining company and compliance with state and Federal
regulatory authorities.
RESPONSE: Comment noted. We agree that mitigation measures required for an actual mine
would be determined through the regulatory process. No change suggested or required.
Roy A. Stein. Ph.D.
Mitigation: Complete Tailings Storage Facility (TSF) l-'ailiire In some ways, some of the
failures reviewed herein are not really subject to mitigation. For example, if a Tailings Storage
Facility (TSF) completely fails, options for mitigation become limited \ cry quickly. With a
complete failure, it is "game over", with toxic sediments flowing into the Nushagak River all the
way to Bristol Bay, thereby destroying the entirety of salmonid spawning hahitilt in this river by
redirecting the channel and inundating the gravel/cobbIe steam bed with sediment (meters in
depth). Mitigation under these circumstances is impossible, in my view. Given this scenario, I
was surprised that the impacts were only assessed 3<> km downstream of the TSF failure; is this
realistic? I think not, given what we know of other mi lies
RESPONSE: We acknowledge the limitations of the hydrologic modeling. We now emphasize
more clearly the potential for the effects of TSI-'failure to extend to Bristol Bay. The potential
for remediation (and associated risks) are discussed in Section 9. 7.
The amount of text dedicated to a TSI failure is large (36 pages) as compared to other failures. I
suggest this section be shortened to bring it more in line with other sections. Briefly
summarizing impacts would focus the text and help the reader appreciate what it would mean to
have a TSI ' failure without ha\ ing to wade through so much text.
RESPONSE: We have revised this section of the assessment. However, given its importance to
stakeholders, local communities, and decision makers, it was not significantly shortened.
Other accidents and failures are now addressed in more detail.
Dry Stacking Mine Tailings Appendix I, page 9. Given the horrific impact of a TSF dam
failure, should mine operators consider a relatively new technique incorporating "paste tailings
technology"? Here, tailings are thickened by water removal (down to 20% water) and filtering;
tailings are then dry stacked onto a lined disposal site. These stacks "have a lower potential for
structural failure and environmental impacts" (Martin et al. 2002). I would encourage the
authors to argue for this substitute for the more traditional TSF for this solves at least two
problems: 1) eliminates the possibility of a TSF failure (a huge gain!) and 2) reduces the amount
of monitoring and maintenance of the waste tailings "In Perpetuity" (another major gain).
Finally, what does "Best Management Practices" have to say about these two approaches to the
storage of tailings waste?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 170
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: We agree that dry stack storage and paste storage are options for mine
proponents to consider. Although the option of using paste tailings is one that might be
explored, the type of tailings disposal that operators propose is generally based on two things:
the type of mine and the quantity of tailings. A balance would need to be considered between
the low risk/high impact of a tailings dam failure and the certainty of increased area of
disposal and the increased cost of tailings handling for paste tailings. A "rule of thumb" for
design of dry stack tailings is to allot 25 acres for every thousand dry tons of tailings per day
over the life of a 20 year operation (SME Mining Engineering Handbook 1973). This would
amount to 4,900 acres (25 *200000 *.99/1000). Since paste tailings are more voluminous than
dry and the scenario contains a 25 year mine life rather than 20 years, the area of a paste
tailings facility would exceed the calculated amount resulting in over a 1/3 increase in the
area occupied by tailings than contemplated in the current mine scenario. With a paste
tailings facility, the monitoring and maintenance of a large dam is removed but that is
replaced by the monitoring and perhaps treatment of the leach ate draining through the
tailings pile (that contains 20% water) as well as the long-term maintenance of the cap that is
placed over the tailings at the end of the mine life. The choice of a tailings disposal
methodology may be affected by the need to avoid or minimize effects to wetlands or the
technical feasibility of constructing the structural components of the project.
Mitigation: Partial TSF Failure. If a TSF partially fails or is discovered beginning to fail, then
I believe mine operators have a chance to save the dam and I hereby protect the river, but only if:
1) the appropriate Standard Operating Procedures (SOP) for an emergency response are in place,
2) the necessary equipment (my presumption here is that heavy, earth-moving equipment would
be required), materials, and supplies are onsile near the facility, and 3) trained personnel
(meaning that they have practiced these repair SOPs in the preparation for such an event) are
available for immediate action. One niighl argue thai these procedures are more proactive than
mitigating and I would agree, reflecting the near impossibility of invoking any mitigation
measures associated with TSI' failure
RESPO N SE: Our scenarios presuppose that there are mitigation measures in place, but
examines the effects of a failure of those measures. EPA agrees that SOPs are part of a
mining operation and plan; however, inclusion of SOPs and details for emergency responses
to a failing dam were outside the scope of this assessment. No change suggested or required.
Dredging: Post TSK failure In the text (pages 6-1 to 6-2), a reference is made to dredging
materials out of the ri\ er post spi 11 I can't imagine this would mitigate any losses of spawning
substrate for salmon. Indeed, because only 5% fines in gravel substrates compromise salmon
reproductive success (and perhaps even selection of these areas for spawning in the first place),
removal of meters of sediment with a dredge doesn't seem to be a solution. Whereas dredging
might, in a best-case scenario, reduce the time to recovery of the substrate, I don't believe it will
hasten recovery significantly. Dredging also serves to bring toxic sediments up into the water
column perhaps compromising all organisms in the system. By not dredging, we allow the
natural system to recover, which, in my view, would be preferable to any sort of "dredging
mitigation". Some reflection by the authors on this issue would be valuable.
RESPONSE: Agreed. Adverse effects of dredging are discussed in Section 9.7.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 171
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Mitigation: Pipeline Failure. With automatic shut-off valves stationed along all four pipelines,
we would expect to know precisely just how much effluent will be spilled during any single
event. With this information in hand, mine operators can easily anticipate spill size, toxicant
characteristics and thereby judge what equipment, materials, and supplies would be necessary for
mitigating any spill. As with the TSF failure, mine operators should have in place: 1) the
appropriate Standard Operating Procedures (SOP) for an emergency response, 2) the necessary
equipment, materials, and supplies onsite near the pipeline (given the length of the road, these
items should be cached at several locations along the road, such that response time is
minimized), and 3) trained personnel (meaning that they have practiced these mitigation SOPs in
the preparation for such an event) are available for immediate action. Shouldn't pipes be double-
walled? Again, what would "Best Mining Practices" say in this context?
RESPONSE: The draft assessment included pipeline failure scenarios that quantified the
magnitude of spill events. The draft included the most commonly used and accepted practice
of using double walled pipes where above ground and water. There is. to the best of our
knowledge, no standard SOP for a product concentrate slurry spill. . I.v is discussed in the
assessment, terrestrial spills would be excavated. Spills to streams might also be excavated,
although the physical damage to habitat would be considerable and the efficacy of excavation
would depend on how far stream flow had spread the material. An assessment of diesel spills
has been added to the revised assessment in Chapter I /. and it refers to a NO A A and API
report for standard remedial practices.
Mitigation: Failure of Water and Leachale Collodion This failure differs from TSF and
pipeline failure, where failures are more akin to catastrophic. Ibr here failure is somewhat more
gradual in coming (my guess is) I'roactix e \ igilance is the watch phrase here where continual,
careful monitoring will indicate when failure hegins Because the "spill potential" is relatively
small (certainly compared to a TSF failure), less urgency is required on the part of mine
operators. Ffowever, as pointed out |">re\ iously. just because the potential is small, overtime the
impacts could be great. Hence, the mitigation undertaken with water and leachate collections
would require (one would hope) just the tweaking of the collection system in place to eliminate
leakage through time. Again, personnel trained in how to respond to these gradual increases in
water and leachate leaks are required to stay ahead of this issue, thus preventing any toxic
materials from flowing downstream into iheNushagak and Kvichak rivers.
RESPONSE: The revised assessment acknowledges that the leachate capture system is not
likely to capture all leachate during routine operations and shows that water quality standards
would be exceeded in a considerable stream length as a result. Mitigation measures that
ensure or maximize the detection of leachate in groundwater before it reaches surface waters
would be vital.
William A. Stubblefield, Ph.D.
Fm sure there are number of technological/engineering measures that could be implemented to
reduce the potential for environmental injury associated with development of mining in the
Bristol Bay watershed. The development of this a priori risk assessment provides useful
information in identifying were potential risks may exist and should provide mine development
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 172
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
professionals with the degree of guidance about the types of risks and potential consequences of
mine activity failures. Perhaps by recognizing the magnitude of adverse consequences associated
with potential failures, steps can be taken to implement safety measures early in the planning
process that would render mine development more acceptable. In addition, using the assessment
to define areas of uncertainty my provide direction for future research that would be beneficial
for the project. Again, because of the lack of detail associated with the hypothetical mine
scenario, it is impossible to estimate the likelihood of success of any mine control activities.
RESPONSE: One purpose of this assessment is to inform future decisions concerning mine
design and required mitigation. The purposes of the assessment are clarified in the revised
introduction (Chapter 1).
Dirk van Zvl, Ph.D., P.E.
Yes, there are reasonable mitigation measures lhal would reduce or minimize the mining risks
and impacts beyond those already described and incorporated by the i-PA in the assessment.
There are a host of measures that are not addressed in the assessment and lists of these are
identified below under the headings of regulatory and engineering. This list is by 110 means
exhaustive.
RESPONSE: See response under the conuiienter \ headings of "Regulatory" and
"Engineering" below.
While the EPA Assessment presents a series of potential mitigation measures in the main report,
the majority were rejected Appendix .1 [sic | to the report also includes a generic discussion of
mitigation measures The Main Report does not address the application or implications of these
in any project specific details, e g compensatory mitigation for wetlands, streams and other
aquatic resources
RESPONSE: The HP. I did not reject the majority of potential mitigation options mentioned in
the report. The mitigation measures proposed within the mine scenarios are those that could
reasonably be expected to be proposed for a real mine. They are a subset of options presented
in Appendix /. were presented as appropriate for the Pebble deposit in Ghaffari et al. (2011),
and are common with mining of porphyry copper deposits. Mine design for the scenarios is
closely based on that presented in Ghaffari et al. (2011) and incorporates the mitigation
measures included in that mine design. Mitigation to compensate for effects on aquatic
resources that cannot be avoided or minimized by mine design and operation would be
addressed through a regulatory process that is beyond the scope of this assessment.
Nevertheless, in response to public and peer comments we have included a discussion of
compensatory mitigation in Appendix J of the revised assessment.
Multi-stakeholder engagement processes, such as Failure Mode and Effects Analysis, can be
used to further expand on these mitigation measures. It is recommended that EPA recognize
these and potentially other measures that may be proposed in the public comments and make a
serious effort in including the potential effects of these on failure likelihoods, consequences and
risk magnitudes. It is an important aspect of improving on the range of potential outcomes.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 173
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The EPA recognizes that other mitigation measures and mine designs have the
potential to address impacts and risks identified in the assessment. Evaluating the risks of all
potential mitigation measures and mine designs is not a goal of this assessment, which
identifies potential risks of concern from plausible mine scenarios based on current practices
and a recent plan (Ghaffari et al. 2011). A permit applicant's mitigation measures that further
reduce the risks identified in the assessment are appropriate for the application process.
Regulatory. The EPA Assessment neglects the typical outcomes resulting from the permitting
and regulatory processes for new mines, where permit stipulations may require specific actions
resulting from discussions, public comments and regulatory frameworks. The following is a
partial list of these:
• Section 404 of the CWA (See discussion under Question .1 above).
• Permitting stipulations and requirements for monitoring I sing permit stipulations for
monitoring can reduce a large number of the consequences identified in the report. For
example, the consequences associated with Mocked culverts, etc can be significantly
reduced by permit monitoring requirements and subsequent enforcement.
• Financial assurance will be required for mine closure. Financial assurance can be a very
beneficial tool during operations, premature closures, mine closure, as well as post-
closure monitoring and maintenance. Experience gained during operations can help
develop the closure and post-closure financial assurance requirements.
RESPONSE: The assessment is intended to identify impacts and risks of concern which may
inform future NEPA and permitting processes. The revised assessment presents basic
discussion offinancial assurance and listings of requirements for permitting a mine in
Chapter 4. The assessment is not a mining plan and is not meant to replace the regulatory
process for development of mining plans for these prospects, but can inform that process
about important risks that should be addressed.
Engineering. A niimher of engineering options are mentioned in the report but discounted in
many cases Many of these engineering mitigations are currently used in the industry. The
following may repeat a number of those mentioned in the report:
• Redundancy, e.g. additional embankments may be considered downstream of the TSF to
contain tailings and supernatant that may be transported as a result of TMF failure. While
this may result in a larger local surface impact, it will protect downstream waters in the
case of a failure resulting in tailings discharge.
• Further processi ng of tai I i ngs to remove the remaining sulfides.
• Tailings management options other than slurry deposition, such as production and
management of filler cake.
• Waste rock management options to reduce releases during operations, e.g. addition of
lime to the PAG rock.
• High standards implemented for road design, construction, monitoring and maintenance.
• Double containment of all pipes containing concentrate and other materials. This is
already required under the International Cyanide Code for all pipelines containing
cyanide solutions.
RESPONSE: The revised assessment presents background information on mining, including
a sub-set of mitigation measures from Appendix I, in Chapter 4 and the scenario-specific
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 174
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
information in Chapter 6. The option of selective flotation to reduce the amount of sulfide in
the bulk tailings, as well as processing of PAG waste rock over the course of operations, have
been included in Chapter 6. In the draft assessment, selective flotation was discussed in the
background information, but was missed in the scenarios. A double-walled pipeline along the
entire length of a pipeline might be desired from a purely environmental protection
standpoint; however, it may not be feasible or cost effective to do this given the length of the
pipeline. Therefore, the assessment proposed the most commonly used (and accepted method)
of double-walled construction over any water bodies. While the EPA opted not to include
redundancy in tailings embankments in the scenarios, this option, as well as full pipeline
double-wall construction, could be evaluated for appropriateness during the regulatory
process for any future permit application.
The likelihood of success of these proposed and other potential mitigation measures can be
evaluated by considering their impacts on the overall project. A range of alternative project
technical alternatives and facility-siting locations will have to be de\ eloped instead of using only
the hypothetical scenario.
RESPONSE: The purpose of the assessment is to describe the potential adverse environmental
effects that could exist even with appropriate and effective site mitigation measures.
Considering alternative options is part of a permit application process and outside the scope of
this assessment. No change required.
Phyllis K. Weber Scunnell. Ph.D.
There are many avoidance or mitigation measures that would be implemented to reduce or
minimize mining risks. I have described some possible approaches when answering the previous
questions. To summarize:
The two most important questions lor reducing or minimizing mining risks are:
• Can a mine in this area be designed for closure?
• Is it acceptable to develop and operate a mine that will require essentially perpetual
treatment''
RESPONSE: EP. I agrees that these are key questions that must be addressed in the regulatory
process. Our purpose in the assessment is to evaluate the risks resulting from a mine operated
with modern conventional mitigation measures for design, operation, monitoring and
maintenance, and closure. The regulatory process addresses significant and unacceptable
risks. Chapter 4 of the revised assessment discusses regulatory andfinancial assurance
requirements for mining in Alaska.
Specific Measures that can be taken to minimize risk include:
Limiting metals contamination and acid drainage:
• Design the mine pit to limit oxidation on pit walls. Where feasible, conduct concurrent
reclamation.
• Develop plans for classification and storage of waste rock, lower grade ore, overburden,
and high grade ore.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 175
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
• Develop and maintain tailings storage facilities with fail-safe provisions. An emergency
discharge of untreated waters from a tailings storage facility could be made to a
collection pond for later treatment or the tailings pond could be engineered to
accommodate a higher flood event so the likelihood of overtopping is minimized.
Consider alternate methods for tailings disposal (dry stack following sulfide removal,
etc.).
• Implement concurrent reclamation of disturbed areas, including stripped areas and mine
pits.
• Collect and treat point and non-point source water.
• Design and implement plans for the quantity and liming of discharges of treated water;
especially if the treated water is high in total dissoK ed solids. Monitor ground water,
seepage water, and surface water.
• Design system for collection and bypass of clean water and collection and diversion of
contaminated water to a water treatment system
• Require stations for truck wheel waslii nu
RESPONSE: Many of the measures presented here were included in the scenarios and in
Appendix I and are mitigation measures commonly included for mining of this type. Other
bullets noted here are good suggestions for things to address during the regulatory process,
should a permit application be submitted.
Protection of Fish Habitat:
• Review all in-stream activities in waters important to the spawning, rearing, or migration
of anadromous and resident fish.
• Design and implement a Momonitoring prouram
• Review every road crossing of fish bearing waters to ensure free passage of fish.
RESPONSE: Suggestions noted for consideration during any future regulatory permitting
process.
Possible Measures to Limit Effects to Wildlife:
• At the planning stages, design aspects of the project to create or enhance wetland and
aquatic habitats for fish, bird, and wildlife species.
• Limit truck traffic on the haul road during migrations.
• Incinerate all kitchen waste.
• Educate workers 011 hear (or other wildlife) safety.
• Limit air traffic and noise during critical times of the year.
RESPONSE: Suggestions notedfor consideration during any future regulatory permitting
process.
Paul Whitney. Ph.D.
Comments on Mitigation. The key word here is "reasonable." What is reasonable to a person not
involved in mining on a day to day basis will likely not be reasonable to a mining company
executive or mining engineer. The likelihood that "reasonable" means different things to
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 176
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
different people is exacerbated by the mixed messages regarding "best" mining practices (e.g.,
page ES-10, five lines from the bottom) versus "not necessarily best" mining practices (e.g., page
4-17, four lines from the top). As mentioned above, both of these statements can't be accurate.
For purposes of this discussion, I am assuming that there are many more mining practices that
could be proposed to address many of the uncertainties mentioned in the assessment. First of all,
most of the mitigation measures mentioned in the assessment provide one line of protection and I
suspect there are many types of redundant mitigation that could be implemented. Redundant
protection such as: double-walled pipes in all sections that cross floodplains (above and below
grade); poly liner/vegetated caps to soak up and capture run off water before it contacts waste
rock; redundant clay, glacial till (waste360.com/mau waste landfillsglacialtill); poly liners;
and secondary liquid collection systems.
RESPONSE: The EPA has revised the assessment to better explain the qualifying terms. What
was intended is that we have assumed modern conventional mining technology and
operations. The terms are qualitative when generally interpreted, or have a regulatory
meaning ("best management practices"), and thus we have eliminated their use in the revised
assessment. Measures for minimizing and controlling sources of pollution in other situations
often are referred to as best practices, state of the practice, good practice, conventional, or
simply mitigation measures. We assume that these types of measures would be applied
throughout a mine as it is constructed, operated, dosed, and post-closure, regardless of the
qualifier that one wishes to place with it. To remove any ambiguity and subjectiveness of terms
"good" or "best", we have removed them in the revision and have added Box 4-1, which
includes definitions for several terms used. Double-walled pipelines were included in the
scenarios for sections above or beneath water bodies. Redundant mitigation measures are
appropriate for consideration during the permitting process.
1 appreciate the assessment's discussion of the potential difficulties of using poly liners, but the
discussion might benefit from a review of Koerner et al. (2005) that provides another perspective
on the life of IIDPI- liners The relationship of liner life to temperature is presented in their Table
2 and a summary in the text indicates that co\ ered liners have a "half life of 446 years at 20
degrees Centigrade." This is a lot different from the 20 to 30 year estimate of service life cited in
the assessment on page 4-1 1 (last para). Appendix I does cite a 2011 version of the Koerner et al.
(2005) paper and appears to misquote it (page 9, last para, last full sentence). The Koerner et al.
(2005) paper estimates a "halflife of 446 years" not a "lifetime" of 446 years, as cited in
Appendix I. Perhaps Koerner updated his 2005 estimate in the 2011 version.
RESPONSE: In the 2011 Koerner reference, there is conflicting text - the 446years is listed
as the "lifetime", and later also as "half-life"; "449"years is also stated as the half-life and
one is referred to the Table in the report with the title of "Lifetime", where there is no such
number. It appears the author is using the terms half-life and lifetime to mean the same thing.
Later in the white paper, there is text that states "...its predicted lifetime (as measured by its
half-life)...". The following text has been used in Appendix I and is consistent with the
reference: "...Koerner et al. (2011) presents that a nonexposedHDPE liner could have a
predicted lifetime ("as measured by its half-life") of 69 years at 40 °C to 446years at 20 °C."
The discussion of liners has been moved to Chapter 4 in the revised assessment and includes
69 to 600 years as half-lives, based on Rowe 2005 and Koerner et al. 2011. The personal
communication reference has been removed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 177
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
I assume a mining engineer (if asked) could design a series of smaller impoundments or
innovative lined impoundments that could avoid a lot of the problems cited in the assessment.
The trade-off might be increased loss of natural resources due to a larger footprint and increased
construction cost, but the risk cited in the assessment and the risk of a catastrophic failure might
be greatly lowered.
RESPONSE: An engineer could design a series of smaller impoundments that would likely
decrease the risk of catastrophic failure, but would impact a larger footprint and increase
construction costs. However, cumulative risk of failure would increase. More than one dam,
even if they are smaller, means a higher risk offailures, even if they contain less volume, as
well as more points where water management systems could fail, even if the dam itself does
not.
I would be interested to know what THE BEST mining piaclices are If the mitigation measures
mentioned in Appendix I are, in fact, the best, il should be so stated and taken into consideration
in the assessment. For example, page 4-21 of llie assessment indicates the TSF would be unlined
and not have an impermeable barrier between the tailings and groundwater Appendix I, page 9
indicates TSFs can be lined if problems are expected. There is a lot of good in formation in the
Koerner et al. (2005) paper and Appendix 1. It seems the types of mitigation measures in
Appendix I could be better captured in the main report.
RESPONSE: The terms "best", "good" or other are qualitative when generally interpreted, or
have a regulatory meaning (best management practices when referring to storm water
measures). Measures for minimizing and controlling sources of pollution in other situations
often are referred to as best practices, state of the practice, good practice, conventional, or
simply mitigation measures. The multiple measures presented in Appendix I are options that
might be "best", depending on the site-specific conditions and any other constraints. More
discussion on mitigation measures has been added to the scenarios in Chapter 6 of the revised
assessment.
Compensatory mitigation and reclamation are briefly mentioned in the assessment. A more
detailed discussion of the opportunities and feasibility of reclamation and compensatory
mitigation might reduce the likelihood of potential impacts of the example mine plan (see
response to Question .1)
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment.
Question 13. Does the assessment identify and evaluate the uncertainties associated
with the identified risks?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 178
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
David A. Atkins, M.S
The Assessment states: the 'range of failures is wide, and the probability of occurrence of any of
them cannot be estimated from available data.' Uncertainty is addressed throughout the report,
typically with a qualitative discussion. There is a high degree of uncertainty with respect to how
the mine would be developed, operated and closed, as well as how any impacts would be
mitigated. This large uncertainty makes assessing risk difficult.
RESPONSE: No change suggested or required.
Steve Buckley, M.S., CPG
The assessment identifies some of the uncertainties associated with the identified risks but does
not evaluate these in great detail.
RESPONSE: No change suggested or required.
Courtney Car others, Ph.D.
The report includes specific sub-seclions lo discuss uncertainties for the risks associated with
habitat modification (Section 5.2.4), pollutants (5 .1 4). and water collection and treatment failure
(6.3.4). Uncertainties related to abundance and distribution of lisli in the watershed draining the
mine site, road and si renin crossings, salmon-mediated effects 011 wildlife, salmon-mediated
effects on human welfare and Alaska Native cultures, tailings dam failure, pipeline failure, and
road and culvert failures are not discussed in separate sections; however, several uncertainties
related to these risks are noted throughout the report, and in summary sections (Sections 8.5 and
8.6).
RESPONSE: He have added separate uncertainty sections to each of the major topics in the
risk assessment chapters.
The "sensili\ ily relative lo o\ era 11 results" of the key assumptions and uncertainties presented in
Table 4.8 in Appendix E (pp I '¦>}-195) would be a helpful model to employ in the main report.
For non-experts in the technical dimensions of mine construction and operation, uncertainty
rankings would be useful I or example, "We are 'highly uncertain' about the accuracy of these
predictions given this unknow 11 factor," or "We expect this uncertainty has a negligible effect on
the model we employ to calculate this risk."
RESPONSE: The assessment has been restructured to consider certain impacts resulting from
the mine footprint (Chapter 7) separately from less certain impacts resulting from potential
failures (e.g., Chapters 8 and 9). For uncertainties considered in each chapter, we also have,
when possible, indicated a general categorization of (1) the level of uncertainty, and (2) the
sensitivity of our conclusions to this uncertainty.
Dennis D. Dauble, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 179
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
The most likely scenarios and probabilities of failure are described based on assumptions of
project size and magnitude. For the most part, estimated risks are conservative (i.e., effects are
stated as "likely" if no further information is available). A weakness of Integrated Risk
Characterization (Chapter 8 of the main report) is having a long list of identified uncertainties,
which leads one to speculate, "so what do we know?" Not being familiar with the formal risk
assessment process, it appears this "assessment" (which is loosely based on a risk assessment
framework), falls short of providing something with any degree of certainty.
RESPONSE: The assessment does not state that effects are likely if no further information is
available. The assessment, as far as the available information allows, identifies potential
events and their effects, their probabilities of occurrence, and possible ranges. Uncertainties,
inherent in any risk assessment or mine plan, are clearly identified in the assessment.
Gordon H. Reeves. Ph. IX
Uncertainties and limitations are explicitly identified and acknowledged lor topics that I am
familiar with (fish and aquatic ecology and fish habitat) throughout the report These are
summarized succinctly and clearly and the consequences to the findings are articulated.
RESPONSE: No change suggested or required.
Charles Wesley Slaughter, Ph.I).
Yes. The authors fairly attempt (pp l-S-24-2(\ and in each chapter) to note the various
uncertainties and assumptions incorporated into the Assessment. Sections 8.5 and 8.6 briefly
summarize those uncertainties A question remains concerning the "uncertainties" associated
with assigning probabilities to \ arious failure scenarios; I remain unconvinced that those
probabilities have real meaning or signilicance for decision-making (see response to Question 5,
above)
RESPONSE:, l .v explained in the assessment (particularly in Chapter 14), the probabilities
have various sources and different interpretations, which we have tried to make clear. Some of
them are more useful in decision making than others, but all are the best values that could be
derivedfrom available information. Most are based on empirical frequencies, but some are
engineering goals or regulatory targets.
John D. Stednick, Ph.D.
The uncertainties are presented adequately.
RESPONSE: No change suggested or required.
Roy A. Stein. Ph. I).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 180
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Uncertainties. I think that the Environmental Risk Assessment did a nice job of identifying
uncertainties surrounding this presentation, but a relatively poor job of quantifying them (at least
partially due to a lack of information). As a consequence, I found this section more than
disconcerting. Certainly, the authors have worked hard to present an accurate portrayal of the
impact of a large-scale open pit mine in the watershed of Bristol Bay. Even so, upon review of
the list of uncertainties with regard to this effort (pages 8-10 to 8-13), I conclude that we know
little of what the impact of this mine will be in any quantitative sense. Clearly, from the
Environmental Risk Assessment, we do know qualitatively what is likely to occur when a mine
of this size and type will be put into operation in this en\ iionment
RESPONSE: The goal of the assessment is to evaluate risks from large-scale mining to
salmon populations and secondary effects of salmon tosses. II here applicable information was
available, we quantified our uncertainties to the extent possible. Hit ere information was
lacking, we provided a qualitative assessment. This approach also reveals missing information
that will be neededfor any future environmental assessment.
However, from the list of uncertainties, we arc operating al the outside edge (and beyond in
many cases) of the semi-predictive models used in anticipating the impacts of the mine footprint,
the routine operations of the mine, and the impacts of failures of TSF, pipelines, and
water/leachate collections on extant salmon populations. And our knowledge of the baseline
populations of the seven species of salmonids is 110 better, for we do not know the size, diversity,
distribution, or vital rates (i.e., recruitment, growth, and survival across life stage) of these fishes.
RESPONSE: The EPA agrees that we do not know everything we would like to know about
fish populations in the project area, but we do know the approximate extent of various species,
and we do know that spawning and rearing by salmon. Dolly Varden, and rainbow trout occur
in the project area. We summarize the types of risks to fish from plausible mine scenarios.
Couple these two sets of uncertainty and the prognosis outlined in the report is suspect, at the
very least, and soniew hat anticipatory at best (1 cannot bring myself to use the word
"predicti\ e") I fully realize that these are the cards the authors were dealt (I do applaud the
authors lor making the best of an information-poor environment), but it seems to me that we are
on tenuous ground when we attempt to predict the impact of the Pebble Mine on salmon,
associated wildlife, and Native Alaskan cultures in the Bristol Bay Watershed.
RESPONSE: EPA agrees that the assessment is not a definitive quantified prediction of all
impacts to salmon, wildlife, and. ilaska Native cultures of large scale mining. However, the
revised draft does develop more quantitative estimates of the implications of the mine
scenarios than the first review draft. The point of a risk assessment is to make the best
prognostication possible given the available information. The complexity of a large mine and
of the receiving ecosystems, as well as the unpredictability of the natural processes and human
errors that would challenge the integrity of a mine, make prediction tenuous even with the
most detailed mine plan. However, the assessment evaluates the likely impacts of a set of
plausible mine scenarios, thereby highlighting potential risks and impacts of concern. This
will inform decisions going forward and any NEPA or permitting processes in the future.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 181
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
William A. Stubblefield. Ph.D.
The risk assessment attempts to identify and evaluate the uncertainties associated with each of
the recognized potential risks. The authors have, for the most part, successfully identified a
number of uncertainties that may affect the accuracy and conclusions of the risk assessment.
Clearly, this information should provide a basis for prospective mine planners and regulatory
authorities to focus their efforts to minimize potential environmental risks. In some cases the
uncertainties identified are probably best addressed through the development of additional data
and this should guide future research efforts undertaken prior to mine development and
operation.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify the risks, only the likelihood of occurrence and the
consequences. See discussion under Question 4 aho\ e about risk. Uncertainties are identified and
evaluated for the likelihoods of occurrence and in some cases lor the consequences. However,
because the magnitudes of the risks are not expressed, llicir uncertainties are also not explicitly
expressed.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows.
The report identifies uncertainties in a number of sections, including in Chapter 8. In many cases,
these uncertainties arc expressed in qualitative terms and are not quantified. The biggest
uncertainly \ ariahi lil\ in the e\ alualion of a hypothetical project is associated with the potential
range of design features, waste management options and operational details that could be
included This was completely ov erlooked in the analysis by assuming a specific design for the
hypothetical mine. The failure likelihoods and consequences on salmonid fish are very
dependent 011 the assumptions lor the hypothetical mine. These uncertainties are neither clearly
identified nor included in the e\ aluations. This is a major shortcoming of the present analysis.
RESPONSE: The comment is correct in stating that the scenarios are a major source of
uncertainty with respect to what might be implemented if an actual mine were developed.
However, the scenarios are intended to represent a mine using modern conventional practices
and are based on the preliminary mining plan put forward by Northern Dynasty Minerals as
"permittable" (Ghaffari et al. 2011). Any future plans put forward would undoubtedly differ,
and a permitted mine plan would differ from that put forward by the mining companies. The
actual operation of a mine, in turn, inevitably differs from the approved plans. Rather than
attempt to estimate the uncertainties in predicting all of those changes in plan, the EPA chose
to put forward reasonable and typical scenarios, estimate some associated risks, and describe
the uncertainties associated with the risk estimation. Identification of the risks associated with
a typical mine can inform initial mine design and the NEPA and permitting processes. The
revised draft makes this clear.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 182
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Phyllis K. Weber Scannell, Ph.D.
The important features of the Environmental Assessment are to describe the fish, wildlife, and
human use of the subject area and to define possible risks from development of a large porphyry
copper mine. There are many uncertainties associated with the identified risks and most were
identified in the document. The document could be strengthened by putting a greater emphasis
on sources of contamination (such as mine seepage, poorly designed collection systems, exposed
pit walls, etc.) in relation to the permeability of the soils.
RESPONSE: The revised assessment includes more emphasis on sources of contamination
(e.g., diesel pipeline failure, quantitative wastewater treatment plant failure, and a refined
seepage scenario) and their potential hydrologic transport including through permeable soil
and rock.
The 5th bullet on page 8-11 outlines important uncertainties for protecting fish species. These
uncertainties include life-stage-specific sensili\ ities to temperature, habitat structure, prey
availability, and sublethal toxicities. These factors must be considered should a mining project
go forward.
RESPONSE: The EPA agrees with this comment. So change suggested or required.
The 6th bullet on this page discusses the preliminary nature of leaching test data. These tests
must be sufficiently comprehensive to predict both short term and long term water quality from
all sources, including l\\(i and NAG waste rock, pit walls, and pyritic tailings.
RESPONSE: The EP. I used the available leaching test data and agrees with the comments
about future tests.
Paul Whitney. Ph.D.
Uncertainty summary The discussion of uncertainties in the assessment is, in most cases,
appropriate. It seems that one could use these discussions as a scope for additional work needed
prior to an assessment that would properly assess risk of the example mine. As the uncertainty
discussion appears i 11 the assessment, this reader wonders what to make of it. There is a lot of
uncertainty in this world that we find acceptable; the ultimate goal seems to determine if the
cumulative uncertainly is acceptable or not. Such an evaluation remains to be made and I'm not
sure how it could be made based 011 the level of information presented in the assessment and the
current state of the uncertainty discussions.
RESPONSE: The goal of the risk assessor is to describe the uncertainty, not determine
whether it is acceptable. The acceptability of existing uncertainty is a judgment made by the
risk manager, who must decide whether to make a decision or defer the decision until more
information is available.
The summaries of uncertainty included in Sections 8.5 and 8.6 could be improved if some sort of
realistic and useful conclusion(s) could be presented. The Section 8.6 summary seems to "pile
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 183
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
on" uncertainties, rather than summarize the uncertainties in the assessment. While piling on is
informative, it is not the sort of summary I was looking for. Conclusions in the Section 8.5
summary of conclusions remind us that the effects of mining on fish populations could not be
quantified and, as a substitute, the effects on habitat were used as a surrogate. So we are left with
an estimate of 87.5 to 141.4 km of streams that would be removed and this would cause an
adverse effect. Based on this very general risk conclusion, we learn that "In summary, it is
unlikely that there would be significant loss of salmon subsistent resources related to the mine
footprint" (page 5-77, last paragraph). First, it's not clear how this conclusion was reached.
Second, if such a conclusion was possible for subsistent resources based on the data available,
why couldn't such a conclusion be reached for sport and commercial fisheries? Third, does it
follow that no significant loss to salmon subsistence resources would result in no significant loss
to wildlife that utilize this resource? If so, such an indirect analysis is not informative regarding
an environmental assessment for the example mine. Alternatively, a direct assessment of the loss
of habitat using habitat-based population models for both fish and wildlife would be much more
informative.
RESPONSE: Uncertainties are now discussed in each chapter of the assessment so they can
be more closely associated with a particular topic.
Additional information on the geographic scope of subsistence use in the area has been added
to the assessment and the conclusions clarified accordingly. . \ lthough there is some
(incomplete) information on subsistence use areas, similar data are lacking for the geographic
scope of sport and commercial fishing in the watershed.
There is not sufficient data to quantify the loss of salmon as a subsistence resource under
various scenarios or to quantify the fish-mediated effects to wildlife. Discussions of these
topics have been expanded to illustrate the complexity of the interactions and provide further
qualitative assessment. Direct effects on wildlife from habitat loss related to large-scale mining
are outside of the scope of this assessment.
Adapli\c Manaucmcnl I lolling (1978) in his Adaptive Environmental Assessment and
Management book discusses political uncertainty and how adaptive management might be able
to address the issue. Considering that the mine being proposed is a multi-century system, it's
poignant to realize that Alaska w as owned by Russia about 150 years ago and Oregon was being
claimed by the Spanish about 2<><> years ago.
RESPONSE: The commenter is correct in pointing out the inapplicability of adaptive
management to decisions like whether to permit large scale mine development in the Bristol
Bay watershed Not only is the time horizon inappropriate, but mine permitting decisions
cannot be remade iteratively like the resource management decisions for which Hollings and
Walters developed adaptive management (e.g., setting annual fishing limits). However, careful
monitoring and revision of permit conditions, although not literally adaptive management,
could have some of its benefits. Monitoring requirements and provisions for modifying permits
are beyond the scope of this assessment.
Adaptive management is a tool designed to deal with uncertainties in risk evaluations (Ruhl and
Fischmann 2010). If implemented properly, with testable hypotheses of risk, adaptive
management may be something to consider for the example mine. My experience with adaptive
management is that adequate funds are seldom allocated to learn by doing, to test hypotheses and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 184
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
to implement new management if hypotheses are not met. Formalizing financial instruments to
ensure that funds are available is equally difficult to negotiate. Nonetheless, I agree with Ruhl
and Fischmann that the theory of adaptive management is sound and may be the only way to deal
with uncertainties such as climate change. Considering the number of uncertainties identified in
the assessment, for the scope of work to clearly state hypotheses, to address the uncertainties, to
fund studies to test the hypotheses, and to fund alternative management if hypotheses are not met
is cumulatively daunting. For example, there are about 50 state variables in each of the Site
Model Figures (3-2A, B, C and D). All totaled, that's about 200 (50 x 4 figures) state variables
for salmon alone. Considering there are 100s more species, that's about 20,000 (200 x 100
species) state variables. Then there are fluxes/linkages between the state variables and that is
another 20,000 fluxes/linkages that should be monitored. For adaptive management to work,
clear goals and hypotheses to assess whether observed data meet the goals should be stated for
approximately 40,000 variables and linkages. Then there is the task of defining the monitoring
methods and statistics to determine whether or not goals for the \ ariahles and linkages are being
met. I acknowledge that there are probably ways lo trim down the monitoring effort, but one can
start to imagine the enormity of implementing a monitoring plan for adaptive management. Then
there is also what to do if goals are not realized I am not aware of any alternative to adaptive
management other than contingency planning which often lacks the "learn In doing" feature of
adaptive management.
RESPONSE: See response to previous comment.
Important wordsmithing. So many of the uncertainly e\ alualions make statements about certain
parameters that "could not be predicted" (page l-S-2<). para 4). "could not be quantified" (e.g.,
page 6-11, second full para), or are "unpredictable" (page 5-44. para, line 8). These are just a few
of many examples. It would he more acceptable, at least to me, to state that estimates were not
included in the assessment This type of wording occurs in some parts of the assessment and
might be more accurate
RESPO NSE: Changes have been made in Chapter ~ (formerly Chapter 5) and Chapter 9
(formerly C hapter (>).
Vague wording The assessment includes a lot language that seems vague, at least to me. The list
is long but includes "highly pure water", "other ecological responses"; "key wildlife"; "essential
wildlife"; "overall ecosystem functioning"; "serious population-level consequences"; "different
thermal characteristics", "could he locally significant"; and "very long time." Lackey (2001)
acknowledges the need lor scientists to communicate with the public using normative science but
expresses concerns that normative descriptors such as ecosystem health are subject to wide
interpretation. He suggests that the most direct alternative to using normative science is to simply
and clearly describe what is being discussed. The assessment would benefit if the normative type
words used above (any many more) were quantified with estimates, a range or some type of
measureable or testable parameter.
RESPONSE: The authors have tried to be precise with the language of the assessment and
avoid normative terms. Changes to the text of the revised assessment have been made with this
comment in mind.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 185
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Question 14. Are there any other comments concerning the assessment, which have
not yet been addressed by the charge questions, which panel members would like to
provide?
David A. Atkins, M.S.
Long-term risks from development of an open pit have not been characterized. It is difficult to
predict the chemistry of the lake that will form in the open pit. hut there is some potential that
water quality will be poor, which may be exacerbated In pil hacklilling with waste rock. The pit
lake could impact waterfowl and may have some impact 011 ground water if there is outflow when
the lake reaches an equilibrium level.
RESPONSE: The potential pit was not assessed because it is not anticipated that salmon or
other fish (the primary assessment endpoint) would reach it. Therefore, it is out of scope.
However, a pit lake as a potential source of water to streams is briefly considered in the revised
Chapter 8. The scenarios no longer include the placement of acid generating waste rock in the
pit at closure.
Steve Buckley, M.S., CPG
None.
Courtney Car others. Ph. I).
None.
Dennis I). Daub/e, Ph.I).
Based on public comments and discussions that took place by panel members in Anchorage
August 7-9 of this year, this report confuses in both intent and approach. Is the intent of EPA's
assessment to characterize potential impacts to the Bristol Bay watershed (title) or does it address
a more defined portion of the Nushagak River and Kvichak River watersheds (objective
statement)? Was the approach an "assessment" (a fairly broad term) or an "ecological risk
assessment" (suggests a specific scientific framework was applied to the risk/effects analyses)?
These shortcomings should be addressed in the final assessment document.
RESPONSE: We have revised the discussion of purpose, scope and endpoints in response to
this and other similar comments (see Chapters 1 through 5 of the revised assessment). The
assessment addresses multiple spatial scales, as detailed in Chapter 2. It is an ecological risk
assessment, but that term is usually shortened to assessment to make the document more
readable.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 186
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Gordon H. Reeves. Ph. IX
The major issue that was not considered in the assessment was the potential impact of climate
change, particularly regarding the form and timing of precipitation. Admittedly, there is
uncertainty about the magnitude of changes that will result from climate change, which makes it
difficult to consider. However, the potential consequences of climate changes on such topics as
tailing site facilities, water availability, and culvert failure seem appropriate. It will also be
important to consider potential impacts of climate change so their signal can be distinguished
from potential mine impacts during any monitoring thai occurs
RESPONSE: Climate change projections, including increases in precipitation, are addressed
in Chapter 3 of the revised assessment. Projected precipitation increases are used as the basis
for a qualitative discussion of the link among changes in precipitation, hydrology, and mining
including site facilities, water availability, and culverts (Box 14-2). This box also discusses the
importance of monitoring impacts of climate change to distinguish between potential impacts
of climate change and potential impacts from the mine.
Charles Wesley Slaughter, Ph.D.
I would simply re-emphasize that a truly comprehensi\ e assessment of the potential
consequences of a large-scale mineral extraction project, be it the "hypothetical" Pebble-like
project or a different cndea\ or. should fully consider both the immediate project-specific
impacts, and the long-term watershed-wide consequences of "ancillary" developments - such as
other mines, which might become economic once the primary project's infrastructure is in place.
There should be full recognition of the irreversible nature of such developments, and of the
potential and limitations of possible reclamation or mitigation measures for the full suite of
resources and ecosystem "services" involved, both short-term and long-term.
RESPO N SE: Ecological risk assessments generally have more limited scopes than
comprehensive environmental impact statements. This document is not meant to be a
comprehensive evaluation of all impacts potentially stemming from a large-scale mine, and its
scope has been more clearly defined in Chapter 2. We have added discussion to the cumulative
assessment chapter (C liapter 13) to emphasize the long-term watershed-wide consequences of
ancillary or induced development. We have added a discussion of compensatory mitigation in
Appendix J.
John D. Stednick, Ph.D.
There are several references to streamflow measurements that would be especially helpful to
better characterize the site. The US Geological Survey has some streamflow gauging stations and
precipitation records that would complement the analysis. Annual precipitation values were
derived apparently from a computer model used to analyze global climate change at University
of Alaska Fairbanks. How do these data compare to field measurements? The prediction of a 10,
50, 100, or larger event using a short-term precipitation record, results in a larger error term on
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 187
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the predicted streamflow. How common is the occurrence of rain on snow (ROS) streamflow
events?
RESPONSE: U.S. Geological Survey and PLP data are presented in Chapter 7 of the
assessment. We believe the SNA P data used in the assessment to be the best available for
summarizing precipitation inputs to large areas; these data have been updated from the May
2012 draft of the assessment. We used accepted, published data to estimate the size of the
probable maximum flood. We do not make estimates of 10, 50, or 100 year events.
Dust production and transport: A variety of mining processes will generate dust. What are the
wind patterns, chemical composition, and opportunity to land in surface waters or wetland areas?
What potential is there for metal or toxin transport? Ovcrhni clen removal will require explosives
that leave nitrate, ammonia, and often sulfur in the air. \\ hal ahout this transport? Or rain out?
RESPONSE: Stressors evaluated in the assessment were based on how they would
significantly affect our primary endpoint of interest (the region \ salmon resources) and their
relevance to EPA's regulatory authority and decision-making context. . Mthough fugitive dusts
from mining operations are a potential source of contamination to streams, and should be
considered in a regulatory permitting process, they are not considered as significant as other
stressors chosen for the assessment and are not regulated under the Clean Water Act.
Discussions of transport dust and associated potential impacts are now included in Chapter
10.
The literature cited is often dated or lacking The technical i e\ iew panel has proposed numerous
references that can be used to strengthen the document
RESPONSE: U e have reviewed all references suggested by the peer reviewers to the extent
possible, and have incorporated information where applicable.
Roy A. Stein. Ph.D.
• IN PERPETl ITY
Sustainable Salmon vs. One-Time Mine. Some irony exists as one considers the trade-off
between salmon and this mining operation (and make no mistake, we cannot have both
mining and prockicti\ e salmon stocks in the Bristol Bay watershed). We are trading
sustainable salmon slocks that, with science-driven management, rigorous regulatory
oversight, and limited exploitation, should provide salmon literally 1000s of years into the
future against the de\ elopment of a mine that will provide minerals in the relative short term
(within 25 to 78 years). As a result of the mining operation, the government (and likely it
will be the state or Federal government) will be saddled with a 1000 years (at a minimum,
based on the assessment) of monitoring and maintenance of this closed-mine site.
RESPONSE: No change suggested or required.
• MINERAL NEEDS
• Strategic Needs. I was surprised that no section of the Environmental Risk Assessment
included any justification for why copper (does mining copper fulfill a strategic need for the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 188
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
United States; most all sites I researched do not list copper as a strategic mineral for our
country), gold, molybdenum, and some additional rare earth elements were needed within the
context of our economy. Are other sources available? Are these specific elements in short
supply? Are they required for the United States to compete in worldwide markets regarding
cell phones, other electronic devices, or solar panels? For example, rhenium is used in the
aviation industry and some web sites suggest it is critical to our defense industry. Some
justification would have helped me understand this huge undertaking.
RESPONSE: The EPA agrees that the strategic needfor these resources should be
discussed in the broader NEPA environmental assessment process, but this topic is beyond
the scope of this assessment.
• ORGANIZATIONAL ISSUES
Page 5-59. I struggled throughout the document with organizational issues. As I read more
and more text, I had the sense that I had read these I acts or these perspectives previously. I
mention this above but it is here on page 5-5l> that the issue is nicely summarized. Note the
text in the last paragraph of the page, where it discusses all of the important issues associated
with roads and stream crossings. My suggestion would be that these sections (for every
topic) be combined such that one section would exist for Roads, one for the Pipelines, etc. In
so doing, the reader can capture all of the relevant information about a specific aspect of the
mine in a single section of the report I believe this would improve impact, readability, and
shorten the report substantially (and also ser\ e to reduce w hat seems to be a fair amount of
redundancy).
RESPONSE: H e have reorganized the document in response to this and similar comments
to improve readability and clarity.
Pages 6-10 to 6-11. These pages relied another example of redundancy. Text to this point
discussed and reviewed just how long we might expect the fine sediments to persist in rivers
and streams post tailings dam failure. Yel, here again, on pages 6-10 to 6-11, these numbers
are reiterated. Combining these sections would help the reader and reduce redundancy.
RESPONSE: The TSF failure analysis was formerly split between two chapters and has
now been combined into one (Chapter 9).
Section 6.3. Pages 6-36 lo 6-42. Not to beat a dead horse, but in this section, nearly all of
the citations to tables and ligures are to those tables and figures that are found in sections of
the report other than Section (x This organizational scheme is what makes the report
cumbersome to read and follow the logic and the argument.
RESPONSE: The assessment has been completely reorganized, as suggested by reviewers.
William A. Stubblefield. Ph.D.
None
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 189
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment mentions twice that "interactions with regional stakeholders" and
interactions with members of the Intergovernmental Technical Team were used to refine the
analysis, etc. (p. ES-2 and p. 3-6). A robust stakeholder process includes careful documentation
of the stakeholders identified during the project (which may include stakeholder mapping),
records of meetings (attendee lists, meeting notes, etc.), resolution of differences, etc. The EPA
Assessment does not contain any references to any such materials, which implies to me that the
stakeholder process was informal and not robust.
RESPONSE: Details on stakeholder involvement in the assessment process have been added
in Box 1-1.
Phyllis K. Weber Scannell, Ph.D.
At present, Pebble remains a prospect and there is no plan of operations lor the mine. Should the
project move forward to development of a mine, il will be necessary to develop an in-depth
mining plan of operations. The mining plan should include the following:
• Transportation - of equipment and personnel and for shipping ore. Transportation of
ore, including loading facilities, wheel washing, and oilier measures to prevent ore
spillage and contamination.
• Siting of mine facilities. including tailings ponds, waste rock storage areas,
concentrate storage area, bypass systems for clean water, and collection systems for
contaminated water
• Mill operations, including a description of the process for concentrating ore.
• Chemical and fuel storage and Spill Prevention and Contingency Plans.
• Personnel housing, including handling of domestic waste (sewage, garbage).
• Wilier treatment plant Processes that will be used, anticipated concentrations of
metals and I DS, anticipated discharge volumes, and predicted mass loadings.
• Monitoring plans for seepage from tailings ponds, waste rock storage areas, etc.
Monitoring likely will include a series of wells and possibly, a pump-back system.
• Predictions for acid rock generation and measures that will be put in place during
mining 1o minimize future seepage from the mine site.
• Plans for concurrent reclamation and future closure of the mine.
• Specifications lor sufficient bonding to provide site stabilization and water treatment
in the event of a premature or temporary shut-down and reclamation at closure.
RESPONSE: The commenter is correct that these are important points that should be
considered in evaluation of a mining plan once submitted No change suggested or required.
After the Mine Plan of Operations is developed, an environmental assessment plan should be
developed that identifies potential effects to fish and wildlife and their habitats from specific
components of the mine (as listed above). In addition, the assessment should include cumulative
effects of nearby mines (if appropriate) on fish and wildlife habitats and water quality.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 190
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The commenter is correct and we would expect these things to be considered
during the regulatory permitting process. No change suggested or required.
Among the most important issues that must be addressed are transportation, potential for acid
mine drainage and metals leachate, control of point and non-point pollution, and developing the
mine for future closure.
RESPONSE: The commenter is correct and we would expect these things to be considered in
more detail during the regulatory permitting process. No change suggested or required.
Paul Whitney. Ph.D.
Here are a number of other comments that do not fit neatly under the first 13 questions:
1. Unquantifiable wetland and riparian loss There are many parameters that are supposedly
unquantifiable. An example "unquanliliable" parameter is the area of riparian floodplain
(page ES-14). Another example is the "unquantifiable area of riparian floodplain and
wetland habitat that would be lost..(page S-2. first sentence). There are many methods
to characterize or delineate riparian floodplain and wetland. For example, information in
Table 5-23 and associated text on page 5-69 characterize wetland loss within 100 and 200
meters of the road. This is a rather crude method but it is at least an estimate. Perhaps it
would be accurate to say that the area of riparian floodplain and wetland habitat lost was
not characterized. Such a statement would be more internally consistent with statements
that the natural system is "incompletely characterized" (page ES-25). Better yet,
characterize the wetland and riparian habitat losses using one of the many existing
methods. Section S 2 is a good start. An explicit list of wetland and riparian loss estimates
seems important lor an N\\ re\ iew. Once a list of potential wetland and riparian losses is
tallied, the next question is whether or not it would be possible to compensate for such an
impact If it is not possible, it might not he possible to achieve no net loss of wetland
resources
RESPONSE: The EP. I agrees that greater quantification of wetland and riparian
losses is essential for the SEPA and permitting processes that oversee compensatory
mitigation and "no net loss". This assessment is based on existing information
(including wetland inventories) and, as the commenter has pointed out, reveals the
areas where existing information is lacking.
2. Combinations and Permutations of analyses. The review of the assessment is somewhat
complicated by the relationship of certain analyses to following analyses and so on. Take,
for example, the use of the Mount St. Helens eruption as a surrogate for a tailings dam
failure. One could assume that such an analogy is inappropriate and recommend that it be
deleted from the assessment. If this recommendation was followed for an initial baseline,
all subsequent information on distance moved, toxicity, remediation, and duration of
impact would no longer apply and would presumably be deleted. Alternatively, EPA
might further develop the Mount St. Helens eruption analogy and demonstrate why it is a
useful part of the assessment. If this is the case, the related following (i.e., secondary)
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 191
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
information presented on distance moved, toxicity, remediation, and duration would be
retained. Then, it is appropriate for a reviewer to comment on the adequacy of the
following secondary information. Then, it is a possibility that the secondary or following
information is not accurate and the reviewer has to consider that it could be dropped or
retained. Soon, a reviewer of the assessment document is wondering how to address an
unwieldy and hard to follow combination of initial, secondary, and tertiary events. More
likely than not, most reviewers do not consider all the combinations and
recommendations. As a result, the reviewers' comments could fall short of adequately
addressing the restructuring of the main assessment. In short, EPA should not assume that
following the reviewers' comments will result in an assessment that addresses the next
iteration of initial, secondary, and tertiary comments.
RESPONSE: The EPA believes that this comment raises an important point. We did
delete the analogy to Mount Saint Helens ash. as recommended by other peer
reviewers. However, as the comment indicates, that leaves the assessment with no
analysis of recovery of salmonid streams from the deposition of large volumes offine
particulate material.
3. Address all levels of ecology. Ecological resources can be characterized al many levels of
organization. Populations are often characterized In birth and death rates Communities
are often characterized by species diversity, succession, and associations of species.
Ecosystems are often characterized by structure, function, nutrient cycling, and energy
flow. From a wildlife perspective, the assessment does a line job of discussing marine-
derived nutrients but concludes that ".. .the lish mediated risk to wildlife - cannot be
quantified gi\ en a\ ailable data..." The assessment could be improved if information
regarding community and ecosystem parameters were quantified (or at least addressed at
the level that marine-derived nutrients were addressed). Woolington (2009) comments on
the importance of certain serai stages for wildlife. 1 am confident that an interview with
him. reclamation specialists, and others at ADF&G could provide a lot of information on
community succession, plant diversity, and wildlife habitat relationships. This
information, in addition to the insight provided in Appendix C, would provide a much
better understanding of possible mine impacts and opportunities for compensatory
mitigation. Tt is also possible that traditional knowledge of villagers might provide insight
to understanding plant community and ecosystem parameters.
RESPONSE: Ecological risk assessments are not intended to address all levels of
biological organization. Rather, as described in the Guidelines for Ecological Risk
Assessment (USEI* I / ')')$), they focus on a limited number of assessment endpoints
that are most likely to influence a decision because of their importance and
susceptibility. The endpoints were selected in consultation with stakeholders and
decision makers. Further, even if more consideration of the effects of marine-derived
nutrients on terrestrial plant communities were included, the salmon-mediated effects
of the mine on wildlife still could not be quantified. We have clarified the assessment
scope, and recognize that numerous other important ecosystem components also could
be affected.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 192
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
4. Ecosystem evaluation. The Ecosystem Integrity Section (2.3.5) seems to miss the mark. It
mentions the "nearly pristine conditions," with the caveat that approximately 70% of
salmon returning to spawn are commercially harvested. This is then described as a
managed and sustainable landscape. Maintaining a sustainable resource is not an accurate
characterization of a nearly pristine ecosystem. First of all, sustainability means a lot of
different things to different people. For example, forests in the Pacific Northwest are
managed in a sustainable way but the ecosystem is hardly pristine. Hilborn (2005)
discusses the multiple definitions of sustainability. Instead of deciding which definition
he likes the best, he indicates that sustainabilily is like good art - it's hard to describe but
we know it when we see it. Second, Hilborn (2<)i)5) stales "The record shows clearly that
almost all forms of human activity - agriculture, forestry, urbanization, industrialization,
and migration - reduce biodiversity of natural llora and lamia This is almost certainly
the case with fishing as well." Hilborn acknowledges the sustainability of the Bristol Bay
fishery but I doubt if he would claim that there is no impact of the lishery on the
ecosystem of Bristol Bay.
RESPONSE: Additional information on commercial fisheries management has been
added to Chapter 5 of the revised assessment. However, the purpose of this assessment
is not to assess the relative effects of potential mining and commercial fishing—it is to
evaluate potential effects on endpoints if a mine it 'ere to be developed, given existing
conditions and activities in the region.
Rather than emphasizing the nearly pristine conditions and sustainability of the fishery, I
suggest that other measureaMe characterizations of the aquatic and terrestrial ecosystem
be measured and quantified I-'or example, a description of plant communities and
succession would |iro\ ide the reader with a better understanding of how plant
communities are naturally maintained, and subsequently, how possible mining activities
might alter the successional processes I 'ailure to address and understand such
relationships led to unexpected consequences for down stream plant communities,
wildlife diversity, and village residents in the Peace Athabasca (Cordes 1975) Delta when
the Bennett Dam was built hundreds of miles up stream on the Peace River. The drastic
changes in the ecosystem function were related to changes in just a few inches of water in
the delta at certain times of the year. The Bristol Bay assessment would be improved if
the expected changes in the hydrographs and sediment transport were related to the
successional processes that maintain early successional plant associations such as the
alder-willow association, which is important for moose. The write up for moose
(Appendix C, page 33) indicates that these early successional associations are maintained
by bank scouring. It would be good to know the role of ice dams and anchor ice on
current levels of scouring and how changes due to mining might alter these important
sustainable functions. The assessment's emphasis on fish and fish mediated impacts runs
the risk of missing possible impacts on the terrestrial environment and down stream
sedimentation that would not only influence human culture but also the aquatic
environment. The impact on terrestrial resources and wildlife is not inconsequential to
aquatic resources, at least in my opinion. Throughout my evaluation, I cite references that
discuss the importance of specific associations between fish and wildlife.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 193
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: A comprehensive assessment of potential impacts of large-scale mining
would need to consider all levels of ecological organization. The scope of this
assessment, however, is much narrower, as described throughout the problem
formulation chapters (Chapters 1-6). We clarified the assessment scope, and recognize
that numerous other important ecosystem components also could be affected.
5. Multi-directional fish and wildlife relationships. The assessment's emphasis on marine-
derived nutrients and the reduction of this salmon-derived resource only looks at a one-
way fish to wildlife interaction. Wildlife to fish functions, such as beaver dam building,
are very important wildlife to fish interactions. Both salmonid fish, as well as forage fish,
receive benefits from beaver functions such as tree felling, dam building, and food
storage (Snodgrass and Meffe, 1997; Schlosser and Kallemeyn, 2000). This issue is
briefly mentioned in Section 5.2.1.2 (para 1) and discussed on pages 5-19 and 5-22, but
needs to be expanded to address the benefits for lish. The dynamic process of dam
construction and dam decay is important, not only for mo\ ing streams across the flood
plain, but also for creating a mosaic of plant associations and wildlife (e.g., moose)
habitat in and near the floodplain. In addition, such activities create a mosaic of habitat
for forage fish. An accurate characterization of the impact of a potential mine and road
necessitates, not only an assessment of the loss of lish on wildlife, but also the loss of
wildlife and their functions on llsli The influence of wildlife and terrestrial processes and
functions on fish discussed in the Cedei holm papers needs to be included in the problem
formulation step. The dynamic process of heaver dams causing creeks to move across the
floodplain should also be a criteria tor determining if and where culverts , versus bridges
or causeways) are installed for a potential road (pages 4-36 and 4-63).
RESPONSE: Additional text has been added discussing the functional role of beaver
dams in Chapter 7. The discussion of wildlife in Chapter 5 has been expanded to
acknowledge the complexity of fish wildlife interactions in the watersheds. Please note
that the assessment does not use beaver activity as a criterion for determining what type
of structure (bridge v.v. culvert) would be constructed at each crossing. The revised
assessment assumes that crossings over streams with mean annual flows greater than
0.15 m3/s would be bridged. However, the actual decision as to what type of structure
would be constructed at each crossing would be made by industry engineers in
consultation with state permitting staff.
6. (i.e.Measurinu lish w i Id I il'e interactions. Understanding and addressing the multi-
directional interaction of fish and wildlife are facilitated by an approach developed by the
Northwest Habitat Institute, introduced in Johnson and O'Neil (2000) and elaborated on
the Northwest Habitat Institute web site (nwhi.org). The Interactive Biodiversity
Information System (IBIS) database allows an assessment of fish and wildlife
interactions and functional characterizations (i.e., ecosystem functions). The IBIS
database is a logical extension of the Jack Ward Thomas Wildlife Habitat Relationships
and the GAP analysis for the Pacific Northwest. Such an analysis could build on
information in the Alaska GAP database and the Natural Heritage Program database in
Alaska.
RESPONSE: A comprehensive assessment of potential impact of large-scale mining
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 194
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
would need to consider all levels of ecological organization and interaction. The scope
of this assessment, however, is much narrower, as described throughout the problem
formulation chapters (Chapters 1-6). We clarified the assessment scope, and recognize
that numerous other important ecosystem components also could be affected.
7. Ecological risk assessment for toxic chemicals. The amount of information in the
assessment pertaining to ecological toxics risk is impressively large. The amount of time
and team expertise needed to adequately review this information is well beyond the scope
of the proposed review. If a team was assembled to review this information, I would ask
them to consider:
a. The applicability of a probabilistic risk assessment to address some of the
uncertainty associated with the deterministic information presented in the
assessment;
b. Stressors in addition to toxics (e g . habitat loss, noise, human disturbance, light,
and water warming in the winter), and
c. The relationship of stressors to populations in addition to no observed effect
levels.
I have worked on large scale impact assessments u here the stress of habitat loss far
exceeds the potential stress of toxics I would think l-IW would want to know whether or
not this is the case for the example mine in their assessment.
RESPONSE: The revised assessment evaluates the potential impacts from habitat loss
and toxic releases from plausible mine scenarios based on existing information about
the watersheds. The review pane! includes an expert on aquatic toxicology who did
provide an excellent review of that material.
8. My only regret in this re\ iew is that I could not visit the proposed mine site and
watersheds during the summer and the winter. 1 have flown over the Bristol Bay
watersheds w hi I c working lor other mines in Alaska, but I have never been on the ground
in these watersheds. 1 am concerned that if 1 did visit the proposed sites my assessment
points above might change and even change dramatically. It is hoped that all the EPA
staff and biologists working on the assessment are able to visit the watersheds in the
summer and winter.
RESPONSE: \o changes suggested or required
3. SPECIFIC OBSERVATIONS
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
David A. Atkins. M.S.
1. Global: All significant figures should be reviewed to make sure they are reflective of the
level of uncertainty (i.e., using an estimate of 141.4 km of streams eliminated when this
value is probably realistically +/- 50%).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 195
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: All significant figures in calculations and measurements made by EPA have
been reviewed and changed to reflect the uncertainty of the analysis. For example km of
stream is reported to the km Significant figures from cited documents are as reported in
the document.
2. Global: Many references cited in the text are not included in the reference list.
RESPONSE: References have been updated.
3. Global: The executive summary, main report, and appendices, in many instances, present
different information with sometimes different implications These three levels of detail of
information should be more cohesive.
RESPONSE: We have re-written the executive summary and main report to make them
more consistent with each other and the appendices.
4. Page ES-24 (P2): Suggest changing 'Cumulati\ e Risks' to 'Cumulali\ e F.ffects of Multiple
Mine Development.'
RESPONSE: The title of the chapter (now C liapter 13 was changed.
5. Page 3-2 (PI): The justification lor excluding ancillary de\ elopment from the assessment
should be better explained. In sonic instances, opening up an area for natural resource
development has had as much or more impact on the en\ ironmeiit and ecosystems as the
development itself (for example, oil and gas development in some areas of the Amazon
Basin)
RESPONSE: Discussion of ancillary mining infrastructure (in addition to the pit, TSFs,
waste rock piles, roads and pipelines) has been added to Box 6-1, and discussion of
induced development (development which follows initial development in an unimproved
setting) lias been added to C liapter 13.
6. Page 3-7 to 3-1 I The conceptual models are quite helpful, but are not referenced or utilized
sufficiently when discussing impacts It would be helpful to more fully incorporate them into
the assessment
RESPONSE: Sub-components of the conceptual models have been included at the outset
of the risk analysis chapters, to better frame the pathways evaluated in each chapter.
7. Page 4-1 (PI, L9): Why does the assessment describe current 'good' and not 'best' practice?
The rationale for this decision needs to be described. In addition, it is likely that anything
other than 'best' practice would not be permitted in this context.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. The terms are qualitative when generally
interpreted, or have a regulatory meaning. The term "best management practices" is a
term generally applied to specific measures for managing non-point source runofffrom
storm water (40 CFR Part 130.2(m)). Measures for minimizing and controlling sources of
pollution in other situations are referred to as best practices, state of the practice, good
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 196
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
practice, conventional, or simply mitigation measures. We have added a text box in the
revision Chapter 4 to discuss terms. Mitigation measures consideredfeasible, appropriate,
and 'permittable' (as per Ghaffari et al. 2011) were considered in the assessment, and
these are measures common to other copper porphyry mines.
8. Page 4-11 (P3, L10): The liner lifetime is quite low, and given the importance of this
assumption, would warrant more than a personal communication that does not appear in the
references (North pers. comm.).
RESPONSE: The discussion of liners has been expanded and moved to Chapter 4 (Section
4.2.3.4) in the revision. The personal communication reference has been removed.
9. Page 4-21 (PI, L4): Is the assumption about the TSI locations from the authors or from the
Wardrop 43-101 report?
RESPONSE: The location of the TSFs is a combination of alternative sites described in
Ghaffari et al. 2011 (the Wardrop report), in the NDM 2006 Water Rights Application (see
reference list for Chapter 6), and our knowledge of site characteristics suitable for tailings
impoundments.
10. Page 4-26 (S4.3.7-Water Managment): It would help to pn>\ ide appropriate ranges for
numbers (e.g., the precipitation at the mine and TDF is S<)3 and 804 mm/yr, respectively,
which implies an unrealistic degree of certainty). It would also be helpful to include a
diagram. I am uncertain why 'cooling tower' water losses would be included in the mine
water balance since power generation would likely be at a remote location and other impacts
from power generation are not considered in this assessment.
RESPONSE: In general, uv have presented our best estimates for the parameter values
discussed in the assessment and have used these estimates in our analyses and
calculations. . \ full uncertainty analysis is beyond the scope of this assessment, so we have
provided ranges only for critical parameters such as the mine size. We followed the
Ghaffari et al. (2011) mine plan in placing power generation on the site, which is the
reason for the natural gas pipeline. Therefore, water use for cooling is appropriate.
Net precipitation at the mine site has been recalculated. The monthly mean flows for each
gage were summed across the year, producing an area-weighted average of860 mm/yr.
Ghaffari et al. (2011) describe the construction of a combined-cycle natural gas-fired gas
turbine power plant at the mine site and estimate the cooling tower evaporative and drift
losses from the plant. The cooling tower losses were included because they would
constitute a substantial consumptive water use.
11. Page 4-31 (P3, L6): How was the filling time of 100 to 300 years for the pit estimated? What
constitutes full (e.g., within x% of the pre-mining water table)?
RESPONSE: The filling time was calculated from the estimated inflow rate to the empty
pit. Most of the groundwater infiltration would come from the uppermost 100 m and the
direct precipitation rate would be constant, so the filling rate would not decrease
significantly until the water level was close to the surface compared to the pit depths. We
recognize that as the water level approached the surface, the cone of depression would
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 197
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
shrink and the filling rate would decrease. We have not attempted to refine our analysis to
account for this decrease in rate and the available data do not merit such a refinement.
The filling time was simply calculated by dividing the pit volume by the inflow rate. In
practice, when the pit was full the pit level would be maintained below the overflow level by
pumping if treatment were required or be allowed to establish a natural outlet at its low
point if treatment was not necessary.
12. Page 4-52: The PMP is based on the Miller (1963) reference (not included in the reference
list). How was it estimated?
RESPONSE: The PMP was based on Technical Paper So. 4~
(http://www.nws.noaa.gov/oh/hdsc/PMP_documents.TP4~.pdf). This PMP was used as the
precipitation input to a Type-1 (for Alaska) HEL -IIMS simulation to calculate a runoff
hydrograph. 291 cms was determined to be the peak discharge from this 24-hour design
storm
13. Page 5-22: Should refer to 'recapture efficiency" rather than 'recoM.-ry rale'. How were the
values of 16% and 63% derived?
RESPONSE: The words "recovery rate" were meant to be descriptive and not as a defined
technical term. There is no standard term in widespread usage for the parameter reported
(i.e. the percentage of excess water available for reintroduction into the streams versus the
total amount captured by the mine processes). In the revised assessment, we have changed
the wording to "reintroduced" (Table 6-3, Section 6.1.2.5). The reintroduction percentages
were calculated as part of the water balance. The water balance totaled all the sources of
water captured per year and then subtracted all of the annual consumptive water losses.
The remaining annual volume of water is excess water that the mine does not need for
operations and which is available for reintroduction to the streams. The ratio of the
annual reintroduced volume to the annual captured volume yields the reintroduction
percentage.
14. Page 5-23 The mean annual unit runoff values are not reproducible from the values given for
drainage area and measured mean annual flow.
RESPONSE: These values have been corrected.
15. Page 5-32 to 5->> Tables showing flow changes for different mine sizes and figures for
minimum mine size are difficult to interpret. A different presentation method and/or narrative
description would help.
RESPONSE: Tables, figures, and text for this section have been revised
16. Page 5-45 (S5.2.3): The preceding section (Section 5.2.2) focuses on 'Effects of Downstream
Flow Changes' and Section 5.2.3 focuses on wastewater treatment. It is not clear why this is
the sole focus of Section 5.2.3. In addition, only a short paragraph is included in this section.
Certainly there are other possible risks beyond water treatment, and even this discussion is
too cursory given the importance of the issue.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 198
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Discussion of water treatment has been expanded in Chapter 8 (a new
chapter in the revised assessment).
17. Page 5-45 (S5.2.4): This section states a number of assumptions that could result in under or
over estimating impacts on stream flow from the mine footprint. This section leaves the
impression there is a lot of uncertainty, both with assumptions behind the estimation and with
how successful any attempts at mitigation may be. Therefore, we are left with very large
error bars on estimates that should be reflected in the numbers presented for loss of length of
streams and areas of wetlands.
RESPONSE: We have improved the rigor of the analyses regarding the impact of mining
on streamflow. We have attempted to be very explicit about our assumptions and
approaches, and we believe the analyses are appropriate and defensible.
18. Page 5-46 (P3): This paragraph discusses the possibility that estimates of stream length
blocked by mine construction may be oveiesli mated if engineered diversion channels are
successful. This is an important form of mitigation that needs to be e\ aluated further. It
would be helpful to evaluate mitigation efforts in similar types of systems to determine if
reconstructing streams is feasible and could he successful
RESPONSE: Appendix J includes a discussion of the efficacy of constructed spawning
channels. There is little evidence in the scientific literature to suggest that such channels
are effective at creating suitable spawning habitat. Furthermore, there is nothing in the
scientific literature that suggests salmon streams could be successfully reconstructed in
landscapes where significant alternations in area soils, hydrology and groundwater flow
paths have occurred, such as would be the case with the mine scenarios described in the
assessment.
19. Page 5-59 (Bullet 4) Whole effluent toxicity (WET) testing and downstream biotic
community monitoring would likely he part of any discharge permit. This requirement would
not preclude de\ eloping a better understanding of protective discharge chemistry and
temperature requirements before permitting, especially given the quality of the receiving
water.
RESPONSE:. \ppropriate II 7:7" testing and downstream biotic monitoring would be
desirable if a mine were permitted Additional studies of site specific chemistry and
temperature tolerances may also be desirable. No change required.
20. Page 5-60 (S5.4.2) The statement that mine traffic will not be a large enough volume to
affect runoff needs support. Do we know the road will only be used for mine traffic? Can we
estimate the volume of mine traffic for a mine of this size and then look at runoff from an
analog system?
RESPONSE: This statement has been deleted from the assessment. However, we found no
analogous data that would allow us to quantify runoff.
21. Page 5-60 (S5.4.4.1): How was the number of stream crossings determined (e.g., what
criteria were used to define a stream vs. a channel)?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 199
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: All USGS designated streams were included. No artificial channels were
identified and all natural channels are streams.
22. Page 5-65 to 5-68 (S5.4.8.2): Text states that 240 km of stream upstream of the transportation
corridor has a gradient greater than 10% and, therefore, is likely to support fish. Should this
be less than 10% (as stated in the header to Table 5-22 and Table 6-9)? If so, how was the
<10% value chosen?
RESPONSE: The comment is correct, though in the revised assessment we use <12% as
the criterion and cite a relevant source.
23. Page 5-74 (S5.5): Salmon-mediated effects on wildlife seems under-analyzed in the report,
particularly when compared to the information presented in Appendix C.
RESPONSE: The discussion of potential salmon-mediated effects on wildlife has been
expanded (Chapter 12).
24. Page 5-75 (S5.6): As above, salmon-media led effects on Alaska lSati\e Cultures seems
under-analyzed in the report, particularly when compared lo the information presented in
Appendix D.
RESPONSE: The discussion of potential salmon-mediated effects on Alaska Native
cultures has been expanded (Chapter 12).
25. Page 6-3 (S6.1.2 I) I don't find the Ml St 1 lei ens analogy useful.
RESPONSE: It has been removed.
26. Page 6-9 (PI, L I) Why would "present' resident and anadromous fish not suffer habitat loss
in the event of a TSI failure' Are they upstream of the area inundated?
RESPONSE: Yes. these populations are upstream of the TSF-affected area. However, they
may be impacted by barriers to seasonal movement or fragmented by loss of connectivity.
This has been clarified in the text.
27. Page 6-2S (So I 5): The "Weiuhing lines of Evidence' section is not well developed and not
particularly useful It also does not inform the risk characterization and uncertainty
discussion.
RESPONSE: The EP. I believes that it is important to weigh all relevant lines of evidence.
An explicit weighing allows readers and reviewers to see what was weighed and how it was
weighted. Such transparency is desirable in general. Therefore, the weighing of evidence
has been expanded and the method is better explained.
28. Page 6-30 (PI, L2): The statement that risks of failure of the gas and diesel pipelines are not
considered because they are not particularly associated with mining makes no sense. Without
the mine, there would be no need for the pipeline. And if the types of failures and risks are
well known, then this is one of the areas that could actually be assessed with some degree of
certainty.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 200
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Diesel pipeline failure has been added. A gas pipeline is still not assessed
because it is judged to not pose a potentially significant risk to fish (see Chapter 11).
29. Page 6-36 (S6.3, P2): Designating closure as 'premature closure,' 'planned closure,' and
'perpetuity' with water treatment ceasing immediately, continuing until permits are
exhausted or water is nontoxic, or until institutions fail does not seem reasonable. Any of
these closure scenarios would be planned, and would involve regulatory compliance reviews,
bonding, etc. Walking away without continuing to collect and treat water would be an
unlikely scenario. The issue of treatment in perpetuity is a larger issue that needs to be
treated in detail.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed. It highlights the need for adequate bonding and for plans that
include closure by the bond-holding agency.
30. Page 6-37 (P4, L4): The report states thai (acid generating) waste rock could be left in place
in the event of premature closure. This scenario should be addressed in the mine closure plan
and during the closure bonding process.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed. It highlights the need for adequate bonding and for plans that
include closure by the bond-holding agency.
31. Page 8-1 to 8-2: Given that this chapter discusses o\ erall risk to salmon, it would be helpful
to put the estimates of km lost in terms of the total stream or watershed available habitat. We
need some context and metric lor assessing significance
RESPONSE: In Table 2-1. uv have explicitly expressed the relative areas of each spatial
scale as % of the scale above it (e.g.. the % of the entire Bristol Bay watershed made up of
the Sushagali and Kvichak River watersheds, the % of the Nushagak and Kvichak River
watersheds made up of the mine scenario watersheds, etc.).
32. Page K-2 (SK 1.2) The locus on a lew types of catastrophic failures does not reflect the
current typical mining scenario Based on experience at other mines, it is more likely that
smaller-impact failures and accidents would occur during the mine life. It would be helpful to
use some current case studies to illustrate this point.
RESPONSE: The number of possible types, magnitudes and combinations offailures and
accidents is virtually infinite. Therefore, we used a bounding strategy. The potential effects
are bounded by those of routine operations and a realistic severe failure or accident. We
also cite case studies from recent reviews of mining failures and individual failure cases to
indicate some of the possibilities.
33. Page 8-3 (T8-1): Why would most concentrate pipeline failures occur between stream and
wetland crossings?
RESPONSE: Most of the transportation corridor is not on or adjacent to a stream or
wetland. Therefore, most failures would occur in those areas between stream and wetland
crossings.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 201
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Steve Buckley, M.S., CPG
1. Page xii: ICF is referred to in the document page xvi, but not listed as acronym or
abbreviation.
RESPONSE: ICF is the name of one of the contracting companies that has worked on the
assessment, and is only referred to as ICF in the introductory material (thus is not
included as an acronym or abbreviation).
2. Page ES-2 (P2, L5-6): "altered by geologic processes In would not degrade..." is unclear
RESPONSE: Metals do not degrade but their chemical form may be altered (via
speciation, binding, formation of salts, etc.). This is explained in the revised Executive
Summary.
3. Page ES-24 (P4, L2): "geologic defects" in unclear.
RESPONSE: This was clarified in the revised assessment.
4. Page 1-1 (P3, LI): "17 existing mine claims " should read "existing claim blocks"
RESPONSE: Corrected
5. Page 4-11 (P3, I. I <>) "\orlh pei s comm "" There is no reference, date, or information on
North.
RESPONSE: The discussion of liners has been expanded and moved to Chapter 4 (Section
4.2.3.4) of the revised assessment, and includes additional material and references on liner
lifetime. The personal communication reference has been removed.
6. Page 5-2
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
10. Page 7-7 (P2, P4): As above
RESPONSE: Corrected.
11. Page 8-1,2 (P2, P3): In the discussion of removal of stream kilometers and wetlands it would
be helpful to express these numbers in a percentage of the overall watershed stream
kilometers and wetlands to put perspective on these numbers.
RESPONSE: In Table 2-1, we have explicitly expressed the relative areas of each spatial
scale as % of the scale above it (e.g., the % of the entire Bristol Bay watershed made up of
the Nushagak andKvichak River watersheds, the "<> of the Xushagak andKvichak River
watersheds made up of the mine scenario watersheds, etc.).
Courtney Car others, Ph.D.
1. Page ES-2 (P3): "wildlife and the Alaska \ali\ e cultures of this region "
RESPONSE: The Executive Summary has been rewritten to incorporate this change.
2. Page ES-5 (P2): "Chief among 1hese resources are workl-duss commercial, sport, and
subsistence fisheries for Pacific sal moil
RESPONSE: Language added to the revised assessment.
3. Page ES-8 (last P) I Should Alutiic.| (SnupiiK|) cultural group also be included? Alutiiq
residents noted in Igiugig and kokhanok (Appendix I), p I 5). 2. Change 2nd sentence to: "In
contrast, the salmon hasc upon which indigenous peoples in the Pacific Northwest depend is
severely threatened "
RESPONSE: /. . Mthough there are residents of other cultural groups in the watersheds,
the predominant cultures are Yup 'il< and Delia "ma. so we have focused on these groups
but have added mentioned of the. Meut/Alutiiq. 2. This sentence has been edited.
4. Page (PI) "Salmon are integral to the entire way of life in these cultures as subsistence
food Jishtng and subsistence-based livelihoods, and as the foundation for..."
RESPONSE: Language added to the revised assessment.
5. Page ES-9 (P2): "52" n of the subsistence harvest, although for some communities this
proportion is substantially higher (e.g., noted to be as high as 82% on pg 93 of Appendix
D).
RESPONSE: Language added to the revised assessment.
6. Page ES-10 (PI): Could also add replacement value for subsistence resources or for salmon,
and the range of estimates for economic valuation of subsistence presented in Appendix E,
noting of that economic valuations do not fully capture the value of these practices.
RESPONSE: Language added to the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 203
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
7. Page ES-14 (#1): Are these all the fish spp. at risk, or only the one deemed to be
commercially, recreationally valuable? Subsistence spp also include others. Should make
clear what the focus is.
RESPONSE: This sentence refers to the sockeye being particularly at risk because they
spawn and rear below the road. Commercial or recreational value is not considered
8. Page ES-23 (P3): As noted above, other mines in Alaska (e.g., Red Dog) and oil and gas
development studies on North Slope may be useful to include predictions about how
subsistence practices will change with mining development and perceived impacts. Including
citations with these statements would be helpful.
RESPONSE: Chapter 12 has been expanded to include discussions of and references to
studies of impacts to subsistence practices from resource extraction industries.
9. Page ES-23,24 (last P): "if salmon quality or quantity is ad\ eisely affected (or perceived to
be affected)"
RESPONSE: Language added to the revised assessment.
10. Page ES-26 (last bullet): There is much data on cultural disruptions caused by the Exxon
Valdez oil spill, and cumulate e effects of oil and gas de\ elopment in North Slope region,
current salmon shortages in Yukon-kuskou kini Clearly subsistence is not about lost food,
but about lost lifeways, loss of practices, loss of teaching learning, and loss of identity. This
point could be made more forcefully While the specific impacts may not be entirely
predictable, there are likely outcomes that could he included based on experiences in other
regions of the state and or world
RESPONSE: The point suggested was made more forcefully in Chapter 12 of the revised
assessment.
11. Page 1-2 (P2) "this assessment does not pro\ide an economic or social cost/benefit
analysis
RESPO NSE: Language added to the revised assessment.
12. Page 2-15 (PI) Other important subsistence fish spp not listed in Table 2-5, e.g., whitefish
and winter freshwater lisli are listed as integral subsistence species in Appendix D. Again
make focus here clear
RESPONSE: Scope of the assessment and the specific assessment endpoints considered
have been clarified in Chapters 2 and 5. Table 5-1 has been incorporatedfrom Appendix B
to give a more complete view of the fish species found in the watershed.
13. Page 2-18 (S2.2.4): The net economic valuation ranges presented in Table 73, Appendix E
would be helpful to include here.
RESPONSE: Because the assessment is not meant to be a cost-benefit analysis, we only
briefly consider the economic value of the assessment endpoints. In addition, Table 73
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 204
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
deals with the economic value of moose, caribou, and brown bear, which are only
considered in terms of fish-mediated effects.
14. Page 2-19 (last sentence): "because no alternative food sources are economic viable." This is
a bit of a misrepresentation. The point is that people choose to live subsistence lifestyles.
Even if food at the stores was cheap, many would choose not to substitute for subsistence
hunting, fishing and gathering. This narrow economic framing misses the cultural and
lifestyle component of subsistence, and frames it merely as food procurement. This is not the
case throughout the document, but in this instance I would suggest changing this sentence to
reflect the irreplaceability of the subsistence lifestyle (dependent on access to high-quality
foods) rather than the economic viability of substituting alternative food sources.
RESPONSE: The revised assessment expands the discussion of the cultural and lifestyle
importance of subsistence (Chapters 5 and 12). The economic viability of alternative foods
is mentioned in Chapter 12, but in the context of a variety of benefits of subsistence.
15. Page 2-20 (first sentence): Here and in Appendix I), the legal framework for federal and state
definitions of subsistence should be clarified Se\ eral times in Appendix D an indigenous
subsistence priority is noted (e.g., pg 88: "No oilier state in the United Stales so broadly
grants a subsistence priority to wild foods to indigenous peoples as does Alaska."). The
authors should clarify what they mean by indigenous preference (i.e., as opposed to rural
preference?) in state and federal subsistence management They should include particular
references and additional clarifying in formation
RESPONSE: Citations were included in Appendix I) in response to this comment.
16. Page 3-2 (PI) "would be benign or have no effect on the environmental or social systems,"
RESPONSE: Discussion of scope has been rewritten (Chapter 2).
17. Page 3-4 (PI) " pro\ ide subsistence lor Alaska Natives and others." Particularly because
subsistence is defined as a rural light in Alaska, all subsistence users should be included as
potentially affected groups.
RESPONSE: References to non-. Maska Natives who practice a subsistence way of life
have been included in the assessment. However, the focus remains on Alaska Natives.
18. Page 3-11 (Fig 3-21-) This conceptual model appears less developed than the others. It
would interesting to work on expanding it out to include missing dimensions; e.g., add health
and healing activity (in addition to nutrition), cultural continuity (alongside social relations
and linked to language and traditional ways of teaching). With a decrease in economic
opportunities comes an increase in reliance on transfer payments. Overall it is a nice
illustration, but strikes me as less complete than the others.
RESPONSE: The conceptual model has been revised and expanded in response to this
comment.
19. Page 4-15 (Table 4-3): Estimation of 200,000 metric tons of ore processed per day is much
higher rate than any of the other mining operations listed in Table 4-4. Is this due to the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 205
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
low/moderate quality of the ore?
RESPONSE: Initially, EPA utilized the ore production rate that was available in Ghaffari
et al. (2011), but we have added a smaller sized scenario to the revised assessment to be
similar to the worldwide median-sized porphyry copper mine. Processing throughput is
generally a function of the quality of ore and the cost of mining and processing the ore.
Ghaffari et al. (2011) state: "The process is based on conventional grind-crush-float
technology, using proven plant equipment of the largest sizes that have been industrially
installed", which suggests that mining rates are limited also by the size of proven mineral
processing equipment.
20. Page 4-21 (last P): 208 m high dam is "much higher than most existing tailings dams." What
are average dam heights? Or how much higher than most existing tailings dams? Does this
high height affect probability of failures?
RESPONSE: Table 4-1 presents dam heights at other Alaska mines, all of which are
significantly lower than 209 m. Higher dams mean larger impoundments, storing more
tailings; thus, if failure occurs, more waste material can be released.
21. Page 4-23 (P2): "a well field spanning the valley floor." This is unclear. Could it be added to
Fig 4-7? How often would ground water be monitored0
RESPONSE: The revision clarifies this sentence, adding that installation would be at the
downstream base of all tailings dams to monitor groundwater flowing into the valley. It
was not added to the revised figures primarily to keep the figure from becoming too
"busy".
22. Page 5-48 (P1): "effluents would be required to meet criteria." How different is treated
discharged water from unaffected water0
RESPON SE: Estimates of wastewater composition have been added to the assessment
(Table 8-9). Background surface water data are presented in Table 8-10.
23. Page 5-5l> (2nd bullet) Is there any information available on ore processing chemicals, how
much are used, and likely toxicities?
RESPONSE: The toxicity of sodium ethyl xanthate, the primary processing contaminant
of concern, is now considered (Section 8.2.2.5). However, its concentration in tailings and
concentrate slurries cannot be estimated with any reasonable confidence due to the lack of
publicly available information.
24. Page 5-76 (bullet list): The list of cultural factors that may be negatively impacted could
include others: individual, community, and cultural identity; sense of place and place
attachments; community sustainability; cultural unity/conflict avoidance.
RESPONSE: The list of cultural factors that could be negatively impacted was expanded
in the revised assessment.
25. Page 6-46 (bullet list): In addition to the two listed, another should be added noted that
subsistence practices (harvesting, processing, sharing, consuming) are important for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 206
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
psychological, social, emotional, and cultural health and well-being.
RESPONSE: This section was clarified to focus on nutritional benefits. The discussion of
health benefits from subsistence practices was expanded elsewhere.
26. Page 6-47 (PI): ".. .the physical, psychological, social, and cultural benefits of engaging in a
subsistence lifestyle.
RESPONSE: Revision made.
27. Page 6-47 (PI): References should be added (and were included earlier in the report) for the
statement: "would likely employ a small fraction of Alaska Natives."
RESPONSE: This section has been revised and references the experience of other Alaska
Native communities.
28. COMMENTS SPECIFIC TO APPENDIX I)
a. Single-space for consistency with the rest of appendices
RESPONSE: Appendix D has been single spaced.
b. The title is a bit misleading Only eight pages in the report discuss traditional
ecological knowledge, and here not in much depth
RESPONSE: The TEK section was expanded
c. The research design, methods, and data analysis should be described in more detail.
Clarify sampling procedure (both lor communities and individuals). For example, it is
unclear if younger generations, particularly acti\ e subsistence harvesters, were
targeted as well as elders and culture bearers Interview protocol should be included
clearly as ail Appendix
RESPO.XSE: The methodology section has been expanded. Villages were asked to
identify Elders and culture bearers for interviews. Younger generations and/or
subsistence harvesters were not targeted but could have been included by villages
for interviews.
d. This section may make a lew o\ ^ statements (e.g., "only in Alaska are wild salmon
abundant")
RESPO NSE: comment noted and taken into consideration during revisions.
e. P12 - "Ihose outside of the state." Change to "outside the region," as many urban
Alaskans are not familiar with subsistence communities.
RESPONSE: The text has been revised.
f. P12 - "Since the questions dealt with a cultural standard, there were few alternative
points of views." Should cultural agreement be a matter of investigation rather than
assumed? This statement needs to be justified. Perhaps with the authors' 40+ years of
experience working with these communities they have come to expect cultural
agreement, especially among elders. If this is the case, that should be clarified. To
what extent did group interviews (2-6 people interviewed together, except for one
single interview) also contribute to cultural agreement? These details are important
given that the results are given on an agree/disagree format.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 207
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The text has been revised to clarify this point.
g. P17 - 2,738 is listed in Table 2 and 2,329 is listed here.
RESPONSE: The discrepancy has been corrected.
h. P19 - Here is perhaps another example of overstatement - 100% of the population
has access to waters of the rivers and lakes. What is meant here? For subsistence, this
access depends upon having transportation and gear or social relations. Do 100% of
people have this in this region?
RESPONSE: The text has been revised and a table added to clarify that the topic is
access to drinking water.
i. P20 - Reword "the archaeological work is largely due lo five projects."
RESPONSE: The text has been revised to clarify.
j. P26 - "located along a salmon stream indicates salmon were likely a primary
resource."
RESPONSE: Text has been revised.
k. P31-32 - Several of these quotes focus on social changes (e.g., elimination of dog
teams, relationships to commercial fishing changing overtime). People likely harvest
less fish now because they do not support dog teams, yet now they need more money
for fuel and equipment. These are important considerations for understanding
contemporary mixed economy. These points are mentioned in this cultural
characterization, but perhaps could be made a bit more clearly. At times, even the
contemporary characterization reads a bit like "timeless"' traditional cultural
relationships to the land and resources, yet it is important to accurately characterize
the subsistence-based communities in their full contemporary realities and
complexities.
RESPON SE: The quotes in the section focus on history and culture.
I. P34 "Large disruptions to the population have not been documented to occur until
epidemic
RESPONSE: Text lias been revised to clarify,
m. P34 5 Both kashgee and qasgiq are used for men's house - it is also defined three
times o\ er these lust lew pages of this section.
RESPONSE: The text was revised.
n. P35 - "earlier bow and arrow wards" should either be explained or omitted.
RESPONSE: Reference omitted from the revised text.
o. P38, first full paragraph, last sentence - What is meant by "observe the practice?"
This general statement is not adequately supported. Authors should provide specific
instances or more discussion if this point is to be included. As written, it risks
conveying a static view of TEK and practice and culture. Many indigenous
communities in Alaska, e.g., Kodiak villages, while exploited by a colonial economic
system, also strategically adapted to benefit from those systems in ways compatible
with their village lifestyles (e.g., cannery and village co-dependencies that elder
fishermen in this region remember fondly; Carothers 2010). It would be helpful to
have more information on this context in this region (e.g., Hebert 2008; Donkersloot
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 208
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
2005).
RESPONSE: Text omitted from the revised draft.
p. P40 - More information would be useful on Alaska Native participation in
commercial fishing in this historic period up through the present.
RESPONSE: Additional reference to Alaska Native participation in commercial
fishing has been added to the assessment text.
q. P47-48 - Ellam yua and tnughit are defined twice.
RESPONSE: Comment noted. No change made.
r. P81-84, Table 9 - Second/third part of questions no explained. Since this is an
agree/disagree table, remove other questions lor u liich no information is presented.
All questions would ideally be contained in an inler\ ieu protocol attached as an
appendix.
RESPONSE: This change has not been made. The authors chose to both identify
the interview questions and the summary responses in this table. As identified in
the introduction the semi-structured interview format was not a "survey" in which
one would ask everyone the same set of questions (as required by structured
interview protocol), but the protocol called for respondents to talk about the subject
elicited by the question. This is a very standard interview technique in
anthropology.
s. P87 - 'non-monetize' - but important to note that the modern subsistence economy
now depends upon cash inputs (ATYs. boats, snow machines, gas, parts, repairs,
guns, nets, etc )
RESPONSE: The relationship between the modern subsistence way of life and the
market economy has been acknowledged in the appendix and the assessment text.
t. PSS I'irst full sentence, last sentence poorly worded.
RESPONSE: This section has been slightly rewordedfor clarity.
u. PS1)-1)!) The subsistence discussion is confusing.
RESPONSE: No suggestions for revisions were made.
v. Pl)2-l).v Tables I pdate with recent data if possible.
RESPOSSE: Tables 12 and 13 contain the most recent data available to the public.
w. P100 - If percentage of working age population not in labor force is a better measure,
it should be included rather than official employment rates (or in addition to).
RESPONSE: These data were added to the appendix.
x. PI 10 - "Villagers in the study also eat store-bought foods, but do not prefer them" -
make clear again that most residents interviewed were elders or identified culture
bearers. A concern for many subsistence villages in other regions of Alaska is the
displacement of younger generations from fish camp and other subsistence practices,
and preferences for store foods, particularly candy and soda. If this region is unique in
that regard, make that clear here.
RESPONSE: The text was revised to make it clear that the interviewees expressed
preference for subsistence foods. The section on nutrition was also expanded.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 209
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
y. Section C "Physical and Mental Well-being" - Subsistence for emotional/mental
health should be added as a sub-section here. Given the high rates of social problems
in Alaska Native villages (e.g., suicide, violence, addiction), many cultures talk about
subsistence practices as being healing activities or producing emotion, spiritual and/or
mental health. This important aspect isn't covered in the other sub-sections.
RESPONSE: The discussion of suicide was expanded and a discussion on
behavioral and mental health was added to the "Social Relations" section. The text
of the assessment was expanded to acknowledge the social and spiritual aspects of
the subsistence way of life.
Z. PI 13 - Makhoul et al. is listed as 2010 in references
RESPONSE: This error was corrected.
aa. PI 14 - Change Local Wild Fish and LocalI'radices. and "ecologically, socially,
culturally, spiritually, and possibly e\ en e\ olutionarily "" Point is that subsistence
salmon are not just vehicles for protein and nutrition, but form the basis of incredibly
important subsistence ways of life thai are irreplaceable.
RESPONSE: The revision was made and the discussion of social and spiritual
values of the subsistence way of life was expanded in the appendix and the
assessment.
bb. PI 15 - Add 'cultural and social disruption" to the list of risks.
RESPONSE: This revision was made.
cc. P152, 2nd and last bullet points These are risks of mining development, not of
decreased quality/quantity of lisli (defined as outside the scope of this assessment).
The last bullet point would apply to fish-effects if reworded - some community
members may decide it is not sale to eat lish, causing factions of those who express
concern and those who do not Others to possibly include: cultural loss as younger
generations do not learn the practices of subsistence; stress on other areas and
communities of the region where people may target subsistence resources; and health
risks of eating contaminated fish.
RESPONSE: C oiiuiient noted. These topics were discussed in the assessment text
(C hapter 12).
dd. PI50 Sing to sign
RESPO N SE: C orrection made.
ee. Several grammatical errors throughout.
RESPONSE: Every effort was made to correct grammatical errors.
29. COMMENTS SPECIFIC TO APPENDIX E
a. P9 - Components of total value should include indigenous homeland for Alaska
Native cultural groups.
RESPONSE: The value of indigenous homeland and the value of being able to
bequest that homeland and cultural traditions onto future generations is included
in the net economic value analysis presented in Section 5. Section 5.2 discusses
and attempts to quantify the cultural value of engaging in a subsistence livelihood,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 210
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
referred to as the "activity value". Additionally, Section 5.6 discusses the
importance of the existence and bequest value of the resource, and the challenges
encountered in estimating these values.
b. P12 - Clarify usage of Aleut (Alutiiq/Sugpiaq?).
RESPONSE: The Bristol Bay region includes Alutiiq and Sugpiaq Native
Alaskans. This has been clarified in the text of Appendix E.
c. P22 and 26 - Change Boraas citations to Boraas and Knott.
RESPONSE: This change has been incorporated.
d P32 - Much of recreational use is non-market and could be included in the list at end
of 2nd paragraph.
RESPONSE: Because we do estimate the value of recreational fishing, sport
hunting, and non-consumptive wildlife viewing, these are not identified as non-
market values in this paragraph.
e. P96 - Citation for typical crew share of 10%?
RESPONSE: This estimate is based on the author's interviews with crew members.
Appendix E has been revised to make this clear.
f. P122 - Reasons for differences i n earnings helu een local residents and others is
important. The mixed subsistence-cash economy and cultural ideas about commercial
work in this region may offer an explanation. See koslow 1986, Langdon 1986,
Carothers 2010.
RESPO NSE: These references have been added to , 1 ppendix E and a brief
discussion has been incorporated in the document.
g. P134 -1 gashik. Egegik. and South \aknek have over 30.
RESPO N SE: Thank you for noting that these communities have over 30 permit
holders per /00 residents. II e have emphasized this in the text.
Ii. P13(\ last paragraph This paragraph seems abrupt/misplaced. A more thorough
discussion is needed here to include these points.
RESPONSE: This paragraph describes information in Table 39, which appears
immediately below the paragraph. The earnings data in Table 39 is from the US
Census as well as the. \ laska Commercial Fisheries Entry Commission. As
earnings data is limited and the reasons for variation in earnings are not apparent,
an additional discussion was not incorporated.
i. PI78, Section 4 3 No discussion of role of regional and village Native corporations
or the Community Development Quota program for federally-managed fisheries.
RESPONSE: Section 4.3 provides a broad overview of the economy of the Bristol
Bay Region. Although specific entities and programs like the Native corporations
and Community Development Quota program are not named, the relative
importance of government at a state, regional, and local level is considered. There
are many other entities that participate in the economy and government programs
that support it. This model of the regional economy does not identify specific
government entities or programs.
j. P191 - While the majority of formal sector jobs are taken by nonresidents, may want
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 211
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
to note that local economy - subsistence - is all local and highly dependent on
resources of the region.
RESPONSE: We appreciate this comment, and this section of Appendix E
discusses the local and regional importance of subsistence in paragraph 3. Because
this information is covered in detail in paragraph 3, additional information was not
incorporated into the paragraph summarizing jobs captured by the market.
k. PI93 - 2009 is mentioned as an unrepresentative year and given a sensitivity ranking
of "high." More information should be included on the anomalous 2009 - in what
direction should we expect to interpret data from this year compared to more average
years, or those at other ends of the extremes''
RESPONSE: The yearly variation in the commercial fishing industry is described
in detail in Chapter 3. The 2009fishing data is given a sensitivity rating of high as
the data presented in Chapter 4 estimating commercial fishing wages andjobs rely
on inputs from the 2009fishing season.
There is no reason to believe that 2009 is inherently an unrepresentative year for
the fishery, but it is impossible to know if it is a representative year for the future
performance of the fishery. In order for the wage and job estimates presented in
Chapter 4 to be representative of future performance, the 2009 data used to
represent the base year would need to also be representative of future performance
of the commercial fishery.
To clarify that this is a potential bias as the future of the commercial fishery cannot
be predicted, the text associated with this item has been rephrased.
I. P195 -dumber of households engaged in subsistence - ADF&G data should provide
estimates.
RESPONSE: The analysis relies on data from ADF&G; however, the item listed
here identifying the potential bias of the number of households engaged in
subsistence acknowledges that the data from ADF&G is purely an estimate. Should
the data relied on from ADF&G be found to be inaccurate for some reason, this
analysis that relies on those data would be biased accordingly.
m. PI ATV, snow machines, should be added to "boats and trucks"; work by Robert
Wollc and others (Wolfe et al. 2009) suggests that about one third of households in
Alaska Native villages harvest the majority of subsistence foods (and share,
especially with the least; active households). How does this finding affect these
estimates'
RESPONSE: II e recognize that ATVs and snow machines are frequently used in
the subsistence harvest. The original data source used to develop these expenditure
estimates did not specify expenditures for these machines, and thus they are not
identified in the quantified expenditures. In this analysis the total expenditure
estimates drive the analysis, not the composition of what these expenditures
purchased. Differences in the composition of expenditures would have a negligible
bias, as identified in Section 4.8.
The finding from the Wolfe 2009 citation provided above would not affect these
findings. This analysis relied on the assumption that all native households are
participating in the subsistence harvest, and estimated expenditures based on that
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 212
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
assumption. If some native households are more active than others, the
expenditures by less active native households will be lower than those estimated
here, but this will be offset by higher expenditures by more active native
households.
n. P202 - Explain why % of adults with 4+ years of college is used in this model? The
model was not explained clearly enough for me to understand it.
RESPONSE: The percentage of adults with four or more years of college
education is used in the model along with other relevant regional variables to
predict adjusted gross income per capita for each community. This instrumental
variable is included in the analysis because it is a variable that is relevant for
predicting adjusted gross income per capita in each community. For additional
information on the analysis used, please see Duffield (1997).
o. Some fisheries, e.g., crab fisheries, are not included in the economic analysis, yet
depend in part of Bristol Bay ecosystems, as discussed in Appendix F.
RESPONSE: As noted in the assessment, there are other species present that
support a commercial fishery, such as the crab fishery the commenter notes.
Although Appendix E is predominately focused on the importance of the salmon
fishery, some considerations for the commercial and subsistence value of these
fisheries is included.
Section 4.3 of Appendix E provides an overview of the regional economy, which
fish harvesting and processing. Table 53 estimates employment in the fish
harvesting and processing sector. As identified in the footnote to Table 53, these
employment estimates include fisheries in addition to the salmon fishery.
Appendix E also acknowledges and considers that subsistence harvests may include
species other than salmon. In Section 2.2 of Appendix E the general distribution of
subsistence harvest by Bristol Hay residents is presented. As identified in this
section. Bristol Bay subsistence harvest includes salmon, non-salmon fish, birds
and eggs, vegetation, marine mammals, marine invertebrates, and land mammals.
Subsistence expenditures are estimated based on the number of native households
and expenditure estimates from Goldsmith (1998). Goldsmith's expenditures
consider subsistence expenditures more broadly, thus these other species are
included in the economic analysis.
p. References Peterson et al. 1992 and Brown and Burch 1992 not included in
references.
RESPONSE: C liange has been incorporated.
30. COMMENTS SPECIFIC TO APPENDIX G
a. Mitigation measures are largely concluded to be ineffective. Would be helpful to
compare mitigation measures and their success/failure in other mining examples.
RESPONSE: Appendix G contains available information about mitigation
measures generally related to roads and pipelines. Mitigation measures are not
specific to mining operations.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 213
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
31. COMMENTS SPECIFIC TO APPENDIX H
a. P7 - Exposure of groundwater and waterfowl to chemical contaminants are listed as
main environmental concerns from tailings storage facilities. Impacts to human health
from ingesting contaminated water or birds. Clarify in report that direct risks to
human health are not accessed (only through reduction or elimination of subsistence
harvests?).
RESPONSE: The assessment language has been clarified to acknowledge that
there are potential direct risks to human health, but that the scope of the
assessment is limited to potential effects to indigenous cultures related to salmon.
An evaluation of human health effects from ingestion of contaminated non-salmon
subsistence foods is not within the scope of the assessment. EPA expects that a full
evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
direct effects.
Dennis D. Dauble, Ph.D.
1. Appendix B, Page 30: Table 1. I suggest adding a column lo indicate relative abundance, for
example, if individual fish species listed are abundant, common or rare. Also, are there
known differences in distribution and abundance for the Nushagak and Kvichak watersheds
relative to those watersheds unlikely to be affected In mining activities?
RESPONSE: This information, when known, has been added to Appendix B.
2. Appendix A, Page 42 The statement that diminished salmon runs present a "negative
feedback loop" where spau nor abundance declines, appears to conflict with the last
paragraph on page 41.
RESPOXSE: The text describes bottom-up effects ofMDNon stream ecosystems and
points out that these linkages have not been empirically established The negative feedback
loop mentioned in the following paragraph could result from reductions in salmon-based
resources that promote either bottom-up (i. e., a reduction in salmon-derived N and P) or
direct (i.e., a reduction in salmon eggs andflesh that can be consumed by fish) benefits to
juvenile fishes. Further, the fact that bottom-up effects ofMDNhave not been firmly
established does not negate considerable circumstantial evidence for bottom-up nutritional
deficits in Columbia Basin spawning streams. Edits have been made to help clarify.
3. Appendix F, Page 3: Is there significant sediment transport from the Bristol Bay watershed to
the Nushegak and Togiak Bays/estuaries?
RESPONSE: A new section was added to Appendix F that discusses the importance of
estuary habitat to salmon populations.
4. Appendix F, Page 7: What is the juvenile salmon resident time in Bristol Bay? How quickly
(and at what size/time of year) do they move from shallow nearshore to offshore habitats?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 214
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: A new section was added to Appendix F that discusses the importance of
estuary habitat to salmon populations.
5. Page 8-15 (L2): Suggest deleting "likely." There will be impacts.
RESPONSE: We believe this language is consistent with the uncertainties explained in the
assessment. No change has been made.
Gordon H. Reeves. Ph.D.
1. Page 6-9 (PI): Why would resident fish not "suffer" immediate loss of habitat as a result of
dam failure like anadromous fish would?
RESPONSE: This statement refers to fish located upstream of TSF failure and affected
downstream reaches. This has been clarified in the text.
2. Couldn't fish move to another stream if a cul\ ert is blocked and pre\ enls upstream
movement? This may eliminate the fish from a particular stream for a year but: it might not
reduce the overall productivity.
RESPONSE: This comment is based on the assumption that spawning and rearing habitat
are not limiting. We believe that they area.
Charles Wesley Slaughter, Ph.I).
1. Global: Precision of lull color \ ersions of all figures would have been helpful to this review.
The selected color \ ersions supplied here u ere useful - we should have had them all.
RESPONSE: Comment noted.
2. Page 2-4 Color codes are confusing use of different colors for same "moisture state" in the
five regions doesn't make sense (to me)
RESPONSE: Figure 2-2 (Figure 3-1 in the revised assessment) depicts hydrologic
landscapes across the Nusliagak and Kvichak watersheds separated into their
physiographic divisions as defined by Wahrhaftig (1965), combined with their climate
class as defined by Feddema (2005). Each color group (e.g., yellow, green, red, etc)
represents a separate physiographic division while color intensity (e.g., light green to dark
green) represents the climate class within a given physiographic division. The use of
different colors is intended to show the variety of climate classes within a given
physiographic division, while also depicting similarities in climate class across different
physiographic divisions (e.g., darkest red and darkest green show that two areas having
different physiographic characteristics are alike in their climate class).
3. Page 3-4 (S3.4, LI): ...when mine is active
RESPONSE: Correct (now in Section 4.2.4).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 215
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
4. Page 4-5 (last P): Refers to Fig. 3-1, but "existing road segments" listed are not shown on
Fig. 3-1, nor are several cited locales: Williams Port, Pile Bay, King Salmon, and Naknek.
RESPONSE: This reference has been deleted.
5. Page 4-11 (first P): "The vast majority of tailings dams are less than 30 m in height.
DOES THIS REFER TO TAILINGS DAMS AT ALL KNOWN MINING OPERATIONS,
OR TAILINGS DAMS ENVISIONED OR PROBABLY TO BE USED IN BRISTOL BAY
WATERSHED?
RESPONSE: The sentence refers to all known mining operations. The text was changed
from "The vast majority of tailings dams..." to "The majority of existing tailings dams..."
(Chapter 4).
6. Page 4-11 (first P): "Although upstream construction is considered unsuitable for
impoundments intended to be very high or to contain large volumes of water or solids... .this
method is still routinely employed." ARE Till- TSFs SUGGESTED IN THIS
ASSESSMENT CONSIDERED "UPSTREAM. " "DOWNSTREAMOR
"CENTERLINE"? Para. 1, Section 4.3.5 (p 4-21). slates that "the most plausible sites" for
TSFs are "the higher mountain," which suggests that these TSFs would be "upstream"
facilities, therefore "considered unsuitable
RESPONSE: The TSFs suggested in the scenarios would he constructed using the
downstream method and then switched to a centerline method. It is true that in this
location, the upstream construction would he unsuitable. The discussion of the upstream
method was intended for background and has been moved to Chapter 4 in the revision,
with the scenario-specific suggestions discussed in Chapter 6.
7. Page 4-23 (P2) Text suggests that a monitoring well field downslope from the TSF (and
presumably from all hypothetical TSI's) would detect seepage; such seepage would then be
intercepted and either relumed lo the TSF or "treated and released to the stream channel."
Either action presupposes adequacy oI"monitoring seepage and subsurface flow (both
spatially and temporally); returning such water to the stream further presupposes fully
adequate treatment to meet both regulatory and aquatic biota requirements for water quality
and flow regime
RESPONSE: It'.v. the well fields would be downstream from all embankments for all TSFs
and this has been clarified in the revision. It is assumedfor the scenarios that mitigation
measures are operated appropriately. Text has been clarified in the new Chapter 6 to
indicate that water would be treated to meet permit requirements. Scenarios in the revised
assessment evaluate effects from uncaptured seepage during routine operations from both
the TSFs and waste rock piles.
8. P4-26 to 4-28 (S4.3.7): This "Water Management" section seems cursory, highly
generalized, and optimistic. Statements such as "uncontrolled runoff would be eliminated";
"water from these upstream reaches would be diverted around and downstream of the mine
where practicable"; and "Precipitation.. .would be collected and stored... " do not indicate
actual (proposed) practices or techniques, nor inspire confidence that actual runoff events
during "normal" conditions, let alone during hydrologic extremes (such as a rain-on-snow
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 216
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
event with underlying soils still frozen) would be planned for or actually managed
adequately.
RESPONSE: Water management measures are more clearly described and discussed in
the revision (Section 6.1.2.5), and in sub-sections for the mine components in the
scenarios. Measures are standard and common to existing mine sites.
Collection and diversion structures would need to be designed and built to handle the
anticipated flows over the life of the mine, including during extreme weather events. The
details of these collection and conveyance systems are beyond the scope of this assessment
and are more properly in the domain of the permitting process when a specific mine plan
is proposed.
9. P4-29 (PI): Suggests that 20% more water than a\ailable would be required "during startup,"
and that difference would be satisfied "from water stored in the TSF"; if 20% more than is
available would be needed, where would it come from to be a\ ail able from the TSF?
RESPONSE: The water balance has been revised in Chapter 6. The mine operation would
capture more water than needed during all phases of operation. II 'titer could be stored in
the TSF as soon as the starter embankment is operational. Ghaffari et al. (2011) further
explain: "Once the TSF starter embankment construction is complete, site water will be
diverted into this facility to ensure adequate make-up water for process plant start-up. At
this time, advanced open pit dewtitering will commence. This water will either be treated
and discharged or diverted to the TSF. depending on environmental requirements."
10. Page 4-29 (P3): Assumptions are \ cry generalized and optimistic: "assuming no water
collection and treatment failures" and "excess captured water would be treated.. .and
discharged to nearby streams..." - this assumes both "no failures" over the life of the
operation, and that such treated "excess captured water" could be successfully treated before
release to fully meet both regulatory water quality criteria and the possibly more sensitive
biological requirements of individual invertebrates and fish stocks (Appendices A & B).
RESPONSE: Water management measures are more clearly described and discussed in
the revision in Section (>. 1.2.5. and in sub-sections for the mine components in the
scenarios. Measures are standard and common to existing mine sites. The revised
assessment considers water treatment failure quantitatively and includes refined scenarios
with new data for seepage from TSFs and waste rock piles.
11. Section 4.3.8: This and the previous section mention (but in my view, do not adequately
stress) the extremely long time frame for post-mining active management and oversight.
Many hundreds of years of active management is a longer time than many industrial,
corporate or governmental entities are capable of really embracing - witness the current US
Congressional practice of "kicking the can down the road" - a human trait.
RESPONSE: This issue is addressed in Section 4.2.4 of the revised assessment.
12. Page 4-32 (S4.3.8.2, P3): Suggests that pyritic tailings could be "shipped off site" - i.e., to
where? Deep ocean dumping, or Yucca Mountain?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 217
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: For smaller amounts of tailings, the option of shipping them to an off-site
location for disposal might be an option; however, this is less likely with large mines with
high volumes of tailings, so has been removed in the revised assessment.
13. Page 4-34 to 4-37 (S4.3.9): This reviewer finds the short Transportation Corridor sub-chapter
to be succinct, but inadequate and superficial in view of the long-term consequences of
imposition of the transportation corridor as portrayed. These deficiencies are addressed, in
part, in other sectors of the Assessment, most comprehensively in Appendix G.
RESPONSE: The transportation corridor analysis has been substantially expanded, and is
now found in its own chapter (Chapter 10).
14. Page 4-38 (Box 4-3, P2): Para. 2 states that the southwest extension of the Lake Clark Fault
is currently understood to extend to perhaps 16 +/- km li om the Pebble ore deposit; however,
this is not reflected in Figure 4-11, which suggests that the Lake Clark Fault terminates
perhaps 100 km northeast of the Pebble locale Elsewhere in Box 4 3. there is
acknowledgement that, while there is no e\ idence of recent tectonic activity in the immediate
Pebble vicinity, there is relatively little site-specific data or long-term historical seismic
record. I infer that any predictions concerning seismicity or earthquake occurrence of any
magnitude would have very high uncertainly
RESPONSE: Figure 4-11 (now Figure 3-15) lias been revised to more accurately illustrate
the current understanding of the terminus of the Lake Clark Fault. Section 3.6 describes
the uncertainties in predicting seismicity in the Pebble region.
15. Page 4-41 (Box 4-4) Note that in each of the four tailing dam failure examples, the failed
structure was roughly an order of magnitude snisiller (in height) than the hypothetical TSF-1
structure, yet those failures had major negati\e consequences.
RESPONSE: Text was added to Section l). 1.1 in the revised assessment to indicate that the
dams in these failure examples were significantly smaller than the dams proposed in our
mine scenarios."
16. Page 4-45 to 4-47 (S4 4 2 2) The prohiihiIit\ approach to tailing dam failure is unpersuasive
as presented It is difficult to relate to a number like "0.00050 failures per dam year," or to
the implication (p. 4-47) that one can expect a tailings dam failure only once in 10,000 to one
million "dam years "" This could suggest to the casual reader that failure of the hypothesized
TSF1 dam (for which one "dam year" is one year) should not be anticipated in either the time
of human occupation of North America, or the span of human evolution.
RESPONSE: The proposed dams, if designed, built, and maintained to current
engineering best practices, would be anticipated to have a low annual probability offailure
but the failure probability would not be zero. We concur with the commenter that these low
probability numbers may be difficult for the casual reader to grasp, so we also present the
estimates ofprobability in terms ofprobability failure over different time periods. For
example, an annual probability offailure of 0.0005 equates to a 5% probability offailure
over 100years and a 39% probability of failure over 1000years. Text was also added to
clarify the basis for our failure rates.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 218
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
17. Page 4-48 (Box 4-6): Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-
magnitude event at the Pebble locale has an estimated return period of 200 years. Such a
return interval probability is difficult interpret, given the lack of historical seismic record for
the region; in any event, such a return period estimate is in no way predictive of future
seismic activity, in year 2012 or year 2212.
Box 4-6 does note that "The return periods stated in Alaska dam safety guidance are
inconsistent with the expected conditions for a large porphyry copper mine developed in the
Bristol Bay watersheds, and represent a minimal margin of safety..
RESPONSE: The return periods used are consistent with the Alaska Dam Safety
Guidance, but the operator could include an additional margin of safety in the design for
critical structures. The return period and seismic safety factors do not inform the failure
analysis in this assessment.
18. Page 4-50 to 4-60 (S4.4.2.4): The modeled hydrologic consequences of overtopping/flooding
of the hypothetical TSF1 dam/reservoir seem reasonable, given the relatively limited
hydrologic data set available for model input Probable Maximum Precipitation and Probable
Maximum Flood results should be approximately "correct" and within the same order of
magnitude of potential storm and flood events The potential consequences outlined (peak
flow volume, sediment transport and deposition, length of stream corridor impacted) appear
realistic for the scenario. I suggest that this topic and hypothetical result should be given
more visibility and emphasis in the assessment
RESPONSE: The dam failure was modeled as an overtopping event since the possibility
could exist. However, the magnitude of the PMP run-off as compared to the resulting
failure flood wave is very small. II e agree that an actual dam design will consider
additional information, but this assessment did not investigate the hydrology of the PMF
in greater detail.
19. Page 4-(->2 to 4-(\> (S4.4.4). W hile accurate, this section does not adequately address the
road stream crossing/culvert issue. Given the projected transportation corridor, Pebble locale
to Cook Inlet, and the inevitability of a further network of "minor" roads in the mine and TSF
locale, plus additional infrastructure linkages, road/culvert/stream crossings are a major
concern lor aquatic habitat and fisheries. This issue receives more attention in Sections 5.4
and 6.4, and is mentioned elsewhere in Volume 1 (e.g., Table 8.1, Box 8-1, para. 8.1.2.4.).
Readers of the Assessment should be directed to Frissell and Shaftel's Appendix G for a
more comprehensi\ e discussion of this important topic.
RESPONSE: This issue is now analyzed in more detail in its own chapter (Chapter 10).
The appendix is cited.
20. Chapter 5: Assumes scenario of "no failure" for entire project, over complete project life. Is
this a realistic scenario, given experience with industrial developments in real-world settings
subject to vagaries of equipment, landscape, geology, weather, local climate, and human
judgment and decision making/execution?
RESPONSE: The "no failure" scenario has been eliminated in response to several
comments, and is now considered in terms of mine footprint impacts in Chapter 7.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 219
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
21. Page 5-1 (S5.1.1): Question: is "sampling extensively for summer fish distribution over
several years" adequate for characterizing fish populations, given the wide fluctuations in
salmon escapement and return note elsewhere in the Assessment (e.g., Table 5-1, para.
5.1.2)?
RESPONSE: The answer to this question is of course dependent upon the objectives of the
characterization. For the purposes of this section of the assessment, identifying species
distributions within the study area, sampling was adequate for minimally characterizing
the distribution of salmon, Dolly Varden, and rainbow trout within the project area,
including known spawning areas.
22. Page 5-12 to 5-48 (S5.2): Estimates of habitat, wetland and stream blockage or loss seem
reasonable, but, as noted in the text, are probably conser\ ati\ e or "at the low end."
RESPONSE: Agreed. No change suggested or required.
Estimates of probable streamflow diminution (p 5-25) seem reasonable, but make no
reference to seasonality.
RESPONSE: We assume a constant seasonal demand, thus streamjlow changes follow
baseline seasonal patterns under this analysis. II e clarify that water managers may alter
holding and release of water to address environmental flow needs.
23. Page 5-29 to 5-30 (thermal regimes). This section makes no mention of aufeis or "nalyds,"
ice accumulations which can exert major control on spring and early summer habitat
availability and thermal conditions I-'or examples, see Slaughter (1990), among many other
references.
RESPONSE: Citation and text have been added in Section 7.3.2.
24. Page 5-30 to 5-3 I Concur that maintenance of natural flow regime is the desirable target; the
"sustainability boundary" approach is a way to attempt managing within the "natural" bounds
of variability Note I'iuiire 5-l-> is map of streams and wetlands lost, not predicted flow
alteration hydrograph (see last sentence, p 5-3 I)
RESPONSE: Figure references have been updated.
25. Page 5-37 (Figure 5-1 ny I TI<>0D - predicted flow is ALWAYS below lower 20%
sustainability boundary I TI'iOC, UT100C1, UTC100B -predicted flows are always within
the 20% +/- sustainability boundaries.
RESPONSE: Figure has been revised, and data now presented in Table 8-1.
26. Page 5-38 (Figure 5-11): Predicted flow for two upper gages (SK100G, SK100F) is always
below the 20% sustainability boundary. Predicted flow at other gages appears to be near or
within the 10% and 20% lower sustainability boundary.
RESPONSE: Figure has been revised, and data now presented in Table 8-2.
27. Page 5-39 (Figure 5-12): Predicted high flow for NK119A is far below the lower 20%
sustainability boundary throughout the open water season. From onset of snowmelt, flow at
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 220
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
other gages is roughly at or within the lower 20% sustainability boundary.
RESPONSE: Figure has been revised, and data now presented in Table 8-3.
28. Page 5-41: Gage NK119A - is the estimated decrease of streamflow (minimum mine size)
63% (Table 5-13) or 73% (text)?
RESPONSE: Values have been updated
29. Page 5-41: In any case, text is clear: predicted flow reductions for North and South Forks
Koktuli River are materially below the lower 20% sustainability boundary. Text suggests
that while upper Talarik Creek would be essentially obliterated in this hypothetical scenario,
lower gaging stations on Talarik Creek might have partial augmentation of reduced flows,
from small tributary flows and groundwater; without suppoi'iiny data, this suggestion seems
unsupportable.
RESPONSE: Revised text illustrates important trans-watershed groundwater contributions
to lower Upper Talarik Creek.
30. Page 5-42 to 5-45: These pages fairly summarize the potential for substanti\ e alterations to
streamflow regime and surface water 'groundwater relationships.
RESPONSE: No change suggested or required.
31. Page 5-44 (P3): "Once the mine is no longer a net consumer of water, we assume that flow
regulation through the water treatment facility could he designed to somewhat approximate
natural hydrologic regimes, which could pro\ ide appropriate timing and duration of
connectivity with off-channel habitats "" I suggest that this is a highly optimistic
assumption, and does not address water quality questions (which are raised elsewhere in the
Assessment)
RESPONSE: II e emphasize that the ability to manage flows will be dependent upon
sufficient infrastructure and flexibility in water management.
32. Page 5-45 (S5.2.3): This entire paragraph should receive greater emphasis.
RESPONSE: This section (now Section 7.3.3) has been expanded.
33. Page 5-46 (P4) Ignores \ ariaMe-source-area concepts, which are widely accepted in
hydrologic and watershed analysis.
RESPONSE: We respectfully disagree. The variable source area concept presents the idea
about the expansion and contraction of saturated areas that are the immediate source of
streamflow during storms. In does not invalidate the notion that the total drainage area
contributes with long flowpaths to streams. These long flowpaths ultimately contribute to
saturated zones that support perennial streamflows..
34. Page 5-46 (P5): Assumes requirement for more water than is available, but leaves hanging
the question of where that more water might be sourced. Given that the site is a watershed
headwaters, what might be tapped as additional water supply, and what might be the impacts
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 221
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
on that source(s)?
RESPONSE: The revised water balance indicates that the mine would have the capability
to capture more water than it needs during all operational phases, including start-up, from
within the mine footprint.
35. Page 5-59 to 5-63 (S5.4): See earlier cautions concerning stream crossings, culverts. Note
that road cuts and culverts are particularly susceptible to development of aufeis ("icings"),
often resulting from blockage or alteration of subsurface water movement during cold
conditions, as witnessed by long-standing AKODT maintenance issues - Richardson, Steese,
Dalton highways, for example.
RESPONSE: The potential for culvert blockage by aufeis is discussed in Chapter 10 of the
revised assessment.
36. Page 5-60 (S5.4.2, PI): The statement thai " it is unlikely thai a mine access road would
have sufficient traffic to significantly contaminate runoff with metals or oil" is unsupported;
it might be instructive to look at traffic loads lor the access road from the Steese Highway to
the Ft. Knox mine, a much smaller operation than the proposed Pebble de\ elopment.
RESPONSE: This statement no longer occurs in the assessment.
37. Page 5-60 (S5.4.2, P2): First sentence is correct. Second sentence is unsupported and
probably incorrect (see Appendix (i) Yes. runoff from roads is location-specific; that does
not mean that runoff from roads would he insignificant to salmonids, given the very large
number of streams (perennial, intermittent, and ephemeral), and wetlands, which would be
intersected by the total road system of the IVhhle project. This also seems to be contradicted
by Section 5.4.3
RESPONSE: This issue is now given more attention in Chapter 10.
38. Page 5-(¦><) (S5 4 4 I): There are many more water or seep crossings than 34 - see USGS
topou sheets. ,\(MI- Mapper. or(iooule l-arlh
RESPONSE: The number of crossings has been updated in the revised assessment.
39. Page 5-61 (S5 4 4 2 and 5 4 4 3): Development of aufeis ("icings") consequent to partial or
full culvert blockage, or induced by soil mantle compaction (i.e., by roads or off-road vehicle
traffic) can partially or u holly block stream channels. Such blockage, in association with ice
on the streambed, may last long past snowmelt and persist well into early summer, possibly
affecting fish movement.
RESPONSE: The potential impact of aufeis on stream channels and potential existence
past breakup is discussed in Chapter 10 of the revised assessment.
40. Page 5-62 (S5.4.5): While I don't have the specific citations at hand, there are published
analyses of dust effects associated with the North Slope haul road and Prudhoe Bay road
network. Obvious effects include accelerated snowmelt along the road corridor, and nutrient
or pollutant contributions to road corridor environs.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 222
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The revised assessment includes information from published literature on
dust effects in the Arctic (Chapter 10).
41. Page 5-63 (S5.4.6.3): Should it read .impacting 270.3 km of stream..."? (Incidentally, it
is interesting that this coarse assessment finds it possible to state impact to within 100-meter
resolution.)
RESPONSE: The revised assessment notes that risks to salmonids from de-icing salts and
dust suppressants could be locally significant, but we cannot state that the entire stream
length between potential road crossings and Iliamna Lake would be affected.
42. Page 5-65 (S5.4.7.3 and 5.4.8.3): Viewing the transportation corridor landscape, via maps,
Google Earth or ACME Mapper, gives me the impression lluil the estimate of 4.9 km2
wetlands directly impacted is a very low number. Il is easy to play with the numbers given
on pp. 5-69 - 5-70, regardless of their accuracy. (->6 km of road impacting wetlands, assuming
* 2
a 10-meter roadway footprint, yields 0.6r> km" of wetlands "under the road" (vs. 0.18 km in
the text). The 200-meter proximity to wetlands cited, over 66 km of road, yields some 13
2 2
km of wetland impact (vs. 7.3 km in the text). SO km of road within 2'm m of streams or
wetlands yields 16 km2 of road/wetland impact Since these are all assumptions and
estimates, it is not possible to conclude that any of these figures would be the "'true" area
impacted.
RESPONSE: The 0.18 km refers to the area of wetlands which would be filled by the
roadbed "intersecting" wetlands (19.4 km length of roadbed x 9.1 m width of roadbed),
and does not include roadbed adjacent to wetlands (note that this area is 0.11 km in the
revised assessment due to a slight change in methodology for measuring road length). The
4.9 km2 (not 7.3 km2) area of wetland within 200 in of the length of road (on both sides)
within the study area is based on actual Sational Wetland Inventory (NWI) data (the
calculation methodology is described in Box 10-1 of the revised assessment).
Though NWI data were utilized in this analysis. the 200 m road buffer was derivedfrom a
literature estimate of the road-effect zone for secondary roads.
43. Page 5-74 (S5 4 10): Should this say impact rather than "risk"?
RESPONSE: "Impact" could be used here, but we chose to use "risk" to note the
probability rather than certainty that salmon would be affected by the corridor-associated
activities/events listed
Text implies that e\ en this "no-failure" scenario will impact salmonids; however, it is
apparently not possible to estimate specific changes or the magnitude of such changes.
RESPONSE: The no-failure scenario has been eliminated in response to several
comments; effects resulting from the mine footprint are now discussed in Chapter 7. The
exact magnitudes of changes in fish productivity, abundance and diversity cannot be
estimated at this time, but the species, abundances, and distributions that would potentially
be affected are summarized in Chapter 10 of the revised Assessment.
44. Page 5-74 to 5-77 (S5.6): This section seems cursory and understated, particularly in view of
the extensive discussion of Appendix D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 223
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: The discussion of fish-mediated effects on Alaska Native culture has been
expanded and is now in Chapter 12.
45. Section 6.1: Concur with general overview statements, and with conclusions regarding
immediate consequences of TSF dam failure, which would likely be as severe as or more
severe than stated in 6.1.2.1.
P. 6-6, first sentence - note that the failure scenario predicts over 70% fines <0,1 mm, vs.
the 6% "natural" fines concentrations.
P. 6-13, last para. - the assumption that overtopping would not occur in winter is not
warranted, as the authors admit when citing the Nixon Fork Mine incident. In the Bristol
Bay environment, a major rain-on-snow event in winter or spring is within the realm of
possibility, and of course human error is, if not inevitable, always possible.
RESPONSE: We clarify that the scenario we chose to illustrate overtopping did not occur
during low-flow conditions, but was in response to a flood We acknowledge that failure at
other times, and in response to other events, is possible (Section 9.3).
46. Page 6-15 (Box 6.2): Box text implies thai human error, lack of timely o\ ersight and
correction was responsible - but never directly says "human error." The apparent
assumption that there is no hydrologic activity after freeze-up (or perhaps, after an ice cover
forms on the pool) was naive and incorrect At least in that case, it appears that both dam and
spillway design (not adequately considering winter and ice conditions) and
operation/inspection (human error) were responsible
RESPONSE: The comment is probably correct. However, uv have no evidence that the
commenter's conclusion that both engineering error and operator error were responsible
was correct.
47. Section 6 14 1' Appropriately recognizes the long time period for exposure, over extended
stream lengths, through both initial deposition and multiple re-mobilization and redeposition
e\ ents
RESPOSSE: Agreed. So change suggested or required.
48. Page 6-28 (So I 5 and Table (->-0). Seems jargon-laden and does not add to strength of the
Assessment
RESPONSE: Opinion has been divided concerning this explicit weighing of evidence.
Given that the analysis of risks to fish from exposure to spilled tailings involves six
separate lines of evidence, the EPA intended to show and discuss how the evidence was
weighed.
49. Page 6-29 (S6.1.7): Concur that remediation "...would be particularly difficult and
damaging..."
RESPONSE: No changes suggested or required.
50. Page 6-30 (S6.2): Even though "We do not assess failures of the natural gas or diesel
pipelines...," those pipelines would be equally susceptible to failure as the slurry line.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 224
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
Concerns with pipelines crossing streams, watercourses and wetlands are similar to those
earlier expressed for the road corridor. Similarly, I suspect that careful inspection would
reveal many more "watercourses," including intermittent and ephemeral streams, than the 70
crossings cited.
RESPONSE: A diesel spill has been added to the assessment, but it was judged that a gas
leak would not pose a potential risk to fish (Chapter 11).
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and
recently in Montana (?), suggesting the probability (with what confidence limits?) that there
"should be" only 1.5 stream-contaminating spills in7K years of operation seems wildly
optimistic.
RESPONSE: These probabilities and frequencies are based on a large data set, not just the
TAPS experience (which is an atypical pipeline design and much larger than the diesel
line) or the spill in the Yellowstone River (which is a single event).
Assuming that any spill (over the 78-year project span) would last only two minutes (pp. 6-
32, 6-34), with a consequent minimal volume of spilled material, also seems highly
optimistic. Even highly-automated systems, with redundant sensors and automatic responses,
are susceptible to error or failure, and the Bristol Bay watershed environment is not benign
with regard to mechanical apparatus
RESPONSE: The EPA agrees with this comment and has increased the response time to 5
minutes.
The authors appear to recognize this with their discussion of the Alumbrera incident.
RESPONSE: No change suggested or required.
51. Section 6 4: Potential road/cul\ ert failures are recognized (again, not that ice issues are not
discussed). Extended periods for repair/rebuilding might be anticipated - witness the
repeated problems with the highway to Eagle, AK over the past several years - and that is a
State of Alaska responsibility, not that of a private company.
RESPONSE: We agree with the commenter that extended periods for repair/rebuilding
might be anticipated. Chapter 10 of the revised assessment notes that long-term fixes to a
road that was damaged by erosional failure of a culvert may not be possible until
conditions are suitable to replace a culvert or bridge crossing. Further, multiple failures,
such as might occur during an extreme precipitation event, would likely require longer to
repair.
Potential for multiple simultaneous or concurrent failures is appropriate. Non-Alaska
examples would be the Pacific Northwest flood events of 1964 and 1996, both major
precipitation events with widespread flooding and road failures, in a region with much more
developed infrastructure and response capacity.
RESPONSE: We agree with the commenter with respect to the likelihood for multiple
simultaneous or concurrent failures during extreme precipitation events. This is alluded to
in the assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 225
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
52. Chapter 7: Recognition of probable additional mining activity, in the wake of a Pebble
project, is appropriate. Assessment is necessarily limited to currently-known potential
mining projects. The cumulative and irreversible consequences of multiple developments,
with associated road, power, housing, communications infrastructure ("secondary
development") should be more heavily emphasized, even though it is not possible to quantify
all those consequences.
RESPONSE: The cumulative impacts of multiple developments, road corridors, and
secondary or induced development have been emphasized through additional discussion in
Chapter 13.
53. Chapter 8: Section 8.1.2 - note many potential failure modes not analyzed; the lack of
analysis in this Assessment should not be taken to mean thi.il such failure could not or will
not occur.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The focus of the assessment was on
potential effects to salmon. The revised assessment includes more failure scenarios (e.g.,
diesel pipeline failure, quantitative water treatment failure, and refined seepage scenarios)
and explains why these particular failure scenarios were chosen. Section 14.1.2 in the
revised assessment explains that the assessment only presents a few failure scenarios. The
EPA agrees with the commenter that lack of analysis should not imply a failure could or
will not occur.
54. Table 8-1 (Row 2): The reasoning behind the statement that "Most [product concentrate
pipeline] failures would occur between stream or wetland crossing [sic] and might have little
effect on fish" is hard to understand; stream crossings, whether via elevated utilidors or via
sub-channel borings or utilidors, are locales of angular change, piping connections and
joints, and subject to stresses of hydrologic extreme events - so why would such sites be less
subject to potential pipeline failure?
RESPONSE: Most of the corridor is not at or adjacent to a stream or wetland. Therefore,
failures are most likely to occur between streams or wetlands rather than at streams or
wetlands. The statement is not about probabilities per unit length.
55. Page 8-4 (Box S-l) As noted elsewhere, the probability arguments for TSF dam failure are
not persuasive, and seem designed to imply that a TSF dam failure would not occur within
the next 10,000 years (or 3.<)()i) to 300,000 years with three TSFs operational). This
implication is difficult to square with information on actual past failures presented in Box 4-4
and Table 4-8.
RESPONSE: The discussion of this issue has been expanded to clarify that the failure rate
is a design goal and is not based on empirical evidence.
56. Chapter 8: The potential risks and impacts are fairly and succinctly stated. Given the
extremely long-term nature of the projected Pebble project, and the irreversible changes
which would be imposed to the region, the risks seem, if anything, understated. I attribute
this to the decision to focus this Assessment on salmon and anadromous fisheries, with some
attention on salmon-mediated impacts - i.e., effects on indigenous culture, on wildlife other
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 226
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
than salmon, etc.
RESPONSE: No change suggested or required.
John D. Stednick, Ph.D.
1. Page ES-9: Economics of Ecological Resources - section seems weak.
RESPONSE: Economic data were included as background only. Economic effects of
mining are not assessed
2. Page ES-14: Overall risk to salmon and other fish \e\ or really separates fish species out in
other discussions. Dolly varden more sensitive to metal s'
RESPONSE: The distributions of the sahnonid species can be and are distinguished by
species, but not by their sensitivities to metals, lor copper and some other metals, data are
available for rainbow trout that can be applied to other salmonids. Dolly Varden have not
been used as a test species.
3. Page 2-3 (P3): Four climate classes Why this classification system? Perhaps easier to
identify by watershed maps?
RESPONSE: Climate classes are used as defined from Feddema (2005) to give the average
annual moisture index. The Feddema Index was used as it broadly characterizes a given
landscape location on a spectrum of wetter to drier conditions due to the effects of
precipitation and evapotranspiration. C limate classes are used independent of watershed
boundaries, as this classification system allows one to compare regions that are either
alike or dissimilar across watershed boundaries and highlight that, depending on scale,
watersheds can be quite heterogeneous in their degree of water availability from
headwaters to mouth. The revised assessment now has five classes ranging from semiarid
to very wet as shown in Figure 3-1.
4. Page 2-5: Precipitation values - significant figures?
RESPONSE: Precipitation values as shown in Figure 2-3 (Figure 3-2 in the revised
assessment) are calculated from SNA P data. SNA P data report monthly values of
precipitation to the nearest mm. During calculations, all precipitation values are rounded
to the nearest whole mm.
5. Page 2-23: Monthly values of streamflow. Would be nice to see average or daily streamflows
somewhere.
RESPONSE: The purpose of this figure is to show the general pattern of annual
streamflows in this region. More detailed information (e.g., daily flows) are readily
available for the USGS gages, but those data are not necessary to illustrate general annual
trends.
6. Pages 3-7 to 3-11: Would like to see more discussion of conceptual models. If a picture is
worth a 1000 words
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 227
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: Conceptual models have been revised and new models developed, and these
models now appear in each of the risk analysis chapters to clarify the pathways considered
in each chapter.
7. Page 4-1 (PI): ...represent current good, but not necessarily best, mining practices. Why not
use the best methods or state of the art methods?
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. The terms are qualitative when generally
interpreted, or have a regulatory meaning. The term "best management practices" is a
term generally applied to specific measures for managing non-point source runofffrom
storm water (40 CFR Part 130.2(m)). Measures for minimizMg and controlling sources of
pollution in other situations are referred to as best practices, state of the practice, good
practice, conventional, or simply mitigation measures. We have added a text box in the
revised Chapter 4 to discuss terms.
8. Page 4-18: Shaded relief. Perhaps contour lines in another figure?
RESPONSE: This figure is showing shaded relief of the landscape in Figure 4-6 and is
consistently used across multiple figures as a common base map for figures in this
assessment. It was decided to use shaded relief as opposed to contour lines to keep the
figure(s) cleaner in appearance, while also showing general topographic features. The
focus of this figure is on significant mineral deposits within the assessment watersheds and
not topography, therefore uv have decided to keep the background as currently shown.
9. Page 4-21 (P1): most plausible sites given geoiechnical, hydrologic, and environmental
considerations. Can this lx- elaborated?
RESPONSE: This sentence was intended to convey that these locations have similar
geotechnicai hydrologic. and environmental conditions to the site of TSF1. A more
comprehensive overview of the geologic and geomorphic setting of the Nushagak and
Kvichak River watersheds is now provided in Chapter 3.
10. Page 4-21 (P2) fhe IS/- would be unlined other than on the upstream dam face and there
would be no m>/>eri>ieah/e harrier constructed between tailings and underlying groundwater.
Is this correct? 1 thought I read the whole TSF would be underlain by liner?
RESPONSE: The assessment states that TSF embankments would have an engineered
liner, but the impoundment is not lined in our scenarios. A full liner is possible, but with
very large TSFs, stability with a liner may become an issue and benefits must be weighed
against other potential adverse outcomes, as well as costs. This is something that could be
considered in a regulatory permitting process. No change required.
11. Page 4-24: Leachate Recovered. This refers to only the leachate collected from the dam face?
RESPONSE: No, this box refers to leachate from tailings and waste rock. The text
"leachate recovered" in the formula included in the box refers to the leachate recovered in
testing conducted by the PLP. No change required
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 228
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
12. Page 4-26: Water management. This is confusing. Collect precipitation for processing, yet
divert upstream waters around the mine and not use? Where are the leachate recovery wells,
and are they just a safeguard?
RESPONSE: Water management measures are now clearly described and discussed in the
revision (Section 6.1.2.5), and in sub-sections for the mine components in the scenarios.
Upstream water sources would be diverted around a mining site to minimize losses of that
water source to the environment, as it could remain uncontaminated. The leachate
monitoring wells are to monitor groundwater downstream from the TSFs and waste rock
piles that would experience infiltration of their surfaces by precipitation falling on them.
Those wells could be converted to recovery wells, or new recovery wells placed, if it were
found that seepage from these structures had contaminated the groundwater. Measures
are standard and common to existing mine sites.
13. Page 4-28: Significant figures on precipilalion estimates? What is the KT and how is it
calculated?
RESPONSE: The water balance is now discussed in detail in Section 6.2.2 and the water
inputs discussed in Section 6.2.2.1. In the revised assessment, we used the three known
USGS gages draining the Pebble deposit site fl\hl00.1 (L SGS 15302250) on the North
Fork Koktuli, SK100B (USGS 15302200) on the South Fork Koktuli, and UT100B (USGS
15300250) on the Upper Talarikl to calculate monthly mean flows at each gage over their
period of record. Monthly mean flows for each gage were summed across the year and
then averaged across the three gauges weighted by their watershed area to produce an
area-weighted average of net (Precipitation — FT) yearly runoff of860 mm/year for the
general deposit area. . ill data were presented to the nearest whole mm.
14. Page 4-31 (P3Y the nunc />// would take approximately 100 to 300 years to fill. From
ground waler inflow only' Why such a large range of the estimate?
RESPONSE: The time for the pit to fill was based on groundwater infiltration and direct
precipitation into the pit and was presented as an approximate range. The lower end of the
range of estimates pertained to the Pebble 2.0 scenario and the higher end pertained to the
Pebble 6.5 scenario. Current estimates are approximately 20 years, 80 years, and 300 years
for the Pebble 0.25. Pebble 2.0, and Pebble 6.5 scenarios, respectively.
15. Page 4-50 (P3): This peak flow calculation and discussion is confusing and needs
clarification.
RESPONSE: Presentation of the peak flow calculations and hydrologic modeling of the
TSF failure scenarios has been revised and clarified. This is now presented in Section 9.3.
16. Page 4-52: What is the recurrence interval of the 356 mm?
RESPONSE: The recurrence interval is not reported in the reference cited The PMP 24-
hr is greater than the reported 100-yr 24-hr (see Technical Paper No. 47).
http://www. nws. noaa.gov/oh/hdsc/studies/pmp. html
http://www.nws.noaa.gov/oh/hdsc/PMP_documents/TP47.pdf
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 229
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
17. Page 4-60 (P4): Why a geometric mean using three values?
RESPONSE: The three values were chosen because they represented subsets of the
available data that better represented the conditions of the proposed pipelines (i.e.,
pipelines less than 20 cm in diameter, pipelines in northern climates, and pipelines run by
small operators). Focusing on these three subsets reduced the range of the estimated
probability of failure. The geometric mean was used because the comparison involved
ratios with potentially significant differences in magnitude in the denominators. Using the
geometric mean typically reduces the bias that a small dataset can introduce into an
arithmetic mean. In this case, the difference between the two means (0.0010 vs. 0.0011) is
immaterial and small with respect to the range of failure probabilities.
18. Page 5-10: Define highest reported index spawncr
RESPONSE: Section 7.1.2 defines this as the highest count of salmon reported in the EBD
for selected index reaches.
19. Page 5-20 (PI): ...salmon abundance relata! to pool size .and beaver points provide
particularly large pools. Are data a\ ailahlc lo characterize the stream type' Arc beaver
present?
RESPONSE: We now provide a characterization of stream gradient, mean annual flow,
and % flatland to characterize general stream and valley characteristics. Beaver are
present in the area (Chapter 7).
20. Page 5-22 (P2):. Issuniing that no natural flow or uncontrolled runoff would be generated
from the mine footprint. Is all precipitation is intercepted or does this refer to the subsurface
streamflow generation mechanisms0
RESPONSE: In our scenarios. uv assume that all precipitation that falls within the mine
footprint (except the fraction of the TSi' and waste rock leachate that bypasses the
groundwater interception and collection systems) is managed by the operator by collection
and routing leading to either storage in a TSF or testing, potential treatment, and release
to the environment. Any water consumed or stored on site would reduce the total
streamflow below baseline conditions, although some stream reaches would receive
increased stream flows due to discharges from the wastewater treatment plant.
Groundwater and streamflows are inextricably interdependent throughout the site, so one
could anticipate that areas with lower streamflows would also have lower groundwater
levels.
21. Page 5-23: It appears that the mean annual unit runoff is calculated incorrectly.
RESPONSE: This has been corrected.
22. Page 5-24: Table shows flow returned from footprint. Does not fit with page 5-22?
RESPONSE: Water balance descriptions, tables, andfigures have been extensively
revised.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 230
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
23. Page 5-29 (P2): Groundwater-surface water connectivity. Are data available to show this
connection throughout the watersheds or does the groundwater only return to the hyporheic
in the low gradient areas? Similarly, where are the temperature data that suggest the lake and
groundwater connection or this reference by incorporation?
RESPONSE: Figure 7-14 is now included to illustrate areas of modeled upwelling
strength. We provide several examples of groundwater-moderated temperatures from the
EBD data. But water temperature data are not generally summarized here and we refer the
reader to the extensive data provided in the EBD.
24. Pages 5-32 to 5-39: The tables and hydrographs illustrating the potential flow changes are
difficult to appreciate or interpret. Another means of presentation?
RESPONSE: We now include figures (7-10 through ~-l ~) illustrating the locations, extent
and magnitude of flow changes.
25. Page 5-41 (Table 5-13): The value of 0.15 km affected by the maximum mine size is
questioned.
RESPONSE: Stream lengths are now rounded to / km
26. Page 5-52 (Table 5-17): Can we see summary statistics on water quality, not just means? Plot
of concentrations vs. streamflou'
RESPONSE: Coefficients of variation have been added to the table (now Table 8-10). Plots
of concentrations versus other variables are available in Pl.P's Environmental Baseline
Document, Chapter l).
27. Page 5-55: These effluent spec/he values are higher than those for background surface water
because of the higher eonieni of mineral ions. This sentence needs clarification.
RESPO.\SE: The sentence has been rewritten to clarify that it refers to differences in the
BLM-derived criteria.
28. Page (•>-(•> (TaMe 6-1V l.ast line What is the +/» value after the mean?
RESPONSE: + or - one standard deviation are reported for the basin-wide samples; this
has been clarified in the table.
29. Page 6-13 (lJ3): ...an intense local storm. Why use the Type la distribution for precipitation
distribution?
RESPONSE: This has been corrected to Type 1 per SCS guidance for Alaska.
Roy A. Stein. Ph.D.
1. Global (see Table 5-3 as example): Some thought should be given to significant figures for
the numerical values given in the report, especially so for Chapter 5, where much uncertainty
exists regarding stream lengths blocked (for example) by the mine footprint: ".. .34.9 km of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 231
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
first- through third-order streams..will be eliminated. Rounding in this context makes
sense for we really do not know this impact to the nearest 0.1 km. I would encourage a
thorough review of these values throughout the report.
RESPONSE: Significant figures for calculations and measurements by EPA have been
changed to reflect uncertainty (for example, stream length is reported to the km).
Significant figures from literature are not revised
2. Page 2-17: Salmon populations are closely managed by Alaska Department of Fish and
Game; how closely and what do we know of the populations that are managed regarding
numbers, resilience, variability, etc.? A box summarizing the fishery management practiced
by Alaska Department of Fish and Game would help put Molouical resources in perspective.
RESPONSE: Box 5-2 has been added in Chapter 5. providing an overview of commercial
fisheries management in Bristol Bay.
3. Page 2-17: The importance of salmonids lo murine predators is likely not an issue, given the
very large numbers of salmon stocked in the Pacific.
RESPONSE: No changes suggested or required.
4. Page 2-20: Nushagak and Kviclink ri\ eis contain 5S.<)oo km of streams; 13% is
anadromous fish habitat, but this is likely underestimated I low do we know this is an
underestimate? What proportion of this 5K.<)<)<) km has Ixvn surveyed?
RESPONSE: A DFX (i estimates that the current listing of II aters Important for the
Spawning, Rearing and Migration of. I nadromous Fishes represents "less than 50% of the
streams, rivers and lakes actually used by anadromous species" (ADF&G AWC database,
available at: www.adfg.alaska.gov/sf S. IRR \ II C/)). This clarification has been added to
the assessment in Box ~-I.
5. Page 2-25 (hollom of page) I wonder if a bit more couldn't be written about the idea of a
salmon sanctuary, ileshing out the ideas of Rahr and Pinsky here.
RESPONSE: The creation of salmon sanctuaries in the Bristol Bay region is a
management decision, and thus outside the scope of this assessment; therefore, no change
is necessary.
6. Page 3-2: With all the items in the first full paragraph eliminated from consideration (e.g.,
power generations, worker housing, Cook Inlet Port), might the analysis herein be considered
minimal impact on the Bristol Bay watershed?
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. It is true that this assessment only considers a subset of
potential impacts from large-scale mining.
7. Pages 3-7 to 3-11: These conceptual models might best be placed in the chapter to which
they refer; in so doing, it is easier for the reader to follow along. In turn, these models did
not seem to be discussed in text to the extent that they drove the impacts generated. Add
more explanatory text.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 232
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
RESPONSE: In the revised assessment, conceptual models are presented and discussed in
the appropriate chapters.
8. Page 4-5 (bottom of page): Typically, when citing a figure in another chapter, i.e., one not
near the current text, the format should be "see Figure 3-1" to keep the reader on track.
RESPONSE: Comment noted. Figure references have been retained in parentheses.
9. Page 4-8: Is there some chance of "block caving" here? Some text clarifying this point here
would be appropriate.
RESPONSE: Additional information on block caving is provided in Chapter 4 (Section
4.2.3.1 and Box 4-4). Block-caving is a possibility for mining deposits that are deeper.
Sources of hazards assessedfor risks, however, would be similar, as block-caving and
surface mining both create tailings, waste rock (block-caving does have a smaller amount
of waste rock), and sources of contamination to the water.
10. Chapter 4: Could these data and insights be productively moved to Chapter 5, thus reducing
redundancy? The organization would be 1) a description of the mine features relevant to the
text, then 2) a discussion of salmon and the mining impacts on their habitat. Separating these
into chapters seems artificial.
RESPONSE: Revision of the assessment has included organization of material into
additional chapters and the EP. I believes this new organization is easier to follow.
11. Page 5-20 (P3): What does the phrase, "lYce-waler area" mean ' Ice free, perhaps?
RESPONSE: Removed.
12. Pages 5-2<~) to 5-21 Paragraphs on these two pages reflect some of the redundancy that I saw
throughout the report. In the 3ul full paragraph on page 5-20, text begins with
" groundwater inputs may be critical "" and in turn on page 5-21 in the first full paragraph,
the topic sentence begins " gioundwalcr-inlliienced stream flow...likely benefit fish..."
Both deal with the same topic Hence, these two paragraphs could easily be combined,
serving 10 shorten the text, reduce redundancy, and improve readability.
RESPONSE: These paragraphs have been combined
13. Page 5-22: Similarly, the text on page 5-22 includes citations (at the end of the same
sentences) to both sections and tables and figures in Chapter 4. This suggests to me that the
text is not only overlapping, it is redundant and better overall report organization would serve
its presentation well.
RESPONSE: Chapters have been reorganized, but some cross-chapter referencing
remains necessary. In this case, it is important that the reader be able to refer back to the
mine scenarios and operations chapter if desired, to better understand the complex water
management system and budget that drives the analysis in Chapter 7.
14. Page 5-23 (Table 5-5): Some explanation in the table title would benefit the reader, as to
where these gages are placed in the stream. The labeling is arcane at best UT100D; letters
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 233
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
suggest Upper Talarik Cr., but what is 100D. Now, I figured out from text that the first ones
were high in the watershed and then they proceeded downstream (increasing drainage area
gave me a hint as well). Use better descriptors (such as SKI through SK5 from low to high
stream order) or explain the ones that are being used.
RESPONSE: New figures illustrating gage locations are included. Gage names remain
consistent with EBD usage for consistency
15. Page 5-26 (Table 5-7): I cannot find any bold in this table that would reflect the pre-mining
condition. Is that the same column as "Pre-Mining" as it is in Table 5-8 on page 5-34 (and
the next few tables as well)?
RESPONSE: Corrected
16. Page 5-27: Cite Figure 5-8 in text any time the stream gages are mentioned. The reader then
has the ability to easily refer back to stream gage locations.
RESPONSE: Corrected
17. Pages 5-32 to 5-39: How do these tables and figures differ'' Might they he I <>% reduction in
flow, 11-20% reduction inflow and 20% reduction in flow, as suggested in text? If so, then
these table and figure titles need lo relied this information and be better described in the
legends of the figures. Finally, do we need both tables and figures?
RESPONSE: Figures have been removed, with exception on one illustration to clarify
process.
18. Page 5-31 • The Richlcr et al (2dI I) reference, which underpins this section is incomplete in
the Literature Cited (Chapter ^). suggesting only March as the publication date. Update in
any re\ ision
RESPONSE: Citation has been updated.
19. Page 5-3 I It should he made clear that the Richter et al. (2011) sets quite specific bounds for
all ri\ ers regarding ecosystem function and does not provide any specific insight into salmon
production (made somewhat clear in the last sentence of the next to the last paragraph on
page 5-43). 1 think an additional caveat stating this explicitly on page 5-31 would improve
the text.
RESPONSE: Additional clarification has been added
20. Page 5-46: Stream flow losses due to the mine footprint are discussed first as underestimated,
then as overestimated, and then a conclusion that they are underestimated (i.e., because the
mine will need more water than is available from surface run-off). If this is the case, why go
through the other scenarios.. .to seem even-handed? I am not sure all of this text is required.
RESPONSE: Uncertainties and assumptions sections have been revised
21. Page 5-46: Mine start-up will require more water than is available from the footprint of the
mine itself. Hence, water will be captured from other streams than just those associated with
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 234
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
the mine, which will influence stream flow and groundwater supplies. Are there estimates of
the amount beyond the surface water that will be required?
RESPONSE: An extensively-revised water balance description and accounting are now
used Indirect effects of groundwater loss to pumping are incorporated.
22. Pages 5-53 to 5-58: I am a little confused by this section. The implication throughout is that
copper toxicity will be based on the response of aquatic invertebrates (which would then be
protective of direct effects on salmon). However, near the end of this section, there is a
discussion about zooplankton being most sensitive and a comment about the reliance of
juvenile sockeye rearing in lakes on these zooplankton Yet. there is no resolution of what
criteria will be used for toxicity values.. .will it be aquatic i n\ ertebrates or the more sensitive
zooplankton? Clarification is required here.
RESPONSE: Zooplankton are aquatic invertebrates. The same criteria apply to both
invertebrates in the benthos and those that drift in the lake.
23. Page 5-59: "Discharge permits for mine in the Bristol Bay watershed should include relevant
whole-effluent toxicity testing and monitoring of hi otic communities in recei\ ing streams".
Does this quote solve the problem mentioned in the pie\ ions comment'.' Is this a realistic
expectation for mine operators before permits are issued '
RESPONSE: The statement reflects the hopes of the assessment authors. They do not
constitute an obligation upon the State of. Maska. This is clarified in the revised
assessment.
24. Page 5-68 (Table 5-22) 10".. rather than lo"„ in the table title.
RESPONSE: The reference was correct in the draft. Table 10-8 now shows streams with
gradients <12% likely to support salmon.
25. Chapter (•> Ri\ers is lower case when multiple ii\ers are listed.
RESPONSE: Comment noted.
26. Page 6-1 Why assume that only 2<)% of the tailings stored would be mobilized with any dam
failure? Is there a justification for this important assumption? Yes, see Appendix I, page 14
for citation to Dalpatram (2011); this should be cited in the main report in Chapter 6.
RESPONSE: Citations of Dalpatram (2011) and Azam and Li (2010) have been added,
drawn from Appendix 1.
27. Page 6-4: Add map showing the impact of the failed TSF 1, i.e., distribution of sediments and
impact downstream.
RESPONSE: Because we are not predicting specific sediment depths in this revision, we
have determined that this information was best presented by a description in the text and
tables.
rd
28. Page 6-42: The sentence in mid-paragraph (3 complete paragraph on this page) ".. .multiple
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 235
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
failures such as might occur..My guess is that an example should follow the phrase "such
as."
RESPONSE: Culvert failures are described in the preceding paragraphs (in greater detail
in the revised assessment), so an example is not needed here.
29. Page 7-2: The sentence at the end of the last full paragraph on the page makes little sense:
"The overall consequences are diminished and extinct salmon populations."
RESPONSE: This sentence has been clarified.
30. Page 8-6 (last sentence of 2nd complete para): "further" should he "farther"; sorry, I just
couldn't help myself...
RESPONSE: Text revised
William A. Stubblefield, Ph.D.
None
Dirk van Zvl, Ph.D., P.E.
I have indicated a number of specifics in the le\l abo\ e and do not ha\ e any others to list.
PhyllisK. Weber Scannell, Ph.D.
1. Page 4-1 I The document slates "ucomcmbianes are generally estimated by manufactures to
last 2<) lo .ii) years when co\ered by tailings (North pers. comm...)." Unless North is a P.E.
with experience in geomembranes. the statement needs a stronger reference. For example,
Erickson et al (200K):;: discuss the quality issues with geomembranes related to manufacture,
installation and application of a soil-based cover (such as bentonite).
RESPONSE: The reference is appreciated, but others were chosen for use related to time
of life for liners. The discussion of liners has been expanded and moved to Chapter 4
(Section 4.2.3.4) in the revision, and includes additional material and references on
lifetime. The personal communication reference has been removed
2. Page 4-23 (LI): This first sentence is confusing and implies that oxygen has low solubility
because it is in the tailings pond. Suggested change: eliminate the first phrase "In a TSF".
RESPONSE: The change has been made in the revised assessment.
3. Page 6-36 (P3): Last sentence states: [water] treatment would continue until institutional
failures ultimately resulted in abandonment of the system, at which time untreated leachate
discharges would occur. This statement is not supported by any documentation and is not
clear what is being implied. Failure of governments? As stated in my response to questions,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 236
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Mr. Steve Buckley
DELIBERATIVE - FOR EPA USE ONLY - NOT FOR PUBLIC RELEASE - 06-11-13 DRAFT
any mine plan must include sufficient bonding and plans for reclamation, including necessary
water treatment.
RESPONSE: Bonding and a perpetual trust could result in treatment for some period, but
it possible that long-term funding will not last to the end of time.
4. Page 6-37 (P4): End of paragraph states "premature closure could leave waste rock piles in
place." Again, there is a need for plans for mine closure, concurrent reclamation and
sufficient bonding.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed It highlights the need for adequate bonding andfor plans
that include closure by some agency.
5. App G, Page 5 (PI, last line): Document stales " other short road segments connect
Dillingham to Aleknagik and Naknek to King (I 'igin e 1)."
Shouldn't King be King Salmon?
RESPONSE: Corrected
6. App G, Page 5 (P3): Dolly Varden should be capitalized throughout the document. The list
of fish species should contain scientific names or reference a table of common and scientific
names.
RESPONSE: Corrected
The document states "In the most comprehensive published field inventory, Woody and
O'Neal (2010) reported "" Because the authors have not reviewed other documents on field
inventories, the phrase "most comprchensiv e" should be changed to "a comprehensive"
RESPONSE: C orrected.
Paul Whit new Ph.D.
A lot of page and paragraph comments are included the above responses to charge questions. I
have no further comment
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 237
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft]
-------
Dr. Courtney Carothers
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Courtney Carothers rmailto:clcarothers@alaska.edul
Sent: Thursday, July 18, 2013 1:21 AM
To: White, Jessica; Thomas, Jenny
Subject: Carothers' summary evaluation of EPA's response to 2012 peer review report
Dear Jessica and Jenny,
Please find attached my letter summarizing my evaluation of the EPA's response to 2012 peer
review report.
I did not include the editorial corrections in my letter, but note them below.
Could you please confirm receipt of this email and document and please let me know if I need to
provide anything else.
Best regards,
Courtney Carothers
Editorial corrections:
Page vi: Affects to Effects
Page ES-2, ES-25, (and others): Alaska native to Alaska Native (consistent usage throughout
document)
Page ES-9: first full paragraph, first full sentence, rewrite to: "The subsistence way of life in
many Alaska Native villages is augmented with participation in the cash economy."
Pages 1-6; 14-13; fisherman to fishermen
Page 5-2; 8watersheds
Page 5-2; "These resources generate significant benefit for...., and provide sustenance for
Alaska Native populations and other rural residents."
Page 12-9: second full paragraph, period missing at end of first sentence, incorrect comma.
-------
Dr. Courtney Carothers
Summary Evaluation of the U.S. EPA's Response to September 2012 Peer Review Report
Date: July 17, 2013
Jessica White
Contracting Officer
USEPA Headquarters
Ariel Rios Building
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460
Dear Ms. Jessica White:
As per the terms of my peer review contract and supplemental clarification of the scope of work, I
provide here my summary evaluation of the U.S. EPA's response to: 1] the key recommendations in
the Executive Summary of the Peer Review Report that pertain to my areas of expertise and 2] the
individual comments and suggestions that I provided in the Peer Review Report. This summary is
based upon my review of the following documents:
1. The April 2013 version of the draft Bristol Bay Assessment (referred to as "Revised
Assessment" in this letter]. This document is available at:
http://www.epa.gov/ncea/pdfs/bristolbav/bristol bay assessment erd2 2013 voll.pdf
2. Revised Appendix I (Conventional Water Quality Mitigation Practices for Mine Design,
Construction, Operation and Closure] and new Appendix J (Compensatory Mitigation and
Large Scale Hardrock Mining in the Bristol Bay Watershed], Other appendices are not
considered part of the evaluation being conducted under this contract. Appendix I is
available at: http://ofmpub.epa.gov/eims/eimscomm.getfile7p download id=513560.
Appendix J is available at:
http://ofmpub.epa.gov/eims/eimscomm.getfile?p download id=513561
3. Final Peer Review Report, External Peer Review of the EPA's Draft Document, An Assessment
of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska, dated September 17,
2012 (referred to as "Peer Review Report" in this letter]. This document is available at:
http://www.epa.gov/ncea/pdfs/bristolbay/Final Peer Review Report Bristol Bav.pdf
4. An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska,
External Review Draft, dated May 2012. EPA 910 R 12 004a.
5. The draft response to comments document detailing how the EPA responded to
comments provided by the external peer reviewers in the Peer Review Report (referred to
as "EPA Response" in this letter],
I. Comments on the U.S. EPA's Response to Key Recommendations in the Peer Review Report
(relevant to my areas of expertise)
Scope of the Document
KEY RECOMMENDATION: "Explain why the scope for human and wildlife impacts was limited to
-------
Dr. Courtney Carothers
fish--- mediated effects, as well as why fish-- mediated effects on humans were limited to Alaska
Native cultures. Reviewing effects beyond fish- - mediated ones (e.g., potential for complete loss of
the subsistence way of life) would improve the assessment."
EPA RESPONSE: "The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. Throughout the assessment we acknowledge that direct effects
of large- - scale mining on wildlife and Alaska Native cultures may be significant, but that these
direct effects are outside the scope of the current assessment."
Overall, the scope of the assessment has been more clearly articulated. Figure 2.1 is helpful to
convey the focus points of this assessment, as well as other important impacts and groups that
fall outside of this scope. The Revised Assessment explains why the scope for human impacts
was limited to fish mediated effects; however, the EPA Response also notes the prevalence of
direct human impacts discussed in public comment of the May 2012 draft watershed
assessment. These public comments, in addition to peer review comments, prompted an
expanded discussion of direct impacts of mining on people in the Revised Assessment,
particularly in Chapter 12. Given the narrow focus of the majority of the assessment on fish
mediated human impacts, the important addition of a discussion of direct effects in Chapter 12
could be more clearly contained in a separate section to avoid confusion. The section should
also be prefaced with the clarification that it is a partial review of direct effects that have been
identified in previous case studies, rather than a review, summary, or study of all possible
impacts.
The Revised Assessment also clarifies the focus on Alaska Native cultures in at least two places:
"Fish-- mediated effects on Alaska Natives were considered because sustainability of the
region's fish populations is critical to the future of Alaska Natives in Bristol Bay, and
because concern about the region's fishery resources prompted original requests from
Alaska Natives that USEPA examine issues in the Bristol Bay watershed" ('1---3)
"Although Alaska Natives are not the only people that would potentially be affected by
mining in the region, Endpoint 3 focuses on Alaska Native populations because of the
centrality of salmon and other salmon-- dependent resources to their way of life and
well-- being, and because this assessment was initiated in response to requests from
federally recognized tribal governments to restrict large-scale mining the watersheds"
(5—2)
Both of these statements indicate that the assessment focuses primarily on Alaska Natives in the
region for two reasons: 1] the centrality of salmon to the indigenous cultures of the region, and
2] the original request for the assessment came from Alaska Natives and federally recognized
tribal governments. It would helpful to also clarify if the EPA had special obligation to conduct
this review given the government to government relationship of these tribes with the federal
government. The addition of this second point (that Alaska Native cultures are considered an
endpoint in this assessment in part because tribal governments asked for the assessment to be
conducted] should also be added to the Executive Summary. As written, the Executive
Summary's discussion of scope (ES 2] discusses wildlife and Alaska Native cultures together as
endpoints given their dependence on salmon. As discussed in the first peer review process, this
grouping is awkward and potentially offensive. Sections that link discussion of these two
endpoints (e.g., Section 13.4.2] should be divided into separate sections.
Technical Content: Human Cultures
KEY RECOMMENDATION: "Use case histories to provide insight and anticipate mining impacts on
2
-------
Dr. Courtney Carothers
Alaska Natives (e.g., those exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil
and gas development in the North Slope region, and social impacts related to mining development
in Alaska)."
EPA RESPONSE: "Examples from applicable case studies, including the Exxon Valdez oil spill,
are cited in Chapter 12 of the revised assessment."
The inclusion of these case studies has strengthened the report. These case studies provide insights
into both fish mediated effects to cultures and communities, as well as direct effects from resource
exploration and development.
KEY RECOMMENDATION: "As noted above (Scope of the Document), clarify why the scope was
limited to fish--- mediated effects. The potential direct and indirect impacts for human cultures
extend far beyond fish-- mediated impacts (e.g., potential complete loss of the subsistence way of
life). The rationale for this narrow focus should be fully explained. In addition, a clear explanation
should be given for why fish- - mediated human impacts focused only on Alaska Native cultures."
EPA RESPONSE: "The assessment focuses on a specific, limited set of endpoints as defined in
Chapter 5. We have added text to explain both why these endpoints were selected, and that
responses other than those considered in the assessment are likely but are outside the scope of
the assessment. The assessment was expanded (Chapters 5 and 12) to acknowledge that there
are a wide range of potential direct and indirect impacts to indigenous culture, but they are
outside of the scope of this assessment. The discussion of potential effects to indigenous
cultures was expanded to explain that a loss of subsistence resources would extend beyond a
loss of food resources to social, cultural, and spiritual disruption. The text has been expanded to
acknowledge the strong cultural ties of many non--Alaska Natives to the region, and potential
effects on all residents from loss of a subsistence way of life. However, the focus of the
assessment rem ains on effects on indigenous cultures resulting from effects on salm on."
See comments above under "Scope of the Document." The expanded discussion of subsistence
and recent research in the region in Chapter 5 has greatly strengthened the assessment. The
revisions of the language in the Revised Assessment to reflect the whole suite of physical, social,
cultural, economic, and spiritual aspects of subsistence have also adequately addressed earlier
concerns raised.
Editorial Suggestions
KEY RECOMMENDATION: "The appendices contain detailed and useful information that should
be summarized and included in the main document (e.g., Appendix E: Economics, Appendix G:
Road and Pipelines, and Appendix I: Mitigation). Additionally, consider expanding the preface to
include information on the use of the appendices. If the information is not included in the main
report, then justify its absence."
EPA RESPONSE: "More information from the appendices was brought forward into
appropriate chapters of the revised report. The purpose of the appendices—to provide the
detailed background characterization necessary for the ecological risk assessment—has also
been clarified in Chapter 2. The document no longer contains a preface because that material
has been incorporated into Chapters 1 and2."
3
-------
Dr. Courtney Carothers
More information from Appendix E (as well as Appendix D] has been included in the main body of
the report, but as described below, more information from Appendix E is relevant to include,
particularly Alaska Native engagements in commercial fishing and fish based tourism in this region.
Research Needs
KEY RECOMMENDATION: "What are the locations of subsistence areas and can these
areas be characterized and differentiated by collecting local environmental and
ecological knowledge (e.g., fish overwintering areas, climate change, ecological shifts,
etc.]?"
EPA RESPONSE: "The revised assessment incorporated current data on subsistence use areas
available from ADF&G. EPA acknowledges that these data are incomplete and would
encourage additional collection of subsistence data and Traditional Ecological Knowledge."
The Revised Assessment includes more detailed information about subsistence harvests and use
areas (e.g., Figure 5.12], The necessary caveats that these data are coarse and incomplete, and that
additional studies of traditional ecological knowledge (especially in times of social and
environmental change] are greatly needed, are important to highlight in the main text (e.g.,
coarseness and limitations of data explained in Box 5.1 should also be repeated in main text],
II. Individual Responses to Charge Questions
1. General Impressions: Comments here are addressed above in response to clarification of
scope in Section I, Key Recommendations.
2. Question 1: Additional methodological detail was added to Appendix D, but the revised
Appendix D was not included for review and was "not considered part of the evaluation
being conducted under this contract" (Statement of Work, p 1], so could not be reviewed.
I found evidence throughout the Revised Assessment that supports EPA's response that "the
assessment text regarding the importance of the subsistence way of life has been expanded to
recognize the centrality of subsistence to the social, cultural, and spiritual well being of the
indigenous cultures" (EPA Response, p 40],
Suggested references on subsistence harvests, use areas and local context were included in
the Revised Assessment. Section 5.4 of the Revised Assessment (updating section 5.6 of the
original report] is substantially improved from the first draft assessment.
I disagree with the EPA response that Appendix E on the economics of salmon in the region is
not relevant to explore the salmon mediated effects to Alaska Native cultures (EPA Response,
p 39], Just as salmon are central to culture, they are central to the mixed (market and non
market] economies of the Alaska Native communities of the region. The dependence of the
local communities and cultures on commercial fishing and tourism, as well as the links
between these commercial engagements and subsistence fisheries are important aspects to
consider in evaluation salmon mediated effects to Alaska Native cultures. Some additional
information from Appendix E has been added to the report. More of this information,
particularly that information relevant to the Alaska Native communities in the region (e.g.,
community engagement in, and dependence on, commercial fisheries] would enhance the
report. The fish mediated potential impacts to commercial fishing would affect Alaska Native
cultures; commercial fishing has long been a part of the communities and cultures in this
region.
4
-------
Dr. Courtney Carothers
Additional references on subsistence and mixed economies would strengthen the revisions
(e.g., p 5 36, last paragraph, no references provided; Krieg et al. 2007, Wolfe and Walker
1987, among others would be helpful references to add],
Krieg, T.M., J. A. Fall, M. B. Chythlook, R. LaVine, and D. Koster. 2007. Sharing, Bartering, and
Cash Trade of Subsistence Resources in the Bristol Bay Area, Southwest Alaska.
Alaska Department of Fish and Game, Division of Subsistence Technical Paper No
326, Juneau.
Wolfe, R.J. and R. J. Walker. 1987. Subsistence Economies in Alaska: Productivity,
Geography, and Development Impacts. Arctic Anthropology 24(2]: 56 81.
3. Questions 2 9; 12, 13: EPA responses are sufficient.
4. Question 10: While the EPA Response notes that additional discussion of the connections
between subsistence and commercial fishing in this region has been added to the report, I
did not find much evidence of these additions. One instance of this addition in the Revised
Assessment includes a short paragraph (5 36] only briefly mentioning these connections.
Chapter 5, Section 5.4, for example, could include a discussion of Alaska Native
engagements in, and dependence on, commercial fishing and fish related tourism
businesses in the region. Similarly, Chapter 12, Section 12.2 could include a section that
discusses how decreased commercial fishing and tourism opportunities would affect Alaska
Native communities and cultures. Some of this information is contained in Appendix E and
could be brought forward into the main body of the report.
As noted above, the report makes clear the scope of the assessment of human impacts is
limited to fish mediated effects. However, the EPA Response notes that Chapters 5 and 12
have been expanded to acknowledge the range of direct and indirect impacts of mining to
human cultures and communities. At times, there is some slippage between the defined
scope and the partial inclusion of direct impacts. For example, the list of salmon mediated
effects on Alaska Natives (section 12.2, p 12 6] lists a direct effect of mining on social
systems ("a shift from part time to full time wage employment in mining or mine associated
jobs would affect subsistence gathering capabilities by reducing the time available to harvest
and process subsistence resources."] This is not a salmon mediated effect, but a direct effect
of mining. I agree that these direct effects are highly important to consider and deserve
acknowledgement and discussion in the report. However, because the scope has been
narrowly defined as salmon mediated effects, the interspersing of some direct effects into
sections that describe salmon mediated effects calls into question this focus. As stated above,
the discussion of direct effects could be contained within separate sections in Chapters 5 and
12. These sections should reiterate that a full treatment of these direct effects is outside of
the scope of this assessment, and as such the abbreviated discussions are necessarily cursory
and incomplete. Given the number of public comments that focused on direct effects, this
clarification and acknowledgement is appropriate and may be less confusing if placed in
separate sections within the report.
5. Question 11: Chapter 12 now clarifies that the cumulative effects of multiple mines on
humans would not merely be a loss of food resources, but possibly the loss of subsistence
way of life.
6. Specific observations: EPA responses are sufficient, with the following notes:
5
-------
Dr. Courtney Carothers
In response to specific observation #3, the Revised Assessment now includes the mention of
"Aleut/Alutiiq people" inhabiting the Alaska Peninsula (Revised Assessment, p5 2], Should
this be rewritten as "Aleut/Unangan and Alutiiq/Sugpiaq peoples," or is the cultural group
being referred to only the Alutiit/Sugpiat? If the latter, "Alutiiq/Sugpiaq people" should be
used to be consistent with current published usage. If the former is true, that both
Aleut/Unangan and Alutiiq/Sugpiaq peoples inhabit this region, it should be rewritten as
"Aleut/Unangan and Alutiiq/Sugpiaq peoples" for clarity. While many Alutiiq/Sugpiaq people
refer to themselves as Aleut in the English language, the usage of Aleut/Alutiiq is confusing.
The standard usage of Alutiiq/Sugpiaq is preferable.
Revisions were noted to Appendix D and E. Responses seem sufficient, but as these revised
appendices were not provided for review, I cannot comment on these revisions.
III. Specific Observations of the Revised Assessment
Section 5.2.6 (p 5 27]: The last paragraph of this section should add language about the holistic
fishery systems in this region. In addition to the relatively pristine salmon ecosystems and
populations, the region is occupied by indigenous human cultures who continue to occupy their
homelands, speak their languages, and continue to depend upon these intact fisheries. These
features of the Bristol Bay fishery systems also greatly add to the watershed's global conservation
value.
Table 5.1 (p 5 3]: Some subsistence use species described in Box 5.1 are not listed in Table 5.1 as
harvested (H] species (e.g., least cisco, round whitefish; and broad whitefish, which does not appear
in Table 5.1],
Section 12.2.3 and Box 12.1 (p 12 11]: The text preceding Box 12 1 states that Bristol Bay residents
have expressed mixed opinions about development and mining. Box 12 1 focuses only on concerns
about potential effects of large scale mining in the Bristol Bay watershed. These are effective at
conveying some of the testimony received about concerns; however, missing are excerpts of
testimony representative of those who expressed desires for jobs and development related to large
scale mining. The EPA authors could add in a couple of additional quotes reflecting the full
spectrum of opinions expressed in the testimony received at public meetings. If the excerpts in Box
12.1 represent the bulk of testimony received, this point should be made in the box text. I
also attach several editorial corrections in an email correspondence.
I commend the EPA team for careful and thoughtful revisions of the sections I have reviewed
closely. I appreciate the attention given to the peer review comments throughout this process of
review and revision. Please let me know if I can answer any questions or be of further assistance.
Sincerely,
Courtney Carothers
6
-------
Dr. Dennis Dauble, Washington State University
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: |
Sent: Friday, July 12, 2013 6:29 PM
To: Thomas, Jenny
Subject: Bristol Bay Peer Review Follow On
It]
Hi Jenny
The attached two Word (docx) documents complete my peer reviewer activity on the
Revised Draft Bristol Bay Assessment Peer Review Follow-On (this one needs an
acronymn!), specifically Task 2: Evaluate review materials. The attachments are 1.
comments as inserted into the comment-response document format, and 2. additional
comments not covered in the former approach.
Please let me know at your convenience if there is anything additional I need to do to
complete the invoice. Thanks for involving me in this important process.
DDD
-------
Dr. Dennis Dauble
July 12, 2013
Dennis Dauble
Peer Reviewer Evaluation of Second External Review Draft (April 2013):
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
ADDITIONAL COMMENTS NOT COVERED BY
THE COMMENT RESPONSE FORMAT
In terms of overall organization, the second draft is much improved. One
issue is Section 6.4 (conceptual model) included with descriptions of the mine
footprint and operations did not work for me. I do agree, however, that this section
serves as a useful transition from Mine Scenarios (Chapter 6) to the first risk
assessment chapter (Chapter 7). To resolve this arrangement, I suggest you either
(1) Move Section 6.4 to the end of Chapter 5 and expand the Introduction/Overview
to Chapter 7 to include a brief discussion that it is the first of several risk
assessment chapters, or (2) make Section 6.4 a separate chapter (after the current
Chapter 6 and before the current Chapter 7).
Although the title says "ecosystem", one weakness of the document continues
to be the paucity of ecological information, i.e., the ecosystem does not begin and
end with salmon Nutrient exchange focuses almost entirely on the role of salmon
and marine-derived nutrients. Watershed and terrestrial inputs are rarely
discussed, nor is bottom-up energy flow from algae/periphyton to aquatic
invertebrates to fish. I revisit this point several times in my other comments.
Overall, the second external review draft was impressive in terms of detail,
supporting graphics and tables, and depth of discussion. One challenge to having so
many complex tables was that often little or no narrative text was provided, leaving
the reader to figure out their meaning largely on their own. I strongly suggest
adding one or two sentences for each table in the text prior to the "call out" that
states key results or example data from that table.
One useful addition to the second draft was the summary of evidence tables.
Although somewhat subjective (or qualitative), they helped provide nice summary
of effects for each exposure scenario. I could provide numerous examples of other
useful tables and figures but there were too many to count. Needless to say this
document should serve as a model for completeness. The authors should be
collectively proud of their accomplishment.
1
-------
Dr. Dennis Dauble
The following specific technical and editorial comments are also offered to
improve the Assessment document.
1. p. ES-4, line 4- On what basis would the site be monitored during the post-mining
phase? Would a specific state or federal regulatory requirement be in effect?
2. p. ES-28, bullet 4- There should be an additional statement about uncertainty as it
relates to effects on primary and secondary production, hence the salmonid food
web.
3. p. 5-34, last line of page- The call out for Table 5-1 should come after "non-salmon
fishes.:
4. p. 7-12, section 7.1.2- Please clarify if aerial index counts are estimates of
individual fish or redds.
5. p. 8-29- Please state if avoidance is considered harmful (i.e., fish are excluded
from preferred habitat) or good (i.e., they are able to detect and move to non-
polluted areas).
6. p. 8-31, line 19- Based on what author's experience?
7. p. 8-43, Table 8 19- Is something missing here? No data is presented.
8. p. 8-48, Table 8 22 and p. 8-49, Table 8 23- What does the dash or hyphen (-)
mean in these tables? No data or no effect?
9. p. 8-62, section 8.3.1.2- Heat transfer between atmosphere and stream bottom
will likely over-ride and mitigate any increases in water temperature than might
occur downstream of the confluence of the source.
10. p. 9-19, line 12- Thr reference to Tables 9-3 and 9-4 provides an opportunity to
provide relevant examples of depth and velocity from each scenario.
11. p. 10-20, section 10.3.2- Please clarify the last sentences of the 1st paragraph.
12. p. 11-11, line 11- This sentence implies contaminated groundwater would
expose aquatic organisms to toxic concentrations. Suggest add a caveat, such as "If
toxic concentrations were present in groundwater, areas of upwelling might pose a
risk to aquatic invertebrates and fish eggs/larvae unless sufficient dilution
occurred."
13. p. 11-17, line 16- "non-trivial" is an odd choice of words. Suggest replace with
significant or large or substantial.
14. p. 12-5, line 3- the dietary maximum for pigs and poultry can be used as a
surrogate benchmark for what? All wildlife? Bears and ducks? Please state.
15. p. 12-8, line 14- It's not obvious there is reliance of subsistence users on water
for transportation in the area of probable impacts due to water-withdrawal. Specific
locations should be identified if possible.
16. p. 13-8, line 15- Note that additional mines might also "tier off" the
infrastructure developed for the Pebble 0.25 scenario.
17. Chapters 8, 9 and 13 included insufficient discussion of risk to primary and
secondary production. For example, there would be loss of production due to
sediment deposition and scouring (tailings dam failure), due to floe formation from
leachate (water collection, treatment and discharge), and would multiply across
projects (cumulative effects). Further these alterations in the aquatic community
would impact focal fish species through loss of food, and, ultimately, subsistence
lifestyles.
2
-------
Dr. Dennis Dauble
June 12, 2013
Dennis Dauble
Peer Reviewer Evaluation of Second External Review Draft (April 2013):
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
This letter report was developed in response to Statement of Work (SOW), Revised Draft
Bristol Bay Assessment Peer Review Follow-On, of June 2013. The general approach as
specified by the SOW was to 1. Review key recommendations in the Executive Summary of the
17 September 2012 Final Peer Review Report, External Peer Review of the EPA's Draft
Document, An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay,
Alaska (Draft Assessment), in addition to specific comments and suggestions provided on the
Draft Assessment by myself (Contractor); 2. Review the EPA's response to the Peer Review
Report, Draft of June 12, 2013; and, 3. Review the April 2013 revised Draft Assessment, revised
Appendix I and new Appendix J to evaluate how well these documents incorporated and
responded to comment provided by this Contractor.
On June 26, 2013,1 received an e-mail from Jenny Thomas, EPA Contracting Officer,
that suggested peer reviewer letter reports be organized in a manner to allow EPA to easily see
how our evaluation comments correspond to their responses to the Peer Review Report.
Consequently, I used EPA's draft response to comments document as a template for my
evaluation. My evaluation comments follow EPA's bold italicized responses in blue font.
It should be noted that several of my first-round review comments were responded to by
referring to changes made to appendices to the main report. However, without having these
appendices to examine, it was impossible to assess whether my comments were adequately
resolved in the revision process.
Additional comments of concern to this reviewer that did not neatly fit into the peer
review response to comment document format are attached as a separate document. These
comments are both editorial and technical in nature and should be considered prior to release of
the final Assessment document.
-------
Dr. Dennis Dauble
KEY RECOMMENDATIONS FROM PEER REVIEWERS, EPA RESPONSE AND
DENNIS D. DAUBLE'S CORRESPONDING EVALUATION OF REVISIONS MADE TO
THE APRIL 2013 REVISED DRAFT BRISTOL BAY ASSESSMENT
Scope of the Document:
• COMMENT: Articulate the purpose of the document more clearly via a primer on the
Ecological Risk Assessment process. If the purpose of the assessment is to inform EPA as the
decision maker, then the level of detail should correspond lo this purpose. The authors should
justify and explain what level of detail is required.
•RESPONSE: Additional information on both the purpose of the assessment and
ecological risk assessment (ERA) in general has been added to Chapters 1 and 2, as well as
the Executive Summary. Section 1.2 includes information about the use of the assessment.
The assessment has been reorganized into two major sections (problem formulation, risk
analysis and characterization) to clarify where different chapters fall in the typical ERA
process.
Dennis D. Dauble- The first page of the Executive Summary now provides an excellent
lead-in to the document. The first four paragraphs dearly establish context and intent of
the main Assessment document. Further detail is provided in the first section, Scope of the
Assessment.
• COMMENT: Include a statement upfront about the role of risk managers and other
audiences, such as project managers/engineers, regulators, mine owners/operators. Knowing
their role ensures inclusion of information necessary for any risk assessment by (1)
describing the need lor a risk assessment, (2) listing those decisions influenced, and (3)
characterizing what risk managers require from the risk assessment.
RESPON SE: Section 1.2 of the revised assessment discusses the use of the assessment.
Dennis /). Dauble- This comment was adequately described in Section 1.2 as well as
paragraph 3-5 of the Executive Summary.
• COMMENT I Explain u hy the scope for human and wildlife impacts was limited to fish-
mediated effects, as well as u hy fish-mediated effects on humans were limited to Alaska
Native cultures Re\ iewing effects beyond fish-mediated ones (e.g., potential for complete
loss of the subsistence way of life) would improve the assessment.
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. Throughout the assessment we acknowledge that direct
effects of large-scale mining on wildlife and Alaska Native cultures may be significant, but
that these direct effects are outside the scope of the current assessment.
Dennis D. Dauble-I am comfortable with revisions made to describe the scope of the
assessment. As far as potential effects to Alaska Native cultures, it's my opinion that fish-
mediated impacts should be the focus. Further, risk of large-scale mining activities to their
subsistence lifestyle is covered as well as it can be at this point in time.
2
-------
Dr. Dennis Dauble
• COMMENT: Be more consistent throughout the document in terms of the level of detail
provided for the different scenarios and stressors. For example, the document has devoted 36
pages to the discussion of catastrophic Tailings Storage Facility (TSF) failure, while sections
on the pipeline, water treatment, and road/culvert failures are brief. Indeed, the long
discussion on the TSF failure belies a certainty and understanding of dam failure dynamics
that is inaccurate.
RESPONSE: The final document includes more failure scenarios (e.g., diesel pipeline
failure, wastewater treatment plant failure, and refined seepage scenarios) in Chapter 8. It
also explains why these specific failure scenarios were chosen, and discusses these
scenarios in greater detail than the previous draft (i.e.. to more closely match the level of
detail originally provided only for the TSF failure scenario). . 1 Iso see detailed responses to
comments on Peer Review Question 5.
Dennis D. Dauble- The Revised Draft Assessment is now a more balanced document in
terms of level of detail for individual failure scenarios. I believe the amount of text
devoted to each failure scenario is appropriate for the level of perceived risk andfor the
amount of information available.
Technical Content:
Mine Scenario
• COMMENT: Consider the document lo he a scieeninu-le\ el assessment of all potential
stressors. Focusing on failure mode overemphasizes catastrophic events (e.g., TSF failing),
rather than considering all potential stressors, such as holding mine owners strictly
accountable for their day-to-day activities with regard to best practices.
RESPONSE: Additional information on the purpose and scope of the assessment has been
added to Chapters 1 and 2. A screening of all potential stressors, including individual
chemicals, is presented in Section 6.4.2. Also see detailed RESPONSES to Peer Review
Question 2 on the use of "best practices" and RESPONSES to Peer Review Question 5 on
failure scenarios.
Dennis D. Dauble- While it may appear the document focused on failure mode, these are
the types of events ofprimary concern. Thus, they require more discussion in terms of
both potential effects and uncertainty. One suggestion is to add a statement early in both
the Executive Summary and Introduction that states something like " This document
includes an assessment of potential failure scenarios, as well as effects that may occur
during day-to-day operations of a large-scale mine."
• COMMENT: Reexamine the document's use of historical data and case studies to describe
and estimate the risk of failure for certain mine facilities (including the TSF, pipeline, water
treatment, etc.), as these examples from extant mines may not be an appropriate analog for a
new mine in the Bristol Bay watershed.
RESPONSE: The TSF failure range was, and still is, based on design goals, not the
historical data. The historical TSF failure data are provided as background. The pipeline
failure rates are based on the most relevant historical data from the petroleum industry.
3
-------
Dr. Dennis Dauble
They are directly relevant to the dieselpipeline, and experiences at the Alumbrera mine
(described in the previous draft) and the Antamina and Bingham Canyon mines (added to
this draft) suggest that they also are relevant to the product concentrate pipeline. Water
treatment failure rates were not quantified. However, recent reviews cited in the revised
draft indicate that water collection and treatment failures have been reported at nearly all
analogous mines in the U.S. The estimation of culvert failure frequencies has been revised
and is now based on only recent literature (2002 and later). We believe that these estimates
are appropriate.
Dennis D. Dauble-1 was impressed with how much new literature was added to the
Revised Draft, particularly with the number of relevant examples from mining and other
industry. Ifound the use of "box text" to highlight specific examples to be very
informative.
• COMMENT: Expand the discussion on 1 Ik- use of "best" management practices, as the
document states that the mine scenario employs "good," but not necessarily "best" practice.
For a mine developed in the Bristol Bay watershed, only "best" practice likely would be
appropriate and anything less may not be permi lied. E-ven so, without a track record of "best"
practice (e.g., new technologies), we cannot assume that technology, by itself without
appropriate operational management controls, can always mitigate risk.
RESPONSE: The term "best management practices " is a term generally applied to
specific measures for managing non-point source runoff from storm water (40 CFR Part
130.2(m)). Measures for minimizing and controlling sources of pollution in other
situations often are referred to as best practices, state of the practice, good practice,
conventional, or simply mitigation measures. We assume that these types of measures
would be applied throughout a mine as it is constructed, operated, closed, and post-closure,
and have used the term "conventional modern " throughout the assessment to refer to these
measures. To remove any ambiguity related to the subjectiveness of terms "good" or
"best", we have removed them in the revision and have provided definitions for relevant
terms used in Box 4-1.
Dennis I). Dauble- Box 4. / does an adequate job of resolving this comment. The topic is
further covered in Appendices I and J.
• COMMENT Adopt a broader range of mine scenarios (not only minimum and maximum) so
as to bound potential impacts, especially at smaller mine sizes (e.g., 50th percentile).
Underground mine de\ elopment, with its different impacts, also should be considered and
included in the assessment.
RESPONSE: A third mine size scenario (250 million tons) has been added to the
assessment, to represent the worldwide median sized porphyry copper mine (based on
Singer et al. 2008).
Dennis D. Dauble- The third scenario is an excellent addition. More important, in my
mind, was comparative data on other mines within the region, for example Table 4.1.
• COMMENT: Based on the hypothetical mine scenario, perpetual management of the
geotechnical integrity of the waste rock and tailings storage facilities, as well as perpetual
4
-------
Dr. Dennis Dauble
water treatment and monitoring, will most likely be necessary (i.e., a "walk away" closure
scenario after mining ends may not be possible). Therefore, emphasize how monitoring and
management of the geotechnical integrity of waste rocks and tailing storage facilities should
continue "In Perpetuity" (i.e., for at least tens of thousands of years). Discuss what conditions
would need to be met to allow "walk away" closure in the Bristol Bay environment gaining
insight into these observations from mines where perpetual treatment and monitoring are
ongoing (e.g., the Equity Silver Mine in British Columbia).
COMMENT: RESPONSE: The conditions for closure and the potential needfor perpetual
site management are discussed in general terms in the revised assessment The primary
condition assumed to be required is water chemistry that meets all criteria and permit
conditions and that is stable or improving. However, even though there are some facilities
with "perpetual treatment" conditions in place, there is obviously no information about
how these facilities perform over very long periods of time.
Dennis D. Dauble- Given my experience with assessments of long-term risk of
contaminants from the Hanford Site in southeastern Washington State. I can understand
the suggestion for 10,000 year or "in perpetuity" scenarios. However. / believe this risk
scenario is difficult if not impossible to describe give the huge range of possibilities for
mining development and potential impacts on the environment. I am comfortable with the
general terms provided in the Assessment.
• COMMENT: Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting,
overflight, etc.) were managed. This includes roads, airslrips. helipads, camps, fuel dumps,
and ATV trails thai ha\ e already been de\ eloped or imposed on the watershed, and what
"mitigation" already has been undertaken 011 those sites Assess the consequences/impacts of
these activities in the Cumulative Risks section
RESPONSE: The effects of exploratory activities are outside of the scope of this
assessment.
Dennis /). Ihiuble- / disagree, in part. Exploratory activities are one part of mining
development where Stage 1= exploratory, Stage 2=construction, Stage 3=operation, and
Stage 4-site closure. The Main document should at least provide a brief description of
whether specific exploratory activities such as drill holes, blasting, roads,
airstrips/helipads, camps and trails fit within or are "superseded" by the overall mine
development footprint. / would argue this description would suffice and that exploratory
activities need not be discussed in either a mitigation or cumulative risk chapter.
Risks to Salmonid I 'ish
• COMMENT: Place potential mining impacts in the context of the entire Bristol Bay
watershed by emphasizing the relative magnitude of impacts. For example, of the total
salmon habitat, assess the proportion lost due to mining. Further, reflect on the non-linear
nature of the relationship between habitat and salmon production; 5% of the habitat could be
critical and thus responsible for 20% or more of salmon recruitment. Intrinsic potential,
which measures the ability of particular habitats to support fishes, would lend credibility to
this analysis.
5
-------
Dr. Dennis Dauble
RESPONSE: We are unable to build a complete Intrinsic Potential (IP) model, as this
would require validation and more elaborate construction of metrics appropriate to this
region. Our preliminary characterization provides the building blocks for assessing the
distribution of key habitat-forming and constraining features across these watersheds. We
now include a characterization of the major drivers of habitat potential across the
watershed and place the mine-site specific effects in this context (Chapters 3, 7, and 10).
Dennis D. Dauble-I noted several significant improvements in relating salmon habitat lost
in terms of its rearing and/or spawning potential. Species distribution and habitat use
maps have been "upgraded" throughout the document to reflect life history attributes in
relation to mining development.
• COMMENT: Include a section on the impact of Global Climate Change with explicit
reference to a monitoring program that will allow scientists. if the mine is built, to distinguish
between effects of climate change and mining effects on the physical and biological
components of this ecosystem.
RESPONSE: Climate change projections and potential impacts are now included in
Chapter 3, and as important external factors in the risk analyses presented in Chapters 7,
9,10, and 14. Development of a monitoring program to distinguish between mining and
climate change effects is outside of the scope of the assessment.
Dennis D. Dauble- The topic of climate change was extensively and adequately covered
throughout the revised Assessment, both in terms of additive risk and uncertainty. The
authors are to be commended for their careful attention to this complicated topic.
• COMMENT: Explicitly recognize thai the I mil spoliation corridor and all associated ancillary
development, including future resource de\ elopments made possible by the initial mining
project, will necessarily and ine\ itably ha\ e impacts (hydrologic, noise, dust, emissions,
etc ) These impacts will \ary in duration, intensity, severity, relative importance, spatial
dispersion, and ine\ itably expand geographically through time with further "development."
These impacts should be incorporated into the Cumulative Risks section.
RESPO NSE: The cumulative risk section (Chapter 13) has been expanded to include the
multiple transportation corridors, ancillary mining development and secondary
development associated with multiple mines in a qualitative discussion. The issues
addressed in the assessment of the transportation corridor (Chapter 10) have also been
expanded to include chemical spills, dust, invasive species, and road treatment salts.
Dennis D. Dauble- I am comfortable with revisions made on this topic in both the
transportation and cumulative risk chapters.
• COMMENT: Incorporate current research findings into stream crossing and culvert-design
practices (e.g., arch culverts, bridges, etc.).
RESPONSE: We describe current culvert design practices in a box titled "Culvert
Mitigation" in Chapter 10.
Dennis D. Dauble-1 appreciate that new research findings were incorporated in the
Revised Draft document, primarily in section 10.3.2.3 (which appeared somewhat biased
negatively towards potential magnitude of risks). Given that culvert failure is one risk to
6
-------
Dr. Dennis Dauble
fish populations from development of the transportation corridor, this risk should be
specified in paragraph one of section 10.3 Potential Risks to Fish Habitats and
Populations. I would argue that current culvert design practices are not covered in Box
10.2. Rather, this box (with the exception of the Tier 1-Stream Simulation Design which is
not alluded to in the text) provides useful detail on regulations associated with culvert
design and placement.
• COMMENT: Recognize in the assessment that risk and impact are not equivalent. Risk may
be low, but the potential impact could be huge (e.g., in the case of a TSF failure).
RESPONSE: Risk has been defined in many ways, even by risk assessors. The commenter
seems to define risk as probability. To avoid that potential source of confusion, we use the
term "probability"for that concept. Similarly, the commenter seems to use "impact"
where we use "effect" or "magnitude of effect''. II e use "risk" to refer to both concepts
combined—that is, an event or effect and its probability).
Dennis D. Dauble-1 am comfortable with the terminology and definitions of risk as
provided in the Revised Draft Assessment.
• COMMENT: Recognize and justify chronic l">eha\ ioial end points, such as those potentially
affecting survival and long-term success of fish populations
RESPONSE: The chronic behavioral effects of copper on salmonids, the primary endpoint
of concern, were described in Chapter 5 and are now described in Chapter 8. Although
those effects occur at lower levels of copper than conventional survival, growth and
reproduction end points for salmonids, they are less sensitive than the conventional
endpoints for aquatic invertebrates.
Dennis D. Dauble- / noted additional text related to behavioral effects in Chapter 8. While
of value in providing for perhaps a greater level of protection, behavior (such as avoidance
or olfactory response) is a much more difficult endpoint to measure and extrapolate to the
real world than endpoints such as survival, growth and reproduction. I am comfortable
with the level of detail provided in the Revised Draft Assessment as it relates to potential
behavioral effects on salmonids.
Wildlife
• COMMENT Recognize that the draft assessment did not account for all levels of ecology,
such as the indi\ idual (eg. a bald eagle nest), population, community, ecosystem, and
landscape levels I 'old other levels of organization into the stressors assessment where
appropriate or justify a more limited approach.
RESPONSE: As is appropriate for an ecological risk assessment (as opposed to an
environmental impact assessment), this assessment focuses on a specific, limited set of
endpoints as defined in Chapter 5. We have added text in Chapters 2 and 5 to explain both
why these endpoints were selected, and that responses other than those considered in the
assessment, at multiple levels of ecological organization, are likely but are outside the
scope of the assessment.
Dennis D. Dauble-I agree that the number of individual endpoints and levels of
7
-------
Dr. Dennis Dauble
organization must be limited in an assessment that covers such a broad geographic area.
• COMMENT: Discuss in the document fishes other than salmonids The assessment focuses
on risks to sockeye salmon in the Bristol Bay watershed (and also considers anadromous
salmonids, rainbow trout, and Dolly Varden), but does not account for potential impacts to
other members of the resident fish community. Further, primary and secondary production,
including nutrient flux was not addressed. Expanding the assessment to consider other levels
of organization, including direct as well as indirect effects on wildlife and other fish, would
provide additional context in the assessment of mine-related impacts.
RESPONSE: See response to comment above; uv also incorporated additional
information from Appendices A, B, and C into the C hapter 5 text, to provide additional
detail on the area's biota. We chose our endpoints for reasons described in Chapters 2 and
5. Other endpoints, including indirect effects on fish and wildlife. are now discussed more
explicitly, but are generally considered outside the scope of the assessment.
Dennis D. Dauble-This bullet is actually two separate comments. / am now comfortable
with the amount of information provided in the main document on fishes other than
salmonids. A remaining weakness, however, is lack of discussion of mining leachates on
primary production. While fish are the principal species of concern, the role of
algal/periphyton communities in streams potentially impacted by mining activities should
be acknowledged Specifically, scouring by sediments and flocculation due to leachates are
likely impacts to algal and invertebrate production that would affect fish populations due
to reduced food supply.
Human Cultures
• COMMENT I 'so case histories lo |no\ icle insight and anticipate mining impacts on Alaska
Nati\ es (e g . those exempli lying the l-\\on Vaklex oil spill impacts, cumulative effects of
oil and gas de\ elopnieni in the North Slope region, and social impacts related to mining
de\ elopment in Alaska)
RESPONSE: Examples from applicable case studies, including the Exxon Valdez oil spill,
are cited in Chapter 12 of the revised assessment.
Dennis D. Dauble-The lixxon Valdez oil spill case study, as presented in Chapter 12, is
relevant.
• COMMENT: As noted above (Scope of the Document), clarify why the scope was limited to
fish-mediated effects. The potential direct and indirect impacts for human cultures extend far
beyond fish-mediated impacts (e.g., potential complete loss of the subsistence way of life).
The rationale for this narrow focus should be fully explained. In addition, a clear explanation
should be given for why fish-mediated human impacts focused only on Alaska Native
cultures.
RESPONSE: The assessment focuses on a specific, limited set of endpoints as defined in
Chapter 5. We have added text to explain both why these endpoints were selected, and that
RESPONSES other than those considered in the assessment are likely but are outside the
scope of the assessment. The assessment was expanded (Chapters 5 and 12) to
8
-------
Dr. Dennis Dauble
acknowledge that there are a wide range of potential direct and indirect impacts to
indigenous culture, but they are outside of the scope of this assessment. The discussion of
potential effects to indigenous cultures was expanded to explain that a loss of subsistence
resources would extend beyond a loss of food resources to social, cultural, and spiritual
disruption. The text has been expanded to acknowledge the strong cultural ties of many
non-Alaska Natives to the region, and potential effects on all residents from loss of a
subsistence way of life. However, the focus of the assessment remains on effects on
indigenous cultures resulting from effects on salmon.
Dennis D. Dauble-Chapter 5 outlines (using EPA jargon not necessarily translatable to the
general public, e.g., Section 5.1), how assessment endpoints were selected and how they
relate to Alaska Native cultures. Chapter 5, while emphasizing fish-mediated effects on
Alaska Natives, also includes several graphics and supporting test that relates to
recreational fisheries.
Water Balance/Hydrology
• COMMENT: Better characterize water resources and assess the potential effect of mine
development on these resources by (1) generating a diagram similar to the conceptual models
beginning on page 3-7 to illustrate the potential effects of mine construction and operation on
surface- and ground-water hydrology. (2) developing a t|Lianlitative water balance and
identifying water gains and losses. {3) identifying seasonality of hydrologic processes,
including frozen soils and their associated \ alues (e.g., mni/yr) for each component of the
water balance; (4) incorporating these processes into a landscape characterization; (5)
evaluating how global climate change will influence these hydrologic processes and rates;
and 6) using this characterization to demonstrate the expected hydrologic modification
associated with the mine scenarios and infrastructure development.
RFSP().\SF: The original Figure 4-9 (new Figure 6-5) has been revised to more clearly
show water management in the assessment's mine scenarios. In addition, three schematics
illustrating water flows under each of the mine size scenarios (Figures 6-8 through 6-10)
have been added to Chapter 6. as have quantitative water balances for each mine size
scenarios. . I qualitative discussion of climate change is included in Chapters 3 (Section
3.8) and 14 (Hox 14-2).
Dennis D. Dauble- Figure 6.5 is a excellent general schematic of water management and
water balance processes for the mine scenarios. Figures 6.8-6.11 are quite helpful in
understanding local processes related to groundwater and surface water movement. The
summary of water balance flows in Table 6.3 provides additional specific detail. Overall,
the discussion of water balance/hydrology is much improved.
• COMMENT: Demonstrate the interconnectedness of groundwater, surface water, hyporheic
zone, and its importance to fish habitat. Address how interconnectedness changes over time -
seasonally, and with varying weather (e.g., wet vs. dry summers or years, and over the long
term as climate changes).
9
-------
Dr. Dennis Dauble
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
(Chapter 7) and qualitatively discuss the potential role of climate change (Chapter 3).
Dennis D. Dauble- Simply acknowledging that these interconnections exist and their
relative importance to fish habitat such as spawning and rearing satisfies this reviewer.
• COMMENT: Provide information on all rivers, including ephemeral and intermittent
streams, and first-order to main-stem streams that could be potentially influenced by the
proposed mine, its ancillary facilities, and the transportation corridor.
RESPONSE: Due to lack of consistent coverage, ttv rely on the NHD hydrography layer in
this analysis, and can only address ephemeral and intermittent streams qualitatively
(Chapter 7).
Dennis D. Dauble-1found the amount of detail provided on the stream network to be
sufficient given the limitation of comprehensive stream survey and other hydrographic
data.
• COMMENT: Emphasize the importance of a thorough characterization of ilie leaching
potential of acid-generating and non-acid generating waste rock and tailings, given the low
buffering capacity and mineral content in the streams and wetlands that could receive runoff
and treated water from the proposed mine. Recognize that collection and treatment of runoff
and leachate generated will be critical to maintain baseline water chemistry in these streams
and wetlands.
RESPONSE: M e agree that these are important issues, and the discussion of leachate
from waste rocks and tailings has been expanded in the revised assessment (Chapter 8).
Dennis D. Dauble- / appreciate that issues associated with water collection, treatment and
discharge are now discussed separately in Chapter 8. Although failure of the tailings dam
would be catastrophic, the Hklihood of such an event is much lower than those impacts
occurring from normal operations. C onsequently, the expanded discussion of impacts of
leachates from all possible sources is relevant.
Geochem i stry/M etal s
• COMMENT Reference the most current geochemistry data on potentially acid-generating,
non-acid generating, and metal leaching so as to describe any potential effects of seepage and
changes to surface- and ground-water quality via non-catastrophic failure.
RESPONSE: We used the geochemistry data in PLP's Environmental Baseline Document,
as summarized by the USGS in Appendix H. The effects of seepage on water quality are
analyzed in Chapter 8 of the revised assessment.
Dennis D. Dauble-These changes are noted and are appropriate.
• COMMENT: Explain how contaminants/metals were selected (and others ignored) by EPA
as causes for concern. Information should be included on additional metals and their toxicity
so as to assess impacts of potential leachates. The Pebble Limited Partnership baseline
10
-------
Dr. Dennis Dauble
document presented additional metals that might be useful to include in the assessment.
RESPONSE: The revised assessment describes the selection of contaminants and other
stressors of concern in Section 6.4.2. Additional metals, process chemicals and dissolved
solids are now included
Dennis D. Dauble- Agreed that other contaminants of concern are now mentions. But,
why was section 6.4 Conceptual Models placed in Chapter 6 Mine Scenarios? While I
agree it might serve as a transition or setup to the risk assessment chapters that follow, it is
too important to be included as a add-on to Chapter 6.1 think this section (which is mostly
methodology) should either be moved to the end of Chapter 5 Endpoints or made a
separate chapter. This change would only require slight re-wording of text in the
paragraph on p. 6-36 that begins with "In this section.... "
Mitigation Measures
• COMMENT: Incorporate the critical mitigation information from Appendix I into the main
report's mine scenarios. Include standard mitigation measures that could provide insight into
how well they might work in this context Tf this information is not included in the main
report, then justify its absence.
RESPONSE: Mitigation measures incorporated into design and operation to minimize
potential impacts were included in the assessment, as were some reclamation measures for
closure; these measures are made clearer in the revised assessment. These mitigation
measures were a sub-set of those presented in Appendix I. The assessment assumes that
measures chosen for the scenarios would be effective. Mitigation to compensate for effects
on aquatic resources that cannot be avoided or minimized by mine design and operation
would be addressed through a regulatory process that is beyond the scope of this
assessment. Nevertheless. in RESPONSE to public and peer comments we have included a
discussion of compensatory mitigation in . \ppendix J of the revised assessment.
Dennis D. Dauble-Appendix I lacks a summary or conclusion section, otherwise it reads
well. Appendix J provides a good description of regulatory requirements and methodology.
Considerations and potential mitigation measures were carefully and clearly described.
The opportunities for mitigation (against the regulatory backdrop) were summarized
within the watershed and challenges outlined. One suggestion is to strength the last
statement within the Conclusion section. Revise to perhaps something more along the
lines of "No viable alternatives could be identified to address impacts of this type and
magnitude."
• Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in the main
report, as they ultimately influence the range of mining impacts and consider time frames of
mitigation or reclamation measures (e.g., immediate RESPONSE, long-term reclamation).
RESPONSE: See response to previous comment. Mitigation measures are discussed at
greater length in the revised assessment report (e.g., Chapter 4 and Appendix J).
Dennis D. Dauble- The discussion of mitigation measures in Chapter 4 and Appendices I
and J seems adequate.
11
-------
Dr. Dennis Dauble
Uncertainties and Limitations
• COMMENT: Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of
uncertainty and (2) assessing the certainty of some mine impacts. Discuss data limitations in
the context of uncertainty.
RESPONSE: The individual analysis chapters and the revised Integrated Risk
Characterization (Chapter 14) discuss certainties and data limitations to a greater extent,
as suggested
Dennis D. Dauble- Chapters 8 through 12 each have uncertainty sections. Chapter 7
includes a section on uncertainties and assumptions. The Integrated Risk Assessment
(Chapter 14) covers uncertainties and data limitations. Thus. I believe this topic is covered
• COMMENT: Articulate early in the document how much uncertainty is acceptable. The
assessment provides little insight with respect to the decisions the document is intended to
support.
RESPONSE: Acceptable levels of uncertainty can be defined prior to an assessment if a
decision and a decision maker are identified and if data will be collected by a specified
design to implement a specified model, as described in the EPA '.s Data Quality Objectives
process, However, because this assessment is based on available data and is intended as a
background scientific document rather than a decision document, it is not possible to
specify the amount of uncertainty that is acceptable. Rather. the available data determine
the uncertainty and if the assessment is subsequently used to inform a decision, the
decision maker must determine whether the level of uncertainty is acceptable.
Dennis D. Dauble- I agree with the EPA response.
Editorial Suggestions:
• COMMI-NT The title of the document leads one to believe that the assessment addresses the
entire liristol Bay watershed; rather, the report deals with two major rivers and their
watersheds, the Nushagak and Kvichak. Thus, the title should be changed to reflect the
emphasis on these two rivers and their watersheds. A possible title may be "An Examination
(or identification) of the Potential Impacts of Mining and Mining Associated Activities on
Salmon Ecosystems in the Nushagak River and Kvichak River watersheds, Bristol Bay."
RESPONSE: The assessment addresses multiple scales: the Bristol Bay watershed, the
Nushagak and Kvichak River watersheds, the watersheds of the three streams draining the
Pebble deposit, and the watersheds crossed by the transportation corridor. These multiple
scales, and how they are used throughout the assessment, are described more clearly in the
revision (Chapter 2).
Dennis D. Dauble-1 was not one for changing the title, rather suggested an expanded
description of geographic scope. One issue I have with the revised Assessment was the use
of the phrase "multiple spatial scales" (secondfull paragraph p. ES-2). This term is
commonly used by stream ecologists to describe different measurement scales of habitat
assessments (i.e., individual, reach, watershed). As such, the term suggests a hierarchical
approach, which was not the case for this assessment. I prefer the term first used on p. 1-3
(third full paragraph Chapter 1 of the main document) "multiple geographic scales." I
12
-------
Dr. Dennis Dauble
suggest relevant wording in the E.S. be revised to be consistent with the description used
throughout the Introduction of the main document. In addition, please revise text under
section 2.2.2 Spatial Scales and Table 2.1 to show that data was considered across five
"geographic scales."
• COMMENT: Revise the Executive Summary to more precisely reflect the findings in the
document.
RESPONSE: The Executive Summary has been rewritten to reflect the revised assessment
findings.
Dennis D. Dauble-I was impressed with the revised Executive Summary (ES.) This section
might be the most important part of the revised Draft. I ssessment since due to being the
only thing some people will read. Thus, it is important that it be an accurate, yet synthetic,
version of the main document. Supporting graphics, e.g., Tables ES.3, ES.4 and Figure
ES.6 are excellent choices. Note, however, that topics in the E.S. are not presented in
parallel format or in the same sequence of topics (in terms of heading structure) with those
in the main document. Headings in the E.S. include the word "risks " while comparable
sections in the main document do not. The E.S. uses the term common-mode failures,
which I infer is the same topic as those discussed in C hapten 8 of the main document.
• COMMENT: The appendices contain detailed and useful information that should be
summarized and included in the main document (eg. Appendix E: Economics, Appendix G:
Road and Pipelines, and Appendix 1. Mitigation) Additionally, consider expanding the
preface to include information on the use of the appendices If the information is not included
in the main report, then justify its absence
RESPONSE: More information from the appendices was brought forward into
appropriate chapters of the revised report. The purpose of the appendices—to provide the
detailed background characterization necessary for the ecological risk assessment—has
also been clarified in Chapter 2. The document no longer contains a preface because that
materia! has been incorporated into Chapters 1 and 2.
Dennis /). Dauble-I noted several examples where specific Appendices were cited in the
main document as a source for more detailed information. It was also apparent that
material was brought forward where appropriate.
• COMMENT: Discuss in more detail the instructive and well-thought-out conceptual models
(pages 3-7 to 3-11) illustrating the impacts of mining on Bristol Bay ecosystem processes.
Also, consider expanding the conceptual models to include wildlife, fish-wildlife
interactions, vegetation/terrestrial habitat, and hydrologic processes. Allow them to guide the
text because they appear detailed and complete.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual
models presented in Chapter 6 (Chapter 3 in the first draft) have been broken into their
relevant component parts throughout the risk analysis and characterization chapters, to
better frame the specific pathways addressed in each chapter. Additional conceptual
13
-------
Dr. Dennis Dauble
models considering impacts on wildlife, Alaska Native populations, and cumulative effects
of multiple mines have been added to Chapters 12 and 13.
Dennis D. Dauble-This reviewer considers the conceptual models as an attempt to provide
a visual representation of a more much complicated (and largely unquantifiable) process
than can be described in words. It was helpful to include a separate model for separate
chapters. This representation simplified the thought process. One challenge is that Box
2.1 (so-called guide to conceptual model diagrams) was presented well before the first
(complicated) model in Chapter 6. Thus, the reader is forced to refer to the box to sort out
elements that composed the various pathways, modifiers, etc. One thing that would help
Box 2.1 is to provide specific examples of modifying factors (note the description is not
exactly in layman's terms), for example, temperature. Also, perhaps a short version of
what is included under "When viewing these diagrams.... " could be added to the legend of
each individual conceptual model to aid the reader. For example "Arrows leading from
one shape to another indicate a cause-effect relationship with bold lines, arrows and
outlines indicating high-priority pathways that were evaluated. Finally, there are up/down
arrows within most boxes. The legend should state what these mean.
• COMMENT: Incorporate the information contained in the conceptual models into a formal
framework, such as a Bayesian or other decision-analysis models.
RESPONSE: This is an excellent suggestion for future efforts, but is beyond the scope of
the current assessment. Creating a Bayesian Belief Network would require that the Agency
convene experts to subjectively estimate the probabilities of each transition in the
conceptual models. In contrast, this assessment is intended to elucidate the risks from
potential mining based on available data and analyses of those data. A Decision Analysis
would require that alternative outcomes be specified, the utility of each outcome for a
decision maker be defined and the probabilities of each outcome be estimatedfor each
possible decision so that the expected utilities of each outcome can be calculated. Because
this assessment is not a decision document, these requirements are not feasible or
appropriate.
Dennis /). Dauble- I agree with EPA's comment.
• COMMENT: Generate a standard operating protocol for significant figures and use it
throughout the document
RESPONSE: The authors have carefully addressed this issue. Numbers from the literature
or from the PLPEIil) retain the number of significant figures in the original. Numbers
derivedfor this assessment have the appropriate number of significant figures given the
precision of the input data and uncertainties due to modeling and extrapolation.
Dennis D. Dauble-1 am convinced proper homework was done on this topic.
• COMMENT: Remove all references to Mount St. Helens as a surrogate for a TSF failure.
Using a non-human-caused release of material into the ecosystem as an analogue for a mine
failure is not comparable in terms of likelihood or risk for a human-caused release. It would
be more appropriate to extrapolate from the impacts of known mine failures.
14
-------
Dr. Dennis Dauble
RESPONSE: We are puzzled by this comment. The Mount Saint Helens data were used
strictly to address the rate of benthic habitat recovery from a massive deposition of fine
mineral particles. The hydrological processes that determine the recovery of substrate
texture and the requirements of fish or aquatic invertebrates are not known to depend on
whether mineral particles were from a natural event or an anthropogenic event. We have
reviewed the literature on known mine failures. They studied tailings spills in terms of
toxicity but not in terms of physical habitat effects, which is why we used Mount Saint
Helens data. Nevertheless, we have removed references to Mount St. Helens in the revised
assessment to eliminate concern.
Dennis D. Dauble- The issue was raised primarily because of concern raised by some
members of the peer review panel that more relevant examples were available and should
be used. I have no issue if these references are used in the context of deposition of mineral
particles similar in nature to those that could be released during a TSF failure.
• COMMENT: Ensure that the draft assessment remains part of the puhlic record, allowing the
document history to remain intact.
RESPONSE: All drafts of the watershed assessment will remain part of the public record.
Dennis D. Dauble- No comment.
Research Needs:
• COMMENT: What are the acute and chronic impacts of mixtures of contaminants, including
metals, acid mine drainage, etc , on the fauna and llora of the Nushagak River and Kvichak
River watersheds ' What species are most sensitix e and might surrogate species exist for
those for which u e do not ha\ e data ' Re\ iew the European literature and regulatory
requirements for additional data
RES POSSE: The acute and chronic impacts of contaminant mixtures, including metals
and acid mine drainage (i.e.. metals in low pll-waters) were addressed using concentration
additivity models in the leachate chemistry tables in Chapters 5 and 6 (now Chapters 8 and
11). . Ulditional toxicity data were obtained by searches of the EU and OECD database
eChem. the EPA's ECOTOX and the Environment Canada site. More metals are now
included. In general, metals are most toxic to aquatic arthropods rather than fish, as
discussed for copper.
Dennis D. Dauble- What I found in the Revised Assessment was considerably more toxicity
data, including information from the most recent open-literature publications. The
potential risk of exposures to contaminant mixtures (a complicated topic) was addressed
using the additive model, which is about all you can do without having specific mixture
data.
• COMMENT: Can an inventory of nutrients, total organic carbon, and dissolved organic
carbon inputs to aquatic environments be developed that demonstrates their relative
magnitude and spatial variation from headwaters to Bristol Bay? What is the relative
importance of marine-derived nutrients relative to other nutrients from watershed and
terrestrial sources? What is the current atmospheric input of nutrients?
15
-------
Dr. Dennis Dauble
RESPONSE: These data would be very useful in the risk assessment, but are not currently
available for the Bristol Bay region. We agree this is a research need.
Dennis D. Dauble-Given the emphasis in this document on the importance of marine-
derived nutrients from salmon populations on the Bristol Bay watershed, it would be
instructive if an example or two was provided from the literature on the relative magnitude
and/or importance of nutrient inputs to watersheds from anadromous fish populations,
terrestrial, atmospheric and/or autochthonous sources.
• COMMENT: What are the locations of subsistence areas and can these areas be
characterized and differentiated by collecting local environmental and ecological knowledge
(e.g., fish overwintering areas, climate change, ecological shifts, etc.)?
RESPONSE: The revised assessment incorporated current data on subsistence use areas
available from ADF&G. EPA acknowledges that these data are incomplete and would
encourage additional collection of subsistence data and Traditional Ecological Knowledge.
Dennis D. Dauble- No comment.
• COMMENT: What impact might mining ha\ e on oilier important wildlife species in the
basin (e.g., freshwater seals in lliamna Lake)?
RESPONSE: The scope of the assessment is focused on potential risks to salmon from
large-scale mining and salmon-mediated effects to indigenous culture and wildlife. Direct
effects on wildlife from large-scale mining are likely to be important and Appendix C (now
a stand-alone US Fish and Wildlife report) provides useful information for a future
evaluation of direct effects on wildlife from large-scale mining. We agree that this is an
important area for future research.
Dennis D. Dauble- No comment
• COMMENT: What is the comprehensive hydrologic regime of the specific project mining
area, and the broader watershed system as characterized by baseline monitoring, spatial
distribution, and quantitativ e flow of surface- and ground-waters?
RESPONSE: Comprehensive spatial estimates of mean annual flow are now presented in
Chapter 3. Quantification of spatial and temporal patterns of groundwater flows is an
acknowledged highly desirable product, but it not feasible within the scope of this
assessment. Results of an independent groundwater-surface water modeling effort are
described in Chapter ~.
Dennis D. Dauble- It is apparent that the hydrological description was much expanded.
These additions are appropriate and informative.
• COMMENT: What is the cumulative impact of commercial fisheries on the Bristol Bay
watershed, especially in an ecosystem context as related to marine-derived nutrient and
energy flow? Acknowledge that commercial fishing has had an impact on the amount of
marine-derived nutrients returned to the watersheds.
RESPONSE: The impact of commercial fisheries on the watershed is not within the scope
of this assessment. Information on commercial fisheries management has been added in
16
-------
Dr. Dennis Dauble
Box 5-2. However, the purpose of this assessment is not to assess the relative effects of
potential mining and commercial fishing—it is to evaluate potential effects on endpoints if
a mine were to be developed, given existing conditions and activities in the region.
Dennis D. Dauble-1 am satisfied with the additional information on commercial fisheries
management practices included in the assessment.
17
-------
Dr. Dennis Dauble
WRITTEN PEER REVIEW COMMENTS- DENNIS D. DAUBLE
1. GENERAL IMPRESSIONS
• COMMENT: Overall, the main report and each of the accompanying appendices were well
written. I was unable to identify major inaccuracies or bias in the material as presented.
There were shortcomings in the main report, however. For example, some topics would
benefit by being expanded (Sections 5.6 and 8.7), while others have more detail than
appeared necessary (Section 6.1). The assessment effectively addressed three appropriate
time periods: (1) operation, (2) post-closure, and (3) perpetuity Potential effects are
bounded by a minimum and maximum mine size, u liich is also appropriate. Inclusion of
inference by analogy strengthened the conclusions reached in the assessment and helped
validate results obtained from model predictions.
RESPONSE: Previous Section 5.6 (wildlife and culture) has now been expanded and
treated as a stand-alone chapter (Chapter 12). The summary of risks from the mine
scenarios (previous Section 8.7, now Chapter 14) has been expanded to include fish-
mediated risks to wildlife and culture, and more numerical results are included
Dennis D. Dauble- These changes are all improvements to the May 2012 Draft
Assessment. Each chapter now begins with a clear description of what it includes. On a
more editorial note, Chapters 7. 8. 1 /. 12. and 14 begin with an overview of what follows,
but have no heading. In contrast. Chapters 9 and 13 begin with an Introduction or
Overview heading. These differences may have been due to different authors and different
writing styles.
• COMMENT: Most figures and laMes were useful The conceptual models and
accompanying illustrations of potential habitat effects (Figs 3-2A and C) are important
because they provide a view of complicated pathways and relationships among potential
acti\ ities and en\ ironniental attributes. I lo\\e\ er. these relationships are not revisited in any
detail later in the document I recommend discussing the conceptual models in more detail in
the main report (Section 3 (•>) and summary section in Chapter 8.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual
models presented in C liapter (> (previously in Chapter 3) have been broken into their
relevant component parts throughout the risk analysis and characterization chapters, to
better frame the specific pathways addressed in each chapter.
Dennis D. Dauble- Having separate (and less complicated) conceptual models relevant to
specific risk scenarios is an improvement.
• COMMENT: The Integrated Risk Assessment (Chapter 8) did a creditable job of
summarizing habitat losses and risks from mine operations. What is missing, however, are
quantitative descriptions of habitat lost relative to total habitat available in the larger
watershed and individual systems. Habitat loss should be further discussed in terms of
salmonid life stage and productivity (i.e., not all stream miles are equal).
RESPONSE: Unfortunately, no salmon habitat characterization is available for the
region. The State of Alaska has not even identified all anadromous streams in the region.
Productivity data are not available, even for the streams studied by the PLP. However, the
18
-------
Dr. Dennis Dauble
revised Chapter 14 contains tables summarizing habitat loss in stream lengths and wetland
areas.
Dennis D. Dauble- My question may have been misinterpreted (or poorly worded). What I
found in the revised Draft Assessment is a much improved description of habitat use by
salmonids of concern, including stream miles and locations for both spawning and rearing
(where known). I did not expect that habitat information was available for the entire
watershed There was some information on productivity in terms of the range of returning
adults.
• COMMENT: If anything, the conclusions could be strengthened. The summary of
uncertainties and limitations (Section 8.5) dwells 011 things that "could not be quantified" due
to lack of information, model limitations, or insufficient resources. Thus, this reader was left
somewhat in limbo as to the potential magnitude of effects IV0111 mining activities. (Note that
this "neutral voice" is carried throughout the Executiv e Summary) Many people might
interpret such statements of uncertainty as no proven effect My point is that probable
environmental consequences of mining acti\ ities are much greater than this report alludes to,
given that consequences are likely, even if their magnitude is "uncertain."
RESPONSE: We use a neutral voice throughout the document to convey the neutral
scientific perspective of this scientific assessment. II e tried to convey the qualitative
likelihood of occurrence when quan titative probabilities were not obtainable. This section
has been edited in Chapter 14 of the revised version to make the relationship between
uncertainty and probability of occurrence dearer.
Dennis D. Dauble- I am content with changes made to Chapter 14.
• COMMENT: Section S 7 is perhaps the most important section of the report. It should be
comprehensive, i e . co\ er all resources and he more quantitative. Missing from the
summary were impacts 011 wildlife, human culture, resident fish, and other ecological
resources l-ssential details from Appendices A. ('. li, F, and 1, for example, could be
synthesized and mined into the main report.
RESPONSE: The summary of risks from the mine scenarios (Chapter 14 in the revised
version) has been expanded to include fish-mediated risks to wildlife and culture and more
numerical results.
Dennis D. Dauble- Chapter 14 is a much improved version of what was formerly in
Chapter 8.
2. RESPONSES TO ( M ARGE QUESTIONS
Question 1. The EPA's assessment focused on identifying the impacts ofpotential
future large-scale mining to the fish habitat and populations in these watersheds. The
assessment brought together information to characterize the ecological, geological, and
cultural resources of the Nushagak and Kvichak watersheds. Did this characterization
provide appropriate background information for the assessment? Was this
characterization accurate? Were any significant literature, reports, or data missed that
would be useful to complete this characterization, and if so what are they?
19
-------
Dr. Dennis Dauble
• COMMENT: As noted in the approach, characterization of and risk to ecological resources
emphasized salmon and other important sport and commercial fish species. Consequently,
the description of non-salmonid species generally lacked estimates of population size, except
for sport and subsistence catch statistics. There was a long list of other resident fish in
Appendix A, but their role in the Bristol Bay watershed (including the Nushagak River and
Kvichak River watersheds) is not described in any detail there or in the main report.
Available data on known or perceived ecological interactions among salmonid and resident
fish should be included in the assessment.
RESPONSE: The assessment endpoints—salmonid fishes and their effects on wildlife and
Alaska Native cultures—have been clarified in Chapters 2 and 5; other fish species are
thus outside the scope of the assessment. However, uv recognize in the text that other
fishes (as well as other biota) are important components of the ecosystem, and have
included a table of all documented fish species in the region in Chapter 5 to better reflect
the fish fauna in the region.
Dennis D. Dauble- While I understand why many fish species are outside the scope of the
assessment, what's still missing is information on known ecological interactions (i.e.,
predation, competition) within the fish community. It's possible these interactions have as
much influence on the salmonid fish population as mining impacts or global climate
change might have, for example. Table 5. 7 is a start. . \nother table or brief description of
the ecological role of each important or abundant fish species in the Bristol Bay
water shed-in relation to salmonids- would be useful
• COMMENT: Another limitation to the salmon-centric assessment is that risk assessment
endpoints, described in Chapter 3 of the main report. do not address other aquatic ecological
resources. Consequently, while there was acknowledgment of ecological dependencies
among salmon, other fishes, and land mammals, very little information was provided on
primary and secondary production processes of aquatic communities. For example, the
relati \ e importance of marine-derived nutrients (MDN) in the form of salmon eggs and
carcasses is discussed, hul there is only brief mention of aquatic insects in the diet salmonid
species What nutrient le\ els occur in these stream systems with and without MDN?
RESPO N SE: We recognize that nutrient status, and more important prey availability, is a
critical component of habitat capacity for fish in these systems, and may be strongly driven
by salmon derived nutrients. We concur that more information is needed regarding
potential limiting factors for salmon productivity and capacity, and that food availability
may be one such factor. The role of aquatic invertebrates in the diet of salmonids receives
more attention in the revised draft, and is an essential part of the risk assessment for water
treatment and discharge, given the relatively high sensitivities of aquatic invertebrate taxa
to metals. However, because water chemistry data may not provide a complete picture of
trophic status, particularly where direct consumption of salmon flesh, eggs, and fry is of
such high importance as it is in many of the area streams, we determined that nutrient
status of area streams is outside the scope of this assessment.
Dennis D. Dauble-1 defer to EPA's judgment on this topic.
• COMMENT: A description of major groups of aquatic invertebrates in terms of biomass
and seasonal abundance should be included in the main report. Further, aquatic and terrestrial
food webs and linkages need more embellishment. One approach might be to add narrative
20
-------
Dr. Dennis Dauble
text with the conceptual model discussion, including descriptions of community structure,
function, and biomass.
• RESPONSE: Additional detail on food webs is beyond the scope of this assessment (as
detailed in Chapters 2 and 5). Further, available data are inadequate to assess risks at that
level of specificity. For example, there are no acute copper toxicity data for any aquatic
insects and only one old chronic value for a caddisfly.
Dennis D. Dauble-1 don't agree that detail on food webs as it relates to salmonids is
beyond the scope of the assessment. No acute copper toxicity data for any aquatic insects?
Note that effects on primary production were also ignored in the document.
• COMMENT: More detail on river and lake limnology would be helpful. For example, the
hydrology of the watershed is mainly limited to a brief discussion of salmonid habitats. The
geology of the basin emphasizes geology of mining ureas and mineral processes. A more
landscape-based description is warranted gi\ en the importance of geology to surface water
processes and groundwater movement. The report would benefit lYom having a summary
table listing lake size/volume and river length discharge for watersheds potentially affected
(and not affected) by mining activities.
RESPONSE: We now include maps of geology and estimated mean annua! flow for the
study region (Chapter 3).
Dennis D. Dauble- River length can be easily inferred from new, detailed maps of the
watershed and specific water bodies. Mean annua! flows are also now presented in
Chapter 3. Consequently, this comment was largely resolved.
• COMMENT: Also missing were specific habitat ret|Lii remenls for rearing of juvenile
salmon. A brief description of where pink and chum salmon spawn and rear in the Bristol
Bay watershed relati\ e to other salmon species should be included in the main report. There
was nothing in Appendix A on u here colio. pink, and chum salmon reside within the Bristol
Bay watershed
RESPONSE: Identified spawning and rearing habitats for the five Pacific salmon species
are reflected in Figures 5-3 through 5-8, and additional text on salmon life histories has
been included in Chapter 5.
Dennis D. Ihiuble- Figures 5-3 through 5-8 and supporting narrative text are excellent
additions to the main document.
• COMMENT: luich appendix has a wealth of supporting information and could serve as a
stand-alone document I lowever, having to work back-and-forth between the main report
and appendices to interpret critical aspects of the assessment presents a challenge. Don't
assume the average reader will read (and interpret) these appendices. To help remedy, the
authors of the main report should strive to directly cite relevant information (and/or a specific
appendix) that supports their conclusions.
RESPONSE: Additional information from Appendices A and B has been pulled into
Chapter 5 of the main assessment. In addition, the purpose of the appendices has been
clarified in Chapter 2.
Dennis D. Dauble- This approach is now apparent throughout the main document and is
both informative and helpful to the reader.
21
-------
Dr. Dennis Dauble
Question 2. A formal mine plan or application is not available for the porphyry copper
deposits in the Bristol Bay watershed. EPA developed a hypothetical mine scenario for
its risk assessment, based largely on a plan published by Northern Dynasty Minerals.
Given the type and location of copper deposits in the watershed, was this hypothetical
mine scenario realistic and sufficient for the assessment? Has EPA appropriately
bounded the magnitude of potential mine activities with the minimum and maximum
mine sizes used in the scenario? Are there significant literature, reports, or data not
referenced that would be useful to refine the mine scenario, and if so what are they?
• COMMENT: The hypothetical mine scenario initially appeared realistic and useful in terms
of potential project scope. However, it was apparent during the public hearing, and upon
further discussion between members of the panel, that assumptions on mine size should be
revisited based on deposit characteristics and extraction potential. Also, assumed practices
and operations should be verified against current best-practice and State of Alaska permitting
guidelines.
• RESPONSE:: The revised assessment includes a smaller sized mine that is based on the
median-sized porphyry copper mine on a worldwide basis. The State of. Maska does not
have permitting guidelines that address the size of a mining operation. Land use activities
were previously subject to stipulations meant to minimize surface damage or disturbance
under 11 Alaska Administrative Code (AAC) 96.140, but this regulation was repealed in
December 2002. The State does have statutory and/or regulatory requirements for an
approved Plan of Operations (11 AAC 86.800). a Reclamation Plan (Alaska Statute (AS)
27.19.30) and appropriate financial Assurance (AS 27.19.040).
Dennis D. Dauble- The three mine scenarios as presented in the revised Assessment cover
a realistic range of potential development.
• COMMI-NT Referenced literature pro\ ides appropriate context, however, I cannot help
belie\ e that information on environmental impacts from past mining activities conducted in
the Rocky Mountain metal belt would be relevant to this assessment in some cases. It is also
possible that recent published information from Holden Mine in northern Washington State
would help establish context lor effects of leachates and model results that predict
downstream transport of tailing material in a wilderness setting, for example.
RESPONSE: Environmental impacts from historic mining are the basis for understanding
that risks from hazards of mining need evaluation. Modeling of tailings transport was
based on the expected characteristics of tailings for the Pebble deposit. There is an
expanse of literature on Superfund sites and interactions of metals associated with
sediments and their leaching. We included a number of selected sites in our background
information, but to include all possible sites would get further away from the scope of the
assessment, which was to evaluate potential effects within the Bristol Bay watershed.
Dennis D. Dauble- In this reviewer's opinion, the writers did an adequate job ofproviding
comparative information from other mineral development sites.
Question 3. EPA assumed two potential modes for mining operations: a no-failure
mode of operation and a mode involving one or more types offailures. Is the no-failure
22
-------
Dr. Dennis Dauble
mode of operation adequately described? Are engineering and mitigation practices
sufficiently detailed, reasonable, and consistent? Are significant literature, reports, or
data not referenced that would be useful to refine these scenarios, and if so what are
they?
Dennis D. Dauble. Ph.D.
• COMMENT: The description of the no-failure mode for mine operation appears adequate
in terms of potential mitigation measures that might be employed. I have limited knowledge
of current engineering practices and subsequent risks lo the en\ ironment from best practices
of modern mines, including those operating under optimal conditions. However, it would be
helpful to include a short discussion on which mitigation measures would be most applicable
to mining activities in the Bristol Bay watershed
RESPONSE: Standard design mitigation measures considered feasible, appropriate, and
'permittable' (as per Ghaffari et al. 2011) were considered and are discussed in Chapter 6
and Appendix I of the revised assessment: these are standard measures common to other
copper porphyry mines. Evaluation of measures that would be proposed for an actual mine
would occur through the regulatory process. H hether these same measures would be
appropriate for all locations within the Bristol Hay watershed would depend on the given
site's specific characteristics.
Dennis D. Dauble- The topic of mitigation is adequately covered in Chapter 6 as well as
Appendices I and J of the revised assessment.
Question 4. Are the potential risks to salmonid fish due to habitat loss and
modification and changes in hydrology and water quality appropriately characterized
and described for the no-failure mode of operation? Does the assessment appropriately
describe the scale and extent of risks to salmonid fish due to operation of a
transportation corridor under the no-failure mode of operation?
• COMMENT: The assessment describes the number of stream miles impacted under each
mode of operation, including miles blocked and eliminated. Less specific were descriptions
of impacts due to sedimentation and leachates. What is lacking is quantitative estimates of
spawning and rearing habitat that would be lost relative to the total habitat available. Having
this information would help provide perspective of overall risk to individual watersheds and
the Bristol Bay watershed as a whole. Risks to salmonid fish due to changes in water quality
(i.e., toxic materials) need to consider differences in sensitivity and behavioral response
according to salmonid life stage.
RESPONSE: Stream habitat losses are now characterized in relation to the distribution of
habitat conditions throughout the larger watersheds (Chapters 3 and 7). The assessment of
risks from aqueous toxicity distinguishes overt toxic effects on early life stages from
behavioral effects on adults.
Dennis D. Dauble- Stream habitat losses and toxicity are well-described in terms of life
stage and species for each of the identified risk areas described in Chapters 7 through 11.
One problem is that many tables (e.g., 8.4-8.8) lack narrative text. These tables are
somewhat complicated and not self-explanatory and should not be introduced to the reader
23
-------
Dr. Dennis Dauble
as a citation without supporting narrative text. For example, what are quotients? Does a
low quotient value relative to the CMC or CC indicative of no risk? Not until Box 8.3 (or
20 pages after tables 8-4 to 8-8 is an explanation provided so that readers might know what
the data in these tables means. Strongly suggest that Box 8.3 be movedforward in this
chapter and that narrative text be added to the document.
• COMMENT: Surface water characteristics of site watersheds within the area of probable
impact are detailed in Table 5-17, but not so for other streams and lakes in the broader
watershed. More information should be presented where available. It is not clear whether
potentially affected streams and lakes might be nutrient limited (seems that they might be
given their dependence on MDN). For example, include N or P concentrations and some
discussion about primary and secondary productivity
RESPONSE: We recognize that nutrient status, and more importantly prey availability, is
a critical component of habitat capacity for fish in these systems, and may be strongly
driven by salmon derived nutrients. We concur that more information is needed regarding
potential limiting factors for salmon productivity and capacity, and that food availability
may be one such factor. However, because water chemistry data may not provide a
complete picture of trophic status, particularly where direct consumption of salmon flesh,
eggs, andfry is of such high importance, and because nutrient status is a water quality or
habitat parameter not directly in fluenced by mining operations as outlined in our
conceptual models (e.g., Figure 7-1). ire determined that nutrient status of area streams is
outside the scope of this assessment.
Dennis D. Dauble- The amount of additional detail in C liapter 8 on stream hydrology is
impressive and useful for this reviewer. II hile nutrient status of the watershed relative to
mining impacts may be outside the scope of the assessment, this information should be
included as part of the characterization process.
• CO\fMEXT I found risks lo salmonid fish due to operation of the transportation corridor
well-described with respect lo spatial distribution offish and their habitats.
RESPONSE: ,\n changes suggested or required.
Dennis I). Dauble- No comment.
Question 5. Do the failures outlined in the assessment reasonably represent potential
system failures that could occur at a mine of the type and size outlined in the mine
scenario? Is there a significant type of failure that is not described? Are the
probabilities and risks of failures estimated appropriately? Is appropriate information
from existing mines used to identify and estimate types and specific failure risks? If
not, which existing mines might be relevant for estimating potential mining activities in
the Bristol Bay watershed?
• COMMENT: My experience in system failure of mines of the size and type outlined in the
scenario is limited. However, what does seem to be missing is the long-term effects of
leachates to receiving water bodies in any type of risk scenario, including both non-failure
and failure modes. That is, assuming no catastrophic failure, how might tailings constituents
24
-------
Dr. Dennis Dauble
interact with aquatic habitats seasonally, such as during periods of snowmelt and severe
rainfall events?
RESPONSE: The original draft assessment contained a scenario in which tailings
leachate was not fully contained and reached a stream (Section 6.3 in the May 2012 draft).
The revised assessment includes estimates of leachate escaping from the TSFs and from
the waste rock piles, bypassing the collection systems, and entering the streams (Chapter
8). The estimated loadings of copper and other elements from these leachate flows are
included in stream concentration estimates. The assessment discusses the impacts of these
concentrations on the aquatic habitat and biota. The commenter is correct that we did not
include a scenario in which the dam does not fail, but snowmelt and severe rainfall would
result in overtopping and release of untreated water. That is very plausible, but there are
just too many possible failure scenarios to include more than a few of them.
Dennis D. Dauble- If the scenario is plausible, it should be included in the assessment.
However, if the potential risk is covered based on a similar or parallel scenario, then it
should be stated as such.
Question 6. Does the assessment appropriately characterize risks to salmonidfish due
to a potential failure of water and leachate collection and treatment from the mine site?
If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
• COMMENT: More information on local hydrology, including seasonal runoff patterns
(e.g., peak flows) and groundwater movement would be useful. I found no description of
existing water quality characteristics of potential receiving waters, except what is included in
Tabic 5-17 of the main report. Are these values (such as hardness, which moderates metal
toxicity) consistent throughout the watersheds, including downstream lakes? Other questions
include What \ oliimes of leachates might he collected and treated versus volumes not
captured and subsequently released downstream? Is copper the only constituent of concern to
aquatic animals? Are there processing chemicals that would also be toxic?
RESPONSE: Monthly flow patterns for area streams are now presented in Chapter 7. The
three streams described in Table 7-17 are the three potential receiving waters for any site
effluents. The water balance, including leachate volumes, is now described in the
assessment (Chapter 6). Copper is the primary contaminant of concern. Others are
described in the new Section 6.4.2.3 and discussed in Chapters 8 and 11.
Dennis D. Dauble-1 noted hardness values were now included in Chapter 3. Also, several
new maps (Figures 61 to 6 3 and 6 8 to 611) provide sufficient detail of regional and local
hydrology that flow paths for receiving waters are clearly understandable.
• COMMENT: The assessment should also consider and discuss relative risk to aquatic
ecosystems from downstream transport of sediment-bound metals to Iliamna Lake, if deemed
probable.
RESPONSE: Although metals in aqueous emissions would partition to sediment, and the
sediment would mobilize during high flows and eventually reach the lake, this route is not
25
-------
Dr. Dennis Dauble
judged to be significant. Toxicity is caused by dissolved metals, and concentrations from
release of metals from transported sediment to lake water are likely to be minor.
Dennis D. Dauble-1 defer to EPA's judgment on this topic.
Question 7. Does the assessment appropriately characterize risks to salmonidfish due
to culvert failures along the transportation corridor? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so what are
they?
• COMMENT: Mitigation practices, such as new ail\ en design, was well described, as was
bridging of roadways and porous fills to mitigate risks due lo culvert failure along the
transportation corridor. This assessment also included appropriate risk characterization for
both the no-failure and failure scenarios. There should be literature a\ ailable from the
Washington State Department of Transportation on fish passage relali\ c to culvert placement
and design. Otherwise, I have no suggestions for improvement.
RESPONSE: We have examined literature on fish passage relative to culverts from the
Washington State Department of Transportation. C u I vert failure frequencies from their
2012 paper, in which approximately (>2"» of culverts were identified as total or partial
barriers, were not used in the assessment because we could not determine the age of
examined roads.
Dennis D. Dauble- I had no serious issues with the assessment ofpotential culvert failures.
However, I can't help but believe that data on timing of road construction for WA State
DOT projects would be available. In terms of aquatic invasive species (Section 10.3.6.1),
no plausible vector was described for their transport. What types of construction
equipment or mining infrastructure would be expected to lead to the introduction of
aquatic invasive species? Typical examples would be recreational vehicles, such as boat,
but that does not seem applicable. Suggest you introduce the topic but tone down the risk
relative to mining operations.
Question 8. Does the assessment appropriately characterize risks to salmonid fish due
to pipeline failures? If not, what suggestions do you have for improving this part of the
assessment? A re significant literature, reports, or data not referenced that would be
useful to characterize these risks, and if so what are they?
• COMMENT: The risks lo salmonid fish due to release of pipeline concentrate/slurry and
leachates (as return water) are well described. However, risks of a diesel fuel spill are not.
More detail could be provided on reclaimed water. For example, what toxic constituents
(and at what volumes) would be released to the environment if these pipelines failed?
RESPONSE: A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11 in the revised assessment. New data on concentrate leachate in the slurry have
been added, and it is assumed to also describe the return water.
Dennis D. Dauble-1 appreciate the addition to the assessment. What's missing, however, is
an explanation of which mine scenario was selected for the pipeline discussion and why.
26
-------
Dr. Dennis Dauble
Not until section 11.3.4.4 was it stated that pipeline failures would be likely under the
Pebble 6.5 scenario. What about Pebble 0.25 and 2.0? Were they considered? If not,
please state why. Also, the last two bullets under 11.5.5 Uncertainties appear as conclusion
statements, not uncertainties.
Question 9. Does the assessment appropriately characterize risks to salmonid fish due
to a potential tailings dam failure? If not, what suggestions do you have for improving
this part of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
• COMMENT: Tailings deposition is described in ("haplcr 4 of the main report, but I could
not find anything that described potential risks to fishes, including effects to aquatic food
webs and loss of fish spawning and rearing habitat.
RESPONSE: The TSF failure description and the assessment of risks were presented in
separate chapters in the original draft assessment, but have now been moved into one
location (Chapter 9) for clarity
Dennis D. Dauble- This change is a big improvement over the previous draft. What I like
is the inclusion of analogous sites: Clark Fork. Coeur d'Alene River, and Soda Butte
Creek to the discussion. The summary of evidence concerning risks (Table 9.11) was also
helpful.
• COMMENT: As noted in the text, the sediment transport model used could only simulate
sediment transport and deposition -30 km downstream of the mine site. Thus, potential
effects to fish habitats were not well quantified for the main stem Koktuli River (and beyond),
in addition to the Mulchatna and Nushagak rivers. Is there likelihood that any tailings
material might reach I.ake Iliamna0 If not, say so in the document. It is equally useful to say
where impacts will not occur (as it relates to sensitive habitat) as it is to describe where
impacts are likely and reasonable
RESPONSE: None of the 3 TSl's are in the watershed of Iliamna Lake. The hydrology of
the site lias been clarified and better maps added to clarify this issue. Risks to Iliamna
Lake from water treatment failures, even if they occur in the Nushagak drainage, are now
noted (e.g., transport of toxic leachate from the South Fork Koktuli to Upper Talarik
Creek via groundwater exchange between theses basins; Chapter 8). However, sediment
(i.e., tailings) would not follow that route.
Dennis D. Dauble- The new maps are a welcome addition. It was useful to learn tailings
sediments would not likely reach Lake Iliamna and that transport of toxic leachates from
groundwater exchange would be a minor pathway.
• COMMENT: The assessment deemed that it was "not possible" to determine how far the
initial slurry deposition would extend, how far re-suspended sediments would travel, and
how long erosion processes would continue. I believe information from other mine closure
sites could be included by assessment authors to infer effect by analogy. The statement
alluding to potential sediment run out distance at the bottom of page 4-56 of the main report
should be included in the summary of effects. This is an important point.
27
-------
Dr. Dennis Dauble
RESPONSE: The revised assessment now includes a clearer description of the magnitude
and duration of effects. We apply the runout distance equations of Rico et al. (2008) to
conclude that under Pebble 2.0 scenario dam failure conditions, runout distance exceeds
307 km (190 miles), reaching the marine waters of Bristol Bay (Section 9.3.2).
Dennis D. Dauble- The new description is now sufficient, given the limitations of
modeling.
Question 10. Does the assessment appropriately characterize risks to wildlife and
human cultures due to risks to fish? If not, what suggestions do you have for
improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
• COMMENT: There is considerable detailed information in Appendices D and E relating to
impacts of the project to the economy. This information includes how salmon affect all
segments of the population, such as cultural resources of Native Peoples. However, not
addressed in detail were long-term impacts to Native Peoples that might occur after losing a
way of life that includes salmon. The description of potential impacts to their health and
welfare should be expanded. There are numerous examples of how Columbia River tribes
have been negatively impacted due to loss of fish resources (and fishing as a lifestyle) as a
result of dam construction. These impacts go beyond simple economics.
RESPONSE: EPA recognizes that there is a great deal of information about cultural and
health effects on indigenous populations from loss offish resources. We have referenced
some case studies in the assessment (with a primary focus on Alaska), and have expanded
the discussion of potential effects from a loss of salmon resources in Chapter 12.
Dennis D. Dauble- The case studies are helpful. I agree that Alaska's native peoples are a
unique situation.
• COMMI-NT The report should include a discussion of effects specific to unique user
groups Thai is. some communities rely almost solely on sockeye; kings are more important
to others These impacts could be segregated by watershed, for example. Also, some groups
ha\ e more option for subsistence gathering if sockeye and Chinook salmon resources are
impacted Potential impacts of a declining salmon population due to mining operations
would be less lor them than groups "on the edge" who currently rely mainly on salmon.
RESPONSE: The text of the revised assessment (Section 5.4.2.2) has been expanded to
acknowledge the differences between communities with regard to use and reliance on
salmon and other fish resources.
Dennis D. Dauble- Although text was revised in Chapter 5 to acknowledge differences in
reliance between Chinook and sockeye salmon, these differences were not brought forward
to risk assessment discussions in Chapter 12. Discussion in section 12.2.3 should address
the declining state of economy for some upper watershed cultures and specify which
resident group expressed a desire for jobs and development.
• COMMENT: Disturbance of wildlife from noise and roadways should be included with
respect to migration corridors and critical habitat. Highlight species most likely at risk from
human disturbance, habitat loss/displacement (from the project footprint), and loss of salmon.
For instance, do some piscivorous species have the ability to shift their diet to include
28
-------
Dr. Dennis Dauble
another source of protein? If so, how would this shift affect the human culture with reference
to a subsistence lifestyle?
RESPONSE: EPA recognizes that there are potential direct effects on wildlife from large-
scale mining operations, including disturbances from noise and roadways; however direct
effects on wildlife are outside of the scope of the assessment, as explained in Chapter 2. We
would expect that a full evaluation of any future mining permit applications and
subsequent National Environmental Policy Act Environmental Impact Statements would
consider these direct effects on wildlife.
Dennis D. Dauble- It would be informative (and in my opinion within the scope of the
assessment) to include additional narrative text on wildlife species having the greatest
dependency on salmon as a prey base.
Question 11. Does the assessment appropriately describe the potential for cumulative
risks from multiple mines? If not, what suggestions do you have for improving this part
of the assessment?
• COMMENT: Individual risk is described in \ ai\ inu 1e\ els of detail wi ill o\ erall risk or
effects considered to be largely additive. IIo\\e\ ei". the relati\ e magnitude of the effects of
mining each ore deposit is difficult to discern. It is possible that one of the smaller ore sites
could be developed within an acceptable risk scenario, but it is difficult to determine given
that the assessment is largely built on potential impacts of the Pebble Mine. To put things in
perspective (individually and cumulatively), there should be a discussion of habitat lost given
each individual mine footprint, during normal operation (includes water treatment and
withdrawal) and as a result of pollutant exposure. Also. Section 7.4.1 of the main report
provides estimates of stream miles affected due to blockage and elimination, but provides
nothing quantitative for other direct and indirect impacts of mine operation. The cumulative
risk discussion in Chapter 7 could be expanded to link up with the conceptual model
described in Chapter 3.
RESPONSE: H e cannot reliably predict the habitat loss due to additional mines. We have
tried to describe plausible examples of where additional mines could be developed on the
basis of active exploration on existing claim blocks. In the revised assessment (Chapter 13)
we predicted aquatic resource impacts based on a typical mine footprint being constructed
anywhere in the block and an average stream and wetlands density. We also expanded
discussion of the impacts of ancillary mine infrastructure and induced development. This
provides a conservative estimate of the cumulative effects of multiple mines and sheds light
on whether cumulative effects are a significant concern. We have developed a specific
conceptual model for cumulative impacts and used it to enhance the discussion.
Dennis D. Dauble- The expanded discussion on cumulative impacts and re-organization of
the document largely resolves my questions on this topic. However, it would be informative
to define what "more than minimal" means in Table 13.1, and to include the area or
footprint of each individual site. Note that area of each site are included later in narrative
text, but in different subsections. This information should be included in the Description,
not in Potentially Affected Waters. Finally, I think human development in the Pacific
Northwest is a poor analogy with respect to cascading effects on salmon populations (i. e.,
last paragraph of 13.5 Summary). Either remove this discussion or make a stronger
29
-------
Dr. Dennis Dauble
argument that mining development in Alaska is equivalent to hydro development
-^irrigation +deforestation+ over-fishing in the Columbia Basin.
Question 12. Are there reasonable mitigation measures that would reduce or minimize
the mining risks and impacts beyond those already described in the assessment? What
are those measures and how should they be integrated into the assessment? Realizing
that there are practical issues associated with implementation, what is the likelihood of
success of those measures?
• COMMENT: Potential mitigation measures are well described in Appendix I. I have no
suggestions for additional measures. Implementation of mitigation measures is entirely
dependent on the regulatory framework for operations and the oversight and monitoring
practices that would be mandated as a condition of the mi ni nu activity. Thus, some
discussion of how/which mitigation practices would be most applicable in the Bristol Bay
watershed (and limitations thereof), given constraints and characteristics of local hydrology
and geology, is warranted.
RESPONSE: The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine, are a subset of options
presented in Appendix I, and were presented as appropriate for the Pebble deposit in
Ghaffari et al. (2011). Evaluation of alternative strategies (e.g., other options presented in
Appendix I for the mitigation of the same issue) should be done during a permitting
process for a specific mining plan. Permit applicant mitigation measures that reduce the
risks identified in the assessment would be welcome during the application process.
Dennis D. Dauble- The addition of Ghaffari et al (2011) helps substantiate the discussion
of potential mitigation measures. Further, this topic is well-covered in both Appendices I
and J, as well as Chapter 7 of the main document.
Question 13. Does the assessment identify and evaluate the uncertainties associated
with the identified risks?
• COMMI-NT: The most likely scenarios and probabilities of failure are described based on
assumptions of project size and magnitude. For the most part, estimated risks are
conservative (i e , effects are slated as "likely" if no further information is available). A
weakness of Integrated Risk Characterization (Chapter 8 of the main report) is having a long
list of identified uncertainties, which leads one to speculate, "so what do we know?" Not
being familiar with the formal risk assessment process, it appears this "assessment" (which is
loosely based on a risk assessment framework), falls short of providing something with any
degree of certainty.
RESPONSE: The assessment does not state that effects are likely if no further information
is available. The assessment, as far as the available information allows, identifies potential
events and their effects, their probabilities of occurrence, and possible ranges.
Uncertainties, inherent in any risk assessment or mine plan, are clearly identified in the
assessment.
30
-------
Dr. Dennis Dauble
Dennis D. Dauble- Either I have an improved understanding of risk assessment jargon as
a result of reviewing the document or the revised Assessment does a better job of
explaining the process.
Question 14. Are there any other comments concerning the assessment, which have
not yet been addressed by the charge questions, which panel members would like to
provide?
• COMMENT: Based on public comments and discussions lhal took place by panel members
in Anchorage August 7-9 of this year, this report confuses in both intent and approach. Is the
intent of EPA's assessment to characterize potential impacts lo the Bristol Bay watershed
(title) or does it address a more defined portion of the \ushauak River and Kvichak River
watersheds (objective statement)? Was 1he approach an "assessment" (a fairly broad term) or
an "ecological risk assessment" (suggests a specific scientific framework was applied to the
risk/effects analyses)? These shortcomings should be addressed in the final assessment
document.
RESPONSE: We have revised the discussion of purpose, scope and endpoints in response
to this and other similar comments (see Chapters I through 5 of the revised assessment).
The assessment addresses multiple spatial scales, as detailed in Chapter 2. It is an
ecological risk assessment, but that term is usually shortened to assessment to make the
document more readable.
Dennis D. Dauble- My concerns and questions on the intent and approach of the
assessment were adequately addressed.
3. SPECIFIC OBSERVATIONS IJV DKNNIS D. DAUBLE
1. COMMI-YI' Appendix li, Page fable 1. I suggest adding a column to indicate relative
abundance, for example, if incli\ idnal fish species listed are abundant, common or rare. Also,
are there known differences in distribution and abundance for the Nushagak and Kvichak
watersheds relative to those watersheds unlikely to be affected by mining activities?
RESPONSE: This information, when known, has been added to Appendix B.
Dennis D. Dauble- More important, Table 5.1 includes a list of anadromous and resident
fish species reported in the Sushagak and Kvichak River watersheds and their relatives
abundance.
2. COMMENT: Appendix A, Page 42: The statement that diminished salmon runs present a
"negative feedback loop" where spawner abundance declines, appears to conflict with the
last paragraph on page 41.
RESPONSE: The text describes bottom-up effects of MDN on stream ecosystems and
points out that these linkages have not been empirically established. The negative feedback
loop mentioned in the following paragraph could result from reductions in salmon-based
resources that promote either bottom-up (i. e., a reduction in salmon-derived N and P) or
direct (i. e., a reduction in salmon eggs andflesh that can be consumed by fish) benefits to
juvenile fishes. Further, the fact that bottom-up effects of MDN have not been firmly
31
-------
Dr. Dennis Dauble
established does not negate considerable circumstantial evidence for bottom-up nutritional
deficits in Columbia Basin spawning streams. Edits have been made to help clarify.
Dennis D. Dauble- You present a convoluted argument. I could not find specific edits but
recognize the topic is more complicated than some might imagine. As long as it is
acknowledged somewhere that MDN is most important in systems that would otherwise be
nutrient-deficient and that this input may be more important in lentic systems than lotic
systems, I'm happy.
3. COMMENT: Appendix F, Page 3: Is there significant sediment transport from the Bristol
Bay watershed to the Nushegak and Togiak Bays/csUiaiics°
RESPONSE: A new section was added to Appendix /•' that discusses the importance of
estuary habitat to salmon populations.
Dennis D. Dauble- The response did not address my question. . I Iso, note that the peer
review panel did not receive copies of revised. \ppendices. / did. however, find a reference
in the main document about potential for sediments to be transported to estuarine habitats.
4. COMMENT: Appendix F, Page 7: What is the juvenile salmon resident lime in Bristol Bay?
How quickly (and at what size/time of year) do lliev move from shallow ncaishore to
offshore habitats?
RESPONSE: A new section was added to. Ippendix /•' that discusses the importance of
estuary habitat to salmon populations.
Dennis D. Dauble- Hopefully. this new section addresses my question.
5. COMMI-YI' Page N-l 5 (1.2) Suggest deleting "likely"" There will be impacts.
RESPO NSE: II e believe this language is consistent with the uncertainties explained in the
assessment. \o change lias been made.
Dennis I). Ihiuble- So comment.
32
-------
Dr. Gordon Reeves,
USDA Pacific Northwest Research Station
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Reeves, Gordie -FS rmailto:Qreeves@fs.fed.usl
Sent: Monday, August 05, 2013 1:18 PM
To: Thomas, Jenny
Subject: RE: Bristol Bay Peer Review - Confidentiality Reminder
Jenny: Here is my review.
Gordie Reeves
-------
Dr. Gordon Reeves
Department of Service Northwest
Agriculture Research
Station
Date: 5 August 2013
Subject: Comments on the revised EPA Bristol Bay report
To: To Whom It May Concern
From: Gordie Reeves
EPA did a good job with the revision. All of the primary issues that I raised appear to
have been considered but there were a few that I with which I did not agree with the
response (See below).
Response to EPA's response to G. Reeves' comments on first draft:
p. 7. #1. There is no need to build an IP model from scratch. Rather, EPA could draw
from results of studies that have employed the concept. Generally, a small proportion of
the landscape is responsible for a large proportion of the production. See the attached
draft of Bidlack et al., which is in review, on Chinook salmon in the upper Copper River.
Simply stating the amount of miles of stream/habitat that may be affected may not
accurately represent the impact of mine operations.
#2. The discussion of the potential impact of climate change in the revised
document is good. I have two additional suggestions. First, changes in temperature are
presented for air temperature only. You should attempt to discuss what this means for
water temperatures. Generally, a two degree change in air temperature translates to a one
degree change in water temperature. Or, at least state that there is not a corresponding
one-to-one degree change in air and water temperature so that the reader is aware of the
potential change in water temperature. Second, I think that it would be good to include
Fig. 3 from McCollugh (1999. EPA-910-R-99-010). This figure shows the relation
between changes in water temperature and time of emergence of fry, which is potential
impact of increasing water temperatures from climate change and the altered water
temperatures from mining operations.
p. 10. #1. Highlight more prominently in the document (Ch. 7) the fact that the NHD
layer does not accurately represent smaller streams, which may be important
ecologically, and the consequences of this to the analysis.
p. 22.1 agree that Furniss et al. (1991) is a seminal paper on road and their impacts on
aquatic organisms and ecosystems. However, it deals with forest roads that are much
different in many ways from the roads considered in this assessment. The differences
should be made more explicit. Also, you should look at Lee et al. (1997 USDA Forest
Service Gen. Tech. Report. PNW-GTR-405, vol. Ill pp. 1057-1713) for additional
-------
Dr. Gordon Reeves
potential impacts of roads. Again, this pertains to forest roads and limitation of the
application to Bristol Bay should be acknowledged.
Specific comments on revised report:
p. 3-44-45. See comment p. 7 #2 above. You need to talk more explicitly about the
potential impacts of warmer winters.
p. 3-45. Good point about the importance of genetic diversity.
p. 7-16. You should make the point that the percent of stream affected does not
necessarily accurately reflect the potential impact. See comment on p. 7 #1 above.
p. 7-32. I think that is important not to just list potential mitigation measures but to also
assess the potential for success. I think that you would be hard pressed to find examples
where mitigation has even compensated for a fraction of the impact of given activities.
Particularly present assessments from mining.
p. 10-6. This is a particularly good diagram.
-------
Dr. Charles Slaughter, University of Idaho
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From:
Sent: Tuesday, Duly 30, 2013 4:40 PM
To: Thomas, Denny; White, Dessica
Cc: Frithsen, Deff
Subject: CWS review of 2013 draft Bristol Bay docs
Duly 30, 2012
Denny Thomas and Dessica White
Office of Wetlands, Oceans, and Watersheds US Environmental Protection Agency
1200 Pennsylvania Ave NW
Washington, DC 20460
Thomas.DennyQepa.gov
White.DessicaQepa.gov
Dear Denny and Dessica,
Attached are my comments in response to the Statement of Work regarding review of the
documents: (1) EPA Response to Peer Review Report, External Peer Review of EPAs Draft
Document An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska, and (2)EPA 910-R-12-re004Ba, April 2013, An Assessment of Potential
Mining Impacts on Salmon Ecosystems of Bristol Bay Alaska, Volume 1;, Second External
Review Draft
As you had recommended, my comments are appended to appropriate sections of the first
document.
My "review" of the larger main report was limited to checking specific points,
evaluating degree of incorporation of reviewers' 2012 recommendations into the
revision, and gaining an overall impression of the present state of the draft
assessment. The time allotted did not permit a truly comprehensive, paragraph-by-
paragraph review of that document (nor was such review requested).
Nevertheless, I must compliment the authors for their work. The revised Assessment
is a major (almost "quantum leap) improvement over the 2012 effort, incorporating
much new analysis and discussion. The many conceptual models, flow charts, maps,
figures and tables are very useful.
I specifically applaud (1) the decision to add chapters on the transportation
corridor, cumulative impacts, and integrated risk characterization, and (2) the
integration of uncertainty and risk characterization considerations into the topical
chapters.
I trust that these review comments will prove useful to EPA. Please don't hesitate
to contact me [preferred right now, as I am in SE Washington with only
occasional internet access], or
with any questions or concerns about this submission.
Thanks for the opportunity to again review this important body of work.
Sincerely yours,
Chuck
Charles W. Slaughter, PhD
-------
Dr. Charles Slaughter
SUMMARY OF KEY RECOMMENDATIONS FROM PEER
REVIEWERS
This section summarizes the significant general recommendations put forth by the peer reviewers
regarding EPA's draft assessment. In developing these recommendations, peer reviewers provided
input on three major areas of the assessment: (1) scope, (2) technical content and (3) editorial
suggestions. Reviewers also identified research needs for EPA to consider. Please note that this
summary of peer review comments did not reflect a consensus or group perspective, but was
compiled from a discussion of individual peer reviewer recommendations. Additional details,
including references cited, can be found in the reviewers' individual comments in Section III.
CWS Comments July 2013
Scope of the Document:
• Articulate the purpose of the document more clearly via a primer on the Ecological Risk
Assessment process. If the purpose of the assessment is to inform EPA as the decision maker, then
the level of detail should correspond to this purpose. The authors should justify and explain what
level of detail is required.
RESPONSE: Additional information on both the purpose of the assessment and ecological risk
assessment (ERA) in general has been added to Chapters 1 and 2, as well as the Executive
Summary. Section 1.2 includes information about the use of the assessment. The assessment has
been reorganized into two major sections (problem formulation, risk analysis and
characterization) to clarify where different chapters fall in the typical ERA process.
CWS Comment July 2013: Chapter 1 provides a clear statement of the rationale underlying the
Assessment - why it was undertaken, an overview of how the Assessment was developed, and a
brief statement on how it might be used.
• Include a statement upfront about the role of risk managers and other audiences, such as project
managers/engineers, regulators, mine owners/operators. Knowing their role ensures inclusion of
information necessary for any risk assessment by (1) describing the need for a risk assessment, (2)
listing those decisions influenced, and (3) characterizing what risk managers require from the risk
assessment.
RESPONSE: Section 1.2 of the revised assessment discusses the use of the assessment.
CWS Comment July 2013: Section 1.2 discusses possible use of the assessment, but does not
identify or define "risk managers" per se. Section 1.1 mentions "ecological risk assessment",
"risk assessors " , "risk assessments ", "risk analysis and characterization ", "'potential risks "
and "risks to assessment endpoints " — so who or what are "risk managers " -project
engineers, regulatory authorities, actuaries, industry funding entities, indigenous people living in
the watershed?
-------
Dr. Charles Slaughter
• Explain why the scope for human and wildlife impacts was limited to fish-mediated effects, as well
as why fish-mediated effects on humans were limited to Alaska Native cultures. Reviewing effects
beyond fish-mediated ones (e.g., potential for complete loss of the subsistence way of life) would
improve the assessment.
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. Throughout the assessment we acknowledge that direct effects of
large-scale mining on wildlife and Alaska Native cultures may be significant, but that these direct
effects are outside the scope of the current assessment.
CWS Comment July 2013: Despite the primary focus of the Assessment on salmonids, the
authors have done a creditable job of identifying probable (actually, inevitable and irreversible)
consequences for Alaska Native culture of a PLP-style mineral extraction project in the Bristol
Bay watershed. Similarly, consequences for wildlife populations are identified. Both topics
received careful attention in appendices to the 2012 draft of the Assessment; incorporation of
materials from those appendices into appropriate sections of this revised Assessment is
commended.
• Be more consistent throughout the document in terms of the level of detail provided for the
different scenarios and stressors. For example, the document has devoted 36 pages to the discussion
of catastrophic Tailings Storage Facility (TSF) failure, while sections on the pipeline, water
treatment, and road/culvert failures are brief. Indeed, the long discussion on the TSF failure belies a
certainty and understanding of dam failure dynamics that is inaccurate.
RESPONSE: The final document includes more failure scenarios (e.g., diesel pipeline failure,
wastewater treatment plant failure, and refined seepage scenarios) in Chapter 8. It also explains
why these specific failure scenarios were chosen, and discusses these scenarios in greater detail
than the previous draft (i.e., to more closely match the level of detail originally provided only for
the TSF failure scenario). Also see detailed responses to comments on Peer Review Question 5.
CWS Comment July 2013: The more detailed and expanded treatment in Chapters 7, 8, 9, 10
and 11 of the revised Assessment is greatly appreciated. The new Assessment gives more
balanced and appropriate consideration to potential failure scenarios which are in many ways at
least in equal importance with potential TSF failure.
Technical Content:
Mine Scenario
• Consider the document to be a screening-level assessment of all potential stressors. Focusing on
failure mode overemphasizes catastrophic events (e.g., TSF failing), rather than considering all
potential stressors, such as holding mine owners strictly accountable for their day-to-day activities
with regard to best practices.
RESPONSE: Additional information on the purpose and scope of the assessment has been added
to Chapters 1 and 2. A screening of all potential stressors, including individual chemicals, is
-------
Dr. Charles Slaughter
presented in Section 6.4.2. Also see detailed responses to Peer Review Question 2 on the use of
"best practices" and responses to Peer Review Question 5 on failure scenarios.
CWS Comment July 2013: The greater detail and much more comprehensive in-depth
analysis in Chapter 6 and in the subsequent topic-specific chapters greatly enhance the utility
and credibility of the Assessment.
• Reexamine the document's use of historical data and case studies to describe and estimate the risk
of failure for certain mine facilities (including the TSF, pipeline, water treatment, etc.), as these
examples from extant mines may not be an appropriate analog for a new mine in the Bristol Bay
watershed.
RESPONSE: The TSFfailure range was, and still is, based on design goals, not the historical
data. The historical TSFfailure data are provided as background. The pipeline failure rates are
based on the most relevant historical data from the petroleum industry. They are directly relevant
to the dieselpipeline, and experiences at the Alumbrera mine (described in the previous draft) and
the Antamina and Bingham Canyon mines (added to this draft) suggest that they also are relevant
to the product concentrate pipeline. Water treatment failure rates were not quantified. However,
recent review's cited in the revised draft indicate that water collection and treatment failures have
been reported at nearly all analogous mines in the U.S. The estimation of culvert failure
frequencies has been revised and is now based on only recent literature (2002 and later). We
believe that these estimates are appropriate.
CWS Comment July 2013: RE pipeline failures, the revised Assessment does reference both
the 2012 Ft. Knox AK incident, and the 2010 Klamazoo MI (Enbridge Energy) pipeline failure
(p. 11-6). The Alumbrera, Antamina, and Bingham Canyon incidents discussed on pp. 11-14 to
11-16 are instructive, as they are considered modern, current-practice mines. The detailed
analyses of hypothetical Knutson and Chinkelyes creeks spills are well thought out and helpful.
The analysis ofpossible waste-water treatment facility failure in Chapter 8 is useful, as is the
discussion of the Fraser River as an (inappropriate) analogue for Bristol Bay waterways. The
advert discussions in Chapter 10 are appropriately expanded in comparison with the 2012 draft
Assessment.
• Expand the discussion on the use of "best" management practices, as the document states that the
mine scenario employs "good," but not necessarily "best" practice. For a mine developed in the
Bristol Bay watershed, only "best" practice likely would be appropriate and anything less may not be
permitted. Even so, without a track record of "best" practice (e.g., new technologies), we cannot
assume that technology, by itself without appropriate operational management controls, can always
mitigate risk.
RESPONSE: The term "best management practices" is a term generally applied to specific
measures for managing non-point source runofffrom storm water (40 CFR Part 130.2(m)).
Measures for minimizing and controlling sources ofpollution in other situations often are
referred to as best practices, state of the practice, good practice, conventional, or simply mitigation
measures. We assume that these types of measures would be applied throughout a mine as it is
constructed, operated, closed, and post-closure, and have used the term "conventional modern "
throughout the assessment to refer to these measures. To remove any ambiguity related to the
-------
Dr. Charles Slaughter
subjectiveness of terms "good" or "best", we have removed them in the revision and have provided
definitions for relevant terms used in Box 4-1.
CWS Comment July 2013: OK
• Adopt a broader range of mine scenarios (not only minimum and maximum) so as to bound
potential impacts, especially at smaller mine sizes (e.g., 50th percentile). Underground mine
development, with its different impacts, also should be considered and included in the assessment.
RESPONSE: A third mine size scenario (250 million tons) has been added to the assessment, to
represent the worldwide median sized porphyry copper mine (based on Singer et al. 2008).
CWS Comment July 2013: OK
• Based on the hypothetical mine scenario, perpetual management of the geotechnical integrity of the
waste rock and tailings storage facilities, as well as perpetual water treatment and monitoring, will
most likely be necessary (i.e., a "walk away" closure scenario after mining ends may not be
possible). Therefore, emphasize how monitoring and management of the geotechnical integrity of
waste rocks and tailing storage facilities should continue "In Perpetuity" (i.e., for at least tens of
thousands of years). Discuss what conditions would need to be met to allow "walk away" closure in
the Bristol Bay environment gaining insight into these observations from mines where perpetual
treatment and monitoring are ongoing (e.g., the Equity Silver Mine in British Columbia).
RESPONSE: The conditions for closure and the potential need for perpetual site management are
discussed in general terms in the revised assessment. The primary condition assumed to be
required is water chemistry that meets all criteria and permit conditions and that is stable or
improving. However, even though there are some facilities with "perpetual treatment" conditions
in place, there is obviously no information about how these facilities perform over very long
periods of time.
CfVS Comment July 2013: The revised Assessment repeatedly acknowledges that there is no
extant precedent for "'perpetual" monitoring, facility maintenance, and treatment if required to
maintain either ecosystem function or simply maintain acceptable water quality.
• Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting, overflight, etc.) were
managed. This includes roads, airstrips, helipads, camps, fuel dumps, and ATV trails that have
already been developed or imposed on the watershed, and what "mitigation" already has been
undertaken on those sites. Assess the consequences/impacts of these activities in the Cumulative
Risks section.
RESPONSE: The effects of exploratory activities are outside of the scope of this assessment.
CWS Comment July 2013: This statement simply acknowledges that this Assessment is not
truly comprehensive. Exploratory activities (e.g., drillholes, blasting, helicopter andfixed-wing
overflights, etc.) , including "pioneer" roads or airstrips, helipads, camps, fuel dumps, and ATV
trails that have already been developed or imposed on the watershed, are an inevitable precursor to
-------
Dr. Charles Slaughter
project implementation and "pave the way "for the project(s). Such activities are thus an integral
component of a project such as the potential Pebble mining development, and clearly are directly
relevant to this Assessment. I recall hearing reference to such ongoing activities in public testimony
during the 2012 Anchorage meeting, and anecdotal evidence is available from many sources, such as
Carter (2012).
Risks to Salmonid Fish
• Place potential mining impacts in the context of the entire Bristol Bay watershed by emphasizing
the relative magnitude of impacts. For example, of the total salmon habitat, assess the proportion lost
due to mining. Further, reflect on the non-linear nature of the relationship between habitat and
salmon production; 5% of the habitat could be critical and thus responsible for 20% or more of
salmon recruitment. Intrinsic potential, which measures the ability of particular habitats to support
fishes, would lend credibility to this analysis.
RESPONSE: We are unable to build a complete Intrinsic Potential (IP) model, as this would
require validation and more elaborate construction of metrics appropriate to this region. Our
preliminary characterization provides the building blocks for assessing the distribution of key
habitat-forming and constraining features across these watersheds. We now include a
characterization of the major drivers of habitat potential across the watershed and place the mine-
site specific effects in this context (Chapters 3, 7, and 10).
CWS Comment July 2013: This is an appropriate response.
• Include a section on the impact of Global Climate Change with explicit reference to a monitoring
program that will allow scientists, if the mine is built, to distinguish between effects of climate
change and mining effects on the physical and biological components of this ecosystem.
RESPONSE: Climate change projections and potential impacts are now included in Chapter 3,
and as important external factors in the risk analyses presented in Chapters 7, 9,10, and 14.
Development of a monitoring program to distinguish between mining and climate change effects is
outside of the scope of the assessment.
CWS Comment July 2013: Incorporation of potential climate change consequences into the
body of the Assessments is appropriate, and the discussions are well-considered and reasonable.
• Explicitly recognize that the transportation corridor and all associated ancillary development,
including future resource developments made possible by the initial mining project, will necessarily
and inevitably have impacts (hydrologic, noise, dust, emissions, etc.). These impacts will vary in
duration, intensity, severity, relative importance, spatial dispersion, and inevitably expand
geographically through time with further "development." These impacts should be incorporated into
the Cumulative Risks section.
RESPONSE: The cumulative risk section (Chapter 13) has been expanded to include the multiple
transportation corridors, ancillary mining development and secondary development associated
with multiple mines in a qualitative discussion. The issues addressed in the assessment of the
-------
Dr. Charles Slaughter
transportation corridor (Chapter 10) have also been expanded to include chemical spills, dust,
invasive species, and road treatment salts.
CWS Comment July 2013: The addition of Chapter 10 is greatly appreciated. Chapter 13 is
largely successful in attempting a comprehensive overview of cumulative effects; that chapter
will hopefully not be treated as an "afterword" but rather will be viewed as equally important as
the topic-focused chapters that precede it. Induced and cumulative consequences will be
inevitable upon construction of the initial mine project and transportation corridor, and will be
irreversible for the Bristol Bay watershed, ecosystem, and human population (Alaska Native and
others), let alone effects on salmonids.
• Incorporate current research findings into stream crossing and culvert-design practices (e.g., arch
culverts, bridges, etc.).
RESPONSE: We describe current culvert design practices in a box titled "Culvert Mitigation" in
Chapter 10.
CWS Comment July 2013: OK
• Recognize in the assessment that risk and impact are not equivalent. Risk may be low, but the
potential impact could be huge (e.g., in the case of a TSF failure).
RESPONSE: Risk has been defined in many ways, even by risk assessors. The commenter seems
to define risk as probability. To avoid that potential source of confusion, we use the term
"probability" for that concept. Similarly, the commenter seems to use "impact" where we use
"effect" or "magnitude of effect". We use "risk" to refer to both concepts combined—that is, an
event or effect and its probability).
CWS Comment July 2013: To repeat my earlier comment: Section 1.2 discusses possible
use of the assessment, but does not identify or define "risk managers" per se. Section 1.1
mentions "ecological risk assessment", "risk assessors " , "risk assessments ", "risk analysis
and characterization ", "'potential risks " and "risks to assessment endpoints " — so who or
what are "riskmanagers" - projectengineers, regulatory authorities, actuaries, industry
funding entities, indigenous people living in the watershed?
• Recognize and justify chronic behavioral endpoints, such as those potentially affecting survival and
long-term success of fish populations.
RESPONSE: The chronic behavioral effects of copper on salmonids, the primary endpoint of
concern, were described in Chapter 5 and are now described in Chapter 8. Although those effects
occur at lower levels of copper than conventional survival, growth and reproduction endpoints for
salmonids, they are less sensitive than the conventional endpoints for aquatic invertebrates.
CWS Comment July 2013: The detailed discussion of leachate toxicity and exposure-response
for aquatic organisms, including salmonids, in Chapter 8 was very helpful. The careful
discussion of uncertainties in Chapter 8 (and throughout the revised Assessment) is also
appreciated.
-------
Dr. Charles Slaughter
Wildlife
• Recognize that the draft assessment did not account for all levels of ecology, such as the individual
(e.g., a bald eagle nest), population, community, ecosystem, and landscape levels. Fold other levels of
organization into the stressors assessment where appropriate or justify a more limited approach.
RESPONSE: As is appropriate for an ecological risk assessment (as opposed to an environmental
impact assessment), this assessment focuses on a specific, limited set of endpoints as defined in
Chapter 5. We have added text in Chapters 2 and 5 to explain both why these endpoints were
selected, and that responses other than those considered in the assessment, at multiple levels of
ecological organization, are likely but are outside the scope of the assessment.
CWS Comment July 2013: The authors do recognize that the full ecological web of Bristol Bay
and its watershed is the ultimate "receptor " of whatever activities take place. I accept that a more
comprehensive analysis at all levels is "outside the scope of the assessment. "
• Discuss in the document fishes other than salmonids The assessment focuses on risks to sockeye
salmon in the Bristol Bay watershed (and also considers anadromous salmonids, rainbow trout, and
Dolly Varden), but does not account for potential impacts to other members of the resident fish
community. Further, primary and secondary production, including nutrient flux was not addressed.
Expanding the assessment to consider other levels of organization, including direct as well as indirect
effects on wildlife and other fish, would provide additional context in the assessment of mine-related
impacts.
RESPONSE: See response to comment above; we also incorporated additional information from
Appendices A, B, and C into the Chapter 5 text, to provide additional detail on the area's biota. We
chose our endpoints for reasons described in Chapters 2 and 5. Other endpoints, including indirect
effects on fish and wildlife, are now discussed more explicitly, but are generally considered outside
the scope of the assessment.
CWS Comment July 2013: OK — The authors do recognize that the full ecological web of Bristol
Bay and its watershed is the ultimate "receptor " of whatever activities take place. Other fish are
mentioned in the draft Assessment. I accept that a more comprehensive analysis at all levels is
"outside the scope of the assessment. "
Human Cultures
• Use case histories to provide insight and anticipate mining impacts on Alaska Natives (e.g., those
exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil and gas development in the
North Slope region, and social impacts related to mining development in Alaska).
RESPONSE: Examples from applicable case studies, including the Exxon Valdez oil spill, are cited
in Chapter 12 of the revised assessment.
-------
Dr. Charles Slaughter
CWS Comment July 2013: Section 12.2, while relatively short in relation to the importance of
the issue, does provide an overview of consequences for Alaska Natives, and acknow ledges that
those consequences would be long-lasting and effectively irreversible.
~ As noted above (Scope of the Document), clarify why the scope was limited to fish-mediated
effects. The potential direct and indirect impacts for human cultures extend far beyond fish-mediated
impacts (e.g., potential complete loss of the subsistence way of life). The rationale for this narrow
focus should be fully explained. In addition, a clear explanation should be given for why fish-
mediated human impacts focused only on Alaska Native cultures.
RESPONSE: The assessment focuses on a specific, limited set of enclpoints as defined in Chapter
5. We have added text to explain both why these endpoints were selected, and that responses other
than those considered in the assessment are likely but are outside the scope of the assessment. The
assessment was expanded (Chapters 5 and 12) to acknowledge that there are a wide range of
potential direct and indirect impacts to indigenous culture, but they are outside of the scope of this
assessment. The discussion of potential effects to indigenous cultures was expanded to explain that
a loss of subsistence resources would extend beyond a loss offood resources to social, cultural,
and spiritual disruption. The text has been expanded to acknowledge the strong cultural ties of
many non-Alaska Natives to the region, and potential effects on all residents from loss of a
subsistence way of life. However, the focus of the assessment remains on effects on indigenous
cultures resulting from effects on salmon.
CWS Comment July 2013: OK Chapters 12 and 13 do address the broader direct and
indirect effects on Bristol Bay watershed residents.
Water Balance/Hydrology
• Better characterize water resources and assess the potential effect of mine developmvcent on these
resources by (1) generating a diagram similar to the conceptual models beginning on page 3-7 to
illustrate the potential effects of mine construction and operation on surface- and ground-water
hydrology; (2) developing a quantitative water balance and identifying water gains and losses; (3)
identifying seasonality of hydrologic processes, including frozen soils and their associated values
(e.g., mm/yr) for each component of the water balance; (4) incorporating these processes into a
landscape characterization; (5) evaluating how global climate change will influence these hydrologic
processes and rates; and 6) using this characterization to demonstrate the expected hydrologic
modification associated with the mine scenarios and infrastructure development.
RESPONSE: The original Figure 4-9 (new Figure 6-5) has been revised to more clearly show
water management in the assessment's mine scenarios. In addition, three schematics illustrating
water flows under each of the mine size scenarios (Figures 6-8 through 6-10) have been added to
Chapter 6, as have quantitative water balances for each mine size scenarios. A qualitative
discussion of climate change is included in Chapters 3 (Section 3.8) and 14 (Box 14-2).
CWS Comment July 2013: The much more comprehensive hydrologic analyses of the revised
Assessment are welcome. These allow more detailed consideration of the hydrologic aspects of
-------
Dr. Charles Slaughter
the various mine scenarios presented. While there can still remain questions (not detailed in this
review) about the many necessary assumptions involved in those analyses, they do appear
realistic and point out the significant consequences of alternative scenarios.
• Demonstrate the interconnectedness of groundwater, surface water, hyporheic zone, and its
importance to fish habitat. Address how interconnectedness changes over time - seasonally, and with
varying weather (e.g., wet vs. dry summers or years, and over the long term as climate changes).
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
(Chapter 7) and qualitatively discuss the potential role of climate change (Chapter 3).
CWS Comment July 2013: Given the lack of site-specific hard data, the revised Assessment
does fairly acknowledge the importance of surface water groundwater interactions.
• Provide information on all rivers, including ephemeral and intermittent streams, and first-order to
main-stem streams that could be potentially influenced by the proposed mine, its ancillary facilities,
and the transportation corridor.
RESPONSE: Due to lack of consistent coverage, we rely on the NHD hydrography layer in this
analysis, and can only address ephemeral and intermittent streams qualitatively (Chapter 7).
CWS Comment July 2013: The NHD data are necessarily broad-brush; I recognize that on-
the-ground work would be required to adequately understandfiner-scale hydrologic issues
including ephemeral and intermittent streamflow. Nevertheless, such "small" features can be
significant in streamflow generation, in transport of nutrients or contaminants, and in immediate
sensitivity to perturbation, be that human-caused, seasonal weather, or climate change. Simply
glancing at the streams and landscapes shown in figures 3-8 and 3-11 allows one to visualize the
finer-grain flowpaths which cannot be discerned at the scale of those photos or the
accompanying maps.
• Emphasize the importance of a thorough characterization of the leaching potential of acid-
generating and non-acid generating waste rock and tailings, given the low buffering capacity and
mineral content in the streams and wetlands that could receive runoff and treated water from the
proposed mine. Recognize that collection and treatment of runoff and leachate generated will be
critical to maintain baseline water chemistry in these streams and wetlands.
RESPONSE: We agree that these are important issues, and the discussion of leachate from waste
rocks and tailings has been expanded in the revised assessment (Chapter 8).
CWS Comment July 2013: Leachates are extensively considered in the revised Assessment.
Geochemistry/Metals
• Reference the most current geochemistry data on potentially acid-generating, non-acid generating, and
-------
Dr. Charles Slaughter
metal leaching so as to describe any potential effects of seepage and changes to surface- and ground-water
quality via non-catastrophic failure.
RESPONSE: We used the geochemistry data in PLP's Environmental Baseline Document, as summarized
by the USGS in Appendix H. The effects of seepage on water quality are analyzed in Chapter 8 of the
revised assessment.
CWS Comment July 2013: OK
• Explain how contaminants/metals were selected (and others ignored) by EPA as causes for concern.
Information should be included on additional metals and their toxicity so as to assess impacts of potential
leachates. The Pebble Limited Partnership baseline document presented additional metals that might be
useful to include in the assessment.
RESPONSE: The revised assessment describes the selection of contaminants and other stressors of
concern in Section 6.4.2. Additional metals, process chemicals and dissolved solids are now included.
CWS Comment July 2013: To me (lacking expertise in this field) the consideration of contaminants
and stressors in the revised Assessment appears comprehensive and competent.
Mitigation Measures
• Incorporate the critical mitigation information from Appendix I into the main report's mine
scenarios. Include standard mitigation measures that could provide insight into how well they might
work in this context. If this information is not included in the main report, then justify its absence.
RESPONSE: Mitigation measures incorporated into design and operation to minimize potential
impacts were included in the assessment, as were some reclamation measures for closure; these
measures are made clearer in the revised assessment. These mitigation measures were a sub-set of
those presented in Appendix I. The assessment assumes that measures chosen for the scenarios
would be effective. Mitigation to compensate for effects on aquatic resources that cannot be
avoided or minimized by mine design and operation would be addressed through a regulatory
process that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the revised
assessment.
CWS Comment July 2013: The mitigation measures discussed in Appendix I are, as stated,
"conventional'. " Despite the statement above that "The assessment assumes that measures
chosen for the scenarios would be effective ", after again reviewing Appendix I it seems to me
that there is large uncertainty that all those measures would be effective in actual practice in the
Bristol Bay watershed environment. Many of the measures discussed in Appendix I are
generalities - i.e Sections 1.1, 3.1, 5.0,, 7.1, 8.1, 9.1 — with questionable applicability to a
potential Pebble project.
• Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in the main report,
as they ultimately influence the range of mining impacts and consider time frames of mitigation or
reclamation measures (e.g., immediate response, long-term reclamation).
-------
Dr. Charles Slaughter
RESPONSE: See response to previous comment. Mitigation measures are discussed at greater
length in the revised assessment report (e.g., Chapter 4 and Appendix J).
CWS Comment July 2013: See previous comment. Appendix J (to my reading) strongly
indicates that compensatory mitigation is essentially inapplicable to the Bristol Bay watershed.
Uncertainties and Limitations
• Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of uncertainty and (2)
assessing the certainty of some mine impacts. Discuss data limitations in the context of uncertainty.
RESPONSE: The individual analysis chapters and the revised Integrated Risk Characterization
(Chapter 14) discuss certainties and data limitations to a greater extent, as suggested.
CWS Comment July 2013: Throughout the revised Assessment, the discussions of "risk
characterization " and "uncertainty " are well-thought-out and useful - for example, Sections
8.2.5, 8.3.3, 8.3.4, 10.5.
• Articulate early in the document how much uncertainty is acceptable. The assessment provides little
insight with respect to the decisions the document is intended to support.
RESPONSE: Acceptable levels of uncertainty can be defined prior to an assessment if a decision
and a decision maker are identified and if data will be collected by a specified design to implement
a specified model, as described in the EPA '.v Data Quality Objectives process, However, because
this assessment is based on available data and is intended as a background scientific document
rather than a decision document, it is not possible to specify the amount of uncertainty that is
acceptable. Rather, the available data determine the uncertainty and if the assessment is
subsequently used to inform a decision, the decision milker must determine whether the level of
uncertainty is acceptable.
CfVS Comment July 2013: Concur with the above response; the various discussions of
uncertainty in the body of the revised Assessment provide a "feel" for the unknown or
unpredictable, without attempting to quantify specific boundaries or levels.
Editorial Suggestions:
• The title of the document leads one to believe that the assessment addresses the entire Bristol Bay
watershed; rather, the report deals with two major rivers and their watersheds, the Nushagak and
Kvichak. Thus, the title should be changed to reflect the emphasis on these two rivers and their
watersheds. A possible title may be "An Examination (or identification) of the Potential Impacts of
Mining and Mining Associated Activities on Salmon Ecosystems in the Nushagak River and Kvichak
River watersheds, Bristol Bay."
-------
Dr. Charles Slaughter
RESPONSE: The assessment addresses multiple scales: the Bristol Bay watershed, the Nushagak
and Kvichak River watersheds, the watersheds of the three streams draining the Pebble deposit,
and the watersheds crossed by the transportation corridor. These multiple scales, and how they are
used throughout the assessment, are described more clearly in the revision (Chapter 2).
CWS Comment July 2013: OK
• Revise the Executive Summary to more precisely reflect the findings in the document.
RESPONSE: The Executive Summary has been rewritten to reflect the revised assessment
findings.
CWS Comment July 2013: The Executive Summary of the revised Assessment is provides a
comprehensive overview. (I still question the discussion of probabilities in relation to TSF dam
failure (p. ES-21), as noted in my other comments.)
• The appendices contain detailed and useful information that should be summarized and included in
the main document (e.g., Appendix E: Economics, Appendix G: Road and Pipelines, and Appendix I:
Mitigation). Additionally, consider expanding the preface to include information on the use of the
appendices. If the information is not included in the main report, then justify its absence.
RESPONSE: More information from the appendices was brought forward into appropriate
chapters of the revised report. The purpose of the appendices—to provide the detailed background
characterization necessary for the ecological risk assessment—has also been clarified in Chapter
2. The document no longer contains a preface because that material has been incorporated into
Chapters 1 and 2.
CfVS Comment July 2013: I support the decision to incorporate much of the previous
Appendix information into the body of the Assessment (as recommended in 2012).
• Discuss in more detail the instructive and well-thought-out conceptual models (pages 3-7 to 3-11)
illustrating the impacts of mining on Bristol Bay ecosystem processes. Also, consider expanding the
conceptual models to include wildlife, fish-wildlife interactions, vegetation/terrestrial habitat, and
hydrologic processes. Allow them to guide the text because they appear detailed and complete.
RESPONSE: Additional information on the use of conceptual models throughout the assessment
has been incorporated into Chapter 2. The more comprehensive conceptual models presented in
Chapter 6 (Chapter 3 in the first draft) have been broken into their relevant component parts
throughout the risk analysis and characterization chapters, to better frame the specific pathways
addressed in each chapter. Additional conceptual models considering impacts on wildlife, Alaska
Native populations, and cumulative effects of multiple mines have been added to Chapters 12 and
13.
CWS Comment July 2013: The addition of conceptual models andflow charts throughout the
revised Assessment is really appreciated - they add clarity to the issues.
-------
Dr. Charles Slaughter
• Incorporate the information contained in the conceptual models into a formal framework, such as a
Bayesian or other decision-analysis models.
RESPONSE: This is an excellent suggestion for future efforts, but is beyond the scope of the
current assessment.
Creating a Bayesian Belief Network would require that the Agency convene experts to subjectively
estimate the probabilities of each transition in the conceptual models. In contrast, this assessment
is intended to elucidate the risks from potential mining based on available data and analyses of
those data.
A Decision Analysis would require that alternative outcomes be specified, the utility of each
outcome for a decision milker be defined and the probabilities of each outcome be estimated for
each possible decision so that the expected utilities of each outcome can be calculated. Because
this assessment is not a decision document, these requirements are not feasible or appropriate.
CWS Comment July 2013: OK
• Generate a standard operating protocol for significant figures and use it throughout the document.
RESPONSE: The authors have carefully addressed this issue. Numbers from the literature or
from the PLPEBD retain the number of significant figures in the original. Numbers derived for
this assessment have the appropriate number of significant figures given the precision of the input
data and uncertainties due to modeling and extrapolation.
CWS Comment July 2013: OK
• Remove all references to Mount St. Helens as a surrogate for a TSF failure. Using a non-human-
caused release of material into the ecosystem as an analogue for a mine failure is not comparable in
terms of likelihood or risk for a human-caused release. It would be more appropriate to extrapolate
from the impacts of known mine failures.
RESPONSE: We are puzzled by this comment. The Mount Saint Helens data were used strictly to
address the rate of benthic habitat recovery from a massive deposition offine mineral particles.
The hydrological processes that determine the recovery of substrate texture and the requirements
offish or aquatic invertebrates are not known to depend on whether mineral particles were from a
natural event or an anthropogenic event. We have reviewed the literature on known mine failures.
They studied tailings spills in terms of toxicity but not in terms ofphysical habitat effects, which is
why we used Mount Saint Helens data. Neverpotheless, we have removed references to Mount St.
Helens in the revised assessment to eliminate concern.
CWS Comment July 2013: I was not enthusiastic about removal of all reference to Mount St.
Helens, simply because that event does provide a case history of massive sediment loading and
long-term stream and landscape response. However, I think I understand the concern that heavy
reference to Mount St. Helens might imply that the volcanic eruption was analogous to a TSF
dam failure, though the causes would be unrelated. I would be comfortable with again utilizing
the sediment immobilization, channel and habitat "recovery " information from Mount St. Helens
research in generalizing about consequences of a major TSF dam failure.
-------
Dr. Charles Slaughter
• Ensure that the draft assessment remains part of the public record, allowing the document history to
remain intact.
RESPONSE: All drafts of the watershed assessment will remain part of the public record.
CWS Comment July 2013: OK
Research Needs:
• What are the acute and chronic impacts of mixtures of contaminants, including metals, acid mine
drainage, etc., on the fauna and flora of the Nushagak River and Kvichak River watersheds? What
species are most sensitive and might surrogate species exist for those for which we do not have data?
Review the European literature and regulatory requirements for additional data.
RESPONSE: The acute and chronic impacts of contaminant mixtures, including metals and acid
mine drainage (i.e., metals in lowpH-waters) were addressed using concentration additivity
models in the leachate chemistry tables in Chapters 5 and 6 (now Chapters 8 and 11). Additional
toxicity data were obtained by searches of the EU and OECD database eChent, the EPA '.v
ECOTOX and the Environment Canada site. More metals are now included. In general, metals
are most toxic to aquatic arthropods rather than fish, as discussed for copper.
CWS Comment July 2013: This seems to be well covered in the revised Assessment.
• Can an inventory of nutrients, total organic carbon, and dissolved organic carbon inputs to aquatic
environments be developed that demonstrates their relative magnitude and spatial variation from
headwaters to Bristol Bay? What is the relative importance of marine-derived nutrients relative to
other nutrients from watershed and terrestrial sources? What is the current atmospheric input of
nutrients?
RESPONSE: These data would be very useful in the risk assessment, but are not currently
available for the Bristol Bay region. We agree this is a research need.
CWS Comment July 2013: OK
• What are the locations of subsistence areas and can these areas be characterized and differentiated
by collecting local environmental and ecological knowledge (e.g., fish overwintering areas, climate
change, ecological shifts, etc.)?
RESPONSE: The revised assessment incorporated current data on subsistence use areas available
fromADF&G. EPA acknowledges that these data are incomplete and would encourage additional
collection of subsistence data and Traditional Ecological Knowledge.
CWS Comment July 2013: That is probably all that you can do at this stage. Figure 5-12
attempts to show generalized subsistence use.
• What impact might mining have on other important wildlife species in the basin (e.g., freshwater
seals in Iliamna Lake)?
-------
Dr. Charles Slaughter
RESPONSE: The scope of the assessment is focused on potential risks to salmon from large-scale
mining and salmon-mediated effects to indigenous culture and wildlife. Direct effects on wildlife
from large-scale mining are likely to be important and Appendix C (now a stand-alone US Fish
and Wildlife report) provides useful information for a future evaluation of direct effects on wildlife
from large-scale mining. We agree that this is an important area for future research.
CWS Comment July 2013: OK
• What is the comprehensive hydrologic regime of the specific project mining area, and the broader
watershed system as characterized by baseline monitoring, spatial distribution, and quantitative flow
of surface- and ground-waters?
RESPONSE: Comprehensive spatial estimates of mean annual flow are now presented in Chapter
3. Quantification of spatial and temporal patterns of groundwater flows is an acknowledged highly
desirable product, but it not feasible within the scope of this assessment. Results of an independent
groundwater-surface water modeling effort are described in Chapter 7.
CWS Comment July 2013: The revised hydrologic analyses of the Assessment are explained
well and are very useful. Admittedly, much site-specific data would be required to thoroughly
understand the complete surface water groundwater system, in its current state and potentially
during and after mine implementation.
• What is the cumulative impact of commercial fisheries on the Bristol Bay watershed, especially in
an ecosystem context as related to marine-derived nutrient and energy flow? Acknowledge that
commercial fishing has had an impact on the amount of marine-derived nutrients returned to the
watersheds.
RESPONSE: The impact of commercial fisheries on the watershed is not within the scope of this
assessment. Information on commercial fisheries management has been added in Box 5-2.
However, the purpose of this assessment is not to assess the relative effects ofpotential mining and
commercial fishing—it is to evaluate potential effects on endpoints if a mine were to be developed,
given existing conditions and activities in the region.
CWS Comment July 2013: I note that commercial and subsistence harvest are both
considered in Chapter 5. The Assessment thus does reference commercial harvest, even if not
evaluating the long-term consequences of that harvest, with or without mining activity in the
Bristol Bay watershed.
-------
Dr. Charles Slaughter
CWS RESPONSE TO:
DRAFT
EPA Response to Peer Review Report
External Peer Review of EPA's Draft Document
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
Executive Summary - 2013 Revision
CWS Comment July 2013:
p. ES-4, last para: The clarification that " This assessment does not consider all impacts
associated with future large-scale mining in the Bristol Bay watershed" is appreciated. A deep-
water port, electrical generation capabilities (or electricity import via now non-existent
transmission lines from a potential new hydroelectric facility being considered on the Susitna
River), one (or more) airports and associated infrastructure, extended road and ORV networks,
and the gamut "induced development " and support services associated with increased activity
and worker populations (and, later, sport and recreational visitations), will all have a
cumulative and enduring impact on the Bristol Bay watershed and environs and its present
human, wildlife, andfisheries systems, which temporally and spatially will extendfar beyond the
posited PLP operation itself
Thus, the question is left hanging - if this assessment does not or cannot address the full suite of
probable consequences of the PLP project, what agency or entity does have that responsibility?
Maps - the ES maps are clear and are appreciated.
Figure ES-3 - at least on my print (and computer screen) the subtle shadings of blue or green
associated with regions re difficult to distinguish - but the message of the figure is clear that
Bristol Bay is the major sockeye salmon producer, and the Nushagak and Kvichak rivers
dominate that production.
Table ES-4 — (1) The attempted clarification of your use ofprobability analysis for TSF
failure, provided in the main body of the assessment, is appreciated. However, this reviewer still
questions the message sent, both in the Executive Summary and in the larger report, that a TSF
failure has a "recurrence frequency of2500 to 250,000 years. " This strongly implies to a
reader of the ES (and most decision-makers will read or skim only the Executive Summary) that
with 2500 years as the shorter time span stated, no one should worry about a possible TSF
-------
Dr. Charles Slaughter
failure, despite the report's documentation of actual failure incidents elsewhere. I suggest
eliminating reference to "recurrence interval " in this discussion. (Ipersonally have seen a
100-year recurrence interval (based on period of record) flood occur twice in one year in one
watershed, belying any suggestion that recurrence interval is related to prediction of event
occurrence.)
Table ES-4 - (2) Pipeline spill possibilities - based on the "probability " column, if there is a
95% probability of a product concentrate pipeline failure over 25 years, and the same applies to
the return water and diesel pipelines, then there is 95% probability of three pipeline failures
over 25 years, or nine in the 78-year Pebble 6.5 scenario.
p. ES-26—bottom para — The statement that "Although remediation would be considered if
spills contaminated streams, features of the Pebble deposit area would make remediation
difficult" suggests that remediation would be only considered but not required, because
remediation would be infeasible - so what then? The difficulty is acknowledged, but no solution
is suggested - such stream-contaminating spills would simply be accepted as a cost of doing
business, or collateral damage?
The final sentence suggests that unspecified "compensatory mitigation " might be considered, -
but the analysis provided in Appendix J suggests that compensatory mitigation is simply not an
alternative in the Bristol Bay watershed system. Suggest revising or eliminating this sentence.
Finally, I must compliment the authors on the overall comprehensive coverage of the Executive
Summary. As mentioned above, many users of this document will only go to the ES. My own
preference might be for even more emphasis in the ES on the long-term unavoidable cumulative
effects, but I concede that the text does a goodjob of summarizing those consequences in
individual sections.
1. General impressions
CWS comment
The Assessment (Volume 1 - Main Report) provides a fairly comprehensive review of fisheries-
driven issues, from the perspective of salmonids. Appendices (Volumes 2 and 3) are very
-------
Dr. Charles Slaughter
informative. The high significance of the Bristol Bay watershed, specifically of the Nushagak and
Kvichak river systems, for commercial fisheries on the global scale and for sport and subsistence
fisheries at the regional and local scales, was appropriately described.
RESPONSE: No changes suggested or required.
CWS Comment July 2013: OK
The potential risks and impacts are fairly and succinctly stated. Given the extremely long-term nature
of the projected Pebble project, and the irreversible changes which would be imposed to the region,
the risks seem, if anything, understated. I attribute this to the decision to focus this Assessment on
salmon and anadromous fisheries, with less attention on "salmon-mediated" impacts - i.e., effects on
indigenous culture, on wildlife other than salmon, etc.
RESPONSE: No changes suggested or required.
CWS Comment July 2013: OK
Chapter 2 (Characterization of Current Condition) provides only a superficial overview of the
landscape of the Bristol Bay watersheds; a reader would preferably have access to Wahrhaftig (1965)
or Selkregg (1976), as only two (relatively dated) suggestions, to gain a more comprehensive
understanding of the region.
RESPONSE: Additional information on the physical environment of the region (e.g., geology,
vegetation, etc.), along with an expanded treatment of the regional landscape, has been
incorporated into Chapter 3 (e.g., Figures 3-4 through 3-7). Chapter 3 also includes the suggested
citations.
CWS Comment July 2013: Chapter 3 is a great improvement. The detailed maps and discussions
of the physiographic setting, hydrologic landscapes, aquatic habitats, water quality, seismicity, and
existing development of the region are comprehensive and informative. Explanation of methodology,
as in Box 3.1, Box 3.3, and throughout the entire document, is very helpful. This chapter now allows
a reader to gain a more realistic understanding of the Bristol Bay watershed setting.
The "Water Management" section (4.3.7) seems cursory, highly generalized, and optimistic.
Statements such as "uncontrolled runoff would be eliminated"; "water from these upstream reaches
would be diverted around and downstream of the mine where practicable"; and
"Precipitation... would be collected and stored..." do not indicate actual (proposed) practices or
techniques, nor inspire confidence that actual runoff events during "normal" conditions, let alone
during hydrologic extremes (such as a rain-on-snow event with underlying soils still frozen), would
be planned for or actually managed adequately.
RESPONSE: Water management measures are more clearly described and discussed in Section
6.1.2.5 of the revised draft, and in sub-sections for the mine components in the scenarios. The
assessment no longer contains a no failure scenario, so complete water collection is not longer
assumed. Rather, standard and common practices are incorporated.
CWS Comment July 2013: The revised Section 6.1.2.5 does provide a better discussion of water
handling. Since this is all hypothetical, we must accept such statements as ''If contaminated
groundwater was detected, monitoring wells would be converted to collection wells or new recovery
-------
Dr. Charles Slaughter
wells would be installed, and water from the well field would be pumped back into the TSF or treated
and released to stream channels. " (p. 6-12). Easier said than done!
The last sentence on p. 6-13 states "Runoff at the port site would be pumped to the mine site in the
return water pipeline, contributing to the mine's water supply and avoiding the needfor treatment at
the port. " What runoff at the port site is being considered - surface runofffrom impervious surfaces
at the port itself? Or does this suggest that streamflow at or in the general vicinity of the port would
be captured and utilized to augment water supply at the mine? If the latter, then a separate
evaluation of hydrologic resources, and permitting, would be necessary.
Para. 6.2.2.3 - The generality that water released to streams "would comply with permitted
discharge requirements ", but may differ from "natural stream water" in actual characteristics,
locale, rates, or timing, would not be reassuring to issues of salmon sustainability — based on
information provided elsewhere in the assessment concerning extreme sensitivity of salmon (and
other organisms) to seemingly minor or very localized water characteristics.
Para. 6.2.2.3 - The last sentence summarizes percentage of water to be reintroduced to streams; it
would also be appropriate to emphasize that this envisions LOSS of 26%, 60%, and 30% of total
flows under the three scenarios.
Perhaps I missed it, but I found no acknowledgment of the potential presence of or consequences of
perennially frozen soils - permafrost - in the Bristol Bay watershed, or more specifically in the
Pebble ore deposit locale or the proposed transportation corridor. Selkregg (1976), Fig. 136, shows
soils of the Pebble locale as INT/2g, INT/lg - HYP, or SOU/2g-HYP - that is, well-drained gravelly
soils (INT) or well-drained acidic soils (SOU) with interspersed peaty, poorly-drained shallow
discontinuous permafrost. There is abundant literature on the influence of permafrost on engineered
structures, roads, hydrology, etc. Even if the bulk of the terrain involved in the proposed Pebble
mine, road and infrastructure project is founded on well-drained gravelly soils, any interspersed
permafrost-underlain terrain can prove problematic in terms of landscape stability, potential erosion,
and consequent structural, engineering, hydrologic and water quality issues. See Specific
Observations for a few suggested references in.
RESPONSE: We have expanded our characterization of the soils and permafrost distribution in
the Bristol Bay watershed in Chapter 3 of the revised assessment. As part of this expansion, we
summarize the nature and distribution of permafrost by physiographic region.
CWS Comment July 2013: The inclusion of detailed maps and information from Selkregg (1974) is
appreciated. Figure 3.4 now acknowledges the possibility of isolated permafrost presence in both
the Pebble locale and through the entire proposed transportation corridor.
Note that presence ofperennially frozen ground (permafrost) is not necessarily detrimental to
construction; fine-grained (silts) high ice content frozen soils pose major problems, but coarse-
grained soils with low frozen water content can be stable upon thawing.
-------
Dr. Charles Slaughter
While there is extensive discussion of a proposed transportation corridor, there was no mention of
construction of a major airfield. A project of this magnitude would undoubtedly require development
of a facility in close proximity to the mine(s) capable of handling CI 30 and commercial j et passenger
and cargo traffic, at least to the 737 class, if not 747.1 don't know what the footprint for such an
airfield would be, but it would be substantial, and with requisite roads, fuel handling, etc., would be a
major project in itself. This would seem to be a logical component of a comprehensive assessment of
the potential Pebble project.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and construction and
operation of a new airport is considered outside the scope of the assessment. We would expect that
a full evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these effects if a new
airport is proposed.
CWS Comment July 2013: Again - if this assessment does not or cannot address the full suite of
probable consequences of the PLP project, what agency or entity does have that responsibility?
Is this just putting off the responsibility for truly comprehensive evaluation to (presumed) future
NEPA review ?
As noted in the Executive Summary, the Assessment does NOT address several major components of
the (hypothetical) Pebble project, including electrical generation and transmission, a deep-water port,
or "secondary development" and associated infrastructure which would follow an initial mining
project. A truly comprehensive analysis should incorporate full analysis of these aspects. This
Assessment is thus inadequate in terms of considering potential broader consequences for the Bristol
Bay watershed system.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and we have stated
throughout the text that areas outside of scope may also be important factors.
CWS Comment July 2013: Again - if this assessment does not or cannot address the full suite of
probable consequences of the PLP project, Is this just putting off the responsibility for truly
comprehensive evaluation to (presumed) future NEPA review?
RESPONSES TO CHARGE QUESTIONS
Question 1. The EPA's assessment focused on identifying the impacts ofpotential
future large-scale mining to the fish habitat and populations in these watersheds. The
assessment brought together information to characterize the ecological, geological, and
cultural resources of the Nushagak and Kvichak watersheds. Did this characterization
provide appropriate background information for the assessment? Was this
characterization accurate? Were any significant literature, reports, or data missed that
would be useful to complete this characterization, and if so what are they?
-------
Dr. Charles Slaughter
CWS comment
If only Volume 1 (the Main Report) is considered, the characterization of some aspects of the
Nushagak and Kvichak watersheds would have to be termed cursory. Chapter 2, Volume 1
(Characterization of Current Condition) provides only a superficial overview of the landscape of the
Bristol Bay watersheds; a reader would preferably have access to Wahrhaftig (1965) or Selkregg
(1976), as only two (relatively dated) suggestions, to gain a more comprehensive understanding of
the region. Similarly, Volume 1 provides a relatively superficial discussion of non-fish wildlife
concerns, or human/cultural concerns
RESPONSE: Additional information on the region's physical environment from Selkregg (1974)
has been included in Chapter 3. We have also clarified that our discussion of biological
communities focuses on the assessment endpoints, as defined in Chapters 2 and 5.
CWS Comment July 2013: Chapter 3 is a great improvement. The detailed maps and discussions
of the physiographic setting, hydrologic landscapes, aquatic habitats, water quality, seismicity, and
existing development of the region are comprehensive and informative. Explanation of methodology,
as in Box 3.1, Box 3.3, and throughout the entire document, is very helpful. This chapter now allows
a reader to gain a more realistic understanding of the Bristol Bay watershed setting.
By contrast, the information provided in Appendices A-H appears to be comprehensive and complete
for each subject field. (Appendix I appears to be a general "template" summary, not tailored to the
Bristol Bay watershed environment).
RESPONSE: The purpose of the appendices vs. the main assessment document has been clarified
in Chapter 2. Appendix I is not meant to be specific to any given region, but discusses options that
are possible and notes that their applicability is dependent on site-specific constraints. What would
be chosen for the Bristol Bay watershed environment, given a mining plan and permit application,
also would be dependent on regulatory decisions.
CWS Comment July 2013: OK The clarification is appreciated, as is the inclusion of much
material from the appendices into the narrative of the revised Assessment.
As noted in the Executive Summary, the Assessment does NOT address several major components of
the (hypothetical) Pebble project, including electrical generation and transmission, a deep-water port,
or "secondary development" and associated infrastructure, which would follow an initial mining
project. A truly comprehensive analysis should incorporate a full analysis of these aspects.
RESPONSE: The scope of this assessment was tailored to its purpose, as clarified in the first two
chapters. We would expect that a full evaluation of any future mining permit applications and
subsequent National Environmental Policy Act Environmental Impact Statements would consider
these components.
CWS Comment July 2013: Again - to me, this simply means that this assessment does not (or
cannot, presumably because of limited EPA authority) address the full suite ofprobable
-------
Dr. Charles Slaughter
consequences of the PLP project. Permit application and NEPA review hopefully would be
more adequate and comprehensive.
Question 2. A formal mine plan or application is not available for the porphyry copper
deposits in the Bristol Bay watershed. EPA developed a hypothetical mine scenario for
its risk assessment, based largely on a plan published by Northern Dynasty Minerals.
Given the type and location of copper deposits in the watershed, was this hypothetical
mine scenario realistic and sufficient for the assessment? Has EPA appropriately
bounded the magnitude of potential mine activities with the minimum and maximum
mine sizes used in the scenario? Are there significant literature, reports, or data not
referenced that would be useful to refine the mine scenario, and if so what are they?
CWS comment
Given the available information base for the ore deposits of the Bristol Bay watershed, and the
publicity which has attended the Pebble planned development over the past several years, the
Assessment's hypothetical mine scenario seems fairly realistic. Further, it is appropriate that the
Assessment consider the probable impacts of other future mineral development projects once an
initial entry (presumably Pebble-Northern Dynasty Minerals) has been accomplished. Such
subsequent development - "cumulative effects over a long time period" - could (and should) receive
more emphasis than is accorded in the Assessment.
RESPONSE: The assessment of cumulative effects of multiple mines is given more emphasis in the
new Chapter 13.
CWS Comment July 2013: Given the present relatively un-altered state of the Bristol Bay
watershed, the inevitable and irreversible cumulative effects - on salmon and on the entire human-
landscape-wildlife biota complex — of additional mines and ancillary and induced development are
certainly as significant in the long term as is the initial (proposed) Pebble mine scenario(s). Chapter
13 is a great improvement to the Assessment, and does recognize many probable consequences and
cumulative effects of multiple mines and "induced development". The attempt to provide an
overview of the primary current mining claims is appreciated. This leads to another question: Does
EPA, or any other State or Federal entity, have legal or "moral" standing to more fully evaluate
such long-term cumulative impact on this large, undeveloped, presently controversial piece of the
United States?
Question 3. EPA assumed two potential modes for mining operations: a no-failure mode of
operation and a mode involving one or more types of failures. Is the no-failure mode of
operation adequately described? Are engineering and mitigation practices sufficiently detailed,
reasonable, and consistent? Are significant literature, reports, or data not referenced that
would be useful to refine these scenarios, and if so what are they?
-------
Dr. Charles Slaughter
CWS comment
Based on the actual history of other major resource extraction projects in Alaska and throughout the
world, a "no failure" assumption seems unrealistic. Rather, the assumption should be that there will
be failures, of varying modes and magnitudes, over the life of the project. This reality is recognized
in several sections of text.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly, and
to separate those effects from a scenario where systems periodically failed. The "no failure"
chapter has been eliminated. The revised assessment differentiates between potential effects
from the footprint of a mine (Chapter 7), water treatment (Chapter 8), TSF failures (Chapter 9),
the transportation corridor (Chapter 10), and pipeline failures (Chapter 11).
CWS Comment July 2013: The revision appropriately recognizes thai in any human-designed and
engineered system there will be glitches or failures; I applaud the decision to eliminate the no-failure
scenario. The chapters cited do more adequately outline potential consequences of those specific
features (mine footprint, water treatment, and so on).
In some sections in the Assessment, presumed "mitigation practices" are either cursory, optimistic, or
so general as to be un-supported. Examples include Section 4.3.7's cursory, generalized statements
about handling water: "Uncontrolled runoff would be eliminated... The mine operator would capture
and collect surface runoff and either direct it to a storage location... or reuse or release it after testing
and any necessary treatment"; "... water from these upstream reaches would be diverted around and
downstream of the mine where practicable"; "precipitation would be collected and stored..."; and
"Assuming no water collection and treatment failures, this excess captured water would be treated to
meet existing water standards and discharged to nearby streams, partially mitigating flow lost from
eliminated or blocked upstream reaches." Other examples from Chapter 6: "... assuming no water
collection and treatment failures" and "excess captured water would be treated... and discharged to
nearby streams..."
RESPONSE: Water management (mitigation) measures are more clearly described and discussed
in the revised Section 6.1.2.5, and in sub-sections for the mine components in the scenarios.
However, the intent of the assessment is not to specify technologies, beyond those already
identified by the existing preliminary mining plans. Rather, the assessment focuses on the
environmental outcomes of conventional modern mining practices and effluents.
CWS Comment July 2013: The revised assessment does more clearly articulate possible
hydrologic (water management or ''water balance") issues, as in Section 6.2.2.3, Section?'.3.2, and
Chapter 8. Given the relatively limited quantitative data available, the writers are to be commended
for their efforts. However, I must confess to a remaining impression of optimistic generalities
concerning some aspects of the water issues (perhaps inevitable because this is all speculation on a
possible project, rather than evaluation of a specific proposal). Examples: in Section 6.2.2.3, "This
released water may differ from natural stream water in chemistry and temperature, but would
comply with permitted discharge requirements. Water may be reintroduced at locations, flow rates,
-------
Dr. Charles Slaughter
or times of year that differ from baseline condition. " And in Section 6.2.2.4, . This water would
be expected to be relatively clean and, if properly managed t control turbidity, could most likely be
released without chemical treatment to maintain or augment streamflow. " And Section 6,3.4, (Post-
closure), assumptions about monitoring runofffrom NAG waste piles, establishment of constructed
wetlands, and a presumably adequate well field downstream of the TSF to monitor water quality.
Question 4. Are the potential risks to salmonidfish due to habitat loss and
modification and changes in hydrology and water quality appropriately characterized
and described for the no-failure mode of operation? Does the assessment
appropriately describe the scale and extent of risks to salmonidfish due to operation
of a transportation corridor under the no-failure mode of operation?
CWS comment
Yes, the risks to salmonids are well characterized with regard to the hypothetical mine operation
itself. However, I suggest that the concept of "no failure," if taken as applying to the entire operation
from inception through operation, is not realistic.
RESPONSE: The "no failure" scenario was not meant to represent a realistic scenario. Rather, it
was meant to illuminate the effects that would occur solely from a mine footprint, even in the
absence of accidents or failure. The revised assessment no longer uses the term "no-failure", but
simply presents effects from scenarios having higher probability and lesser magnitude (e.g., failure
to collect or treat leachate water) and those having lower probability and higher magnitude (e.g.,
TSFfailure).
CWS Comment July 2013: OK
The Assessment makes a fair start toward considering the risks to salmonids from the potential
transportation corridor. However, the many issues regarding stream and wetlands directly or
indirectly affected by roads and pipelines are not fully explored. The extent (length, area) of streams
and wetlands affected, as outlined in the text, should be considered a very optimistic lower estimate.
The specific issues mentioned, such as bridge or road maintenance, culvert blockage or failure,
erosion from cuts, fills, and the roadway itself, are all significant. I simply suggest that the potential
consequences of imposition of the (hypothetical) transportation corridor, and future expansions
consequent to ancillary infrastructure development and further additional resource extraction
projects, would be broader, more severe and of more consequence (and thus should receive more
emphasis) than the Assessment indicates. I suggest more fully incorporating Frissell and Shaftel's
Appendix G into the body of the Assessment.
RESPONSE: The revised assessment notes that the characterization of both stream length and
wetland area affected likely represents a conservative estimate of the potential effects of the
transportation corridor on hydrologic features of this area. The cumulative risk section (Chapter
13) has been expanded to include the transportation corridor, ancillary mining development and
-------
Dr. Charles Slaughter
secondary development Additional information from Appendix G is incorporated into the main
text, and the appendix is referenced in a number of places.
CWS Comment July 2013: Acknowledgement of the conservative (low) estimate of wetland and
stream impacts is appropriate. Chapter 13 is a good addition to the Assessment. As noted
previously, the transportation corridor and "ancillary " or "induced" development activities are,
over the long term, as problematic for the Bristol Bay watershed and region as is the initial PLP
(proposed) mine.
Question 5. Do the failures outlined in the assessment reasonably represent potential system
failures that could occur at a mine of the type and size outlined in the mine scenario? Is there
a significant type offailure that is not described? Are the probabilities and risks offailures
estimated appropriately? Is appropriate information from existing mines used to identify and
estimate types and specific failure risks? If not, which existing mines might be relevant for
estimating potential mining activities in the Bristol Bay watershed?
CWS comment
Potential failures seem reasonable, based on history of other mining operations. However, the
consequences of hydrologic extremes during winter (frozen soil) conditions are not adequately
addressed. The possibility of the mining operation and the transportation network encountering
discontinuous permafrost is not mentioned, although at least some soils maps indicate permafrost
presence.
RESPONSE: Warhaftig (1965) reports that permafrost is sporadic or absent in the Nushagak-
Bristol Bay Lowland (Table 3-1). If permafrost were detected during project development or
construction, the designs would need to address any potential impacts on the infrastructure and
potential impacts of the infrastructure on the permafrost. Frozen soil could improve vehicular
access to parts of the project site and minimize the disturbance from such access.
The occurrence of an extreme hydrologic event, such as heavy rainfall on frozen soil or heavy rain
on an existing snowpack, could produce unusually heavy runoff and higher than normal stream
flows. We analyzed the impact of a tailings dam failure during the Probable Maximum
Precipitation (PMP) event, thereby generating the Probable Maximum Flood (PMF). The
increased flow due to precipitation was, in most of the scenarios, small compared to the flow
released from the TSF. Any increases in the peak runoff due to frozen soil or melting snowpack
would also be small relative to the TSF release, so the modeled scenario can be considered a
reasonable bounding estimate. Relative to a TSFfailure flow release, agree.
Designs for the components of the mine infrastructure would need to consider the natural cold
region conditions and incorporate appropriate design features and safety factors to achieve an
acceptable level of performance. Ghaffari et al. (2011) says "The Pebble deposit is located under
rolling, permafrost-free terrain in the Iliamna region of southwest Alaska..." and "The deposit is
situated approximately 1,000ft amsl, in an area characterized by tundra, gently rolling hills and
the absence of permafrost." Ghaffari et al. (2011) describes the transportation corridor thusly:
"The road route traverses terrain generally amenable to road development.... There are no
-------
Dr. Charles Slaughter
significant occurrences ofpermafrost or areas of extensive wetlands." Nevertheless, if sporadic
areas of permafrost are discovered, the designs will need to address the interactions between the
infrastructure and the permafrost.
CWS Comment July 2013: Recognition of the Warhaftig and SeJkregg references in the revised
Assessment is appreciated. Concern with encountering permafrost involves both engineering
questions - ie, soil stability upon thawing (dependent on grain size and distribution), frequency of
freeze thaw occurrences, and broader "ecosystem " questions of landscape dynamics and ecosystem
function ifpreviously perennially frozen ground warms - as is probable both upon surface
disturbance and consequent to long-term climate warming.
I don't understand your statement, "Frozen soil could improve vehicular access to parts of the
project site and minimize the disturbance from such access." Perhaps you are suggesting that
winter-only off-road vehicular access (with adequate snow cover) would be a management option, as
has been implemented in sectors of the North Slope and along the Dalton Highway? While existence
of a deep snowpack can improve over-snow vehicle travel (snow machines, tracked vehicles), Ifail to
see how either perennially or seasonally frozen soil might either "improve vehicular access " or
"minimize disturbance. " See Slaughter C.W., C.W. Racine, D.W. Walker, L.A. Johnson, and G.
Abele. 1988. Use of off-road vehicles and the consequences for Alaska permafrost environments: a
review. Environmental Management 14(l):63-72. , or Slaughter, C.W., and J.W. Aldrich. 1989.
Annotated bibliography on soil erosion and erosion control in subarctic and high-latitude regions of
North America. PNW-GTR-253, Pacific Northwest Research Statin, USDA Forest Service, Portland
OR. 234 p., among many possible references.
You cite Ghaffari et al. (2011) as saying: "The road route traverses terrain generally amenable to
road development. ... There are no significant occurrences ofpermafrost or areas of extensive
wetlands. "Based on your Figure 3.4, there should be only limited (if any) permafrost encountered,
and of course, engineering measures would be taken with whatever soil or geologic conditions are
actually encountered. But given your estimates throughout the Assessment of the length of streams
and areas of wetlands which would be encountered in the transportation corridor (i.e., Section
10.3.1, Box 10.1) , do you concur with Ghaffari's characterization (almost dismissal) of wetlands
encountered by the road route?
The probability approach outlined for potential TSF dam failure is unpersuasive. It is difficult to
relate to a number like "0.00050 failures per dam year," or to the implication on p. 4-47 that one can
expect a tailings dam failure only once in 10,000 to one million "dam years." This could suggest to
the casual reader that failure of the hypothesized TSF1 dam (for which one "dam year" is one year)
should not be anticipated in either the time of human occupation of North America, or the span of
human evolution.
RESPONSE: The commenter is correct. The proposed dams, if designed, built, and maintained to
current engineering best practices, would be anticipated to have a low annual probability of
failure. However, the failure probability would not be zero. The writers concur with the commenter
that these low probability numbers may be difficult for the casual reader to grasp, so we now also
present estimates ofprobability in terms ofprobability failure over different time periods. The
discussion of this issue has been expanded to clarify that the failure rate is a design goal and is not
based on empirical evidence.
-------
Dr. Charles Slaughter
CWS Comment July 2013: The revised and expanded discussion of Chapter 9 helps. Section 9.1.2
does clarify your reasoning. Box 9.1 and Table 9.1 demonstrate that failures can and do occur. I
remain uncomfortable with employing the terminologies of ''dam year" or "mine year" which allow
suggesting one failure every 2000 mine years, or every 714 to 1754 mine years. A cursory reading
might well allow the impression that this means 2000 or 714 or 1754 years of TSF1 at Pebble, and
therefore possible failure is of no concern to the current or next ten to one hundred generations. See
also Table 14.1, which states that a tailings dam failure has a recurrence frequency of2500 to
250,000 years - allowing the interpretation a Pebble TSF won't occur for the next 2500 years, so
why worry?. Could the text and or an addendum to the relevant tables and boxes, explicitly
differentiate between your "probabilities" and "recurrence frequencies", versus "predictions"? A
calculated recurrence inten'al says nothing about an event actually occurring this year, next year, or
2500 years in the future; I still suspect that some readers and decision makers may not immediately
understand this.
Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-magnitude event at the Pebble
locale has an estimated return period of 200 years. Such a return interval probability is difficult to
interpret, given the lack of historical seismic records for the region; in any event, such a return period
estimate is in no way predictive of future seismic activity, in year 2012 or year 2212. (The suggested
200-year return period should also be viewed in light of the 79-year suggested operating life of the
hypothetical Pebble operation, probable longer-time operations at other mineral extraction sites
which would be developed following implementation of Pebble and building from the infrastructure
associated with Pebble, and also the projected very long persistence of the TSFs following cessation
of active mining).
RESPONSE: Box 9-2 in the revised assessment (Box 4-6 in the May 2012 draft of the assessment)
provided the OBE and return period determined by NDM in the Preliminary Assessment. A
detailed engineering design and safety evaluation is outside the scope of this assessment. The
discussion of seismicity (Section 3.6) addresses the uncertainty in interpreting and predicting
earthquake magnitude and recurrence in the Pebble area.
CWS Comment July 2013: Figure 3.15 is particularly helpful, as is the revised discussion offered
in Section 3.6. Of course, I concur with elimination of the reference to a 200-year return periodfor
a 7.5-magnitude event.
Box 4-6 does note that "The return periods stated in Alaska dam safety guidance are inconsistent
with the expected conditions for a large porphyry copper mine developed in the Bristol Bay
watersheds, and represent a minimal margin of safety."
RESPONSE: The return periods used are consistent with the Alaska Dam Safety Guidance, however
the operator could include additional margin of safety in the design for critical structures. The
return period and seismic safety factors do not inform the failure analysis in this assessment, but
would be important considerations during the review process for any future mine plan.
CWS Comment July 2013: OK
-------
Dr. Charles Slaughter
Question 6. Does the assessment appropriately characterize risks to salmonidfish due
to a potential failure of water and leachate collection and treatment from the mine
site? If not, what suggestions do you have for improving this part of the assessment?
Are significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
CWS comment
No. Text suggests that a monitoring well field downslope from the TSF (and presumably from all
hypothetical TSFs) would detect seepage; such seepage would then be intercepted and either returned
to the TSF or "treated and released to the stream channel." Either action presupposes adequacy of
monitoring seepage and subsurface flow (both spatially and temporally); returning such water to the
stream further presupposes fully adequate treatment to meet both regulatory and aquatic biota
requirements for water quality and flow regime.
RESPONSE: The water treatment and leachate capture discussions have been expanded and are
now detailed and quantified in Chapter 8 of the revised assessment.
CWS Comment July 2013: The extensive revisions and added analysis of Chapter 8 are very
helpful. Qualifications and cautions in the text are also appropriate - such as "Even a small number
of flowpaths with higher than expected hydraulic conductivity could significantly affect the direction
and quantity offlow " (p. 8-11); "Wells would not catch all flows from the mine site given its
geological complexity and the permeability of surficial layers " (p. 8-12); the last sentence in Box
8.1 (p. 8-20); "... although post-closure water treatment failures would be less consequential, they
also would be less likely to be promptly detected and corrected" (p. 8-21); "... it is much too soon to
know whether mines that are permittedfor perpetual water collection and treatment... .can actually
carry out those functions in perpetuity" (p. 8-22); " ...releases of water contaminated beyond permit
limits would be likely over the life of any mine at the Pebble deposit" (p. 8-22).
Assumptions are very generalized and optimistic: "assuming no water collection and treatment
failures" and "excess captured water would be treated... and discharged to nearby streams..." - this
assumes both "no failures" over the life of the operation, and that such treated "excess captured
water" could be successfully treated before release to fully meet both regulatory water quality criteria
and the possibly more sensitive biological requirements of individual invertebrates and fish stocks
(Appendices A & B).
RESPONSE: Water management (mitigation) measures are more clearly described and discussed
in Section 6.1.2.5 of the revised assessment, and in sub-sections for the mine components in the
scenarios. Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly, and
to separate those effects from a scenario where systems periodically failed. However, the "no
failure" scenario is no longer included. The purpose of the assessment is to describe the potential
adverse environmental effects that could exist even with appropriate and effective site mitigation
-------
Dr. Charles Slaughter
measures. The assessment is not intended to duplicate or replace a regulatory process, which is
where required permit discharge limits for water quality would be determined.
CWS Comment July 2013: Agree that the brief discussion in Section 6.1.2.5, in concert with
Chapter 8, more comprehensively treats water management and hydrologic issues. I still suggest
that current or future regulatory water quality criteria may not be adequate to meet and sustain the
possibly more sensitive biological requirements of aquatic invertebrates andfish stocks.
Question 7. Does the assessment appropriately characterize risks to salmonidfish due
to culvert failures along the transportation corridor? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so what are
they?
CWS comment
No. The Assessment does not adequately address the road/stream crossing/culvert issue. Given the
projected transportation corridor, Pebble locale to Cook Inlet, and the inevitability of a further
network of "minor" roads in the mine and TSF locale, plus additional infrastructure linkages,
road/culvert/stream crossings are a major concern for aquatic habitat and fisheries. Readers of the
Assessment should be directed to Frissell and Shaftel's Appendix Gfor a more comprehensive
discussion of this important topic.
RESPONSE: The revised assessment addresses the road/stream crossing/culvert issue in detail
Secondary development is now described qualitatively in Chapter 13 of the revised assessment. The
reader is repeatedly directed to Appendix G for more details.
CWS Comment July 2013: Chapter 10 now provides much better information on the entire culvert
issue. Fish passage, proper (or improper) culvert design and construction, maintenance
requirements, and consequences of failure are more adequately outlined.
The specific consequences of a failure on salmonid habitat and biology are portrayed well.
RESPONSE: No changes suggested or required.
CWS Comment July 2013: OK
Question 8. Does the assessment appropriately characterize risks to salmonidfish due
to pipeline failures? If not, what suggestions do you have for improving this part of the
assessment? Are significant literature, reports, or data not referenced that would be
useful to characterize these risks, and if so what are they?
-------
Dr. Charles Slaughter
CWS comment
No. Concerns with pipelines crossing streams, watercourses and wetlands are similar to those earlier
expressed for the road corridor. On-site investigation may well reveal many more "watercourses,"
including intermittent and ephemeral streams, than the 70 crossings cited; possible pipeline failures
thus may have much wider potential for impacting salmonids than is indicated in the Assessment.
RESPONSE: The document has been edited to indicate that 70 is a minimum value.
CWS Comment July 2013: OK
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and recently in
Montana (?), suggesting the probability (with what confidence limits?) that there would be only 1.5
stream-contaminating spills or two wetland-contaminating spills over 78 years of operation seems
wildly optimistic (and what is half a spill?).
RESPONSE: If you think of the mine as a repeated experiment (i.e., many mines with concentrate
pipelines of that length and duration), we would expect a mean frequency of 1.5 spills. A more
straightforward explanation is that we expect 1 or 2 spills, given the scenario. These frequencies
are based on a large data set, not just the TAPS experience (which is an atypical pipeline design
and much larger than the diesel line) or the spill in the Yellowstone River (which is a single event).
An explanation of the frequency has been added in Section 11.1 of the revised assessment.
CWS Comment July 2013: The narrative of Section 11.1 is helpful. Many uncontrolled variables
inevitably affect the length/duration failure analysis; I can envision an argument that the
transportation corridor pipelines will be in a tightly-controlled (at least in early years) setting and
subject to stringent inspection and maintenance, in contrast to many thousands of miles ofpipeline in
more urban or trafficked locales, with much more opportunity for vandalism, ignorance (ignoring the
"do not dig here " warnings), or accidental damage - so the PLP corridor pipelines should be more
reliable than the historical record might indicate.
Of course, the TAPS line is much larger, insulated, partially above-ground on thermopiles and
partially buried, and atypical in size; it may also be atypical in that at time of construction, it was
designed with state-of-art flow and pressure sensors, automated control systems, regular aerial
patrol inspection, etc - yet still experiences incidents, as does the network of gathering lines on the
North Slope . And of course, the Yellowstone river spill was a single event - as is every spill — so
does not of itself provide frequency analysis.
Given three pipelines (excluding the natural gas line) and your failure probability estimates on p.
11-5, is it fair to say that under the 25-year PLP 2.0 scenario 8 or 9 spills would be expected (3 x
2.8), and under the 78-year PLP 6.5 scenario 24 to 27 spills would be expected (3 x 8.6)? This
would seem in accord with your estimate of virtual certainty (99.9% chance) that at least one
pipeline wouldfail, regardless of mine size.
Assuming that any spill (over the 78-year project span) would last only two minutes (p. 6-32, p. 6-
34), with a consequent minimal volume of spilled material, also seems highly optimistic. Even
-------
Dr. Charles Slaughter
highly-automated systems, with redundant sensors and automatic responses, are susceptible to error
or failure, and the Bristol Bay watershed environment is not benign with regard to mechanical
apparatus. The authors appear to recognize this with their discussion of the Alumbrera incident.
RESPONSE: The scenario has been modified to a 5-minute response. Also, more information has
been added about the possibility of system failure or human error in response to this and similar
comments. The uncertainty discussion in Chapter 11 indicates that the exposure scenario is
predicated on the successful operation of a remote shutoff. There may be extreme weather or
geological events that render the remote shutoff system inoperable. We did not evaluate those
events, so the assessment may underestimate these risks.
CWS Comment July 2013: OK. The hypothesized Chinkelyes and Knutson analyses seem quite
appropriate, and provide good illustrations of the possible consequences of a spill at a water
crossing. I also appreciate the discussion of potential human error, Section 11.3.4.2,, and Table
11.9, and Section 11.5.5. There may well be optimism about SCADA systems, but something or
someone can always mess up!
The specific consequences of a failure on salmonid habitat and biology are portrayed well.
RESPONSE: No changes suggested or required.
CWS Comment July 2013: OK
Question 9. Does the assessment appropriately characterize risks to salmonid fish due
to a potential tailings dam failure? If not, what suggestions do you have for improving
this part of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
CWS comment
Yes. Physical consequences of TSF dam failure are fairly portrayed. I would only suggest that effects
of initial sediment deposition and long-term remobilization and redeposition would extend beyond
the spatial and temporal limits of the modeling used in the Assessment.
RESPONSE: Agreed. We now more clearly state that remobilization and deposition could be
extensive; potentially reaching Bristol Bay.
CWS Comment July 2013: I concur that the revised Assessment provides a much better
characterization of potential TSF dam failure. The difficulties with quantitatively analyzing sediment
deposition and remobilization further downstream are appreciated.
Employing advanced eco-hydraulic modeling tools such as MIKE-11, MIKE-SHE (DHI,
Copenhagen), and consultation with state-of-art practitioners (IAHR-International Association for
Hydraulics Research, UI Center for Ecohydraulics Research, and others), along with improved high-
resolution input data such as LIDAR survey of the complete Kvichak and Koktuli/Nushagak systems,
would allow a more complete estimate of potential hydrologic and sedimentation (and consequently
biotic) consequences of TSF dam failure for the entire river system, headwaters to Bristol Bay.
-------
Dr. Charles Slaughter
RESPONSE: We agree that LIDAR survey data would greatly improve the understanding of the
project area topography. Alternative modeling platforms coupled with LIDAR have the potential
to improve the estimates provided in this assessment, but these data collection and modeling
efforts were not within the scope of this project.
CWS Comment July 2013: Understood - your mandate was not to do such field work and
subsequent analysis, which takes a Jot of time, expertise, dollars, and advanced modeling
capabilities. Such an effort could be a future task as the project (potentially) moves forward.
Question 10. Does the assessment appropriately characterize risks to wildlife and
human cultures due to risks to fish? If not, what suggestions do you have for
improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
CWS comment
No. The Assessment clearly qualified that its objective was to consider risks to salmonids, and only
inferentially consider "salmon-mediated" effects.
RESPONSE: No change suggested or required.
CWS Comment July 2013: The revised Assessment does provide a better consideration of both
wildlife and human culture aspects - at least in part in response to other review comments on the
first draft. Sections 12.2 and 12.3 are succinct summaries, which should be considered by all
decision makers regarding this or other "development "proposals. The final bullet ofp. 12-17,
though it might have been re-worded, appropriately states that mitigation for lost subsistence
resources or cultural values is not feasible.
Appendix C provides a comprehensive discussion of non-fish wildlife and the relation of those
populations to salmon. However, the Assessment itself (Volume 1) provides only a brief summary in
Chapter 2.2.3, which could allow a cursory reader to perhaps conclude that wildlife populations have
little risk of impact from the hypothetical Pebble project. Is this the intent of the Assessment authors?
A more in-depth reading of Appendix C allows inferring potential consequences to wildlife and birds
of "salmon-mediated" impacts of mining development.
RESPONSE: We have clarified assessment endpoints and provided additional background
information from Appendices A and B in Chapter 5. Direct effects on wildlife are outside the scope
of the assessment (as stated in Chapter 2), but their potential importance is acknowledged in
Chapter 12.
The assessment has been expanded to include more information from Appendix C of the draft
report (now an independent U. S. Fish and Wildlife Report).
-------
Dr. Charles Slaughter
The scope of the assessment is focused on potential risks to salmon from large-scale mining and
salmon-mediated effects to indigenous culture and wildlife. EPA recognizes the complexity of
potential direct, secondary, and cumulative effects on wildlife. We would expect that a full
evaluation of any future mining permit applications and subsequent National Environmental
Policy Act Environmental Impact Statements would consider these complexities and utilize the
information in U.S. Fish and Wildlife report (Appendix C of the draft report.)
USFWS RESPONSE: We acknowledge the comment about the comprehensive discussion of
wildlife and the relation of selected species to salmon in Appendix C. Appendix C could be used as
the basis of an assessment ofpotential impacts of mining on wildlife, but decisions related to
assessment scope with regard to potential impacts of mining on wildlife is the responsibility of
EPA.
CWS Comment July 2013: OK. Inclusion of these materials in the body of the Assessment, rather
than simply in an Appendix, strengthens the document.
Appendix D provides a comprehensive and useful discussion of the indigenous people of the Bristol
Bay region and of their traditional ecological knowledge and cultures. Appendix D clearly lays out
the vulnerabilities and risks (summarized on p. 4-5 and pp. 150-153) associated with the
(hypothetical) major resource extraction projects. However, the Assessment (Volume 1) provides
only cursory consideration of these human aspects of the potential project - on p. ES-23, and in
Section 2.2.5. Presumably, this is because the EPA mandate is to conduct an ecological risk
assessment, rather than assessment of consequences for human populations, whether indigenous,
native, resident, non-native, non-resident, or the larger cash economy world as represented by the
State of Alaska, Northern Dynasty Minerals, or Pebble Limited Partnership.
RESPONSE: The scope of the assessment related to Alaska Native cultures is limited to salmon-
mediated effects—that is, potential effects on indigenous culture if there are negative effects on
salmon. The assessment draws on the more comprehensive information in Appendix D to evaluate
these risks. The text of the revised assessment has been expanded to include and discuss the
vulnerabilities and risks to Alaska Native culture from potential large-scale mining.
CWS Comment July 2013: OK. The revised text in Sections 12.2 and 13.4.2, while brief, do add
substance to the Assessment. (I make the assumption that the original Appendices A-H are still a
part of the overall Assessment.)
Question 11. Does the assessment appropriately describe the potential for cumulative
risks from multiple mines? If not, what suggestions do you have for improving this part
of the assessment?
-------
Dr. Charles Slaughter
CWS comment
Yes - but a qualified "yes." The Assessment appropriately outlines the probability of additional
resource extraction projects beyond Pebble itself, and recognizes that additional resource
opportunities (beyond the claims depicted in Figure 4-6), currently unknown or unverified, could
become viable or desirable to some interests in the future. Section 7.4 summarizes many of the risks.
However, the brief coverage (16 pages) accorded the entire subject of "cumulative risks" is not
consonant with the very long-term, spatially dispersed (and presumably linked by transportation and
communication corridors) impacts and risks of multiple mines (and associated infrastructure) in
many different sectors of the Bristol Bay watershed.
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing the
effects of induced development, including transportation corridors, and the effects of increased
overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a single mine).
The cumulative effects assessment is not a definitive, quantitative evaluation. It is, instead, a
plausible example of how a mining district could develop and a conservative estimate of the
impacts to aquatic resources and fish. It is intended to shed light on whether cumulative effects
are a significant concern.
CWS Comment July 2013: Chapter 13 is a welcome and appropriate addition to the Assessment.
The detailed attention paid to other known mining claims or prospects is appreciated. This chapter
properly highlights the potential consequences of those mining developments, which will become
more likely (and economically feasible) if the initial PLP project at Pebble proceeds as proposed.
(It was interesting to me to note the existence of Pebble South, which apparently directly abuts
Iliamna Lake; I don't recall any reference to that potential development during the August 2012
meeting in Anchorage, nor any discussion during public testimony of the likelihood of its proceeding
if the initial Pebble project goes forward.)
While "induced development" is briefly considered (Section 13.3), that text is limited to generalities.
The discussion is appropriate, as far as it goes. However, a major long-term consequence of actually
implementing the PLP proposal, under any of the three scenarios, will be the ancillary and "induced"
activities which will inevitably follow. More in-depth consideration of this broad topic, including
second- and third-order effects on salmon and "fish-mediated" cultural and wildlife values, should be
undertaken in a truly comprehensive assessment of the broader Bristol Bay watershed
Question 12. Are there reasonable mitigation measures that would reduce or minimize
the mining risks and impacts beyond those already described in the assessment? What
are those measures and how should they be integrated into the assessment? Realizing
that there are practical issues associated with implementation, what is the likelihood
of success of those measures?
-------
Dr. Charles Slaughter
CWS comment
If it is assumed that the PLP project, or some similar development, were to go forward, I cannot
suggest mitigation measures beyond those discussed above. Since a major concern for salmonids -
perhaps THE major concern - is with consequences of the transportation corridor, simply having the
mine without the roads/pipelines would alleviate much potential risk. However, there is presumably
no practical, economically feasible way to not have the transportation corridor; air transport of all
materials to and from the site might technically be possible, but would not be economically feasible.
RESPONSE: We agree that it would be infeasible to have a mine in this location without also
having road access. No change suggested or required.
CWS Comment July 2013: OK. One other off-the-wall thought, having recently again used the
Seattle monorail - might it be worth a brief look at life-of-project costs of establishing an elevated
monorail in lien of the road system - thus avoiding many of the stream culvert wetland issues (except
during construction), possibly providing lower total $/ ton-mile freight costs Cook Inlet to Pebble,
considering the many truck trips (>50/day) projected annually, providing tighter control of access,
and concurrently providing a platform for the four pipelines? Just a "Popular Mechanics-style "
thought
Question 13. Does the assessment identify and evaluate the uncertainties associated
with the identified risks?
CWS comment
Yes. The authors fairly attempt (pp. ES-24-26, and in each chapter) to note the various uncertainties
and assumptions incorporated into the Assessment. Sections 8.5 and 8.6 briefly summarize those
uncertainties. A question remains concerning the "uncertainties" associated with assigning
probabilities to various failure scenarios; I remain unconvinced that those probabilities have real
meaning or significance for decision-making (see response to Question 5, above).
RESPONSE: As explained in the assessment (particularly in Chapter 14), the probabilities have
various sources and different interpretations, which we have tried to make clear. Some of them
are more useful in decision making than others, but all are the best values that could be derived
from available information. Most are based on empirical frequencies, but some are engineering
goals or regulatory targets.
CWS Comment July 2013: The revised Assessment does fairly discuss uncertainties associated
with each sector; the authors have done a good job of outlining assumptions, data limitations, and
the suite of "unknowns " contributing to uncertainty in projections. Similarly, as mentioned earlier,
the expanded discussion of use ofprobability analysis has been helpful.
-------
Dr. Charles Slaughter
Question 14. Are there any other comments concerning the assessment, which have
not yet been addressed by the charge questions, which panel members would like to
provide?
CWS comment
I would simply re-emphasize that a truly comprehensive assessment of the potential consequences of
a large-scale mineral extraction project, be it the "hypothetical" Pebble-like project or a different
endeavor, should fully consider both the immediate project-specific impacts, and the long-term
watershed-wide consequences of "ancillary" developments - such as other mines, which might
become economic once the primary project's infrastructure is in place. There should be full
recognition of the irreversible nature of such developments, and of the potential and limitations of
possible reclamation or mitigation measures for the full suite of resources and ecosystem "services"
involved, both short-term and long-term.
RESPONSE: Ecological risk assessments generally have more limited scopes than comprehensive
environmental impact statements. This document is not meant to be a comprehensive evaluation
of all impacts potentially stemming from a large-scale mine, and its scope has been more clearly
defined in Chapter 2. We have added discussion to the cumulative assessment chapter (Chapter
13) to emphasize the long-term watershed-wide consequences of ancillary or induced
development. We have added a discussion of compensatory mitigation in Appendix J.
CWS Comment July 2013: Concur with the response. The entire Assessment is a major step
forwardfrom the 2012 draft. The detail provided in all sectors, particularly in the hydrologic
("water balance "), water quality, and cumulative effects sectors, is really appreciated.
3. SPECIFIC OBSERVATIONS
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
CWS comment
1. Global: Provision of full color versions of all figures would have been helpful to this review. The
selected color versions supplied here were useful - we should have had them all.
RESPONSE: Comment noted.
CWS Comment July 2013: The many figures and maps provided are greatly helpful.
2. Page 2-4: Color codes are confusing - use of different colors for same "moisture state" in the five
regions doesn't make sense (to me).
RESPONSE: Figure 2-2 (Figure 3-1 in the revised assessment) depicts hydrologic landscapes
across the Nushagak and Kvichak watersheds separated into their physiographic divisions as
defined by Wahrhaftig (1965), combined with their climate class as defined by Feddema (2005).
-------
Dr. Charles Slaughter
Each color group (e.g., yellow, green, red, etc) represents a separate physiographic division while
color intensity (e.g., light green to dark green) represents the climate class within a given
physiographic division. The use of different colors is intended to show the variety of climate
classes within a given physiographic division, while also depicting similarities in climate class
across different physiographic divisions (e.g., darkest red and darkest green show that two areas
having different physiographic characteristics are alike in their climate class).
CWS Comment July 2013: The new figures and maps are very well thought out and presented,
throughout the revised Assessment.
3. Page 3-4 (S3.4, LI): ... when mine is active
RESPONSE: Correct (now in Section 4.2.4).
CWS Comment July 2013: OK
4. Page 4-5 (last P): Refers to Fig. 3-1, but "existing road segments" listed are not shown on Fig. 3-1,
nor are several cited locales: Williams Port, Pile Bay, King Salmon, and Naknek.
RESPONSE: This reference has been deleted.
CWS Comment July 2013: OK
5. Page 4-11 (first P): "The vast majority of tailings dams are less than 30 m in height..." DOES
THIS REFER TO TAILINGS DAMS AT ALL KNOWN MINING OPERATIONS, OR TAILINGS
DAMS ENVISIONED OR PROBABLY TO BE USED IN BRISTOL BAY WATERSHED?
RESPONSE: The sentence refers to all known mining operations. The text was changed from
"The vast majority of tailings dams..." to "The majority of existing tailings dams..." (Chapter 4).
CWS Comment July 2013: OK
6. Page 4-11 (first P): "Although upstream construction is considered unsuitable for impoundments
intended to be very high or to contain large volumes of water or solids... .this method is still routinely
employed." ARE THE TSFs SUGGESTED IN THIS ASSESSMENT CONSIDERED
"UPSTREAM," "DOWNSTREAM," OR "CENTERLINE"? Para. 1, Section 4.3.5 (p. 4-21), states
that "the most plausible sites" for TSFs are "the higher mountain," which suggests that these TSFs
would be "upstream" facilities, therefore "considered unsuitable..."
RESPONSE: The TSFs suggested in the scenarios would be constructed using the downstream
method and then switched to a centerline method. It is true that in this location, the upstream
construction would be unsuitable. The discussion of the upstream method was intended for
background and has been moved to Chapter 4 in the revision, with the scenario-specific
suggestions discussed in Chapter 6.
-------
Dr. Charles Slaughter
CWS Comment July 2013: The revisions in Chapter 4 and Chapter 6 are helpfid in clarifying this.
7. Page 4-23 (P2): Text suggests that a monitoring well field downslope from the TSF (and
presumably from all hypothetical TSFs) would detect seepage; such seepage would then be
intercepted and either returned to the TSF or "treated and released to the stream channel." Either
action presupposes adequacy of monitoring seepage and subsurface flow (both spatially and
temporally); returning such water to the stream further presupposes fully adequate treatment to meet
both regulatory and aquatic biota requirements for water quality and flow regime.
RESPONSE: Yes, the well fields would be downstream from all embankments for all TSFs and
this has been clarified in the revision. It is assumed for the scenarios that mitigation measures are
operated appropriately. Text has been clarified in the new Chapter 6 to indicate that water would
be treated to meet permit requirements. Scenarios in the revised assessment evaluate effects from
uncaptured seepage during routine operations from both the TSFs and waste rock piles.
CWS Comment July 2013: The text acknowledges that well fields would not expect to either detect
or capture all seepage waters. Recognition of effects of uncaptured seepage is appropriate. There
remain (possibly unavoidable) generalized and optimistic statements, such as ''If contaminated
groundwater was detected, monitoring wells would be converted to collection wells or new recovery
wells would be installed and water from the well field would be pumped back into the TSF or treated
and released to stream channels. " (p. 6-12), or " ...excess captured water would be treated to meet
existing water quality standards and discharged to nearby streams, partially mitigating flow lost
from eliminated or blocked upstream reaches. " (p. 6-15 -6-16).
8. P4-26 to 4-28 (S4.3.7): This "Water Management" section seems cursory, highly generalized, and
optimistic. Statements such as "uncontrolled runoff would be eliminated"; "water from these
upstream reaches would be diverted around and downstream of the mine where practicable"; and
"Precipitation... would be collected and stored...." do not indicate actual (proposed) practices or
techniques, nor inspire confidence that actual runoff events during "normal" conditions, let alone
during hydrologic extremes (such as a rain-on-snow event with underlying soils still frozen) would
be planned for or actually managed adequately.
RESPONSE: Water management measures are more clearly described and discussed in the
revision (Section 6.1.2.5), and in sub-sections for the mine components in the scenarios. Measures
are standard and common to existing mine sites.
Collection and diversion structures would need to be designed and built to handle the anticipated
flows over the life of the mine, including during extreme weather events. The details of these
collection and conveyance systems are beyond the scope of this assessment and are more properly
in the domain of the permitting process when a specific mine plan is proposed.
CWS Comment July 2013: The revised Assessment devotes much more attention to hydrology and
water management. The many analyses, figures and tables provide a great deal of usefd
information, and undoubtedly would be utilized by PLP engineers if they proceed to permit
application andfinal design. The calculations andfigures of Section 7.3 seem quite reasonable, as
-------
Dr. Charles Slaughter
are the findings of Sections 7.3.2 , the statement of uncertainties (Section 7.3.1.5), and the summary
(Section 7.4) .
9. P4-29 (PI): Suggests that 20% more water than available would be required "during startup," and
that difference would be satisfied "from water stored in the TSF"; if 20% more than is available
would be needed, where would it come from to be available from the TSF?
RESPONSE: The water balance has been revised in Chapter 6. The mine operation would capture
more water than needed during all phases of operation. Water could be stored in the TSF as soon
as the starter embankment is operational. Ghaffari et al. (2011) further explain: "Once the TSF
starter embankment construction is complete, site water will be diverted into this facility to ensure
adequate make-up water for process plant start-up. At this time, advanced open pit dewatering will
commence. This water will either be treated and discharged or diverted to the TSF, depending on
environmental requirements."
CWS Comment July 2013: The revision is appropriate. Given the sensitivity of saJmonids and
aquatic invertebrates to seemingly-minor or low-level perturbations of water quality and timing, the
statement from Ghaffari et al (2011) that "This water will either be treated and discharged or
diverted to the TSF, depending on environmental requirements " is not particularly reassuring.
10. Page 4-29 (P3): Assumptions are very generalized and optimistic: "assuming no water collection
and treatment failures" and "excess captured water would be treated... and discharged to nearby
streams..." - this assumes both "no failures" over the life of the operation, and that such treated
"excess captured water" could be successfully treated before release to fully meet both regulatory
water quality criteria and the possibly more sensitive biological requirements of individual
invertebrates and fish stocks (Appendices A & B).
RESPONSE: Water management measures are more clearly described and discussed in the
revision in Section 6.1.2.5, and in sub-sections for the mine components in the scenarios.
Measures are standard and common to existing mine sites. The revised assessment considers water
treatment failure quantitatively and includes refined scenarios with new data for seepage from
TSFs and waste rock piles.
CWS Comment July 2013: The revisions are appropriate. The revised Assessment devotes much
more attention to hydrology and water management. The many analyses, figures and tables provide
a great deal of useful information.
11. Section 4.3.8: This and the previous section mention (but in my view, do not adequately stress)
the extremely long time frame for post-mining active management and oversight. Many hundreds of
years of active management is a longer time than many industrial, corporate or governmental entities
are capable of really embracing - witness the current US Congressional practice of "kicking the can
down the road" - a human trait.
RESPONSE: This issue is addressed in Section 4.2.4 of the revised assessment.
-------
Dr. Charles Slaughter
CWS Comment July 2013: I have noted that the long time frame, and the potential frailty of
human institutions over that time frame, is now acknowledged in several places in the revised
Assessment.
12. Page 4-32 (S4.3.8.2, P3): Suggests that pyritic tailings could be "shipped off site" - i.e., to
where? Deep ocean dumping, or Yucca Mountain?
RESPONSE: For smaller amounts of tailings, the option of shipping them to an off-site location
for disposal might be an option; however, this is less likely with large mines with high volumes of
tailings, so has been removed in the revised assessment.
CWS Comment July 2013: OK
13. Page 4-34 to 4-37 (S4.3.9): This reviewer finds the short Transportation Corridor sub-chapter to
be succinct, but inadequate and superficial in view of the long-term consequences of imposition of
the transportation corridor as portrayed. These deficiencies are addressed, in part, in other sectors of
the Assessment, most comprehensively in Appendix G.
RESPONSE: The transportation corridor analysis has been substantially expanded, and is now
found in its own chapter (Chapter 10).
CWS Comment July 2013: This revision and upgrading to Chapter status is quite appropriate.
The more detailed considerations found in Chapter 10 more accurately reflect the concerns of this
(and other) reviewers about the potential transportation corridor.
14. Page 4-38 (Box 4-3, P2): Para. 2 states that the southwest extension of the Lake Clark Fault is
currently understood to extend to perhaps 16 +/- km from the Pebble ore deposit; however, this is not
reflected in Figure 4-11, which suggests that the Lake Clark Fault terminates perhaps 100 km
northeast of the Pebble locale. Elsewhere in Box 4.3, there is acknowledgement that, while there is
no evidence of recent tectonic activity in the immediate Pebble vicinity, there is relatively little site-
specific data or long-term historical seismic record. I infer that any predictions concerning seismicity
or earthquake occurrence of any magnitude would have very high uncertainty.
RESPONSE: Figure 4-11 (now Figure 3-15) has been revised to more accurately illustrate the
current understanding of the terminus of the Lake Clark Fault. Section 3.6 describes the
uncertainties in predicting seismicity in the Pebble region.
CWS Comment July 2013: OK. The discussion of Section 3.6 is appropriate and is appreciated.
15. Page 4-41 (Box 4-4): Note that in each of the four tailing dam failure examples, the failed
structure was roughly an order of magnitude smaller (in height) than the hypothetical TSF-1
structure, yet those failures had major negative consequences.
-------
Dr. Charles Slaughter
RESPONSE: Text was added to Section 9.1.1 in the revised assessment to indicate that the dams in
these failure examples were significantly smaller than the dams proposed in our mine scenarios."
CWS Comment July 2013: The revision was noted.
16. Page 4-45 to 4-47 (S4.4.2.2): The probability approach to tailing dam failure is unpersuasive as
presented. It is difficult to relate to a number like "0.00050 failures per dam year," or to the
implication (p. 4-47) that one can expect a tailings dam failure only once in 10,000 to one million
"dam years." This could suggest to the casual reader that failure of the hypothesized TSF1 dam (for
which one "dam year" is one year) should not be anticipated in either the time of human occupation
of North America, or the span of human evolution.
RESPONSE: The proposed dams, if designed, built, and maintained to current engineering best
practices, would be anticipated to have a low annual probability offailure but the failure
probability would not be zero. We concur with the commenter that these low probability numbers
may be difficult for the casual reader to grasp, so we also present the estimates ofprobability in
terms ofprobability failure over different time periods. For example, an annual probability of
failure of0.0005 equates to a 5%probability of failure over 100years and a 39%probability of
failure over 1000 years. Text was also added to clarify the basis for our failure rates.
CWS Comment July 2013: The clarification of probability analysis for TSF failure, provided
in the main body of the assessment, is appreciated. However, I still question the message sent,
both in the Executive Summary and in the larger report, that a TSF failure has a "recurrence
frequency of2500 to 250,000 years. " This might imply to a reader that with 2500 years as the
shorter time span stated, no one should worry about a possible TSF failure, despite the report's
documentation of actual failure incidents elsewhere. I suggest eliminating reference to
"recurrence interval" in this discussion.
17. Page 4-48 (Box 4-6): Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-
magnitude event at the Pebble locale has an estimated return period of 200 years. Such a return
interval probability is difficult interpret, given the lack of historical seismic record for the region; in
any event, such a return period estimate is in no way predictive of future seismic activity, in year
2012 or year 2212.
Box 4-6 does note that "The return periods stated in Alaska dam safety guidance are inconsistent
with the expected conditions for a large porphyry copper mine developed in the Bristol Bay
watersheds, and represent a minimal margin of safety..
RESPONSE: The return periods used are consistent with the Alaska Dam Safety Guidance, but
the operator could include an additional margin of safety in the design for critical structures. The
return period and seismic safety factors do not inform the failure analysis in this assessment.
-------
Dr. Charles Slaughter
CWS Comment July 2013: The discussion in the revised Assessment Section 3.6, along with Table
3.5 and Figure 3.15, is quite helpful. It clarifies known seismic history and acknowledges the "high
degree of uncertainty" on this topic for the Bristol Bay watershed.
18. Page 4-50 to 4-60 (S4.4.2.4): The modeled hydrologic consequences of overtopping/flooding of
the hypothetical TSF1 dam/reservoir seem reasonable, given the relatively limited hydrologic data set
available for model input. Probable Maximum Precipitation and Probable Maximum Flood results
should be approximately "correct" and within the same order of magnitude of potential storm and
flood events. The potential consequences outlined (peak flow volume, sediment transport and
deposition, length of stream corridor impacted) appear realistic for the scenario. I suggest that this
topic and hypothetical result should be given more visibility and emphasis in the assessment.
RESPONSE: The dam failure was modeled as an overtopping event since the possibility could
exist. However, the magnitude of the PMP run-off as compared to the resulting failure flood wave
is very small. We agree that an actual dam design will consider additional information, but this
assessment did not investigate the hydrology of the PMF in greater detail.
CWS Comment July 2013: As initially noted, the hydrologic sectors of the revised Assessment,
including this discussion of PMP and PMF in relation to volumes which would be released by a TSF
dam failure, are reasonable.
19. Page 4-62 to 4-63 (S4.4.4): While accurate, this section does not adequately address the
road/stream crossing/culvert issue. Given the projected transportation corridor, Pebble locale to Cook
Inlet, and the inevitability of a further network of "minor" roads in the mine and TSF locale, plus
additional infrastructure linkages, road/culvert/stream crossings are a major concern for aquatic
habitat and fisheries. This issue receives more attention in Sections 5.4 and 6.4, and is mentioned
elsewhere in Volume 1 (e.g., Table 8.1, Box 8-1, para. 8.1.2.4.). Readers of the Assessment should
be directed to Frissell and Shaftel's Appendix Gfor a more comprehensive discussion of this
important topic.
RESPONSE: This issue is now analyzed in more detail in its own chapter (Chapter 10). The
appendix is cited.
CWS Comment July 2013: Ifully agree with the decision to highlight the transportation corridor
with a stand-alone chapter. The importance is readily indicated, simply by reference the Figures
10.2, 10.5 and 10.6. \; this importance fully justifies the more comprehensive treatment of
transportation corridor issues provided in Chapter 10. Inclusion of issues such as dust control and
invasive species is appropriate.
20. Chapter 5: Assumes scenario of "no failure" for entire project, over complete project life. Is this a
realistic scenario, given experience with industrial developments in real-world settings subject to
vagaries of equipment, landscape, geology, weather, local climate, and human judgment and decision
making/execution?
-------
Dr. Charles Slaughter
RESPONSE: The "no failure" scenario has been eliminated in response to several comments, and
is now considered in terms of mine footprint impacts in Chapter 7.
CWS Comment July 2013: OK
21. Page 5-1 (S5.1.1): Question: is "sampling extensively for summer fish distribution over several
years" adequate for characterizing fish populations, given the wide fluctuations in salmon
escapement and return note elsewhere in the Assessment (e.g., Table 5-1, para. 5.1.2)?
RESPONSE: The answer to this question is of course dependent upon the objectives of the
characterization. For the purposes of this section of the assessment, identifying species
distributions within the study area, sampling was adequate for minimally characterizing the
distribution of salmon, Dolly Varden, and rainbow trout within the project area, including known
spawning areas.
CWS Comment July 2013: Can '1 disagree, since yon qualify this as "minimally characterizing
distribution ".
22. Page 5-12 to 5-48 (S5.2): Estimates of habitat, wetland and stream blockage or loss seem
reasonable, but, as noted in the text, are probably conservative or "at the low end."
RESPONSE: Agreed. No change suggested or required.
CWS Comment July 2013: OK
Estimates of probable streamflow diminution (p. 5-25) seem reasonable, but make no reference to
seasonality.
RESPONSE: We assume a constant seasonal demand, thus streamflow changes follow baseline
seasonal patterns under this analysis. We clarify that water managers may alter holding and
release of water to address environmental flow needs.
CWS Comment July 2013: The text, tables andfigures of Section 7.3 demonstrate the author's
cognizance of seasonality of streamflow. The analysis seems thorough. I note that (as in the earlier
draft), projected streamflows do not meet the 20% sustainability boundary for flow alterations (as in
Figure 7-19).
23. Page 5-29 to 5-30 (thermal regimes): This section makes no mention of aufeis or "nalyds," ice
accumulations which can exert major control on spring and early summer habitat availability and
thermal conditions. For examples, see Slaughter (1990), among many other references.
RESPONSE: Citation and text have been added in Section 7.3.2.
CWS Comment July 2013: Aufeis is not only a concern for culvert blockage, though that is a
major concern for roads and highways - see Figures 1, 3 and 4 in Ashton and Griffiths (1990).
-------
Dr. Charles Slaughter
Aufeis can be widespread in natural channels andJloodplains (see Figures 1, 3, 6, 10 and 11 in
Slaughter (1990)), and may be induced by seemingly-minor modification of natural water flow
during winter conditions.
24. Page 5-30 to 5-31: Concur that maintenance of natural flow regime is the desirable target; the
"sustainability boundary" approach is a way to attempt managing within the "natural" bounds of
variability. Note: Figure 5-9 is map of streams and wetlands lost, not predicted flow alteration
hydrograph (see last sentence, p. 5-31).
RESPONSE: Figure references have been updated.
CWS Comment July 2013: OK
25. Page 5-37 (Figure 5-10): UT100D - predicted flow is ALWAYS below lower 20% sustainability
boundary. UT100C, UT100C1, UTC100B -predicted flows are always within the 20% +/-
sustainability boundaries.
RESPONSE: Figure has been revised, and data now presented in Table 8-1.
CWS Comment July 2013: Table 8-1 does not provide indication of predicted flows in relation to
20% sustainability boundaries. Figures 7-15, 7-16 and 7-17 do depict potential flow alterations
which would exceed the 10% and 20% sustainability boundaries; these flow alterations are also
noted; however, exceedances of sustainability boundaries not highlighted) in the last column of
Tables 7-16, 7-17, and 7-18.
26. Page 5-38 (Figure 5-11): Predicted flow for two upper gages (SK100G, SK100F) is always below
the 20% sustainability boundary. Predicted flow at other gages appears to be near or within the 10%
and 20% lower sustainability boundary.
RESPONSE: Figure has been revised, and data now presented in Table 8-2.
CWS Comment July 2013: See preceding comment.
27. Page 5-39 (Figure 5-12): Predicted high flow for NK119A is far below the lower 20%
sustainability boundary throughout the open water season. From onset of snowmelt, flow at other
gages is roughly at or within the lower 20% sustainability boundary.
RESPONSE: Figure has been revised, and data now presented in Table 8-3
. CWS Comment July 2013: see preceding comment.
28. Page 5-41: Gage NK119A - is the estimated decrease of streamflow (minimum mine size) 63%
(Table 5-13) or 73% (text)?
-------
Dr. Charles Slaughter
RESPONSE: Values have been updated.
CWS Comment July 2013: OK
29. Page 5-41: In any case, text is clear: predicted flow reductions for North and South Forks Koktuli
River are materially below the lower 20% sustainability boundary. Text suggests that while upper
Talarik Creek would be essentially obliterated in this hypothetical scenario, lower gaging stations on
Talarik Creek might have partial augmentation of reduced flows, from small tributary flows and
groundwater; without supporting data, this suggestion seems unsupportable.
RESPONSE: Revised text illustrates important trans-watershed groundwater contributions to
lower Upper Talarik Creek.
CWS Comment July 2013: OK - but the trans-watershed groundwater contributions are still
hypothetical.
30. Page 5-42 to 5-45: These pages fairly summarize the potential for substantive alterations to
streamflow regime and surface water/groundwater relationships.
RESPONSE: No change suggested or required.
CWS Comment July 2013: OK
31. Page 5-44 (P3): "Once the mine is no longer a net consumer of water, we assume that flow
regulation through the water treatment facility could be designed to somewhat approximate natural
hydrologic regimes, which could provide appropriate timing and duration of connectivity with off-
channel habitats." I suggest that this is a highly optimistic assumption, and does not address water
quality questions (which are raised elsewhere in the Assessment).
RESPONSE: We emphasize that the ability to manage flows will be dependent upon sufficient
infrastructure and flexibility in water management.
CWS Comment July 2013: OK I'm not sure that you could do much more, since this is all
potential and hypothetical and remains an optimistic assumption of future "sufficient infrastructure
andflexibility in water management. "
32. Page 5-45 (S5.2.3): This entire paragraph should receive greater emphasis.
RESPONSE: This section (now Section 7.3.3) has been expanded.
CWS Comment July 2013: The risk characterization (Section 7.3.3), while brief, is written well
and provides the recommended emphasis.
-------
Dr. Charles Slaughter
33. Page 5-46 (P4): Ignores variable-source-area concepts, which are widely accepted in hydrologic
and watershed analysis.
RESPONSE: We respectfully disagree. The variable source area concept presents the idea about
the expansion and contraction of saturated areas that are the immediate source of streamflow
during storms. In does not invalidate the notion that the total drainage area contributes with long
flowpaths to streams. These long flowpaths ultimately contribute to saturated zones that support
perennial streaniflows.
CWS Comment July 2013: Of course the total drainage area may contribute to streamflow; the
contributing area may shrink or expand depending on incoming precipitation (or meltout of
seasonal snowpack). Your "long flowpaths " may expand or contract seasonally; Ifail to see any
argument here.
34. Page 5-46 (P5): Assumes requirement for more water than is available, but leaves hanging the
question of where that more water might be sourced. Given that the site is a watershed headwaters,
what might be tapped as additional water supply, and what might be the impacts on that source(s)?
RESPONSE: The revised water balance indicates that the mine would have the capability to
capture more water than it needs during all operational phases, including start-up, from within the
mine footprint.
CWS Comment July 2013: If we accept the revised water balance, then you would not be looking
for additional water supply.
35. Page 5-59 to 5-63 (S5.4): See earlier cautions concerning stream crossings, culverts. Note that
road cuts and culverts are particularly susceptible to development of aufeis ("icings"), often resulting
from blockage or alteration of subsurface water movement during cold conditions, as witnessed by
long-standing AKODT maintenance issues - Richardson, Steese, Dalton highways, for example.
RESPONSE: The potential for culvert blockage by aufeis is discussed in Chapter 10 of the revised
assessment.
CWS Comment July 2013: : Aufeis is not only a concern for culvert blockage, though that is a
major concern for roads and highways - see Figures 1, 3 and 4 in Ashton and Griffiths (1990).
Aufeis can be widespread in natural channels andfloodplains (see Figures 1, 3, 6, 10 and 11 in
Slaughter (1990)), and may be induced by seemingly-minor modification of natural water flow
during winter conditions.
36. Page 5-60 (S5.4.2, PI): The statement that"... it is unlikely that a mine access road would have
sufficient traffic to significantly contaminate runoff with metals or oil" is unsupported; it might be
instructive to look at traffic loads for the access road from the Steese Highway to the Ft. Knox mine,
a much smaller operation than the proposed Pebble development.
-------
Dr. Charles Slaughter
RESPONSE: This statement no longer occurs in the assessment.
CWS Comment July 2013: OK
37. Page 5-60 (S5.4.2, P2): First sentence is correct. Second sentence is unsupported and probably
incorrect (see Appendix G). Yes, runoff from roads is location-specific; that does not mean that
runoff from roads would be insignificant to salmonids, given the very large number of streams
(perennial, intermittent, and ephemeral), and wetlands, which would be intersected by the total road
system of the Pebble project. This also seems to be contradicted by Section 5.4.3.
RESPONSE: This issue is now given more attention in Chapter 10.
CWS Comment July 2013: Again, I concur with the decision to address the transportation corridor
more comprehensively in a free-standing chapter.
38. Page 5-60 (S5.4.4.1): There are many more water or seep crossings than 34 - see USGS topog
sheets, ACME Mapper, or Google Earth.
RESPONSE: The number of crossings has been updated in the revised assessment.
CWS Comment July 2013: OK
39. Page 5-61 (S5.4.4.2 and 5.4.4.3): Development of aufeis ("icings") consequent to partial or full
culvert blockage, or induced by soil mantle compaction (i.e., by roads or off-road vehicle traffic) can
partially or wholly block stream channels. Such blockage, in association with ice on the streambed,
may last long past snowmelt and persist well into early summer, possibly affecting fish movement.
RESPONSE: The potential impact of aufeis on stream channels and potential existence past
breakup is discussed in Chapter 10 of the revised assessment.
CWS Comment July 2013: I can only assume (hope?) thai conditions in the Bristol Bay watershed
may be Jess conducive to aufeis or nalyd ("icings ") formation than in interior and northern Alaska.
Aufeis is not only a concern for culvert blockage, though that is a major concern for roads and
highways - see Figures 1, 3 and 4 inAshton and Griffiths (1990). Aufeis can be widespread in
natural channels andfloodplains (see Figures 1, 3, 6, 10 and 11 in Slaughter (1990)), and may be
induced by seemingly-minor modification of natural water flow during winter conditions.
I
40. Page 5-62 (S5.4.5): While I don't have the specific citations at hand, there are published analyses
of dust effects associated with the North Slope haul road and Prudhoe Bay road network. Obvious
effects include accelerated snowmelt along the road corridor, and nutrient or pollutant contributions
to road corridor environs.
-------
Dr. Charles Slaughter
RESPONSE: The revised assessment includes information from published literature on dust
effects in the Arctic (Chapter 10).
CWS Comment July 2013: The added information in Chapter 10 is appropriate.
41. Page 5-63 (S5.4.6.3): Should it read ".. .impacting 270.3 km of stream..."? (Incidentally, it is
interesting that this coarse assessment finds it possible to state impact to within 100-meter
resolution.)
RESPONSE: The revised assessment notes that risks to salntonids from de-icing salts and dust
suppressants could be locally significant, but we cannot state that the entire stream length between
potential road crossings and Iliamna Lake would be affected.
CWS Comment July 2013: OK
42. Page 5-65 (S5.4.7.3 and 5.4.8.3): Viewing the transportation corridor landscape, via maps,
Google Earth or ACME Mapper, gives me the impression that the estimate of 4.9 km2 wetlands
directly impacted is a very low number. It is easy to play with the numbers given on pp. 5-69 - 5-70,
regardless of their accuracy; 66 km of road impacting wetlands, assuming a 10-meter roadway
footprint, yields 0.66 kimof wetlands "under the road" (vs. 0.18 kimin the text). The 200-meter
proximity to wetlands cited, over 66 km of road, yields some 13 kimof wetland impact (vs. 7.3 km2
in the text). 80 km of road within 200 m of streams or wetlands yields 16 km2 of road/wetland
impact. Since these are all assumptions and estimates, it is not possible to conclude that any of these
figures would be the "true" area impacted.
RESPONSE: The 0.18 knu refers to the area of wetlands which would be filled by the roadbed
"intersecting" wetlands (19.4 km length of roadbed x 9.1 m width of roadbed), and does not
include roadbed adjacent to wetlands (note that this area is 0.11 knu in the revised assessment due
to a slight change in methodology for measuring road length). The 4.9 kni2 (not 7.3 kni2) area of
wetland within 200 m of the length of road (on both sides) within the study area is based on actual
National Wetland Inventory (NWI) data (the calculation methodology is described in Box 10-1 of
the revised assessment).
Though NWI data were utilized in this analysis, the 200 m road buffer was derived from a
literature estimate of the road-effect zone for secondary roads.
CWS Comment July 2013: OK. Since these are aJJ assumptions and estimates (including the NWI
data), it is not possible to conclude that any of these figures would be the "true " area impacted.
43. Page 5-74 (S5.4.10): Should this say impact rather than "risk"?
RESPONSE: "Impact" could be used here, but we chose to use "risk" to note the probability
rather than certainty that salmon would be affected by the corridor-associated activities/events
listed
CWS Comment July 2013: OK
-------
Dr. Charles Slaughter
Text implies that even this "no-failure" scenario will impact salmonids; however, it is apparently not
possible to estimate specific changes or the magnitude of such changes.
RESPONSE: The no-failure scenario has been eliminated in response to several comments;
effects resulting from the mine footprint are now discussed in Chapter 7. The exact magnitudes of
changes in fish productivity, abundance and diversity cannot be estimated at this time, but the
species, abundances, and distributions that would potentially be affected are summarized in
Chapter 10 of the revised Assessment.
CWS Comment July 2013: OK - as noted before, Chapter 10 is a valuable addition to the
Assessment.
44. Page 5-74 to 5-77 (S5.6): This section seems cursory and understated, particularly in view of the
extensive discussion of Appendix D.
RESPONSE: The discussion of fish-mediated effects on Alaska Native culture has been expanded
and is now in Chapter 12.
CWS Comment July 2013: The expanded discussion in Chapter 12 is appreciated.
45. Section 6.1: Concur with general overview statements, and with conclusions regarding immediate
consequences of TSF dam failure, which would likely be as severe as or more severe than stated in
6.1.2.1.
P. 6-6, first sentence - note that the failure scenario predicts over 70% fines <0,1 mm, vs. the 6%
"natural" fines concentrations.
P. 6-13, last para. - the assumption that overtopping would not occur in winter is not warranted, as
the authors admit when citing the Nixon Fork Mine incident. In the Bristol Bay environment, a major
rain-on-snow event in winter or spring is within the realm of possibility, and of course human error
is, if not inevitable, always possible.
RESPONSE: We clarify that the scenario we chose to illustrate overtopping did not occur during
low-flow conditions, but was in response to a flood We acknowledge that failure at other times,
and in response to other events, is possible (Section 9.3).
CWS Comment July 2013: OK - suggest that the brief statement on p. 9-15 that "an overtopping
event is only one potential failure mechanism " might be fleshed out a bit - what other hydrologic
events or conditions might lead to failure?
46. Page 6-15 (Box 6.2): Box text implies that human error, lack of timely oversight and correction
was responsible - but never directly says "human error." The apparent assumption that there is no
hydrologic activity after freeze-up (or perhaps, after an ice cover forms on the pool) was naive and
incorrect. At least in that case, it appears that both dam and spillway design (not adequately
considering winter and ice conditions) and operation/inspection (human error) were responsible.
-------
Dr. Charles Slaughter
RESPONSE: The comment is probably correct. However, we have no evidence that the
commenter's conclusion that both engineering error and operator error were responsible was
correct.
CWS Comment July 2013: OK
47. Section 6.1.4.1: Appropriately recognizes the long time period for exposure, over extended
stream lengths, through both initial deposition and multiple re-mobilization and redeposition events.
RESPONSE: Agreed. No change suggested or required.
CWS Comment July 2013: OK
48. Page 6-28 (S6.1.5 and Table 6-6): Seems jargon-laden and does not add to strength of the
Assessment.
RESPONSE: Opinion has been divided concerning this explicit weighing of evidence. Given that
the analysis of risks to fish from exposure to spilled tailings involves six separate lines of evidence,
the EPA intended to show and discuss how the evidence was weighed
CWS Comment July 2013: OK It seems that the revised Assessment has belter wording and
explanation.
49. Page 6-29 (S6.1.7): Concur that remediation "... would be particularly difficult and damaging..."
RESPONSE: No changes suggested or required.
CWS Comment July 2013: OK
50. Page 6-30 (S6.2): Even though "We do not assess failures of the natural gas or diesel
pipelines...," those pipelines would be equally susceptible to failure as the slurry line. Concerns with
pipelines crossing streams, watercourses and wetlands are similar to those earlier expressed for the
road corridor. Similarly, I suspect that careful inspection would reveal many more "watercourses,"
including intermittent and ephemeral streams, than the 70 crossings cited.
RESPONSE: A diesel spill has been added to the assessment, but it was judged that a gas leak
would not pose a potential risk to fish (Chapter 11).
CWS Comment July 2013: OK
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and recently in
-------
Dr. Charles Slaughter
Montana (?), suggesting the probability (with what confidence limits?) that there "should be" only
1.5 stream-contaminating spills in78 years of operation seems wildly optimistic.
RESPONSE: These probabilities and frequencies are based on a large data set, not just the TAPS
experience (which is an atypical pipeline design and nuich larger than the diesel line) or the spill
in the Yellowstone River (which is a single event).
CWS Comment July 2013: The narrative of Section 11.1 is helpfiil. Many uncontrolled variables
inevitably affect the length duration failure analysis; I can envision an argument that the
transportation corridor pipelines will be in a tightly-controlled (at least in early years) setting and
subject to stringent inspection and maintenance, in contrast to many thousands of miles ofpipeline in
more urban or trafficked locales, with much more opportunity for vandalism, ignorance (ignoring the
"do not dig here " warnings), or accidental damage - so the PLP corridor pipelines should be more
reliable that the historical record might indicate.
Of course, the TAPS line is much larger, insulated, partially above-ground on thermopiles and
partially buried, and atypical in size; it may also be atypical in that at time of construction, it was
designed with state-of-art flow and pressure sensors, automated control systems, regular aerial
patrol inspection, etc-yet still experiences incidents, as does the network of gathering lines on the
North Slope . And of course, the Yellowstone river spill was a single event - as is every spill — so
does not of itself provide frequency analysis.
Assuming that any spill (over the 78-year project span) would last only two minutes (pp. 6-32, 6-34),
with a consequent minimal volume of spilled material, also seems highly optimistic. Even highly-
automated systems, with redundant sensors and automatic responses, are susceptible to error or
failure, and the Bristol Bay watershed environment is not benign with regard to mechanical
apparatus.
RESPONSE: The EPA agrees with this comment and has increased the response time to 5
minutes.
CWS Comment July 2013: OK -1 still think that even 5 minutes is optimistic.
The authors appear to recognize this with their discussion of the Alumbrera incident.
RESPONSE: No change suggested or required.
CWS Comment July 2013: OK
51. Section 6.4: Potential road/culvert failures are recognized (again, not that ice issues are not
discussed). Extended periods for repair/rebuilding might be anticipated - witness the repeated
problems with the highway to Eagle, AK over the past several years - and that is a State of Alaska
responsibility, not that of a private company.
RESPONSE: We agree with the commenter that extended periods for repair/rebuilding might be
anticipated. Chapter 10 of the revised assessment notes that long-term fixes to a road that was
-------
Dr. Charles Slaughter
damaged by erosional failure of a culvert may not be possible until conditions are suitable to
replace a culvert or bridge crossing. Further, multiple failures, such as might occur during an
extreme precipitation event, would likely require longer to repair.
CWS Comment July 2013: Chapter 10 appropriately addresses this.
Potential for multiple simultaneous or concurrent failures is appropriate. Non-Alaska examples
would be the Pacific Northwest flood events of 1964 and 1996, both major precipitation events with
widespread flooding and road failures, in a region with much more developed infrastructure and
response capacity.
RESPONSE: We agree with the commenter with respect to the likelihood for multiple
simultaneous or concurrent failures during extreme precipitation events. This is alluded to in the
assessment.
CWS Comment July 2013: This is mentioned on p. ES-24
52. Chapter 7: Recognition of probable additional mining activity, in the wake of a Pebble project, is
appropriate. Assessment is necessarily limited to currently-known potential mining projects. The
cumulative and irreversible consequences of multiple developments, with associated road, power,
housing, communications infrastructure ("secondary development") should be more heavily
emphasized, even though it is not possible to quantify all those consequences.
RESPONSE: The cumulative impacts of multiple developments, road corridors, and secondary or
induced development have been emphasized through additional discussion in Chapter 13.
CWS Comment July 2013: The addition of Chapter 13 with its detailed analysis of several
possible scenarios, is welcome and appropriate.
53. Chapter 8: Section 8.1.2 - note many potential failure modes not analyzed; the lack of analysis in
this Assessment should not be taken to mean that such failure could not or will not occur.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to choose
a representative set offailure scenarios. The focus of the assessment was on potential effects to
salmon. The revised assessment includes more failure scenarios (e.g., diesel pipeline failure,
quantitative water treatment failure, and refined seepage scenarios) and explains why these
particular failure scenarios were chosen. Section 14.1.2 in the revised assessment explains that the
assessment only presents a few failure scenarios. The EPA agrees with the commenter that lack of
analysis should not imply a failure could or will not occur.
CWS Comment July 2013: I did notice your statements on this, and appreciate the thought given
to Chapters 13 and 14.
54. Table 8-1 (Row 2): The reasoning behind the statement that "Most [product concentrate pipeline]
failures would occur between stream or wetland crossing [sic] and might have little effect on fish" is
hard to understand; stream crossings, whether via elevated utilidors or via sub-channel borings or
-------
Dr. Charles Slaughter
utilidors, are locales of angular change, piping connections and joints, and subject to stresses of
hydrologic extreme events - so why would such sites be less subject to potential pipeline failure?
RESPONSE: Most of the corridor is not at or adjacent to a stream or wetland. Therefore, failures
are most likely to occur between streams or wetlands rather than at streams or wetlands. The
statement is not about probabilities per unit length.
CWS Comment July 2013: I fail to understand your response that " The statement is not about
probabilities per unit length." If you are saying thai failure at stream crossings or wetlands is not
probable, because such crossings are a small proportion of the total length ofpipeline, then you must
be referring to "probability per unit length. " My previous comment holds: stream crossings,
whether via elevated utilidors or via sub-channel borings, are locales of angular change, piping
connections and joints, and subject to stresses of hydrologic extreme events (which are not likely to
occur between stream crossings) - so why would such sites be less subject to potential pipeline
failure?
55. Page 8-4 (Box 8-1): As noted elsewhere, the probability arguments for TSF dam failure are not
persuasive, and seem designed to imply that a TSF dam failure would not occur within the next
10,000 years (or 3,000 to 300,000 years with three TSFs operational). This implication is difficult to
square with information on actual past failures presented in Box 4-4 and Table 4-8.
RESPONSE: The discussion of this issue has been expanded to clarify that the failure rate is a
design goal and is not based on empirical evidence.
CWS Comment July 2013: The revised wording addresses this issue. Section 9.1.2 does clarify
your reasoning. Box 9.1 and Table 9.1 demonstrate that failures can and do occur. I remain
uncomfortable with employing the terminologies of "dam year" or "mine year" which allow
suggesting one failure every 2000 mine years, or every 714 to 1754 mine years. A cursory reading
might well allow the impression that this means 2000 or 714 or 1754 years of TSF1 at Pebble, and
therefore possible failure is of no concern to the current or next ten to one hundred generations. See
also Table 14.1, which states that a tailings dam failure has a recurrence frequency of2500 to
250,000 years - allowing the interpretation a Pebble TSF won't occur for the next 2500 years, so
why worry?. Could the text and or an addendum to the relevant tables and boxes, explicitly
differentiate between your "probabilities" and "recurrence frequencies", versus "predictions"? A
calculated recurrence inten'al says nothing about an event actually occurring this year, next year, or
2500 years in the future; I still suspect that some readers and decision makers may not immediately
understand this.
A final note on this: P. 9-7 states that there were 88 observed or reported tailing dam failures in the
period 1960 to 2010; an average 1.7 failures per year, 1987 to 2007; "2 to 5 major tailings am
failures annually from 1970 to 2001"; and 49 tailings dam failures in the U.S (over an unstated
period). These actual failures occurred within one human life span, not within a span of 714 or 1754
or 2000 years. The calculated recurrence frequencies or probabilities based on "dam years " or
"mine years " can appear reassuring; actual records of failures are not, given the irreversible
consequences of such a failure in the Bristol Bay watershed.
-------
Dr. Charles Slaughter
56. Chapter 8: The potential risks and impacts are fairly and succinctly stated. Given the extremely
long-term nature of the projected Pebble project, and the irreversible changes which would be
imposed to the region, the risks seem, if anything, understated. I attribute this to the decision to focus
this Assessment on salmon and anadromous fisheries, with some attention on salmon-mediated
impacts - i.e., effects on indigenous culture, on wildlife other than salmon, etc.
RESPONSE: No change suggested or required.
CWS Comment July 2013: OK
-------
Dr. John Stednick, Colorado State University
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Stednick,John rmailto:John.Stednick@colostate.edul
Sent: Tuesday, September 03, 2013 3:31 PM
To: Thomas, Jenny
Cc: Frithsen, Jeff
Subject: RE: Revised BBA Peer Review Follow-on
Hi Jenny and Jeff,
As promised by today, attached is my review of the Bristol Bay assessment. Please feel free to ask if any
questions. This summer has been very busy with fire-related research and most of my summer plans
were changed. Thank you for your patience. I look forward to reading the compiled reviewer comments.
Cheers,
John
John D. Stednick, Ph.D.
Professor, Watershed Science
College of Natural Resources
Colorado State University
Fort Collins, CO 80523-1472
970.491.7248
-------
Dr. John Stednick
Date: 26 August 2013
To: Jeff Frithsen, Ph.D.
National Center for Environmental Assessment
Office of Research and Development
US Environmental Protection Agency
Washington, DC
20460
From: John D. Stednick, Ph.D.
Colorado State University
dba/ Mountain River Associates, Inc.
Fort Collins CO
80523
Subj ect: Techni cal revi ew of:
An Assessment of potential mining impacts on salmon ecosystems of
Bristol Bay, Alaska. Second External Review Draft. US Environmental
Protection Agency, Seattle WA. April 2013. EPA 910-R-12-004Ba.
Introduction
This technical review is on the above referenced document. General comments are
presented here, and chapter-specific comments or critiques are presented below. The
document has now been written as an environmental risk assessment and thus has
eliminated the previous confusion as a Clean Water Act 404(c) evaluation. The
theoretical models (as figures) are excellent, yet scarcely referenced or used. As pointed
out in the review of the first draft, the appendices provide very useful information and
data that are often not fully incorporated or included in the text. Admittedly the
appendices vary in quantitative information, but currently the reader is directed to the
appendix rather than having data or main points presented directly. Through out the
document, I found a lack of coordination between the written text and the tables and
figures in the main text body. Tables or figures were casually referenced with no
presentation or little discussion of main points. Tables and figures should stand
independently, that is information is presented in such a manner that the reader can make
the same conclusion as the authors. Similarly, the text should be able to 'stand alone'
whereby enough information is presented in the text as summary statistics or other
quantitative points that the reader makes the same conclusion. I found that considerable
time was needed to understand the figure/table point(s) as related to the text. This can be
addressed with text editing and some reorganization.
Abstract
The abstract states the purpose of the document is a risk assessment evaluating the
environmental risk of mining in the area. The first document was criticized because there
was not a specific mine plan. There needs to be a better description of how the three
-------
Dr. John Stednick
different levels of mining activity were determined. The mine plan alternatives (Pebble
0.25. 2.0 and 6.5) and associated quantitative effects are listed as respectively but are
hard to follow. Sentence structure could be improved. Discussion of this range of
'alternatives' needs to be expanded.
Executive Summary
The Executive Summary order of findings does not parallel the assessment order; i.e.
culvert failure, streamflow modification, TSF failure.
ES-2. Pebble 0.25, 2.0, and 6.5 are not 'defined' until page ES-10. As in the abstract a
'better' defense or justification of these values would be useful.
ES-11 PAG and NAG are not defined in this chapter.
ES-14. Hard to follow the sentences with the 'respectively' areas or km affected by
alternative plans.
ES-15. Where are the data for the copper toxicity? Would be helpful to present these
data.
ES-16. Culvert failure rate seems excessive.
ES-17. The mitigation or prevention of spills is to use impact-resistant containers is an
oversimplification. What about handling and other spill containment facilities?
ES-17. PLP is never defined in this section.
ES-17. Would be helpful to better separate the TSF failures for scenarios Pebble 2.0 and
6.5.
ES-17. Table suggests that there is no copper toxicity to fish under Pebble 0.25.
ES-18. Wetland loss unquantified or unquantifiable? Low culvert failure, yet failure rate
is 47%?
ES-21. The comparison of the TSF height to national landmarks is anthropocentric.
Suggest that reference be deleted.
ES-22. Copper toxicity to fish is less clear. Does this refer to the risk assessment or the
literature? Literature is well established, so why the disconnect?
ES-23. Why would Pebble 2.0 extend further and last longer?
2
-------
Dr. John Stednick
ES-24. Reference for a diesel spill recovery in 1-3 years in cold environments?
ES-25. .. .effects of a large mine on salmon. The use of the word single in this
sentence is confusing.
ES-26. The inclusion of a groundwater drawdown in Pebble 0.25 increased the amount of
wetland loss by 84%. Similar results for the other scenarios. State the approximate
acreage, the use of percentages is often misleading.
ES-28. Standard leaching test data are available for test tailings and waste rocks from
the Pebble deposit, but these results are uncertain predictors of the actual composition of
leachate from tailings impoundments, tailings deposited in streams and on their
floodplains, and waste rock piles. This statement is confusing and does little to support
any findings of the potential effects of metals, especially copper being mobile in the
aquatic environment. Suggest this be reworked.
Chapter 1. Introduction
General comments:
Introduction to purpose of document. Identification of risk assessment endpoints.
Specific comments:
1-1. High hydrologic diversity is apparently a new term as related to a diverse fish
population. Diverse as well as productive?
1-2. Is the document for a 404(c) decision or a risk assessment? Confusing still.
1-3. Salmon is the endpoint. Wildlife and culture are secondary endpoints. Does this
change the effects of stressors on these endpoints?
1-4. Inform the public of the biologic and mineral resources of Bristol Bay. Only
based on the readily available materials. I doubt that PLP is showing all the
mineral deposit data.
1-5. This document is not to invoke CWA 404(c), but is a risk assessment. One cannot
equate open pit mine impact as filling of wetlands only.
Chapter 2. Overview assessment
General comments:
Introduction to assessment, including mine alternatives and data sources.
Specific comments:
2-2. Some minor data sources were used without peer review? Identify.
2-3. Reference to Ghaffari report, please identify where or what data were exclusively
available in said report.
2-4. Endpoint based on decision maker needs. Suggest that the outcome is pre-
determined. Change the endpoints, change the conclusion. Similarly on 2-7.
stressors outside the scope could have environmental consequences.
Reference to Challenger event is not needed and suggest it be deleted.
3
-------
Dr. John Stednick
2-8. Reference to HUC's seems out of place.
4
-------
Dr. John Stednick
Chapter 3. Region.
General comments: Introduction to the study site setting, with a strong emphasis on
physical attributes. A reader unfamiliar with the setting would be hard pressed to
understand all the physical, biological, and cultural resources present. Materials can be
gleaned from the appendices and add to the chapter utility. I would argue for presentation
of more data for site characterization. The surface water -groundwater connection is
significant hydrologically and biologically. Much of the impact assessment is loss or
disruption of these hyporheic flows. I think elaboration on the determination and areal
extent of this is important and should be included in this chapter.
Specific comments:
Figures 3.1 to 3.7 have useful information and data that are not developed in the text per
se. These should be presented to the reader via text as well as figure.
3-12. Based on extensive studies in the Pebble area I am always uncomfortable with
value judgments of this sort. When or what is extensive? Does this imply that the PLP
data are the sole source, the best etc?
3-13. The hydrologic landscape as related to the physiographic landscape elements?
Hydrologic landscapes defined by calculating the water surplus, which is an obtuse
definition of streamflow. Yet the next section discusses the importance of groundwater
exchange for salmonid habitat. And earlier it was stated that: Permeable shallow
aquifers, upwards hydraulic gradients, and strong local relief indicate that local and
intermediate groundwater flow systems dominate regional groundwater flow systems.
Crosscutting faults with high hydraulic conductivities....
Groundwater exchange and flow stability make for more stable habitat? Meaning is
unclear.
Large and small lakes support flows that are more stable than ...other salmon streams.
What is the definition of stable here? What represents a departure from stability?
Mixed messages. Need to characterize the system hydrology including groundwater
components with hyporheic exchanges, then relate to salmonid habitat. This section is
the description of the Region, yet few data are presented and the inference of hydrology
to habitat is unclear. Suggest a chapter reorganization to improve overall chapter flow
and logic.
There are considerable data that are available that would could improve this section. For
example the USGS collects streamflow, precipitation, and temperature data. Daily plots
of these data would provide a better site characterization. For example:
5
-------
Dr. John Stednick
USGS 15302200 KOKTULI R NRILIAMNA AK
"O
c
g 1000
O
0)
¦H
J3
~
O
30
Oct 01Mov 01Dec BIJan 01Feb GHar Blflpr OUay GlJun OlJul Olfiug 01Sep OJDct 0J
2012 2012 2012 2013 2013 2013 2013 2013 2013 2013 2013 2013 2013
Provisional Data Subject to Revision
Discharge
• Flow at station affected by ice
When coupled with precipitation data this better illustrates both snowmelt and rainfall
events. The monthly hydrograph time steps do little to illustrate the system dynamics.
3-18. Differences in hydrology, geology, and climate across... create the regions diverse
hydrologic landscapes ... ultimately shaping aquatic habitats. These diverse habitats
support a high level of biological complexity, in part supported by enhanced ecosystem
productivity associated with salmon runs, contributing to the environmental integrity and
resilience of the watersheds integrity and resilience on the watershed's ecosystems.
Although I do not disagree with this statement as a summary or concluding statement, an
introductory statement it is unsupported. Although citations are given (out of order—list
by date), I think the development of this statement is lacking in this section. This is again
a value judgment that may not necessarily be repeated by the outside reader.
Habitat complexity. Abundant and diverse array of aquatic habitats. Change in
terminology. Are these terms precise?
Section needs to be developed in a more logical fashion, so reader can be lead to the same
conclusions.
3-18. Streamflows are listed as mean annual flow (m /sec). Figure 3.10 shows streamflow
as monthly runoff in mm.
3-18. ..correlated flow paths?
How were these four gradient classes determined? How many are in each class? How do
these compare to physiographic landscape elements and hydrologic landscapes?
3-19. Box 3.1. Why start with the statement that the measured gradient of the NHD flow
lines were not accurate compared to the topography defined by the DEM? We have seen
this inconsistency repeatedly with these comparative methods (including streams that go
6
-------
Dr. John Stednick
up and over the watershed ridge). This text box (and others) read more like the user
manual for the GIS operator and I would submit should only be briefly commented on in
the text. The findings could be better presented in the text box, which is the 'take-home'
message.
3-21. ... indicate significant spatial variability in thermal regimes. Why not define and
present this material?
3-24. Mean annual flow is a metric of stream size. So is peak flow. How the streamflow
classes selected and what were the ranges of the watersheds measured?
3-25. Text box expresses mean annual flow as cubic feet per second—not metric. The
equation makes no sense and uses British units as well.
Below the text box is a paragraph that suggests that the salmonid species propensity for
small streams varies. What is the relation between salmonid habitat and stream size? If
the usage varies, why the exercise in stream classification by MAF? This paragraph does
not flow with the present section organization.
3-26. Specifically define each scale. What is scale 1? Why are scales 3, 4, and 5 listed but
presented later? Edit section.
3-28. Water quality section is wanting. There has been a large water quality data
acquisition effort done by numerous parties in the study area. A summary table of mean
concentrations for 3 large river systems is an underrepresentation of that effort and does
not provide an adequate water quality characterization for risk analysis. These data are
listed as background concentrations. Are all samples above areas of mineralization?
Changes over time now that the area has been drilled? Groundwater chemistry values?
Concentration relations to streamflow? No nutrient data, but nutrient return from
spawning salmon was presented as significant for system productivity. Are metals
dissolved or total? TDS values cannot be reproduced by addition of presented analytes.
What are the precipitation chemistry inputs? What are the changes over time (years?).
Speculation on climate change influence on water quality? If significant spatial
variability in thermal regimes, would the same be expected or seen in water quality?
3-28. Seismicity section is out of place.
3-36. Climate change. A generic description of climate change scenarios, but little
linkage to changes in hydrologic processes in the study area. Are such potential changes
addressed later? Scenarios of changes in precipitation and temperature due to climate
change (using SNAP); what is the existing record and variability in temperature and
precipitation? Streamflow records could be used to better illustrate past variability. Have
salmon runs been related to streamflow conditions? How variable is the system using
7
-------
Dr. John Stednick
existing records? I think this section is too qualitative and was disappointed with the
effort.
Chapter 4. Type of development.
General comments
Overview of mining processes and terminology.
Specific comments:
4-1. Reference to Table 13-1 and Figure 13-1? Could include in this section or consider
repeating same figures.
4-2. Table 4.1 Data source?
4-3. Units? T or Mg, metric ton=Mg
4-5. Chemistry and associated risks? Or potential environmental risks?
Chapter 5. Endpoints
General comments:
This chapter presents the assessment endpoints. A broad discussion of salmon and fish
populations, wildlife, and viability of Alaskan Native cultures. Salmon and other fish
resources are the primary endpoint used.
Specific comments:
5-2. Line 9. 8watersheds. Correct typo.
5-3. Data sources?
Chapter 6. Mine scenarios
General comments: The mine scenarios are reasonable approximations of site activity.
The text reads in generalities, yet the tables and figures suggest a greater certainty given
the numbers, and significant figures used in tables. The conceptual models (Figure 6.12,
13, and 14) are excellent and need more references made to them.
Specific comments:
6-10. Data source is Ghaffari et al., 2011?
6-11. Comparison of TSF height to human-made structures is anthropocentric. Suggest
omit such reference.
6-12. Actual water quality in the tailings impoundments may differ significantly from
what is estimated (Appendix H). This may be true, however this is the best available
technology to estimate water quality changes and sentence should be reworded as the
approximation of leachate water quality and potential uncertainty in the test results as
applied to SW Alaska, rather uncertainty in method itself.
6-13. Water treated to meet effluents. Good opportunity to explain receiving water
quality standards versus effluent discharge limits (permit conditions).
8
-------
Dr. John Stednick
6-15. Data sources for Table 6.3?
Unclear why little change in TSF stored pore water in alternatives 2 and 3?
Similarly, the changes in total consumptive losses for 2 and 3.
Why the large increase in WWTP return flows between 2 and 3?
TSF pore water accounts for >90 percent of the mine operations water demand. How
was this conclusion made, I cannot reconstruct that value.
6-18. Pile Bay road reference? What road standards were applied and what was the
precipitation event? Or is this meant to be geographically significant?
6-23. 6.2.2. Water balance [via streamflow and precipitation] and out [via surface
add ']' after precipitation
6-24. first paragraph is unclear. How were streamflow values used to estimate net
precipitation? Area weighted average of net runoff of 840mm? Methodology is unclear.
6-25. Calculations of inflow agree closely to those provided by Ghaffari et al., 2011.
What were those methods and those results?
6-26. How was the regression line fit to the data?
6-27. 6.2.2.3. Uncontrolled leachate escapes suggests a controlled leachate escape?
6-35. References to premature mine closures? First paragraph under 6.3.5.
Chapter 7. Mine footprint.
General comments: Chapter is mislabeled. This chapter addresses the stream habitat and
flow risks associated with routine operations of the mine.... It considers the unavoidable
environmental effects associated with the footprint of each mine scenario....
Flow should be changed to streamflow to avoid any confusion with energy flow, nutrient
flow or other flows. This chapter like other chapters has considerable quantitative
information and data in tables and figures that are not communicated in the text. Often
the text is written in generalities, yet there are many more specifics in the tables.
Specific comments:
7-13. Given the numerous fish inventories for the study area, why are only 4 years of data
presented (2004-2008)? In general the number of counts was highest in 2008 and thus
had the higher inventory counts. Is this to be considered the baseline?
Figure 15.1-2 as cited in PLP 2011 should be repeated here. The significance of the
stream segments as opposed to watershed level inventories is not developed. Were these
differences included in the approximation of the unavoidable effects from routine mine
operations?
7-15. The cumulative frequency plots are befuddling. I would like to see a better
presentation of this information, both in the text and graphically.
7-16. Separation or evaluation criteria are 5 and 10%, while flow separation criteria are
10 and 20% (7-51). Explain difference and cite appropriate references.
9
-------
Dr. John Stednick
7-27. Headwater streams. Can these be considered all streams with mean annual flows
<0.15 m3/sec?
7-28. Streamflow modification. Streamflows reduced by the loss of the percentage of the
expected area lost to each mien footprint and the water yield efficiency for that basin
(m3/sec/km2)?
7-45. The annual hydrograph figure is monthly time steps. As part of the site
characterization (Chapter 3), daily hydrographs would be more illustrative of hydrologic
responses to precipitation events as snow, rain, or rain on snow.
7-55. Maintaining the connectivity of the stream to off-channel habitats is important;
however the suggestion that the WWTP discharge could be used as compensatory
mitigation may not be the best mitigation. WWTP discharge concentrations need to be
evaluated for toxicity of fish species and life stage. WWTP discharges are finite and
cannot be used to address all lost off-channel habitat connectivity.
7-60. Why the detailed comparison between the EPA results and those from Wobus et al.,
2012? Other EPA efforts in earlier chapters were not compared to other scientific works.
The Wobus work cited is an unpublished review draft from a consulting company.
Chapter 8. Water collection, treatment, and discharge
General Comments: A broad approach to water collection and treatment. Toxicity
analysis based on harness and biological ligand model water quality standards are
presents and discussed. Copper is the primary metal of concern, but other metal toxicities
are evaluated, most often as a toxicity quotient. The tables and figures contain a large
amount of quantitative data, yet the writing, like so many chapters tends to be qualitative.
The text could be improved by incorporating more of the quantitative results.
Specific comments:
8-1. Suggest better identifying water quality standards as being promulgated, and water
quality criteria as recommendations. Discuss the role of the state government and state-
based standards as compared to federal recommendations of water quality criteria.
Standard exceedance may result in civil or criminal penalties, criteria exceedance do not.
8-11. First paragraph discusses the range of hydraulic conductivities and suggestion of
bedrock fractures for groundwater conveyance. The second paragraph then states that the
risk assessment considered no flow below 100m. Why the inconsistency?
The water volumes should be 106—super position the 6.
8-12. It is unclear where the estimates of 84% of PAG and 84% of waste rock leachate
would be captured by the pit and wells of Pebble 2.0. This represents a significant
volume of water moving into the groundwater system, coupled with the mine waste
chemistry; this represents a significant environmental risk already.
10
-------
Dr. John Stednick
The estimated pit wall leachate compared to the mean tertiary NAG waste rock as
percentages is confusing. The percentage values suggest that numeric data are available,
thus a tabular comparison of these numbers would be much more informative.
8-18. Table 8.9 Awkward format in presentation of contaminants of concern, i.e.
323/338/590.
8-25. Table 8.10. More water quality data! The coefficient of variation is helpful, but I
suggest the standard deviation may be a better measure of central tendency. Would like to
see these data in Chapter 3. Description of region.
8-43. Table 8.18 is empty (at least in my version of the assessment).
8-52. Section 8.2.3.7. The analogous mine discussion could be expanded. It appears to be
an extensive data base. Could the mines be separated by geography and mine age, to
glean more information of the efficiency and effectiveness of mine operations on
environmental quality. The assessment takes issue with the comparison of the Pebble
Mine to the Fraser River (Box 8.4), but does not offer a mine alternative.
8-60. These differences of interpretation of data can be an important source of
uncertainty in environmental modeling.
Under the uncertainties section, a discussion is presented comparing the EPA risk
assessment estimate of copper concentrations of waster rock leachate to results obtained
by another study (Wobus et al, 2012) that is a draft report from a consulting company.
Why the comparison? I suspect that there are other reports with other results of waste
rock leachate chemistries. I reviewed the Wobus report earlier. In my estimation there are
3 major issues in that report: 1) the use of a stable copper release rate is an
oversimplification of the chemical dynamics that occur in the tailings/waste rock. The
physical environment in Bristol Bay will significantly affect the chemical weathering
processes and rates; 2) since the humidity cell test did not follow ASTM procedures, the
leachate concentrations was increased by 1 standard deviation. No justification for this
inflation was provided; 3) The report inseparably links copper concentrations to
streamflow rates by stating that copper was treated as a conservative constituent, which is
incorrect given the water quality in the study area and hence speciation of copper. Their
report stated that the addition of 1 standard deviation to the copper concentration was to
compensate for not following ASTM procedures, not because the high values and
potential episodic exposures would be underrepresented by the mean. Why does EPA feel
that they have to defend their results to this unpublished consulting report? Same
comment as above in Chapter 7.
Chapter 9. Tailings dam failure
General comments:
11
-------
Dr. John Stednick
Flood simulation using the SCS runoff method. Significant lack of detail in streamflow
generation assumptions and results. Sediment pulse model simplified TSF release of
sediment with HEC-RAS modeling of deposition downstream.
Specific comments:
9-5. Another excellent conceptual model of potential pathways linking TSF failure to
salmonid risk, but the text does not parallel nor address all pathways.
9-13. flooding to create the overtopping is 2.5% of the total peak flow in the Pebble 0.25
failure and 0.1% of the total peak flow in the Pebble 2.0 failure. How were these
numbers generated? What assumptions made? The reader cannot make the same
conclusion based on the materials presented.
9-14. Chapter 3 needs more data presentation of meteorological conditions. Maps of
annual precipitation and plots of monthly hydrographs do little to prepare the reader to
understand this chapter for a rain driven storm hydrograph. The probable maximum
precipitation of 356mm (14 inches) is for a 24 hour event of presumably a 100 year
recurrence interval. Checking the cited reference (Miller 1963), such a storm depth was
significantly larger than any found for the study area. Please explain the difference. How
much precipitation data re available from the study site, and can these data be used for
comparative purposes on Technical Paper No. 47 (Miller, 1963) for more frequent
precipitation events. Defend the selection of a Type I precipitation distribution. When
do rain storms occur and are they over wet soils? How does permafrost and discontinuous
frozen soils affect the calculation of curve numbers? We read later that the example has a
watershed area of 14 km . Where are the other data? Channel slope, time to concentrate,
Manning's n etc. The storm event cannot be reproduced without these data.
Why no comparison to flood events based on regression equations? Estimating the
Magnitude and Frequency of Peak Streamflows for Ungaged Sites on Streams in Alaska
and Conterminous Basins in Canada. Janet H. Curran, David F. Meyer, and Gary D.
Tasker. 2003.
9-15 .producing a peak flood immediately downstream of the dam of 11,637 m /sec
for the Pebble 0.25... A frightening event, nonetheless how was this peak flow
determined? A comparison of a sediment laden pulse event versus the 'modest' peak flow
event. Watch significant figures.
9-34 to 9-48. This section of the chapter presents important materials, but the quality of
writing needs improvement. I had to reread several sections to understand the point of
each section. The sections do not flow in a logical sequence and the coverage quality is
variable.
Chapter 10. Transportation corridor
12
-------
Dr. John Stednick
General comments: A thorough presentation of the risks associated with the development
and use of a transportation corridor as related to road corridor encroachment on wetlands,
streams, and riparian areas road drainage, road runoff, as related to fish habitat and fish
behavior.
Specific comments: As commented on the first external review draft, the failure rate of
culverts of 47 % seems high and may be the result of using older references. A recent
best management practices audit here showed that culverts and bridge crossings were
over 95% effective in maintaining water quality. The reference to ADOT could be
expanded to include more site specific standards.
10-9. Cumulative frequency of stream channel lengths...
I think this figure will leave many readers confused. Even with colored lines, the take
home message is unclear. The text does little to support the figure as well. The mean
annual flow 'breakpoints' need to be presented and clarified earlier. The reach gradient
breakpoints were not identified earlier, and finally the 5% floodplain potential is
arbitrary. Nonetheless, when and where are their significant differences? Not in MAF.
Were stream gradients? The tables present more detailed data, or numeric values that can
be bettered compared, suggest the writing team reevaluate the 'cumulative frequency
figures'.
Chapter 11. Pipeline failures.
General comments: A thorough presentation of the estimation of risks from pipeline
failure and associated environmental risks as related to concentrate and diesel in the
aquatic environment.
Specific comments: None.
Chapter 12. Fish-mediated effects.
General comments. The primary discussion is the indirect effects on wildlife and Native
cultures that are fish-mediated. A change in the title to better reflect the chapter purpose?
Another chapter with good conceptual models (Figures 12.1 and 12.2 that are not
followed up in the text. In my estimation, the figures are excellent, and deserve more
coverage in the text than currently afforded (often none).
Specific comments:
12-5.1 am certain there are better references to bio-concentration factors in fish than an
unpublished report.
Chapter 13. Cumulative effects of large-scale mining
General comments. Identification of other potential mineral plays in the study area. Using
statistics from the Pebble mine (NWI, NHD, etc.) impacts of other potential mines were
approximated. Vulnerability of salmon stocks was identified with the need for healthy
watersheds and the 4 H's used in the Pacific Northwest.
13
-------
Dr. John Stednick
Specific comments: This chapter used Tables 13.2 to 13.8 (11 pages total) to inventory
the waters, fish, and subsistence uses potentially affected by the various mine prospects.
Comparisons between mines are difficult when using tabular formats. I suggest that a GIS
generated map would be more illustrative of areas affects and would allow for easier
mine comparisons. The waters, fish habitats, and subsidence uses could be identified and
then the mine 'footprint' overlaid. Figure 13.3 is such an example.
Table 13.8 is a compilation of the 6 potential mine sites effect on the various resources. If
all water, and all fishes, and all subsistence uses are equal, this is a fair comparison;
however that was not supported.
The cumulative effects of the mine projects (Section 13.2.7) is a simple reiteration of the
various potential processes as affects by mine development; i.e. loss of fish habitat from
the mine footprint, loss of streamflow waters by water withdrawal. There is little
evaluation or discussion of cumulative effects—cumulative results result from
individually minor but collectively significant actions taking place over a period of time
(p. 13-2). This current risk assessment suggests that cumulative effects would occur with
the presence of the Pebble Mine and additional mine operations would simply add to that
impact. If that is the conclusion of the EPA, then that theme should be consistent.
Wording such as leading to increased release of contaminated waters downstream (p. 13-
31); the relatively ephemeral nature of human institutions (p. 13-3I), more susceptible to
trespassing, poaching, and illegal dumping (p. 13-32) are in contradiction to statements
such as opportunities for employment at mines or mine-related services would contribute
to growth in nearby communities (p. 13-31). Induced development following the advent of
large scale mining would undoubtedly bring very welcome economic development to the
region (p. 13-32). The cumulative impact analysis sections, lacks a defined methodology
and the chapter lacks objectivity.
The last sentence in this chapter states Even in the coastal population centers of Alaska,
hatcheries are supplementing salmon returns (p. 13-35). How is this related to
cumulative effects? I was disappointed in the quality of this chapter.
Chapter 14. Integrated risk characterization
General comments:
This chapter summarizes the risk analysis results organized by endpoint, Limitations and
uncertainties are identified.
Specific comments: It would be useful to have more references to the previous section in
the assessment to more easily direct the reader to earlier sections.
14-1.... local habitat loss leading to losses of local unique populations. I do not
remember reading about genetic uniqueness in the assessment earlier.
Explain differences between mine footprint loss of 38, 90, and 145km and habitat loss of
8, 24, and 35km.
14
-------
Dr. John Stednick
Explain the threshold exceedance of 20% again and better clarify to show a decrease in
streamflow (mean annual flows).
Altered streamflow would result in an unquantified loss of riparian floodplain wetland
habitat, yet next section quantifies the wetland loss and reduced off-channel salmonid
habitat. Seems inconsistent.
Shift in groundwater and surface water balance in streams makes the stream less suitable
for spawning and rearing in winter due to decreased winter flow?
14-3 suggest the needfor additional mitigation measures. Suggest stronger wording
such as more aggressive mitigation.
Possible introduction of invasive species? Plants or animals?
14-4 the lower bound is purely aspirational. If it is aspirational then it should not
be considered. Realistic bounds should be used always.
Last paragraph in this section should be moved up front as identification of assumptions
in the assessment.
14-7 scouring the valley and depositing tailings. Seems inconsistent as written, you do
not scour by depositing sediment. Needs expansion.
14-13. ...an initial outflow beyond the 30km limit of the model. The 30km was the model
boundary used, not the model limitation.
Chapter 15. References.
Better inclusion of references suggested in review of the first draft.
Appendices
A casual read of the appendices showed that certain appendices were edited or updated.
We were specifically asked to review Appendix I: Conventional water quality mitigation
practices for mine design, construction, operation and closure and Appendix J:
Compensatory mitigation and large-scale hardrock mining in the Bristol; Bay watershed.
Appendix I was a well -referenced document explaining conventional water quality best
management practices (BMPs). The document was easy to read and covered the subject
well. Appendix J identified compensatory mitigation from the federal and Alaska state
levels. It addresses specific comments from public hearings on the Pebble Project as
related to compensatory mitigation.
Both of these appendices contain useful information, which needs to be incorporated in
the main text of the assessment, rather than be relegated to an appendix. A concern that I
have frequently expressed.
Final thoughts
The second review draft is a significant improvement from the first draft, but is still in
need of editing rewriting to address review comments. The second draft addressed
several points that were brought about in the first review, including the effects of
atmospheric N inputs from blasting and global climate change. Conversely EPA decided
not to address other points including the effects of precipitation timing on variably frozen
soils on streamflow generation, and the transport of total and dissolved organic matter on
15
-------
Dr. John Stednick
metal transport. Another point to consider in the next version is the potential for
TENORMs—technologically enhanced naturally occurring radioactive materials being
present or causing risk (see EPA report 402-R-99-002).
Process observations
For the second external review, the US Environmental Protection Agency elected to
negotiate and contract with the external reviewers individually. The pay rate and
estimated time to complete the review were significantly less than original estimates. A
concern also shared by other external reviewers. As much as I argued that the
interdisciplinary approach experienced in Anchorage produced a better document review,
EPA was not interested in allowing collaboration or wanting a 'consensus' opinion. The
former I support, the later I never did. The purpose of any collaboration is to better foster
interdisciplinary discussions and appreciation of alternative views, all to make a better
environmental risk assessment. In my estimation, this is an opportunity lost and a lesson
to be heeded in future external reviews.
16
-------
Dr. Roy Stein, The Ohio State University
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Stein, Roy rmailto:stein.4@osu.edul
Sent: Friday, July 12, 2013 9:58 AM
To: Thomas, Jenny
Subject: RE: Bristol Bay Peer Review - Confidentiality Reminder
Friday, June 12, 2012
Good Morning Jenny,
Attached please find my Peer Review Comments on EPA's Draft Document, Second External Review Draft,
EPA 910-R-12-004Ba, April 2013 entitled An Assessment of Potential Mining Impacts on Salmon
Ecosystems of Bristol Bay, Alaska in letter form. I hope the EPA authors find my review useful as they
revise this second draft. Let me know if I can provide an additional information and what the format of
the invoice might be.
Have a good day and a great weekend, Roy
Roy A. Stein, Professor Emeritus
Aquatic Ecology Lab
Department of Evolution, Ecology, and Organismal Biology
The Ohio State University
1314 Kinnear Road, Columbus, OH 43212-1156
Voice: 614/292-1613; Fax: 614/292-10181
e-mail: stein.4@osu.edu
Visit our web page at www.ael.osu.edu
-------
Dr. Roy Stein
Aquatic Ecology Laboratory 1314 Kinnear Road
Columbus, OH 43212-1156
Phone 614-292-1613
FAX 614-292-0181
e-mail: stein.4@osu.edu
July 12, 2013
Ms. Jenny Thomas
Office of Wetlands, Oceans, and Watersheds
US Environmental Protection Agency
1200 Pennsylvania Ave NW
Washington, DC 20460
Dear Ms. Thomas:
I am submitting my Peer Review Comments on EPA's Draft Document, Second External Review
Draft, EPA 910-R-12-004Ba, April 2013 entitled An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska. All of my comments can be found below with
appropriate page numbers and/or table/figure numbers. I did not follow your suggestion in a
recent e-mail to arrange my comments around my initial concerns regarding EPA's First Draft
Document simply because 1) I had already begun my review without following this advice, 2) I
note that the Second Draft has been completely rewritten, reorganized, and revised and therefore
did not neatly follow my initial concerns, and most importantly, 3) the EPA authors dealt fully and
completely with all of my concerns. In fact, I carefully reviewed the EPA Response to our Peer
Review Report, concluding that I am quite satisfied with how the EPA authors revised the text of
this Second Draft in response to my comments. Even so, I still have additional comments that
grew out of my review of the Second External Review Draft.
1. Executive Summary (ES). I liked the summary, especially the tone, the emphasis on
quantitative statements (with supporting tables, maps and figures), the ecological services
that the Bristol Bay watershed provides, the multiple scales at which the problem was
examined, the 4000-year history of subsistence (and sustainable) fishing by Native
Alaskans, the wildlife resources beyond the salmon and trout (some consideration of their
macroinvertebrate food supply as well), the three different scenarios re the Pebble Mine,
the words and probabilities that surrounded the discussion of Tailings Storage Facilities
(TSF) failures (and other failures, as documented from the scientific as well as the grey
literature), the idea that current technologies are simply insufficiently developed (page ES-
15) to capture a sufficient proportion of the waste water to prevent salmon impacts, and
the expectation that the Pebble Mine will likely lead to a "mining district". I believe EPA has
struck just the right balance with regard to summarizing the state of the probable mine
impacts on the ecological and cultural resources of the Bristol Bay Watershed. I have
provided only a few comments for consideration by the EPA authors.
OHIO
SIATE
UNIVERSITY
Department of Evolution, Ecology, and Organismal Biology
College of Arts and Sciences
-------
Dr. Roy Stein
a. Figure ES-2. My presumption here is that the "blue" streams are ones for which we
simply do not know if we have salmon in them or not...likely yes, but no one has
documented their presence. Is this correct?
b. Figure ES-3. This figure could use some improvement. It should be pointed out that
these are not truly run sizes but rather proportions of run sizes for comparative
purposes. I think spending a bit of text comparing Bristol Bay production to other areas
of the world (Part A of Figure ES-3) would be useful for it emphasizes the unique nature
of the Bristol Bay Watershed.
c. On page ES -8, you argue that "the Yup'ik and Dena'ina are two of the last intact,
sustainable salmon-based cultures in the world". And what does "last" mean in this
context? How many more are out there? Some context would benefit the reader and
the report.
d. On page ES-9, do we have any idea what the proportion of Native Alaskans find this
hard-rock mining approach to be preferable over salmon or other subsistence ways of
life? Do we have good polling numbers that summarize the point of view of Native
Alaskans re this project?
e. Page ES-14 talks about "streams known to provide spawning or rearing habitats". Isn't
this quite a conservative way to go? How does the reader account for reality in these
statements?
f. "In perpetuity" seems to be danced about a bit here, but I am happy that it is indeed
discussed and emphasized here in the Executive Summary. Of course, I would like to
see more definitive statements re this issue here in the Executive Summary.
2. Unfortunately, the Abstract on page 1, given its shorter format and lack of detail, simply
misses the mark. Here the same items mentioned in the Executive Summary should be
mentioned in some form here. Clearly, the 29 pages of the ES allows for space to discuss the
most important issues. Within the abstract, I suggest that the EPA authors abstract the current
ES and in so doing work to capture most of the relevant, even critical, points made therein.
Emphasize the ecological and cultural resources in this watershed, the greatest sockeye
population ion the world, the leakiness of the habitat (with groundwater shared among many
streams), meaning that mine impacts will stray far beyond where toxic wastes might first be
generated. I don't think the authors should concentrate on impacts that are "at a minimum";
rather the abstract should discuss failures and their impacts. Add something about Illiamna
Lake, about the TSF having to be maintained "in perpetuity", and the many other points made
in the ES. I recognize that the authors are limited by space considerations within the abstract,
which makes this a difficult section of the report to write. I'm afraid that EPA authors have
done what many of us have done: write the abstract last, after all other sections of the report
have been written, typically having to generate this quite important section under intense time
constraints. I urge the authors to rewrite this section with an eye toward my comments above,
thus allowing the reader to appreciate the high points of this revised draft document.
3. Introduction (page 1-1).
a. Page 1-2. I very much like the idea that this document has been defined as an Ecological
Risk Assessment or ERA; hence, those who understand the role of these documents in
the review process re the Clean Water Act can now place this document in its
appropriate context (i.e., essentially used to inform environmental decision-making).
2
-------
Dr. Roy Stein
How resources of interest (salmon and their habitat) are influenced by stressors
generated by an activity (namely, copper mining). In this case, salmon are the primary
assessment endpoint and wildlife (owing to marine-derived nutrients) and Alaska
Native cultures (owing to their request to have the EPA enter the fray) are secondary
endpoints—some confusion existed about the choice of these variables in the first draft.
We now have an explicit rationale for these choices, which should then permit a better
understanding by readers of the role of this document in the decision-making process
re the Pebble Mine.
4. Chapter 2. Overview of Assessment.
a. Page 2-2. I liked the conceptual model diagrams.
b. Page 2-2. The emphasis on peer-reviewed publications is appropriate for a science-
based assessment, even given that all of the data herein cannot come from this source,
for we know so little about the Bristol Bay Watershed. I like how EPA characterized
under what situations one could use data from the Fraser River mining operations, or
those from Idaho and Montana; context here is immensely useful, helping to guide the
reader through the set-up of the report.
c. Page 2-4. I liked how EPA anticipated the criticism from developers re how risk
assessment from "old" technologies cannot be used to estimate risk for newer
technologies and then used the Challenger accident to show how new, improved
technology can fail too. The justification for using historical failure rates (because new
technologies may have new ways to fail) is compelling. This argument is critical to the
readers.
d. Page 2-8. I liked the various scales used and the fact that EPA provided a number
identified as percentage of the scale above, for it helped me judge absolute and relative
sizes of geographic areas represented in these scales. Table 2.1 should have a title
associated with it and the mine scenario formats should decline in size per the scales in
the table. Just seeing Figure 2.2 does help me envision the situation. I wondered about
the Transportation Corridor in Figure 2.2 inset thinking that indeed it covered the
entire watershed of all streams, rivers, and wetlands influenced by the road from the
Pebble site to Cook Inlet. That was eye-opening and reflects the impact of the road on
the watershed, reinforcing some who believe that the road has an impact far beyond its
seemingly narrow footprint. Figures 2.2 through 2.7 are stellar and a valuable addition
to the ERA.
e. Figure 2.6 (page 2-13): How are we to distinguish 'wetlands' from other water-body
types in this figure. The legend doesn't help us much here but I guess I could envision
(if I squint) wetlands as compared to lakes or rivers/streams. Using colors that exhibit
more contrast could aid in interpretation of this figure.
f. Page 2-14. I appreciated the references to National Hydrography Dataset but worry
that it adds detail that I simply do not get. "HUC 12 Level" is what, indeed? Could there
be one additional inset box that might explain to the novice (especially if this document
is to inform the citizenry), serving to define the terminology?
5. Chapter 3. Region.
a. Pages 3-13 and 3-14. I liked how the text relates groundwater flow and flow stability
and communication among rivers streams and wetlands to the success of salmonid
3
-------
Dr. Roy Stein
fishes. This certainly is a critical feature of these two watersheds. Following this idea
through the document is critical, in my mind, to assessing the impact of the Pebble
development on native salmon.
b. Pages 3-35. I can see how you came to the conclusion that it difficult to predict how
many and what magnitude earthquakes will play in the Pebble Mine development.
c. Page 3-36. I was extremely pleased to see climate change predictions and just how the
environment might change. You compared your conclusions re change to those
generated by other scientists and all fell within a range that lent credence to your
results. Nicely done.
d. I also liked the idea of doing multiple scenarios re the extent of climate change.
e. Page 3-45. The conclusion generated with regard to the impact of climate change is a
relevant one. Salmon will only persist in the face of climate change by using all of the
genetic and life history diversity that they now have. Reducing this variability (i.e., via
the impact of the Pebble Mine) can only lead to loss of stocks.
6. Chapter 4: Type of Development.
a. Figure 4.1. While most all of the figures (and especially the maps) are much improved
from the first draft, Figure 4.1 needs some work. Perhaps the way to handle it is to
divide it into multiple figures, such that the reader could actually see the distribution of
copper-deposit mining throughout the world. I applaud the authors for including this
deep background and context for it helps us understand the "need" for a mine such as
the Pebble.
b. Box 4.1 clarifies the multiple terms that caused our expert panel some angst. The terms
now seem to reflect a good deal of insight into how various practices will be used within
the context of copper mining at the Pebble facility. This improves the presentation
dramatically.
c. Page 4-10, Box 4.3. As I read through this box, I worried just a little about the multiple
jurisdictions involved in mine oversight, such as the Bureau of Land Management, the
Alaskan Department of Environmental Conservation, the Bureau of Indian Affairs (and
all of the associated tribal authorities), the State of Alaska, the Alaska Department of
Fish and Game, the Alaska Department of Natural Resources, Federal EPA, etc. I list
these organizations here as only a partial list of those agencies who have a "horse in the
race". I worry about how they will interact, the extent of their historical interactions,
and just whether they, collectively, with all of their different mandates and associated
stakeholders, can present a cooperative, united front in dealing with hard-rock mining
in Alaska. Or might there be sufficient "turf" issues that Pebble Corporation could take
advantage of these differences to exploit these resources without minding all of the
regulations from all of these regulatory bodies. Text that might serve to assure the
reader that all is well (if this is the case) in this arena would be helpful. Or perhaps
some text to reflect on how these jurisdictional issues have been successfully handled in
other complex environmental projects.
d. Pages 4-16 and beyond. I liked the discussion of TSF's, their construction, their
limitations, etc., though this would have been the place that I would have liked to have
seen some mention of the groundwater issues in the Bristol Bay watershed and how
communication among water bodies (i.e., streams, rivers, wetlands, ponds, lakes, etc.)
can compromise the usefulness of a TSF. Note point 5a above.
4
-------
Dr. Roy Stein
e. Page 4-17. Certainly, given these on-the-ground parameters, a fully lined TSF (even
when there will be multiple ones) should be required. The size of the first TSF to be
built may cause some to say that lining costs would be prohibitive, but I see no
alternative given the hydrology of the Bristol Bay Watershed. Having said all this, the
comments about the length of life of geo-membranes (for example) is short (perhaps as
long as 600 years) compared to the required life of the TSF before water can be
released (30,000 years). Dry stack tailings may be the only feasible alternative to TSF.
Even so, because of the low grade of ore, the immense amount of tailings, the "wet"
nature of the environment, and the acid-producing nature of the tailings, dry stacking
may not be feasible either (and consider the immense area required for storage, see
Response Document. This is real conundrum for the siting of this mine. It might serve
the reader well to emphasize this central limitation to the siting of this mine in the
Bristol Bay Watershed.
f. Page 4-19. I liked the final discussion on "Timeframes". This sets the reader up for
later, more in-depth, discussion of these issues. Nicely done.
7. Chapter 5. Endpoints.
a. Pages 5-1 and 5-2. The justification for making salmon the primary endpoint was
missing in the original draft and now is nicely summarized and justified here.
b. Table 5-1. The list of fishes present in the Bristol Bay watershed is far more complete
and I believe more accurate than in the first draft. Good work here.
c. Figures 5-1 and 5-2 help the reader assess the distribution of salmon relative to the
variety of fisheries that occur in the watershed, though I must say that Figure 5-2 is a bit
difficult to interpret. Revision with attention to colors and symbols used might improve
its message.
d. Additional detail re one topic or another. This practice should be extended to all
references to appendices in the Main Report.
e. Table 5.3, Page 5-11, should include percentages for the numbers coming from each
river system for then the reader could gain some quantitative appreciation for the
contribution of the two river systems that will be impacted by the Pebble Mine.
f. Figures 5-3 to 5-11 nicely summarize what we know about the distribution of
"salmonids" in the Bristol Bay watershed. My only concern here is that given that the
Nushagak and Kvichak rivers provide 50% of the sockeye salmon harvest, can the EPA
now say what proportion of this harvest might be lost due to the construction of the
mine? In other words, use these numbers, percentages, etc., later in the report to
provide context as to what will be lost if the mine is built.
8. Chapter 6. Mine Scenarios.
a. Page 6-15. I find it a little hard to imagine that water could be treated and then
discharged into streams. Would the water quality of this discharge even begin to match
the extremely high-quality water of extant streams prior to mine operation? This
seems a little beyond the ability of an on-site, waste-treatment facility. A bit more detail
here to convince the reader this is even possible would be valuable here.
b. Page 6-31. The idea that post-closure TSF water would need to the treated for 100's to
1000's of years begs the question just a little. Work done suggests that we are talking
10's of thousands of years rather than these shorter time periods. One could certainly
5
-------
Dr. Roy Stein
make the argument that there is no real difference between these two lengths of time,
simply because human institutions do not last this long and the likelihood of TSF failure
sometime during even the 100's of years projected life is nearly inevitable, i.e., because
no one is watching. Of course, having contaminated water flow downstream via
streams in the Bristol Bay Watershed will likely make these streams uninhabitable for
salmon for a very long time.
c. Page 6-35. What incentives exist for any company to continue to invest in the
monitoring and maintenance of tailings storage facilities (but see Box 4.3 on Financial
Assurance, for the nub of the issue is dollars, isn't it?)? If well fields downstream of the
TSF began to demonstrate that contaminated water is flowing from the TSF, then why
would the company work to eliminate this source of contamination, save for an
oversight organization such as a state or federal regulatory institution? I think I get it,
that there is to be financial assurances that will support ongoing monitoring and
appropriate maintenance post-closure. The greater risk seems to evolve from
"...chemical or tailings spills..." (Box 4.3, page 4-10) which these assurances do not
cover; how does one assure the public that these mishaps can be cleaned up?
d. Pages 6-35 to 6-36.1 appreciate the discussion of financial assurances and how they
have not worked particularly well in the past. The commitment to making these sums
actually reflect the costs of remediation is critical to any permitting that might occur via
the State of Alaska. Should this point be made more explicitly?
e. Page 6-36, for example. Each time the authors came to a "what if" scenario, they then
provided the "then" statement, usually followed by an example from the real world.
This is nicely demonstrated on page 6-36 when they discuss a situation when a mine re-
opens after being closed for some time thus leading to a change in mining practices.
The authors cite the example of the Gibraltar Mine in British Columbia in which it was
first permitted as a zero discharge operation and then re-permitted allowing for treated
water to be discharged. These real-world examples lend credence to the "what if"
scenarios and help the reader understand just what might occur with the establishment
of a mine such as Pebble.
f. Table 6-9 helped me understand EPA's regulatory role in applying the Clean Water Act
to the establishment of the Pebble Mine.
g. The conceptual diagrams (Figures 6-12, 6-13, and 6-14) helped me visualize the
interactions that lead from mining impacts to salmon. They were clear and well put
together.
9. Chapter 7. Mine Footprint.
a. In Box 7.1, the authors argue that stream lengths and wetland area reflect a lower
bound on the estimate lost to the footprint of the mine. In turn, the estimates of the
number of salmonids in these same streams are likely underestimated by 50%. Hence,
this suggests that the number of salmonids eliminated by the footprint of the mine is
likely 4X what is reported. I agree. To remedy these underestimates, the authors
should revise both the Executive Summery and the Abstract, pointing out that these are
either lower bounds to these estimates or are increased by some metric to better reflect
true abundance. Or, at the very least, point out in these sections that the numbers
reported are substantial underestimates of what likely is out there.
6
-------
Dr. Roy Stein
Because it is small headwater streams that are eliminated by the mine footprints, it
would be appropriate to argue that Coho Salmon and Dolly Varden are differentially
affected simply because they spawn in these smaller streams (page 7-22). Is there any
way this can be better quantified?
In Table 7-5, Dolly Varden are not mentioned; is there a reason for this? Do we simply
not know the distribution of Dolly Varden in these streams (see page 7-26)?
Pages 7-27 and 7-28. The difference between the first draft of this report and this
second draft is dramatically demonstrated by the number of peer-reviewed references
cited on pages 7-27 and 7-28 when the authors are discussing general salmonid ecology
relevant to the Bristol Bay Watershed. The text becomes so much more convincing
when most all statements re the ecology of these species are supported by multiple
literature citations.
10. Water Collection, Treatment, and Discharge
a. Page 8-3, Figure 8-1. I liked this figure and most all that followed in proceeding
chapters for they aided in setting the stage for the discussion to follow.
b. Page 8-4. The argument that during mine operations, 10-49 million m3 of treated water
would be discharged per year (Table 6-3). Though there is concern about the quality of
the water overall as well as the quality of this water as influenced by accidents, poorly
functioning equipment, environmental conditions that exceed the specifications of the
various wells, etc., there is real concern about the waste leachate and how it might
directly influence salmon populations. I worry that the metals, acid, and sediments in
this released water during mine operations will likely cause major impacts on the
extant salmon populations (with the most extreme being direct mortality on through
less extreme impacts on growth, physiology, migration patterns, distribution
throughout the watershed, etc., etc.). Then, the impact of this waste leachate post-
closure will undoubtedly be dramatic if not in the first 50 years, most certainly in years
beyond the time we (as the mine operators, the State of Alaska, the EPA, Native
Alaskans, Commercial Fishers, etc.) are watching, i.e., monitoring, assessing, and
maintaining water-treatment facilities.
c. Page 8-11. In the last sentence of the first paragraph, the comment about groundwater
conveyances suggests that toxicants can be transported just about anywhere in the
watersheds of the two affected rivers. This feature of the landscape, mentioned
explicitly elsewhere as well, makes the placement of the mine here in the Bristol Bay
Watershed suspect. And, of course, it is this very groundwater that supports the
reproduction of all of the salmonids in the basin.
d. Page 8-11. Given our inability to predict where surface water, collected as wastewater,
might flow (third paragraph on this page) via the groundwater makes this an even more
intractable issue when one is trying to protect water quality in the Bristol Bay
Watershed.
e. Page 8-11. Combine the above two points with the water chemistry of the natural
waters of Bristol Bay Watershed (i.e., extremely high quality but not well-buffered) and
one has the makings of a dramatic negative impact on the salmonids upon siting of this
copper mine. I would like to see an increased emphasis on these aspects of the
potential effects of the Pebble Mine.
c.
7
-------
Dr. Roy Stein
f. Page 8-20. Has there truly been any thought or consideration given to a reverse
osmosis treatment of wastewater at this site?
g. Page 8-21. Given that mines continue to be operated even when those responsible for
operations know they are not meeting water quality criteria for wastewater, what
assurances do we have from Pebble Limited that their operators will differ from the
past (i.e., in a good way)?
h. Page 8-31. And even given that standards now in place for the quality of water that is
discharged from the site during operations, these standards appear not to protect
macroinvertebrates, food for young salmon. Disrupting the food supply will indeed
compromise salmonid success.
i. Table, 8.22, for example. As the mine grows larger and into the economies of scale that
the company requires, the impact on water quality and ergo the fish becomes more and
more dramatic (i.e., more fish kills, etc.)...and this is during normal operations without
any infrastructure failures. Can this truly be tolerated and a permit issued? Sadly, the
monitoring of ongoing operations at extant copper mines has not been done (see first
partial paragraph on page 8-53); so then how can we anticipate the ecological effects of
the Pebble Mine even under the best of conditions?
j. Page 8-55. More than 500 million salmonids will be affected by the Pebble effluent;
doesn't this tell the story in one sentence. Impacts on resident and anadromous
salmonids will be negative, large, and long-term.
k. Page 8-63. Changes in stream temperatures means a disruption in the various life
history characteristics dependent on temperature, such as growth, survival, migration,
movement, etc. Hence, negative impacts on salmonids will indeed be huge, largely
negating the portfolio characteristics of these species (down to specific life histories
within populations) that have allowed these salmon to be so successful over their
history.
1. Page 8-65. Actually, we know quite a lot about how temperature influences salmonid
success. For example, quantitative energetic models have been built and tested in the
lab and field. While we may not have specific information on every species in the
watershed, the fact remains that we have a pretty good conceptual framework backed
up by solid work to help us understand the impact of temperature on salmonid success.
11. Chapter 9. Tailings Dam Failure.
a. Page 9-1. Not only would this failure eliminate spawning grounds and rearing habitat
forever, the method for remediation (and even the wisdom of any sort of remediation,
given constraints) is not immediately apparent. Streams may well be too small to float
a dredge and because this whole area is road-less with long winters and innumerable
streams, rivers and wetlands, actually getting equipment to these remote sites may be
near impossible. I encourage the authors to emphasize these points.
b. Page 9-3. A failure of this sort is not simply conjecture for these failures have occurred
the world over. And these failures, while dramatic, long-lasting, and impactful, reflect
dams that are less than half the height of the planned Pebble TSF at its largest footprint.
Some emphasis on this point would be useful.
c. Pages 9-3 and 9-4. One piece of good news re the "in perpetuity" nature of minding the
TFS is that these facilities seem to stabilize through time, resulting in fewer failures as
compared to active TSFs. On the other hand, what is the longest period that has been
-------
Dr. Roy Stein
monitored for a dam post-operation Likely, this would be less than 100 years and prior
to these dams being built, engineering standards would no longer be applicable. Rather
much of a conundrum, wouldn't you say?
d. Table 9-2, Page 9-8. The classification scheme Alaska uses to determine the amount of
engineering that must be done according to the risk of losing lives seems a bit
inappropriate in this context. The risk here to the livelihood of individuals is
paramount and I worry that Alaska's scheme may not protect the salmonids, the Native
Alaskan subsistence way of life, and the tremendously productive Bristol Bay
commercial and sport fisheries.
e. Page 9-31. TSF failure is postulated to reduce the Chinook Salmon run by 60% and this
run is one that produces a large proportion of the Chinook on the West Coast, and is the
one least influenced (at present) by habitat loss, etc. Protecting this run is paramount
for Alaska, the nation, and the world.
f. Box 9.6, Page 9-35. I like this box and many of the others that provided brief synopses
re case histories applicable to the Pebble Mine. I appreciate the comparisons and
caveats. In addition, based on results from other mines sites, it is clear that tailings
toxicants, traveling downstream, not only reside in the stream for decades to hundreds
of years, they also have been documented to move 210 km in one system. Because
nearly all of these TSFs are smaller than the Pebble one, impacts are even likely to
exceed these measured effects. Making this point more explicitly would enhance the
message.
g. Table 9-11, Page 9-50. I liked the attempt at weighing the different lines of evidence,
then drawing conclusions based on best scientific judgment. This "weighing of
evidence" helps the reader understand the limitations of the data while simultaneously
being able to draw what valid conclusions one might draw from such an analysis.
12. Chapter 10: Transportation Corridor
a. Figure 10-1, Page 10-3 was a little difficult to interpret, given its scale and the colors
(where are the ponds, for example). So, I much appreciated the next figure (Figure 10-
2) that zoomed in on some critical areas and these maps were far easier to interpret
and to gain some perspective on the impact of this road on the aquatic environment.
b. Page 10-8. The work demonstrating divergences in populations in close proximity
(Quinn et al., 2012) suggest that the road indeed will have a major impact if any
migratory routes are compromised, if spills occur that prevent animals from moving via
their historical pathways, etc. Making this point explicit would improve the document.
c. Figure 10-5, Page 10-11 was a useful one demonstrating the overlap between the road
and Sockeye Salmon. This is a much improved presentation from the first draft.
d. Page 10-26. Given the very high failure rate of culverts, would it make sense to install
more bridges or more arch-only culverts to protect salmon populations? It just seems
that traditional culverts are a poor alternative to bridges and arched culverts.
e. General Point. Throughout the text (see first complete paragraph on page 10-37
dealing with what we might expect with invasive species coming to the environment
with the building of the road, as just one example), both in this chapter and others, the
authors have demonstrated a knowledge of the recent literature which is used
throughout to justify their assumptions, confidence, and risks. This allowed me to
9
-------
Dr. Roy Stein
appreciate and evaluate with a more discerning eye the product of this ecological
assessment. Nice work.
f. Calculations of dust production on pages 10-35 and 10-36 reflect a nice quantitative,
with appropriate caveats, assessment, demonstrating that dust could have a real impact
on the salmonid streams in close proximity to the road.
g. Pages 10-40 to 10-42. I liked the uncertainty section that dealt with the transportation
corridor; issues that I did not anticipate were covered nicely.
13. Chapter 11: Pipeline Failures.
a. Pages 11-2 to 11-4. These conceptual figures are useful (throughout the document) to
help the reader envision the majority of the interactions within the context of pipeline
failures.
b. Pagell-5. I liked how the authors quantitatively developed their best estimates for the
probability of pipeline failures, based on similar situations with other mines and with
other pipelines. In turn, the authors typically laid out the caveats re the comparative
nature of the data: was the pipeline similar, did it exist in a similar environment, was it
designed with Best Management Practices in mind, were there other issues that might
allow estimates of pipeline failure to be higher or lower than what might be expected at
the Pebble Mine?. This sort of quantitative, transparent analysis gives the reader faith
that the EPA has no hidden agenda, no ax to grind, and no previously decided-upon
conclusions. I applaud the authors for these revisions.
c. Pagell-6. The whole idea of human mistakes or accidents is intriguing for this suggests
that no matter how good the infrastructure might be, we still will have human error and
it looks to be quite a prevalent cause of failures in the case histories of mines presented.
What I reflected upon was how one can incorporate this knowledge into current
measures of pipeline failures. The authors did not separate human error from
infrastructure failure and that indeed might be quite a useful exercise. Indeed, how do
mining companies work to reduce this source of failure? Training? A safety-first
attitude that includes strong provisions for reinforcing the idea of doing things "right"
in the field, even when time is of utmost importance. And I guess it would be
modification of worker mind-set that would replace "time" as the most important driver
to "infrastructure protection" as the most important one. I would very much like to see
what Northern Dynasty has to say about this issue. Finally, I would have liked to have
seen a discussion of "human error" in all of the sections dealing with failure. Is this as
prominent an issue with regard to the TSF as it is with pipelines?
d. Page 11-15. Though I have seen just a few typos, there is one on page 11-15 that
requires correcting...next to the last paragraph reports 14,00 gallons when it is likely
that this number should be 14,000 gallons. See also a typo in the first paragraph on
page 11-26. See page 11-30, last complete paragraph on the page.
e. Page 11-18. During the course of this chapter, the authors make the point that release
of contaminants into the environment would be more amenable to mitigation at low
flows rather high ones. True, but the authors fail to point out that "dilution" will be
more of a factor at high flows and thereby may reduce or even negate contaminant
impacts. Some recognition of this in the text would improve the presentation.
f. Page 11-19, second bullet point at top of page. Why choose the 5-min spill duration and
not the more common, 11.5 min, as suggested by O'Brien's Response Management
10
-------
Dr. Roy Stein
publication? It certainly seems this would be more defensible than arguing as the
authors do, that their estimates may underestimate the impact of spills on the
environment.
g. Page 11-20. The whole idea of diesel fuel being so variable relative to its constituents
bothered me some. If this is the case, can no one in a position to review permits judge
its composition and therefore judge its potential impact on the environment? This
section was unsatisfying to me re how a failure of this pipeline might influence nearby
wetlands, streams, rivers, ponds, and lakes.
h. Table 11-10, Page 11-3. Though I liked tables of this sort, I became just a little confused
by the nomenclature, what with a "positive" response having negative consequences. I
think I did get it figured out and I may even now understand your usage here, I worry a
little about how a lay audience or even Northern Dynasty will interpret this text. One
example of this confusion can be found on page 11-29, last paragraph. Note the quote
from the text, "Overall, the available lines of evidence for effects of a diesel spill are
positive for the occurrence of acute toxic effects (Table 11-10)." Really? I trust the
authors get my concern here.
i. General Point. One piece of information that I would like to see added to this draft is
some sort of time period for recovery of the negative impacts of a spill. Most studies
(being Masters theses) typically study these impacts for 1-2 years and by this time, 80%
or so of the macroinvertebrates are back (but maybe not the diversity) and fish have
recovered but not completely. If we think about extrapolating these studies out
(certainly there must be one that goes beyond just a year or two), then when would we
expect full recovery? This is an extremely important point and may provide insight into
longer-term impacts of a spill when the effects of the contaminants are relatively short-
lived (i.e., much, much, much shorter than what is predicted for a TSF breach, for
example). Contrasting impacts of pipeline failure with other infrastructure failures
would lend more credence to the work of the EPA authors.
14. Fish-Mediated Effects on Native Alaskan Culture
a. Page 12-1. The title of the chapter seems to me to be incomplete, in other words, I think
the title should be "Fish-mediated effects on Native Alaskan Culture". This better
encompasses the content of the chapter.
b. Page 12-1. I would like the authors to add about a page of text to set the stage for
Native Alaskans, e.g., a culture that is 4,000 years old, a sustainable culture dependent
on subsistence harvesting ( different than a one-time exploitation by a mining
company), the fact that the jobs generated will only last for a generation or a
generation and one half...far less than 100 years. This text would then set the stage for
the chapter and better provide context for this examination as well as the purpose of
this chapter (as stated on page 12-1).
c. Page 12-6. Throughout this chapter I noted the many "trade-offs" that are inherent in
the establishment of a mine of this type and size. Cash economy versus subsistence,
short-term rewards versus long-term ones, culture-enhanced activities versus culture-
negating ones, split-time activities (i.e., part-time work) versus full-time subsistence
activities, etc. Some recognition of this might improve the message being sent with this
chapter.
11
-------
Dr. Roy Stein
d. Page 12-9. The idea that the mine will come in, extract valuable minerals, and then
leave occurs throughout but I think it is insufficiently emphasized that the physical
degradation to the environment stays...typically in perpetuity. Yes, the text makes this
point but it is somewhat buried, in my view. Can the authors increase its impact and
thereby its visibility?
e. Page 12-10, first complete paragraph. The fact that small changes to movement of
terrestrial organisms can have dramatic impacts on the Native Alaskan culture is a good
one. Like the previous point, I would like a bit more explanation as well as more detail,
coupled with a historical example or two. This information would increase the power
of this paragraph.
f. Page 12-11 and 12-12. These are powerful arguments for the whole idea of job
enticement for Native Alaskans, running the gamut from bigotry to compromising
subsistence. This is an impressive argument and takes much away from the argument
that this mining operation will bring a much-needed (in the eyes of the mine operators)
cash jobs to a "depressed" region of the state. Indeed, based on the case history of
North Slope, less than 1% of the work force was made up of Inupiat. These points are
made almost too dispassionately in text (even recognizing that EPA is attempting to
maintain a "neutral" voice throughout the report).
g. Page 12-17. I am not sure how we manage to emphasize even more the confounding
effects of mining and climate change on subsistence fisheries by Native Alaskans for I
think this is extremely important and (as alluded to in text) extremely difficult to
forecast. Couple these effects with the "portfolio" features of the salmonid populations,
and I think we have real reason for concern, i.e., while these seven species of salmonids
have a full range of life history strategies and flexibilities evolutionarily adapted to the
historical environmental variation in the Bristol Bay Watershed, nothing in their history
(or their portfolio) prepares them for either the mining or the climate-change impacts
on the environment. Consequently, even given their tremendous life history variation,
these salmonids are ill-prepared (evolutionarily speaking) for these future changes.
Life history specialists will be completely lost due to mining and its complicated
interactions with climate change.
15. Chapter 13: Cumulative Effects of Large-Scale Mining
a. Page 13-2, Table 13-1. Could the authors add approximate sizes of these mines that
could be developed in the future? I appreciate the fact that the total area, at 1,300 km2,
is provided on page 13-7.
b. Page 13-5. Once again the conceptual model does a nice job of setting the stage for the
sorts of interactions that would likely occur following the initial establishment of the
Pebble Mine in the Bristol Bay Watershed. After a review of this model, the reader is
then prepared for the text to follow.
c. Page 13-7, first paragraph under 13.2. Impossible to predict the future of mine
development, including the order of mines that might come online and their respective
impacts. It is, however, safe to say that additional development of mines will indeed
occur if Pebble is permitted to move forward; this is certainly inevitable. Too many
caveats here may provide fodder for suggesting that the EPA does not know that a
mining district will develop. What would help us here is what has been done in
12
-------
Dr. Roy Stein
previous chapters, i.e., the documentation of case histories where initial mining
ultimately led to a mining district.
d. Page 13-8. Median mine size of 0.25 Pebble is assumed in the context of predicting the
impacts of new mines. Can this assumption be justified a bit more? Do the authors
sense that this is a conservative estimate for new mines, a reasonable one (this is the
sense that I get from the tone of the text), or a liberal estimate?
e. Page 13-9, Box 13-1. Breaking out methods in boxes like this has served this
document well, making it easy for the reader to capture the essence of the conclusions
without getting bogged down in too many methodological details.
f. Tables 13-2 to 13-8. These tables provide a nice combination of qualitative and
quantitative data to support conclusions generated for cumulative impacts of additional
mining development. Good work with limited information.
g. Page 13-23, Big Chunk North, as just one example. I like the fact that the authors did
multiple scenarios when they were unsure as to whether a future mine would use an
already established TSF or one that the future mine would build on its site. In this
fashion, the EPA has bounded the impact of future development, allowing the reader to
see both minimal and maximal effects of new mines, a technique that could
productively be used elsewhere in this chapter.
h. General Point re Chapter 13. On the one hand, the writing here is dispassionate and
objective, as I presume it should be with scientific writing. Even so, I would like the
authors to make some judgment calls re the impacts of these six different mines on the
environment. Those that have high densities of streams, wetlands, and ponds would
undoubtedly experience greater degradation to the salmonids than one that has a far
lower density (is this true?). In turn, one could ask if there are other characteristics of
the environment that make one mine worse in terms of its negative impact than
another. Bringing out these comparative statements would help the reader understand
the individual impacts of new mines. As just one other example of this issue, it is often
noted that streams are "strongly meandering"; one additional statement that would
help the reader would argue something along the lines of "making road development
far more difficult and likely to have a greater impact on salmonid populations".
i. Page 13-28, third paragraph. I think that the last sentence of this paragraph requires
some modification for it now reads "...diminishing the biological complexity of salmon
stocks and lead to the portfolio effect...". The fact is that because of the diversity of life
histories within species across populations within stocks, salmonids do well even as
environmental conditions change from year to year. Here the authors are arguing in
this last sentence that reductions in biological complexity will "lead to the portfolio
effect" but rather this decrease will compromise the portfolio effect by eliminating
stocks with specific life history characteristics. Cleaning up this text will improve its
presentation.
j. Page 13-29, Impervious Surface Increase. No citations are provided here. Do we know
anything about 10% being a useful metric for increasing negative impacts on stream
flow and flooding?
k. Page 13-35. Indeed, we now require salmon hatcheries in some urbanized areas within
Alaska. Is this point worth emphasizing? Are the hatcheries really required because
the nearshore environment has been degraded? A box detailing this example with
13
-------
Dr. Roy Stein
some hard data might go a long way toward making the point that development means
loss of important locally adapted salmon stocks.
1. General Point re Chapter 13. I am not sure how to handle this concern. When the
authors discuss the various impacts of different infrastructural features in a cumulative
sense, they do not review all of the issues each contributes. For example, there is no
mention of dust or invasive species when discussing roads. Similarly, when discussing
pipelines, nothing is said about accidents. The authors certainly cannot be expected to
review all features of all of these infrastructural impacts, but some recognition of all of
their impacts might be warranted, perhaps as a sidebar or briefly highlighted in an
associated box (as done so successfully in previous chapters).
16. Chapter 14: Integrated Risk Characterization
a. Page 14-3. Conventional mining practices may be sufficient for the smallest footprint
for the Pebble Mine, i.e., 0.25 but beyond this size, conventional mining practices do not
protect the environment to the extent required to protect salmonids, Hence the
commercial and sport fisheries, and specifically the subsistence fisheries by Native
Alaskans are threatened. Given this conclusion, permitting of this mine would be
suspect at best.
17. EPA Response Document.
a. General Point I. Overall, I was impressed with how the team of authors dealt with our
reviews. For each point, there was succinct, explicit, and reasoned response. Not
always did our concerns receive full treatment in the revision. However, if they did not,
the EPA response included an explanation that was well crafted and justified.
b. General Point II. Because there are no data on the management of water quality "In
Perpetuity" (i.e., we have no experiences in this area simply because we do not have
mines that have been closed for a sufficiently long period of time (>100 years) to gain
this experience), the EPA cannot write about what might be expected in the "long term".
Even so, no one is optimistic. Without the profit incentive, corporate entities are not
likely to monitor, treat, and maintain these facilities. Indeed, corporations are likely to
go out of business in their current structure and therefore this responsibility will fall to
the State of Alaska, whose record is not the most enviable in this arena. It seems to me
that TSFs will fail, given enough changes in responsibility over time. Impacts on the
salmon, the subsistence Native Alaskan culture, the commercial/sport fisheries, and the
entire ecosystem will be large, long-lasting (at least multiple hundreds of years), and
not preventable. With this certain knowledge in hand, I can only conclude that the
Pebble copper mine and Bristol Bay salmon cannot coexist. The mine will indeed
eliminate the salmon runs in the areas affected by a catastrophic failure of the Tailings
Storage Facility. This mine should not be built in the Bristol Bay Watershed for it will
have nearly irreparable impacts on the water quality of this watershed.
c. General Point III. Though we do not have long-term data from any mines post-closure,
what do the short-term data tell us? They tell us that insidious effects of the facilities
that remain after mining compromise water quality to the extent that it influences
ecosystem functioning eliminates the ability of streams affected to support fishes
intolerant of poor-quality water. We have experience that tells us that this has occurred
and thus has an extremely high probability of re-occurrence in the Pebble area.
14
-------
Dr. Roy Stein
d. General Point IV. Even though we have the experiences noted in items 17a-b, the
Bristol Bay Watershed is even more vulnerable that the other systems studied, in that it
is an extremely wet environment with ground-water operating to communicate among
streams (inter-connectedness). With this inter-connectedness come the increasing
widespread effects of mine pollution and thus an increased impact on the whole of the
Bristol Bay Watershed.
e. General Point V. The Nushagak and Kvichak rivers are exceptional salmonid streams,
even for the Bristol Bay Watershed. Loss of these populations would likely negatively
influence salmon populations that are available for harvest. That the headwaters of
these two rivers lie in the middle of the Pebble Mine impacts argues for a precautionary
approach to the permitting process.
f. Page 13, EPA Response to Peer Review Report. I agree with the team of authors when
they suggest that including the Mount St Helens data re recovery of stream
communities makes sense. At present, they have removed these data; I would like to
see them included as they can help make the case re the impact of a TSF failure on
Bristol Bay Watershed streams.
g. Some inconsistency exists within the "Response Document". On page 7 the team argues
that development of a monitoring program to distinguish between mining and climate
change effects is outside the scope of the assessment". Yet, on page 44, the authors
argue that"... Chapter 3 and Box 14-2 include a discussion of the need for future
monitoring to differentiate climate change effects from large-mining effects". Clearly,
Box 14-2 seeks to present the importance of the monitoring program in light of the
comparative effects of mining versus global climate change. The text of the "Response
Document" should be revised to reflect these goals to maintain consistency.
h. Page 63, third full paragraph, Response Document. I am surprised that the answer to
this question does not include the reference to Box 4.1 in the Second Draft that defines
all of the relevant terms. For the sake of completeness, this box should be referenced
here.
i. As I read through the Response Document (see top of page 79 as just one example of
many), I am troubled by the ongoing conversation that suggests that we might know too
little specific information about the salmonids in the two river systems affected that the
mine proponents can argue that they will have a minimal impact, thus allowing the
mine and its operations to be permitted by the State of Alaska. I have no solution to this
conundrum, but it bothers me that the burden of proof lies with the EPA and other
regulatory bodies to demonstrate negative impacts on the extant salmonid populations.
I do wish the system was reversed, i.e., that the company proposing this development
would be required to demonstrate "no effect" on the salmonid populations before
permitting could occur.
j. Page 100, Response Document. I agree that the EPA has substantially strengthened the
arguments associated with "future technology" as being the solution to all of the issues
that lay before Pebble LLC in Chapter 9. These arguments are critically important to the
assessment of whether the mine will indeed compromise water quality and therefore
salmonid success. To allow the company to argue that historical failure rates of various
portions of the mine are not appropriate because newer technology will prevent these
failures is simply not right, i.e., in the context of any evaluation of the environmental
impact of a proposed mining facility.
15
-------
Dr. Roy Stein
k. Page 130, Response Document. Under comments about the transportation corridor, it
is appropriate to note that 70 crossings have been changed in the revised document to a
minimum of 70 crossings. This particularly seems pertinent to these responses to the
Expert Panel concerns as expressed.
1. Question 10 dealing with risks to wildlife and human cultures, page 139, Response
Document. To this point, I did think that the responses by EPA were complete and well-
thought-out. Reponses to the critique by the Expert Panel to Question 10 were less
well developed without (typically) references to specific passages in specific chapters. I
believe the EPA authors handled most all of the concerns expressed herein, especially
so after sharpening the scope of their Ecological Assessment. Even so, the EPA failed to
explicitly lay out exactly how they handled each of the concerns expressed by the
Expert Panel, which is too bad because I think that they did deal reasonably well with
these issues in their revised draft. Note just a few examples:
i. Page 141, top of page. In the first response, there are no references to the
revised text.
ii. Page 141, middle of page. In the second and third responses on this page,
little detail is provided; the EPA authors only refer to Chapter 12 when
indeed they could have provided an insightful short paragraph that would
explain precisely how they handled these issues (for they did, as I recall, in
the revised text).
iii. Page 151, second response on the page. Here a more complete answer would
have included references to all of the NEW conceptual models that were
developed and discussed in the revised Second Draft.
iv. Given what I have read to this point re responses, I was disappointed in the
this section re EPA responses...for they are rather perfunctory rather than
explicit and detailed, and this is especially true for the responses to Dr. Paul
Whitney's comments and concerns.
v. Finally, one can appreciate my concern by simply comparing EPA responses
to concerns expressed by the Expert Panel regarding Question 10 versus
Question 11.
m. Page 177, Response Document. I appreciate the response the authors have generated
relative to the ideas surrounding uncertainty. Their revised draft considerably
improves their presentation of the topic and the response on the middle of page 177
helps me understand their point of view with regard to uncertainty (see response
comment on page 178 where the EPA argues that putting forward "reasonable and
typical scenarios" to characterize risk and then the uncertainty surrounding that risk (I
agree with this approach)).
18. General comment on my comments. All comments were seriously considered by the
EPA authors and were either appropriately dealt with in their revised text or explained that
the concern did not fall under the "scope of the ecological assessment". Thus, I am pleased
with the revision as it relates to my initial concerns about the first draft of the Main Report.
Though I cannot speak for my colleagues as to the adequacy of EPA's response to their
concerns, my reading of the Response Document demonstrates that I found most all of
EPA's responses well-reasoned, consistent across questions and concerns, and overall well
done (save for Question 10, see comments above).
16
-------
Dr. Roy Stein
Thank you for the opportunity to review EPA's Second External Review Draft, EPA 910-R-12-
004Ba, April 2013 entitled An Assessment of Potential Mining Impacts on Salmon Ecosystems of
Bristol Bay, Alaska. I trust my comments will be useful as EPA finalizes its assessment of
potential mining effects in the Bristol Bay Watershed.
Sincerely,
Roy A. Stein, Professor Emeritus
Department of Evolution, Ecology, and Organismal Biology
The Ohio State University
17
-------
Dr. William Stubblefield, Oregon State University
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Stubblefield, William rmailto:Bill.Stubblefield@oreaonstate.edul
Sent: Thursday, August 01, 2013 5:43 AM
To: Thomas, Jenny
Subject: Stubblefield comments
Jenny
I was not entirely sure about the form that you wanted this document in. I tried to follow your guidance
from the other day. If this is not appropriate or does not address all of the aspects that were requested, let
me know and I will revise it accordingly.
Cheers
Bill
William Stubblefield
Dept of Environmental and Molecular Toxicology
Oregon State University
1007 Agriculture and Life Sciences Bldg
Corvallis, OR 97331
Bill.Stubblefield@oreqonstate.edu
-------
Dr. William Stubblefield
Evaluation of EPA responses to external reviewer (William Stubblefield, PhD)
comments:
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay,
Alaska (April 2013)
General Impressions:
Based on my review of the second external review draft of the Bristol Bay assessment
document it is my impression that the USEPA has committed substantial time and effort to
revising the original assessment document resulting in a greatly improved report. For the most
part, all of the technical and editorial comments made regarding the original document have
been satisfactorily addressed in the current revision. Substantial improvements in the
organization and content of the document were made including:
• Incorporation of background information regarding the use of a risk assessment-based
approach to addressing the issues,
• Inclusion of an explanation for the inception of the assessment and its ultimate utility to
risk decision-makers,
• Addition of a clear discussion of the uncertainties associated with the assessment and
the potential implications of those uncertainties, and
• Inclusion of an evaluation and application of the "state-of-the-science" regarding the
assessment of metals toxicity in the aquatic environment, inclusion of "new" European
REACH regulation-based data, and the importance of consideration of bioavailability
have been addressed.
Although additional changes and incorporation of additional "new" data and recent scientific
literature would further benefit the document, I think, in its current form, it more fairly
represents the current "state-of-the-science." Therefore, I would commend the USEPA for its
efforts in revising the original assessment report.
-------
Dr. William Stubblefield
WRITTEN PEER REVIEW COMMENTS
1. GENERAL IMPRESSIONS
The document, "An Assessment Of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska," is a well-written, comprehensive document that employs a risk assessment-type
approach to an a priori evaluation of potential environmental effects on the ecosystem and
potential receptor species (e.g., salmon) that may be affected by a potential copper mine located
in the Bristol Bay area of Alaska. This document is somewhat unique, in that no actual mine has
been proposed at the location and few site- or project-specific data are available. Therefore, no
specific information about development plans and potential operational and closure activities
associated with the mine are available. Rather, the authors have attempted to develop a
hypothetical mine and attempted to assess possible environmental effects associated with mine
development, operation, and closure. Although interesting, the potential reality of the assessment
is somewhat questionable. It is also unclear why EPA undertook this evaluation, given that a
more realistic assessment could probably have been conducted once an actual mine was
proposed and greater detail about operational parameters available. The approach taken in the
document attempted to be comprehensive and evaluated a variety of scenarios that may affect
aquatic resources in the Bristol Bay region. Given the importance of salmon populations in the
area, both from a financial and societal perspective, it is important that a comprehensive
evaluation of potential environmental effects associated with mine development and operations
be conducted. The authors have attempted to conduct such a comprehensive evaluation and have
attempted to quantify (to the extent possible) the probability of adverse effects occurring.
Implementation of this approach is proper, and with the correct data, can provide a
comprehensive evaluation of potential environmental effects. Unfortunately, because of the
hypothetical nature of the approach employed, the uncertainty associated with the assessment,
and therefore the utility of the assessment, is questionable.
RESPONSE: The EPA respectfully disagrees that the hypothetical nature of the approach
compromises the utility of the assessment. All mining plans are hypothetical They change in
response to the results of assessments, regulatory requirements, public input, and unforeseen
conditions and events. They cease to be hypothetical only after the mine is closed At every step
in the process, assessments of the current plan are useful even though plans will change. This
assessment is based largely on a preliminary plan, published by Northern Dynasty Minerals
(Ghaffari et al. 2011). Although layout of mining components in a future mine plan may differ
somewhat from the preliminary plan or the EPA scenarios, the main components of mining
would remain the same for open-pit mining (and underground mining would face the same
waste issues).
It would appear that on this topic the reviewer and the US EPA authors are likely to
continue to disagree. The hypothetical nature of the approach presented in the original
assessment document was sufficiently vague that the degree of confidence in the
assessment may limit its utility to risk decision-makers due to the high degree of uncertainty
associated with the evaluation. By increasing the information upon which environmental
exposure concentrations and environmental effects concentrations are estimated and
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 2
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
taking into account site-specific environmental conditions, a more robust assessment with
less uncertainty can be developed. This does not suggest that the original "screening level"
approach has no utility; it merely suggests that the degree of certainty associated with the
assessment may not have been sufficient to provide risk decision-makers with sufficient
information upon which to make long-term project decisions.
A variety of uncertainties and data needs were identified as a result of this effort and this alone
may provide sufficient value to justify the document and approach. For example, the authors note
that there is not an abundance of chronic toxicity data considered in deriving the EPA's ambient
water quality criteria for copper and that there is an uncertainty associated with whether the
biotic ligand model (BLM) adequately protects species of concern in Bristol Bay. It would seem
appropriate for EPA (perhaps in concert with industry) to develop the data to improve our
understanding of copper toxicity and to ensure that regulatory standards are, in fact, appropriate
for their intended use. A substantial body of data evaluating copper chronic toxicity has been
developed by the copper industry as a result of regulatory requirements driven by the European
REACH regulations. It may be beneficial for EPA to examine these data, thus resulting in a
reduction in any uncertainty associated with the evaluation of environmentally acceptable metals
concentrations. It should also be noted that similar datasets and biotic ligand models exist for
number of other metals that may be of concern at the Bristol Bay site.
RESPONSE: The EPA has examined the EU's 2008 Voluntary Risk Assessment of Copper
(the relevant REACH document). Although they do derive a chronic species sensitivity
distribution, it is because of the way they include and aggregate data, rather than the
generation of new data. In particular, they have no data for sensitive aquatic insects, so the
EU does not resolve that problem. The BLM was used for copper because copper is the
contaminant of greatest concern and because the copper BLM has been approved by the EPA
Office of Water. Other metals with BLMs, such as zinc and nickel, occur at much lower levels
in leachates.
The EU's copper voluntary risk assessment does, in fact, include additional data than that
considered in the US EPA's original copper Ambient Water Quality Criteria. For example,
data for freshwater mussels represent a "new" species that was included in the EU's
document. The assessment authors are correct, however, in that new data are not available
for additional insect species, especially, representatives from Ephemeroptera, Plecoptera or
Tricoptera orders (i.e., EPT species). Field population data suggest that these organisms
may have greater sensitivity to the effects of copper than those currently considered in the
existing species sensitivity distribution. Currently no standardized (i.e., OECD, ASTM or EPA)
laboratory test methodologies exist for conducting chronic toxicity test with EPT species. US
EPA may want to consider the possibility of conducting or funding a research program
aimed at the development of chronic tests methodologies for these species in order to
assure that they are adequately protected by AWQC.
One suggestion that would improve the document is that EPA should include a basic description
of the risk assessment process and the relationship between the risk assessor and the risk
manager, i.e., the decision maker. They must include a discussion of why the assessment is
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 3
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
being conducted, the decisions that will be informed, and what information they need from the
risk assessor.
RESPONSE: Additional contextual information for the assessment has been included in
Chapter 1, and additional information on ecological risk assessment has been incorporated
into Chapters 1 and 2. The assessment has also been restructured into problem formulation
and risk analysis and characterization sections, to make the assessment's structure as an
ecological risk assessment clearer.
Substantial improvements were made in the revised assessment to address the concerns
raised in my initial review comments. Through a combination of revision, reorganization,
and expansion in Chapters 1 and 2, the authors have provided the appropriate background
and description of the approach used in the preparation of this assessment.
Taken from the USEPA's Guidelines for Ecological Risk Assessment (EPA630/R-95/002F; April
1998). Note 2nd sentence re: the role of the risk manager.
"2.1. THE ROLES OF RISK MANAGERS, RISK ASSESSORS, AND INTERESTED
PARTIES IN PIANNING
During the planning dialogue, risk managers and risk assessors each bring important
perspective to the table. Risk managers, charged with protecting human health and the
environment, help ensure that risk assessments provide information relevant to their
decisions by describing why the risk assessment is needed, what decisions it will
influence, and what they want to receive from the risk assessor. It is also helpful for
managers to consider and communicate problems they have encountered in the past when
trying to use risk assessments for decision making.
In turn, risk assessors ensure that scientific information is effectively used to address ecological
and management concerns. Risk assessors describe what they can provide to the risk manager,
where problems are likely to occur, and where uncertainty may be problematic. In addition, risk
assessors may provide insights to risk managers about alternative management options likely to
achieve stated goals because the options are ecologically grounded. "
RESPONSE: Section 1.2 in the revised assessment discusses uses of the assessment.
I concur, the revised Chapter 1 adequately describes the purpose and use of the
assessment.
2. RESPONSES TO CHARGE QUESTIONS
Question 1. The EPA's assessment focused on identifying the impacts ofpotential
future large-scale mining to the fish habitat and populations in these watersheds. The
assessment brought together information to characterize the ecological, geological, and
cultural resources of the Nushagak and Kvichak watersheds. Did this characterization
provide appropriate background information for the assessment? Was this
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 4
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
characterization accurate? Were any significant literature, reports, or data missed that
would be useful to complete this characterization, and if so what are they?
The EPA's assessment document presents a seemingly comprehensive compilation of the data
associated with the ecological, geological, economic, and cultural resources of the Bristol Bay
area. The characterization as presented seems to provide appropriate background information for
the assessment considering the hypothetical nature of the evaluation. Without having specific
knowledge of the area in question, it is not possible to provide an assessment as to whether the
characterization was accurate. I'm unaware of significant literature, reports, or data that were
specific to the site and would be useful for consideration. The assessment should be expanded to
include greater detail regarding the environmental aspects of the site.
RESPONSE: Additional information on the physical environment of the region and
assessment endpoints has been incorporated into Chapters 3 and 5.
I agree, substantial additional information on the physical environment of the region in the
assessment endpoints has been incorporated into Chapters 3 and 5. This helps in the overall
description of the Bristol Bay environment and helps characterize potential environmental
concerns.
Question 2. A formal mine plan or application is not available for the porphyry copper
deposits in the Bristol Bay watershed. EPA developed a hypothetical mine scenario for
its risk assessment, based largely on a plan published by Northern Dynasty Minerals.
Given the type and location of copper deposits in the watershed, was this hypothetical
mine scenario realistic and sufficient for the assessment? Has EPA appropriately
bounded the magnitude of potential mine activities with the minimum and maximum
mine sizes used in the scenario? Are there significant literature, reports, or data not
referenced that would be useful to refine the mine scenario, and if so what are they?
No comments on this question.
No additional comments on this question
Question 3. EPA assumed two potential modes for mining operations: a no-failure
mode of operation and a mode involving one or more types offailures. Is the no-failure
mode of operation adequately described? Are engineering and mitigation practices
sufficiently detailed, reasonable, and consistent? Are significant literature, reports, or
data not referenced that would be useful to refine these scenarios, and if so what are
they?
It is interesting and appropriate that the EPA has included both modes of operation in conducting
this assessment. This approach provides some degree of "bounding" for the assessment;
however, the degree of accuracy (i.e., predictability) for either scenario cannot be known at this
time. The document appropriately acknowledges that there are a variety of potential mitigating
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 5
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
factors (e.g., acts of God, accidents, market changes) that may render the assumptions used in
this assessment incorrect.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed. There were a
number of comments on this approach so the revised assessment no longer uses the term "no-
failure" and addresses effects from the foot print (i.e., the no failure scenario), failure
scenarios having higher probability and lesser magnitude (e.g., failure to collect or treat
leachate water), and failure scenarios having lower probability and higher magnitude (e.g.,
TSF failure).
No change suggested or required.
I agree with the new approach suggested by EPA.
Question 4. Are the potential risks to salmonid fish due to habitat loss and
modification and changes in hydrology and water quality appropriately characterized
and described for the no-failure mode of operation? Does the assessment appropriately
describe the scale and extent of risks to salmonid fish due to operation of a
transportation corridor under the no-failure mode of operation?
The document appears to adequately address potential questions associated with habitat loss due
to hydrologic changes, especially considering the hypothetical nature of the mine and the lack of
specific detailed information regarding an actual proposed facility and all of the associated
operational details of the facility. The assessment of potential impacts and ecosystem protection
parameters is predominately based upon the publication of Richter et al. (2011). Additional
support and evaluation of these recommendations for fisheries populations in the Bristol Bay
area should be closely evaluated.
RESPONSE: Prompted by this comment, we consulted with regional biologists and
hydrologists to evaluate the suitability of the sustainability boundary approach for flows. We
asked them if there was any reason that fish populations in these streams, or the specific
hydrology of the area, made it exceptional with regard to this approach (e.g., was there any
reason to think that the Richter approach was not applicable here). We received uniform
support for applying this approach to Bristol Bay streams. We strengthen our emphasis that
this is a precautionary approach, and that the detailed hydrologic and habitat modeling work
that PLP contractors have begun will help provide a useful basis for more sophisticated flow-
habitat modeling.
This response seems to adequately address the initial comment.
Question 5. Do the failures outlined in the assessment reasonably represent potential
system failures that could occur at a mine of the type and size outlined in the mine
scenario? Is there a significant type of failure that is not described? Are the
probabilities and risks of failures estimated appropriately? Is appropriate information
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 6
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
from existing mines used to identify and estimate types and specific failure risks? If
not, which existing mines might be relevant for estimating potential mining activities in
the Bristol Bay watershed?
The scope of failures described in the assessment seems to be sufficiently comprehensive and all
likely failure-types are considered. The probabilities and risks for failure seem to be adequately
estimated, given the state-of-the-science; however, these estimates are likely to be very sensitive
to site-specific concerns and operational considerations. Once site-specific information is
available, it is likely that much better estimates of failure potential at a site can be developed.
RESPONSE: We agree that more site-specific information could support more site-specific
estimates. No change required.
I concur, the US EPA authors have acknowledged the benefit of more site-specific information
and until that information becomes available there is little additional that can be done to
estimate failure potential.
Question 6. Does the assessment appropriately characterize risks to salmonid fish due
to a potential failure of water and leachate collection and treatment from the mine site?
If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
The risk assessment attempts to consider the effects of metal discharges for water and leachate
from the mine site. This assessment is based on metals concentrations measured in potentially
"similar" mine waters from other sites; concentrations of metals are likely to differ based on
source material and operational differences. The effects concentrations used in the evaluation
are based on US EPA ambient water quality criteria (AWQC) for metals, and this approach is
appropriate for "screening level" evaluations. It should be noted that exceedence of an AWQC
does not portend the occurrence of adverse effects. Ambient water quality criteria are derived in
such a way that they are intended to represent "safe concentrations." In other words, if
environmental concentrations remain below the AWQC, it is assumed that unacceptable adverse
effects will not occur; exceedence of an AWQC suggests that adverse effects may occur to some
species, but that this must be evaluated more closely. Salmonid species are not the most
sensitive organisms in the copper AWQC species sensitivity distribution (SSD); therefore, direct
effects on salmon are even less likely at concentrations in the range of the AWQC.
RESPONSE: Effluents or ambient waters from mines at other sites were not used. The
leachate concentrations used (except for the product concentrate leachate) are from available
results of material leaching tests from the Pebble deposit. Otherwise, the Agency agrees with
these comments. The assessment used criteria for screening, but then examines the toxicity
data more closely, including field data to determine potential effects (e.g., aversion, sensory
inhibition, mortality and reduced reproductive success of salmonid fish). The greater
sensitivity of aquatic insects was described in the May 2012 draft of the assessment and is
further highlighted in the revised assessment. The protectiveness of the copper criterion is
considered in both the original and revised assessments.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 7
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
Section 8.2.2.1 provides a good discussion of the copper standard used in conducting the Bristol
Bay assessment. It describes the existing state standard, the incorporation of the biotic ligand
model calculations and its effect on site-specific copper criteria, the importance of
"nonstandard" endpoints (e.g., olfactory concerns), and the concerns that the US EPA has about
the "protectiveness" of the existing criteria. The authors identify a group of recent scientific
publications that have "identified effects on aquatic insect populations and invertebrate
communities at concentrations below water quality criteria for the dominant metals (cadmium
copper and zinc)." This is an intriguing finding that may run counter to previous assumptions
regarding the protection afforded by water quality criteria. It could also have far-reaching
consequences for water quality criteria and their derivation procedures. Field-based
assessments are frequently difficult to evaluate due to confounding factors. Because of this
alternative approaches, such as mesocosms, are frequently used to assess community or
population level concerns under more "controlled" environmental conditions. A number of
mesocosm studies have been conducted with copper in recent years and these studies should
be evaluated and considered in assessing the adequacy of the copper criteria.
One statement contained in Chapter 8, Section 8.2.2.1 suggests that the EPA may believe that
the AWQC for copper may be significantly underprotective. That statement indicates that:
"Based on the literature cited above and the author's experience, resolution of this uncertainty
by additional research and testing is likely to lower the chronic criterion by approximately a
factor of 2 and not as much as a factor of 10."
This is a concerning statement especially when one considers the ramifications if this is in fact
correct. For example, the table below presents the BLM copper criteria based on the mean
water chemistries in the mine scenario watersheds (taken from Table 8.11) with consideration
of a 2-10 fold "uncertainty" factor. These values are compared to background copper
concentrations in the same river segments (taken from Table 8.19). All values are in |-ig/L.
Stream
CMC
CMC
(2-10 factor)
CCC
CCC
(2-10 factor)
Background
range
South Fork Koktuli River
2.4
1.2-0.24
1.5
0.75-0.15
0.42-2.4
North Fork Koktuli River
1.7
0.85-0.17
1.1
0.55-0.11
0.31-0.61
Upper Talaric Creek
2.7
1.35-0.27
1.7
0.85-0.17
0.21-0.5
This assessment suggests that if the BLM chronic criteria concentrations are "under-protective"
by the suggested 2-10 factor, a significant portion of the streams would be in violation of the
copper criteria based on only background concentrations.
It is interesting to note that the risk assessment document states that copper is one of the "best-
supported criteria. However, it is always possible that it would not be protective in particular
cases due to unstudied conditions or responses." Further, the document goes on to suggest that
organisms such as mayflies etc. are important to the aquatic ecosystem but are not considered in
the copper AWQC and therefore may not be sufficiently protective. It also suggests that because
an acute-chronic ratio approach is employed to correct the final acute value to obtain a final
chronic value, there may be increased uncertainty associated with the protectiveness of the
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 8
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
chronic criterion. This appears to be an area where EPA might benefit by conducting research
(either alone or in concert with industry) to reduce uncertainty in the criteria to an acceptable
level. In addition, additional chronic toxicity data may be available from research conducted in
response to the European REACH regulations and consideration of this research may reduce the
level of uncertainty in the criteria. Bioavailability correction via the BLM approach is only
considered for copper in the risk assessment; biotic ligand models have been developed for a
number of metals (e.g., zinc, nickel) and these should be considered in the assessment as well.
Finally, the assessment approach seems to use a sum TU-based approach for assessing "metals
mixture" impacts. This is based on an assumption of additive interactions among the metals.
Although this is probably the best assumption in going forward, limited data are available to
support this approach.
RESPONSE: The EPA has examined the EU's 2008 Voluntary Risk Assessment of Copper
(the relevant REACH document). Although the authors could derive a chronic species
sensitivity distribution, that is because the way that they include and aggregate data differs
from the EPA '.s method. They apparently did not generate new test data for the assessment. In
particular, they have no data for sensitive aquatic insects, so the EU assessment does not
resolve that problem The BLM was usedfor copper because it is the contaminant of greatest
concern and because the copper BLM has been approved by the EPA Office of Water. Other
metals with BLMs, such as zinc and nickel, occur at lower levels in leachates relative to their
toxicities, so BLM modeling was not justified.
The US EPA authors are to be commended for taking the time to review the EU's voluntary risk
assessment of copper. This document represents the most recent evaluation of copper
environmental toxicity that this reviewer is aware of. Additional data that were generated as
part of the copper industry environmental research program are reflected in this document
(e.g., data for freshwater mussels). Differences do exist between the US EPA and EU approach
for endpoint assessment (EC10 vs EC20), aggregating data, and statistically deriving
"thresholds." However, the data are available to permit threshold calculation using either
method. EPA is to be commended for the use of the copper BLM in conducting this assessment;
it represents the current state-of-the-science and provides the most accurate method for
deriving thresholds considering site-specific physicochemical parameters. BLMs for other
metals such as zinc, nickel, and cobalt exist in the peer-reviewed scientific literature, but have
not yet gained regulatory acceptance by the EPA Office of Water. It is anticipated that this will
come with time. EPA did not consider the BLMs for these other "lessor" metals because of their
"lower levels in leachates relative to their toxicities." This may be a reasonable approach;
however, when using an additivity-based approach (ITU) when assessing metal mixtures, it may
be advisable to "correct" for bioavailability in even the "lessor" metals, when possible.
Areas where additional research would be beneficial include:
• Mixtures: Information regarding the potential interactive effects of multiple metal
exposures would be useful and would reduce assessment uncertainty.
• Species sensitivity concerns: there is extremely limited data (esp. chronic data) on all of
the salmon species of concern in Bristol Bay
• Additional data, especially chronic toxicity data and data for additional metals for which
no water quality criteria exists, would be extremely helpful.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 9
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
RESPONSE: The EPA agrees that these are good research topics. No change is suggested or
required.
To the extent possible, EPA may want to consider conducting these research programs or
funding others to do so. Also, based on the comments/revisions in the Bristol Bay assessment
there may be concern that the US EPA's current ambient water quality criteria for copper, and
perhaps other substances, do not adequately address potential effects to "sensitive" insect
species (e.g., Ephemeroptera, Plecoptera or Tricoptera orders). Currently no standardized (i.e.,
OECD, ASTM or EPA) laboratory test methodologies exist for conducting chronic toxicity test
with EPT species. EPA may want to develop methods that could be used to assess toxicity to
these organisms.
Question 7. Does the assessment appropriately characterize risks to salmonid fish due
to culvert failures along the transportation corridor? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so what are
they?
Potential effects on salmonid populations were evaluated due to culvert blockage and failures.
Culvert blockages will prevent salmon passage leading to possible effects on reproductive
success. Literature data for the incidence of culvert failures were used in assessing failure
probability. This seems to be an appropriate approach given the hypothetical nature of the mine
used in the assessment; however, this is not my area of expertise and I am not aware of
additional data that should be considered.
RESPONSE: No change suggested or required.
No additional comment on this question
Question 8. Does the assessment appropriately characterize risks to salmonid fish due
to pipeline failures? If not, what suggestions do you have for improving this part of the
assessment? Are significant literature, reports, or data not referenced that would be
useful to characterize these risks, and if so what are they?
Potential effects on salmonid populations were evaluated due to potential pipeline failures as part
of the risk assessment. This evaluation focused on potential failures associated with the pipelines
for the product concentrate slurry and return water. No consideration of the natural gas or diesel
pipelines was presented, stating that such pipelines "are common and the risks are well-known."
Although I would acknowledge the failures in natural gas and petroleum pipelines are common, I
would not discount the potential effects to salmon populations associated with such spills.
RESPONSE: A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11. Natural gas is not a contaminant of concern because it would vaporize and, at
worst, burn, which would not pose a significant risk to salmonidfish.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 10
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
The revised document evaluates the potential effects to aquatic organism populations as a
result of a petroleum diesel fuel spill. The document attempts to examine the available
literature and characterizing the acute and chronic effects of diesel fuel on aquatic organisms.
However, it is apparent is that there is little high quality information available that has been
gathered in any sort of consistent manner that would permit an adequate quantitative
evaluation of the effects of a diesel fuel spill. Inconsistent approaches in characterizing
hydrocarbon exposures (i.e., differing analytical methods) and inconsistent approaches in
conducting toxicity tests make it virtually impossible to apply a quantitative approach to the
evaluation of diesel fuel spill effects at this time. Hydrocarbon toxicity models such as the
Target Lipid Model (DiToro et al. 2000 a,b) or PETROTOX (Redman et al 2012) offer the best
approach to assess hydrocarbon mixture toxicity; however, substantial analytical
characterization of hydrocarbon concentrations are needed. For purposes of the Bristol Bay
assessment the approach taken is probably adequate given that conclusions can be reached
about the potential effects of a diesel spill. It should be recognized, however, that the state of
the science for evaluating hydrocarbon mixture toxicity is somewhat lacking and additional
research is needed.
Di Toro DM, McGrath JA, Hansen DJ. 2000a. Technical basis for narcotic chemicals and polycyclic aromatic
hydrocarbon criteria. I. Water and tissue. Environ Toxicol Chem 19:1951-1970.
Di Toro DM, McGrath JA. 2000b. Technical basis for narcotic chemicals and polycyclic aromatic hydrocarbon
criteria. II. Mixtures and sediments. Environ Toxicol Chem 19:1971-1982.
Evaluation of potential impacts due to a spill of product concentrate slurry or return water was
based on extant data from an existing copper mine in Sweden; to the extent that this slurry and
return water is representative of similar materials coming from the Pebble mine, this approach is
appropriate. The assumptions used in the amount of material that might possibly be spilled seems
appropriate and based on past experience and realistic assumptions; however, these assumptions
need to be reconsidered if and when a real mine plan is prepared.
RESPONSE: TheAitik leachate is no longer used to estimate the aqueous phase of the slurry,
in response to other comments. Because the slurry would be alkaline, appropriate analytical
data were obtained from Rio Tinto (Section 11.3.2.1). However, theAitik data are still used for
the leaching of the concentrate in a stream or wetland where neutral water would be the
leaching agent. No other relevant data are available.
As previously stated, to the extent that the slurry and return water used in the assessment is
representative of similar materials coming from the Pebble mine, this approach is appropriate.
Question 9. Does the assessment appropriately characterize risks to salmonid fish due
to a potential tailings dam failure? If not, what suggestions do you have for improving
this part of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
Potential effects on salmonid populations were evaluated due to tailings dam failures. Tailings
dam failure would potentially result in the release of large volumes of mine tailings and
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 11
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
associated contaminated waters, leading to possible acute and long-term effects on salmon
populations. It is also important to note that direct effects on salmon may be very species
dependent, due to life-cycle differences, and the time at which the dam failure occurs. Potential
effects due to sediment inundation/impaction can adversely affect habitat, leading to decreased
spawning. Evaluation of the potential for tailings dam failure effects considered acute and
chronic risks due to aqueous exposures, chronic risks due to sediment exposures, and risks due to
dietary exposures. All of these seem to be appropriate exposure pathways and all were
adequately considered, although site-specific information will improve risk predictions.
RESPONSE: Agreed. No change suggested or required.
No additional comments are provided.
Question 10. Does the assessment appropriately characterize risks to wildlife and
human cultures due to risks to fish? If not, what suggestions do you have for
improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
Potential effects to wildlife and human cultures are briefly addressed in the risk assessment. No
"quantitative" assessment of potential effects is provided. For the most part, it appears that
potential affects to both wildlife and human cultural endpoints are directly proportional to the
injury suffered by salmon populations as a result of any spills or failures. Given the level of
detail available at this point in time regarding mine operations and closure, that is probably about
as far as any assessment could go. I'm not aware of any literature reports or data that would
assist in further characterization of these potential injuries.
RESPONSE: No change suggested or required
No additional comments are provided.
Question 11. Does the assessment appropriately describe the potential for cumulative
risks from multiple mines? If not, what suggestions do you have for improving this part
of the assessment?
The potential for cumulative risks associated with the development of multiple mines in the
Bristol Bay watershed is not treated with a great degree of detail. Although each of the potential
stressors (e.g., water withdrawal, habitat illumination, road and stream crossings) are
acknowledged and addressed, little quantitative consideration is given to the potential effects
associated with development of multiple mines. This, however, is probably appropriate given the
hypothetical nature of the single mine scenario and the potential for greater impacts associated
with the development of multiple lines. Short of concluding that "failures at one mine could be
bad and failures at multiple mines could be worse," little else could be concluded. It is noted that
the multiple mines scenario leads to multiple tailings impoundments, more roads and culverts,
increased discharge potential of contaminated waters and increased habitat loss and reduction of
water resources and all of these lead to potentially greater environmental injury as a result of
failures.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 12
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
Based upon the available information and some reasonable assumptions, it appears that the
EPA has addressed the comment satisfactorily.
Question 12. Are there reasonable mitigation measures that would reduce or minimize
the mining risks and impacts beyond those already described in the assessment? What
are those measures and how should they be integrated into the assessment? Realizing
that there are practical issues associated with implementation, what is the likelihood of
success of those measures?
I'm sure there are number of technological/engineering measures that could be implemented to
reduce the potential for environmental injury associated with development of mining in the
Bristol Bay watershed. The development of this a priori risk assessment provides useful
information in identifying were potential risks may exist and should provide mine development
professionals with the degree of guidance about the types of risks and potential consequences of
mine activity failures. Perhaps by recognizing the magnitude of adverse consequences associated
with potential failures, steps can be taken to implement safety measures early in the planning
process that would render mine development more acceptable. In addition, using the assessment
to define areas of uncertainty my provide direction for future research that would be beneficial
for the project. Again, because of the lack of detail associated with the hypothetical mine
scenario, it is impossible to estimate the likelihood of success of any mine control activities.
RESPONSE: One purpose of this assessment is to inform future decisions concerning mine
design and required mitigation. The purposes of the assessment are clarified in the revised
introduction (Chapter 1).
I agree with the US EPA comment, Section 1.2 of the document does an adequate job of
clarifying several of the potential uses for this document. The question originally posed asks "if
there are mitigation measures that would reduce or minimize mining risks and impacts beyond
those described in the assessment." In my reading of Section 1.2 I find no mention of this
assessment providing information to industry engineers/environmental scientists/decision-
makers that would help inform them about potential risks (or regulatory/stakeholder concerns)
that may be mitigated by design/practice considerations. This would appear to be an important
use of this assessment. If industry is aware of potential risks and/or concerns early in the
development of a project it may be possible to take steps to incorporate actions to mitigate
identified environmental risks/concerns. My expertise is not appropriate for evaluating possible
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 13
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
mining mitigation measures but I'm sure others on the review panel had that particular
expertise.
Question 13. Does the assessment identify and evaluate the uncertainties associated
with the identified risks?
The risk assessment attempts to identify and evaluate the uncertainties associated with each of
the recognized potential risks. The authors have, for the most part, successfully identified a
number of uncertainties that may affect the accuracy and conclusions of the risk assessment.
Clearly, this information should provide a basis for prospective mine planners and regulatory
authorities to focus their efforts to minimize potential environmental risks. In some cases the
uncertainties identified are probably best addressed through the development of additional data
and this should guide future research efforts undertaken prior to mine development and
operation.
RESPONSE: No change suggested or required.
As discussed in my previous comments, the evaluation and documentation of assessment
uncertainties is critical to any risk assessment. This document has clearly delineated the
uncertainties associated with the various risk scenarios (both in the executive summary and in
the individual chapters). The revisions to the uncertainty section(s) in this draft of the Bristol
Bay assessment have been expanded, where appropriate, to address concerns associated with
revised assessment techniques. If I could offer one suggestion associated with uncertainties
would be that in the executive summary section "Summary of Uncertainties and Limitations in
the Assessment" some more specific mention should be made of the uncertainties associated
with the evaluation a toxicological endpoints and the uncertainties associated with toxicological
benchmarks. Most of these are delineated and discussed in Chapter 8, Section 8.2.5. However,
statements contained in the document like that in Section 8.2.2.1 that states:
"Based on the literature cited above and the author's experience, resolution of this uncertainty
by additional research and testing is likely to lower the chronic criterion by approximately a
factor of 2 and not as much as a factor of 10" suggests that the US EPA believes that ambient
water quality criteria may be significantly under protective of aquatic organisms. If this is the
case, clearly it is a source of major uncertainty and should be identified as such.
Question 14. Are there any other comments concerning the assessment, which have
not yet been addressed by the charge questions, which panel members would like to
provide?
None
No additional comments are provided.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 14
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. William Stubblefield
3. SPECIFIC OBSERVATIONS
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
William A. Stubblefield, Ph.D.
None
No additional comments are provided.
STUBBLEFIELD RESPONSE TO EPA PEER REVIEW COMMENTS on Page 15
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (April 2013 draft]
-------
Dr. Dirk van Zyl, University of British Columbia
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Van Zyl, Dirk
Sent: Sunday, August 04, 2013 9:25 PM
To: Thomas, Jenny
Subject: Re: Bristol Bay Review Check-In
Dear Jenny,
I attach my letter report in fulfillment of the contract. Please let me know if you need anything else at this time.
Please also advise on how to handle the invoicing.
Regards,
Dirk van Zyl
-------
Dr. Dirk van Zyl
Revised Draft Bristol Bay Assessment
Peer Review Follow-On
Review Comments by Dirk van Zyl
1. Introduction
This peer review follow-on of the Revised Draft Bristol Bay Assessment is
provided in fulfillment of an agreement with the US Environmental Protection
Agency. It specifically responds to the following requirement from the Scope of
Work for the task:
The contractor shall provide a letter report evaluating how well the revised
draft Bristol Bay Assessment addresses both the key recommendations
provided in the Executive Summary of the peer review report and the
specific comments provided by the Contractor in the peer review report.
While Volume 1 of the Revised Draft Bristol Bay Assessment was completely
reviewed, the evaluation below focuses narrowly on the aspects required by the
Scope of Work. The report provides some overall review comments in the next
section followed by two sections providing a few comments on the Executive
Summary and the specific comments provided by this author in the 2012 peer
review report.
The Scope of Work did not request or propose any specific format for the
compilation of this review. One approach would be to repeat the original
comments and the EPA responses before providing the comments from this
review. It is clear that this approach will result in a very extensive document. The
author decided on the following approach to reduce the length of this document1:
• In Section 3 below a specific topic and page number will be referred, the
latter as: "EPA Response, p#".
1 Note that for the for the first two bullets it is assumed that the page numbers in the final public
version of the EPA Responses will be the same as the Draft available for this review.
1
-------
Dr. Dirk van Zyl
• In Section 4 below the specific question will be referred to plus the page
number as follows: "EPA Response Q#, p#".
• References to the Second External Review Draft Documents will be to
page numbers as follows: "Second Draft, p#".
2. Overall Review Comments
The restructuring and revisions presented in the Revised Bristol Bay Assessment
resulted in a much improved document that provides better motivation for
selection of specific hypothetical project elements and assumptions. It also
provides further detail information, maps and summaries that improve the logic
and approach of the document and associated analyses.
The document also strives to provide transparent information about the use of
input data, e.g. reports by non-governmental organizations. It is stated that the
"USEPA subjected some of these documents to external review before
incorporating this information into the assessment' (Second Draft, p. 2-3). These
reports and the reviews are available
at http://cfpub.epa.gov/ncea/bristolbav/recordisplav.cfm?deid=182065. While the
reports and reviews are provided for broad review, the value of the described
external review will undoubtedly be questioned for its independence and
credibility. It is also interesting to note that many reports from other credible
sources were provided to the public record in 2012 but that none of them were
apparently considered "useful" to EPA in their preparation of the Revised Draft.
The author questions the objectivity of this process and therefore the overall
scientific value of the review report by EPA.
3. Executive Summary Comments
• Global Climate Change (EPA Response, p. 7): The author would suggest
that the comment about "an explicit reference to a monitoring program" did
not imply that EPA should develop such a plan as part of the Second
Draft. The main insight of this comment is that environmental effects from
2
-------
Dr. Dirk van Zyl
potential future mining in Bristol Bay can be related to climate change as
well as mining and that specific monitoring plans should be developed for
mining projects in Bristol Bay to allow clearly distinguishing between these
effects. It is recommended that this comment be included in the final
document.
• Mount St. Helens as a surrogate (EPA Response, p. 13): The author would
like to provide some further input about the use of Mount St. Helens data
as a surrogate for the "rate of benthic habitat recovery from a massive
deposition of fine mineral particles". The geochemical composition,
particle size and the temperature at the time of deposition of the fine
materials from the Mount St. Helens event must be compared to that of
tailings before this event can be used as a surrogate. The EPA
Responses indicate that "we have removed references to Mount St.
Helens in the revised assessment to eliminate concerns", however on p.
9-27 of the Second Draft there is a reference to Major et al. (2000). On p.
15-40 of the Second Draft this reference is given as: Major, J., T. Pierson,
R. L. Dinehart, and J. E. Costa. 2000. Sediment yield following severe
volcanic disturbance — A two-decade perspective from Mount St. Helens.
Geology 28: 819-822. This seems to still refer to Mount St. Helens in
terms of long-term sediment transport. There seems to be a disconnect
between the statement in the EPA Responses and the Second Draft that
must be corrected.
4. Specific Comments
• EPA Responses Q1, p.31 and 32: no further comments
• Good practices vs. best practices (EPA Response Q2, p. 57 and 58): the
author agrees that from a regulatory perspective "best management
practices" (BMP's) is typically used for non-point source runoff. However,
the term is also in much broader use in practice to refer to other activities
and this is somewhat recognized. Addition of Box 4-1 helps to clarify this
point.
3
-------
Dr. Dirk van Zyl
• Tailings Management Technologies (EPA Response Q2, p. 59): The
response refers to a "rule of thumb" from the SME Mining Engineering
Handbook (1973) about allotment of space for dry tailings. The author is
not aware of any hard rock dry tailings management projects in the
1970's, these really came to being in the 1980's and 1990's. It is therefore
not clear what experience base was used to establish this "rule of thumb".
The author has not reviewed the reference, however assuming 100 pcf
density for the dry tailings the rule indicates a height of tailings stack of
about 130 ft. This clearly is dependent on the topography and there is no
reason not to construct stacks that are higher. There is no reason why the
overall area covered by dry tailings should be larger than the footprint of a
slurry tailings deposit as the dry density of placed and compacted tailings
will be higher than that of slurry tailings. The statement that "dry stack
tailings disposal at this site would create additional surface area loss
versus the scenario's traditional dam" contained in the present response is
not correct and should be reconsidered as a reason to dismiss this option.
Please also note that the comments in the responses are inconsistent with
that in the Second Draft on p. 4-18:
Dry stack tailings management in which tailings are filtered and
"stacked" for long-term storage, is a newer, less commonly used
tailings disposal method. Dry stacked tailings require a smaller
footprint, are easier to reclaim, and have lower potential for
structural failure and environmental impacts (Martin et al. 2002).
However, the high-energy cost of dry stack technology remains a
barrier for mining low-grade ores such as porphyry copper. In
addition, this type of storage is inappropriate for acid-generating
tailings and is less feasible in larger operations, where tailings
impoundments serve to store water as well as tailings. It is most
applicable in arid regions, although dry stacks are also used in wet
climates or in cold regions where water handling is difficult (Martin
4
-------
Dr. Dirk van Zyl
et at. 2002). Currently, the only mines in Alaska that use dry stack
disposal of tailings are underground mines with high-grade ore and
relatively low quantities of tailings (e.g. Greens Creek, a lead,
silver, zinc mine in southeast Alaska; and Pogo, a gold mine in
eastern interior Alaska).
It should also be noted that it is not only higher energy costs but also
capital and other operating costs that make this technology more
expensive. However, if this tailings management method results in lower
environmental risks then is should be seriously considered and potentially
included in project considerations, even if only to enumerate the lower
risks, e.g. the tailings failure scenario will nor occur because there is no
fluid stored and all surface water can be diverted. The issue of acid
drainage can also be addressed in the design of the facility, note that acid
drainage collection and treatment from the tailings impoundment is
included in the present assessment.
• Tailings density (EPA Responses Q2, p. 61 and Second Draft p. 6-33):
The author indicated in his 2012 that the analogy to consolidation of oil
sand tailings is incorrect. The Second Draft includes the following
statements: "Although oil sands are different from porphyry copper
tailings, the principle is the same. Lack of data specific to porphyry copper
tailings suggests a cautious approach, so we do not assume that the
tailings consolidate to a fully stable land form. Thus the system may
require continued monitoring to ensure hydraulic and physical integrity in
perpetuity". The author agrees that the principles are the same, however
where it is estimated to take 200 years plus for a mature fine oil sand
tailings facility to consolidate, a hard rock facility will consolidate during the
operational life (refer to Caldwell et al, 1984; Oliveira and van Zyl, 2006a
and 2006b). While long-term monitoring will be required, especially if the
decision is made to store water on the impoundment, the tailings will be
completely consolidated, i.e. no excess pore pressures will remain.
5
-------
. Dirk van Zyl
• Tailings specific gravity (Second Draft p. 9-12): The report notes that the
specific gravity of the bulk tailings is 2.61 and that of the pyritic tailings is
the same. This is incorrect. The specific gravity of pure pyrite is about 5.0,
however when mixed with other minerals as in tailings one can typically
use a value of 4.2 to 4.5. Please correct this and update the calculations
on p. 9-12.
• EPA Responses Q3, p.73 and 75: no further comments
• Significance of risk: EPA Responses Q4, p. 89: the response indicates
that: "The purpose of the assessment is not to assign significance to the
risks, but to provide information for decision-makers on the consequences
of mining". It is the author's opinion that the decision-makers will also need
some context of comparative information to make a realistic decision. It
seems to the author that providing acute and chronic quotients in Chapter
8 effectively provides significance of risks. Please consider the
consistency of approaches.
• EPA Responses Q5, p. 101 and 102: no further comments
• EPA Responses Q6, p. 112 and 113: no further comments
• EPA Responses Q7, p. 120 and 121: no further comments
• EPA Responses Q8, p. 128 and 129: no further comments
• EPA Responses Q9, p. 136 and 137: no further comments
• EPA Responses Q10, p. 146 and 147: no further comments
• EPA Responses Q11, p. 159 and 160: no further comments
• Financial assurance Chapter 4 (EPA Responses Q12, p. 170): Box 4-3 on
Financial Assurance is a good addition to the report. The author would like
to suggest a reworking of part of the introductory paragraph, namely:
"Thus, regulations also serve to hold an operator accountable for potential
future impacts, through establishment of financial assurance requirements
and imposition of fines for non-compliance with permit requirements". This
sentence can be read to indicate that financial assurance can be used to
non-compliance with permit requirements. The author suggests that the
sentence be modified to clarify the concepts, e.g. to (suggested inserts
-------
Dr. Dirk van Zyl
highlighted): "Thus, regulations also serve to hold an operator accountable
for potential future impacts, through establishment of financial assurance
requirements for closure and reclamation of the mine and imposition of
fines for non-compliance with permit requirements during mine
operations".
• Double-walled pipes (EPA Responses Q12, p. 170): the comment that a
double-walled pipeline along the entire length of the pipeline "may not be
feasible or cost effective to do this given the length of the pipeline" is
puzzling. This seems like a design comment not a risk assessment
comment. Clearly if that is the best option to reduce risks then it becomes
part of the cost of doing business, unless another option having similar low
risks can be proposed. Please reconsider.
• EPA Responses Q13, p. 177 and 178: no further comments
• Stakeholder process (EPA Responses Q14, p. 185): the addition of Box 1-
1 provides very useful information for readers of the Second Draft.
5. References
Caldwell, J.A., Ferguson, K., Schiffman, R.L. and Van Zyl, D.(1984) Application of
Finite Strain Consolidation Theory for Engineering Design and Environmental
Planning of Mine Tailings Impoundments, Sedimentation/Consolidation Models -
Predictions and Validation, R.W. Yong and F.C. Townsend (Ed.), ASCE, pp. 581-
606.
Oliveira, W. and Van Zyl, Dirk (2006a) Modeling Discharge of Interstitial Water from
Tailings Following Deposition - Part 1: Phenomenology and Model description.
Brazilian Journal on Soil & Rocks, Vol. 29, No. 2, August.
Oliveira, W. and Van Zyl, Dirk (2006b) Modeling Discharge of Interstitial Water from
Tailings Following Deposition - Part 2: Application. Brazilian Journal on Soil &
Rocks, Vol. 29, No. 2, August.
7
-------
Dr. Phyllis Weber Scannell
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Phyllis [|
Sent: Monday, Duly 08, 2013 11:14 AM
To: Thomas, Denny; White, Dessica
Subject: transmittal of comments on Bristol Bay Assessment
Attached please find two documents:
A cover letter transmitting my final comments on the Bristol Bay Environmental
Assessment and a document containing the final comments.
Should you have any questions about these documents, please contact me.
Respectfully submitted,
Phyllis Weber Scannell
-------
Dr. Phyllis Weber Scannell
July 8, 2013
Jessica White
USEPA Headquarters
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Jenny Thomas
USEPA
Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20460
Dear Ms White and Ms Thomas:
Attached by email is a document titled FinalCommentsPWS.docx. This comments contains the summary
evaluation of EPA's response to the September 2012 peer review report of the revised Bristol Bay
Environmental Assessment. This document fulfills tasks 2.1, 2.2 and 2.3 of the Statement of Work. Note
that Task 1 was fulfilled by email on June 14, 2013 and EPA's acknowledgement of that email on June
17, 2013.
I have structured the referenced document in a way that I believe is most useful to EPA. The document
contains the initial comment to the earlier Assessment, the EPA response, my response to the comment
and my comments on how the revised Assessment satisfied the initial concerns. Should you have any
questions about the information I have provided, please contact me by email or telephone. Thank you.
Respectfully submitted,
/Ph/Mea?
' (j X
Phyllis Weber Scannell
-------
Dr. Phyllis Weber Scannell
DRAFT
EPA Response to Peer Review Report
External Peer Review of EPA's Draft Document
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
Task 2(1):
Using the final peer review report, review comments and suggestions made on the May 2012
draft. The contractor should specifically review key suggestions provided by the Contractor in
Section III
Task 2(2):
Review the EPA's response to comments document that details how EPA responded to specific
comments and suggestions provided in the peer review report.
Task 2(3):
Review the revised draft Bristol Bay Assessment (Documents 3 [revised Bristol Bay
Environmental Assessment] and 4 [revised Appendix I and new Appendix J]) to assess how
EPA's responses to comments and suggestions were implemented in the revised draft.
Initial Comments by Phyllis Weber Scannell (August, 2012)
Responses by US EPA
follow on comments by
Phyllis Weber Scannell
-------
Dr. Phyllis Weber Scannell
Contents
Summary of Key Recommendations from Peer Reviewers 1
Scope of the Document 1
Risks to Salmonid Fish 6
Wildlife 8
Human Cultures 9
Water Balance/Hydrology 10
Mitigation Measures 13
Uncertainties and Limitations 14
Editorial Suggestions 15
Research Needs 18
Responses to Charge Questions 22
Question 1 22
Question 2 23
Question 3 24
Question 4 27
Question 5 29
Question 6 30
Question 7 32
Question 8 34
Question 10 38
Question 11 39
Question 12 40
Question 13 43
Question 14 45
Specific Observations 47
Page 4-11: 47
Page 4-23 (LI): 48
Page 6-36 (P3): 48
Page 6-37 (P4): 49
App G, Page 5 (P 1, last line): 49
App G, Page 5 (P3): 50
Additional comment on revised Assessment 50
Comments on Revised Appendix 1 52
Comments on New Appendix J 53
-------
Dr. Phyllis Weber Scannell
SUMMARY OF KEY RECOMMENDATIONS FROM PEER REVIEWERS
This section summarizes the significant general recommendations put forth by the peer reviewers
regarding EPA's draft assessment. In developing these recommendations, peer reviewers
provided input on three major areas of the assessment: (1) scope, (2) technical content and (3)
editorial suggestions. Reviewers also identified research needs for EPA to consider. Please note
that this summary of peer review comments did not reflect a consensus or group perspective, but
was compiled from a discussion of individual peer reviewer recommendations. Additional
details, including references cited, can be found in the reviewers' individual comments in
Section III.
Scope of the Document
Initial Comment from Review Team
Articulate the purpose of the document more clearly via a primer on the Ecological Risk
Assessment process. If the purpose of the assessment is to inform EPA as the decision maker, then
the level of detail should correspond to this purpose. The authors should justify and explain what
level of detail is required.
EPA Response
Additional information on both the purpose of the assessment and ecological risk assessment
(ERA) in general has been added to Chapters 1 and 2, as well as the Executive Summary.
Section 1.2 includes information about the use of the assessment. The assessment has been
reorganized into two major sections (problem formulation, risk analysis and characterization) to
clarify where different chapters fall in the typical ERA process.
Reviewer Response
The reviewer believes that this comment has been addressed in the revised Assessment. Of
particular value is the description of the authority under which the Assessment has been
conducted (Section 1.1, page 1-2). The discussion of limitations of the Assessment clarifies
what is addressed and what is not.
Initial Comment from Review Team
Include a statement upfront about the role of risk managers and other audiences, such as
project managers/engineers, regulators, mine owners/operators. Knowing their role ensures
inclusion of information necessary for any risk assessment by (1) describing the need for a risk
assessment, (2) listing those decisions influenced, and (3) characterizing what risk managers
require from the risk assessment.
1
-------
Dr. Phyllis Weber Scannell
EPA Response
Section 1.2 of the revised assessment discusses the use of the assessment.
Reviewer Response
This reviewer is satisfied that the initial comments have been addressed by Section 1.2.
Initial Comment from Review Team
Explain why the scope for human and wildlife impacts was limited to fish-mediated effects, as
well as why fish-mediated effects on humans were limited to Alaska Native cultures.
Reviewing effects beyond fish-mediated ones (e.g., potential for complete loss of the
subsistence way of life) would improve the assessment.
EPA Response
The scope of the assessment has been clarified throughout the document, particularly in Chapters
1 and 2. Throughout the assessment we acknowledge that direct effects of large-scale mining on
wildlife and Alaska Native cultures may be significant, but that these direct effects are outside the
scope of the current assessment.
Reviewer Response
The reviewer understands the limitations of the Assessment and is satisfied.
Initial Comment from Review Team
Be more consistent throughout the document in terms of the level of detail provided for the
different scenarios and stressors. For example, the document has devoted 36 pages to the
discussion of catastrophic Tailings Storage Facility (TSF) failure, while sections on the
pipeline, water treatment, and road/culvert failures are brief. Indeed, the long discussion on the
TSF failure belies a certainty and understanding of dam failure dynamics that is inaccurate.
EPA Response
The final document includes more failure scenarios (e.g., diesel pipeline failure, wastewater
treatment plant failure, and refined seepage scenarios) in Chapter 8. It also explains why these
specific failure scenarios were chosen, and discusses these scenarios in greater detail than the
previous draft (i.e., to more closely match the level of detail originally provided only for the TSF
failure scenario). Also see detailed responses to comments on Peer Review Question 5.
Reviewer Response
As discussed under Peer review Question 5, the additional information on other failures is an
important component of the revised Assessment. I note that failures also are discussed in
2
-------
Dr. Phyllis Weber Scannell
Appendix I. I also commend EPA for including references to possible long-term management of
the site.
Initial Comment from Review Team
Consider the document to be a screening-level assessment of all potential stressors. Focusing on
failure mode overemphasizes catastrophic events (e.g., TSF failing), rather than considering all
potential stressors, such as holding mine owners strictly accountable for their day-to-day
activities with regard to best practices.
EPA Response
Additional information on the purpose and scope of the assessment has been added to Chapters 1
and 2. A screening of all potential stressors, including individual chemicals, is presented in
Section 6.4.2. Also see detailed responses to Peer Review Question 2 on the use of "best
practices" and responses to Peer Review Question 5 on failure scenarios.
Reviewer Response
The additional information that was included in the revised assessment satisfies my
concerns that were stated in the initial comment. I am especially pleased that the revised
Assessment includes discussions of ions other than copper and of chemicals typically used
in this type of mining.
Initial Comment from Review Team
Reexamine the document's use of historical data and case studies to describe and estimate the
risk of failure for certain mine facilities (including the TSF, pipeline, water treatment, etc.), as
these examples from extant mines may not be an appropriate analog for a new mine in the Bristol
Bay watershed.
EPA Response
The TSF failure range was, and still is, based on design goals, not the historical data. The
historical TSF failure data are provided as background. The pipeline failure rates are based on
the most relevant historical data from the petroleum industry. They are directly relevant to the
diesel pipeline, and experiences at the Alumbrera mine (described in the previous draft) and the
Antamina and Bingham Canyon mines (added to this draft) suggest that they also are relevant to
the product concentrate pipeline. Water treatment failure rates were not quantified. However,
recent reviews cited in the revised draft indicate that water collection and treatment failures have
been reported at nearly all analogous mines in the U.S. The estimation of culvert failure
frequencies has been revised and is now based on only recent literature (2002 and later). We
believe that these estimates are appropriate.
3
-------
Dr. Phyllis Weber Scannell
Reviewer Response
The discussion of pipeline failures in the revised Assessment also includes failures from
human error. The revised Assessment contains more comprehensive discussions of
different failures. I am satisfied that the revisions have addressed the initial comment.
Initial Comment from Review Team
Expand the discussion on the use of "best" management practices, as the document states that the
mine scenario employs "good," but not necessarily "best" practice. For a mine developed in the
Bristol Bay watershed, only "best" practice likely would be appropriate and anything less may
not be permitted. Even so, without a track record of "best" practice (e.g., new technologies), we
cannot assume that technology, by itself without appropriate operational management controls
can always mitigate risk.
EPA Response
The term "best management practices" is a term generally applied to specific measures for
managing non-point source runoff from storm water (40 CFR Part 130. 2(m)). Measures for
minimizing and controlling sources of pollution in other situations often are referred to as best
practices, state of the practice, good practice, conventional, or simply mitigation measures. We
assume that these types of measures would be applied throughout a mine as it is constructed,
operated, closed, and post-closure, and have used the term "conventional modern" throughout the
assessment to refer to these measures. To remove any ambiguity related to the subjectiveness of
terms "good" or "best", we have removed them in the revision and have provided definitions for
relevant terms used in Box 4-1.
Reviewer Response
I agree with the changes; however, I also note the valuable information provided in
Appendix I. Appendix I includes discussion of different methods for tailings disposal and
the related stabilities, discussions of conventional practices for different aspects of the
mining operation and discussions of "best" practices that could be used (e.g. selective
flotation, page 10 of Appendix I). The information in this Appendix is a valuable
contribution to the Assessment.
Initial Comment from Review Team
Adopt a broader range of mine scenarios (not only minimum and maximum) so as to bound
potential impacts, especially at smaller mine sizes (e.g., 50th percentile). Underground mine
4
-------
Dr. Phyllis Weber Scannell
development, with its different impacts, also should be considered and included in the
assessment.
EPA Response
A third mine size scenario (250 million tons) has been added to the assessment, to represent the
worldwide median sized porphyry copper mine (based on Singer et al. 2008).
Reviewer Response
The reviewer notes the addition of the 250-million tons mine scenario and the expanded
discussions of other deposits in the region that could possibly be developed.
Initial Comment from Review Team
Based on the hypothetical mine scenario, perpetual management of the geotechnical integrity of
the waste rock and tailings storage facilities, as well as perpetual water treatment and monitoring,
will most likely be necessary (i.e., a "walk away" closure scenario after mining ends may not be
possible). Therefore, emphasize how monitoring and management of the geotechnical integrity
of waste rocks and tailing storage facilities should continue "In Perpetuity" (i.e., for at least tens
of thousands of years). Discuss what conditions would need to be met to allow "walk away"
closure in the Bristol Bay environment gaining insight into these observations from mines where
perpetual treatment and monitoring are ongoing (e.g., the Equity Silver Mine in British
Columbia).
EPA Response
The conditions for closure and the potential need for perpetual site management are discussed in
general terms in the revised assessment. The primary condition assumed to be required is water
chemistry that meets all criteria and permit conditions and that is stable or improving. However,
even though there are some facilities with "perpetual treatment" conditions in place, there is
obviously no information about how these facilities perform over very long periods of time.
Reviewer Response
In the response to the initial comments (Charge questions), I have noted various places in
the document where reference to perpetual treatment and a need for sufficient bonding
has been added. My initial concerns were that this issue had not been adequately
highlighted. The revisions to the Assessment satisfy those concerns.
Initial Comment from Review Team
Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting, overflight, etc.)
5
-------
Dr. Phyllis Weber Scannell
were managed. This includes roads, airstrips, helipads, camps, fuel dumps, and ATV trails that
have already been developed or imposed on the watershed, and what "mitigation" already has
been undertaken on those sites. Assess the consequences/impacts of these activities in the
Cumulative Risks section.
EPA Response
The effects of exploratory activities are outside of the scope of this assessment.
Reviewer Response
No additional comments.
Risks to Salmonid Fish
Initial Comment from Review Team
Place potential mining impacts in the context of the entire Bristol Bay watershed by emphasizing
the relative magnitude of impacts. For example, of the total salmon habitat, assess the proportion
lost due to mining. Further, reflect on the non-linear nature of the relationship between habitat
and salmon production; 5% of the habitat could be critical and thus responsible for 20% or more
of salmon recruitment. Intrinsic potential, which measures the ability of particular habitats to
support fishes, would lend credibility to this analysis.
EPA Response
We are unable to build a complete Intrinsic Potential (IP) model, as this would require validation
and more elaborate construction of metrics appropriate to this region. Our preliminary
characterization provides the building blocks for assessing the distribution of key habitat-forming
and constraining features across these watersheds. We now include a characterization of the major
drivers of habitat potential across the watershed and place the mine-site specific effects in this
context (Chapters 3, 7, and 10).
Reviewer Response
The reviewer notes the additional information provided in Chapters 3, 7 and 10.
Initial Comment from Review Team
Include a section on the impact of Global Climate Change with explicit reference to a
monitoring program that will allow scientists, if the mine is built, to distinguish between
effects of climate change and mining effects on the physical and biological components of this
ecosystem.
6
-------
Dr. Phyllis Weber Scannell
EPA Response
Climate change projections and potential impacts are now included in Chapter 3, and as
important external factors in the risk analyses presented in Chapters 7, 9, 10, and 14. Development
of a monitoring program to distinguish between mining and climate change effects is outside of
the scope of the assessment.
Reviewer Response
The revised Assessment contains different references to climate change and resulting risks to the
aquatic environment. For example, an increase in precipitation and flooding is discussed on page
7-59 of the revised Assessment. Possible risks from increased summer temperatures are
discussed on page 8-63. The reviewer is satisfied that the issue of climate change has been
integrated into the Assessment.
Initial Comment from Review Team
Explicitly recognize that the transportation corridor and all associated ancillary development,
including future resource developments made possible by the initial mining project, will
necessarily and inevitably have impacts (hydrologic, noise, dust, emissions, etc.). These impacts
will vary in duration, intensity, severity, relative importance, spatial dispersion, and inevitably
expand geographically through time with further "development." These impacts should be
incorporated into the Cumulative Risks section.
EPA Response
The cumulative risk section (Chapter 13) has been expanded to include the multiple transportation
corridors, ancillary mining development and secondary development associated with multiple
mines in a qualitative discussion. The issues addressed in the assessment of the transportation
corridor (Chapter 10) have also been expanded to include chemical spills, dust, invasive species,
and road treatment salts.
Reviewer Response
The reviewer is satisfied with these changes.
Initial Comment from Review Team
Incorporate current research findings into stream crossing and culvert-design practices (e.g.,
arch culverts, bridges, etc.).
EPA Response
We describe current culvert design practices in a box titled "Culvert Mitigation" in Chapter 10.
Reviewer Response
The reviewer is satisfied with these changes.
7
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
Recognize in the assessment that risk and impact are not equivalent. Risk may be low, but the
potential impact could be huge (e.g., in the case of a TSF failure).
EPA Response
Risk has been defined in many ways, even by risk assessors. The commenter seems to define risk
as probability. To avoid that potential source of confusion, we use the term "probability" for that
concept. Similarly, the commenter seems to use "impact" where we use "effect" or "magnitude
of effect". We use "risk" to refer to both concepts combined—that is, an event or effect and its
probability).
Reviewer Response
Perhaps including a box with these definitions would be helpful.
Initial Comment from Review Team
Recognize and justify chronic behavioral endpoints, such as those potentially affecting
survival and long-term success of fish populations.
EPA Response
The chronic behavioral effects of copper on salmonids, the primary endpoint of concern, were
described in Chapter 5 and are now described in Chapter 8. Although those effects occur at lower
levels of copper than conventional survival, growth and reproduction endpoints for salmonids,
they are less sensitive than the conventional endpoints for aquatic invertebrates.
Reviewer Response
The reviewer is satisfied.
Wildlife
Initial Comment from Review Team
Recognize that the draft assessment did not account for all levels of ecology, such as the
individual (e.g., a bald eagle nest), population, community, ecosystem, and landscape levels.
Fold other levels of organization into the stressors assessment where appropriate or justify a
more limited approach.
8
-------
Dr. Phyllis Weber Scannell
EPA Response
As is appropriate for an ecological risk assessment (as opposed to an environmental impact
assessment), this assessment focuses on a specific, limited set of endpoints as defined in Chapter
5. We have added text in Chapters 2 and 5 to explain both why these endpoints were selected,
and that responses other than those considered in the assessment, at multiple levels of ecological
organization, are likely but are outside the scope of the assessment.
Reviewer Response
The additional information in Chapters 2 and 5 clarify the use of different endpoints. The
reviewer is satisfied.
Initial Comment from Review Team
Discuss in the document fishes other than salmonids The assessment focuses on risks to sockeye
salmon in the Bristol Bay watershed (and also considers anadromous salmonids, rainbow trout,
and Dolly Varden), but does not account for potential impacts to other members of the resident
fish community. Further, primary and secondary production, including nutrient flux was not
addressed. Expanding the assessment to consider other levels of organization, including direct as
well as indirect effects on wildlife and other fish, would provide additional context in the
assessment of mine-related impacts.
EPA Response
See response to comment above; we also incorporated additional information from Appendices
A, B, and C into the Chapter 5 text, to provide additional detail on the area's biota. We chose our
endpoints for reasons described in Chapters 2 and 5. Other endpoints, including indirect effects on
fish and wildlife, are now discussed more explicitly, but are generally considered outside the
scope of the assessment.
Reviewer Response
My initial comments requested that some of the information in these Appendices be integrated
into the main body of the Assessment. I note that this was done and believe that the expanded
discussions in Chapters 2 and 5 address my initial comments.
Human Cultures
Initial Comment from Review Team
Use case histories to provide insight and anticipate mining impacts on Alaska Natives (e.g.,
those exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil and gas
development in the North Slope region, and social impacts related to mining development in
9
-------
Dr. Phyllis Weber Scannell
Alaska).
EPA Response
Examples from applicable case studies, including the Exxon Valdez oil spill, are cited in Chapter
12 of the revised assessment.
Reviewer Response
Although the area of human cultures is outside my area of expertise, I note expanded discussions
in Chapter 12.
Initial Comment from Review Team
As noted above (Scope of the Document), clarify why the scope was limited to fish-mediated
effects. The potential direct and indirect impacts for human cultures extend far beyond fish-
mediated impacts (e.g., potential complete loss of the subsistence way of life). The rationale for
this narrow focus should be fully explained. In addition, a clear explanation should be given for
why fish-mediated human impacts focused only on Alaska Native cultures.
EPA Response
The assessment focuses on a specific, limited set of endpoints as defined in Chapter 5. We have
added text to explain both why these endpoints were selected, and that responses other than those
considered in the assessment are likely but are outside the scope of the assessment. The
assessment was expanded (Chapters 5 and 12) to acknowledge that there are a wide range of
potential direct and indirect impacts to indigenous culture, but they are outside of the scope of
this assessment. The discussion of potential effects to indigenous cultures was expanded to
explain that a loss of subsistence resources would extend beyond a loss of food resources to
social, cultural, and spiritual disruption. The text has been expanded to acknowledge the strong
cultural ties of many non-Alaska Natives to the region, and potential effects on all residents from
loss of a subsistence way of life. However, the focus of the assessment remains on effects on
indigenous cultures resulting from effects on salmon.
Reviewer Response
The area of human resources is outside my area of expertise; I have no further comment.
Water Balance/Hydrology
Initial Comment from Review Team
Better characterize water resources and assess the potential effect of mine development on these
resources by (1) generating a diagram similar to the conceptual models beginning on page 3-7 to
illustrate the potential effects of mine construction and operation on surface- and ground-water
hydrology; (2) developing a quantitative water balance and identifying water gains and losses;
10
-------
Dr. Phyllis Weber Scannell
(3) identifying seasonality of hydrologic processes, including frozen soils and their associated
values (e.g., mm/yr) for each component of the water balance; (4) incorporating these processes
into a landscape characterization; (5) evaluating how global climate change will influence these
hydrologic processes and rates; and 6) using this characterization to demonstrate the expected
hydrologic modification associated with the mine scenarios and infrastructure development.
EPA Response
The original Figure 4-9 (new Figure 6-5) has been revised to more clearly show water
management in the assessment's mine scenarios. In addition, three schematics illustrating water
flows under each of the mine size scenarios (Figures 6-8 through 6-10) have been added to
Chapter 6, as have quantitative water balances for each mine size scenarios. A qualitative
discussion of climate change is included in Chapters 3 (Section 3.8) and 14 (Box 14-2).
Reviewer Response
The revisions to the draft Assessment help clarify the water management issues. This reviewer is
pleased that water flows are shown for each mine scenario.
Initial Comment from Review Team
Demonstrate the interconnectedness of groundwater, surface water, hyporheic zone, and its
importance to fish habitat. Address how interconnectedness changes over time - seasonally, and
with varying weather (e.g., wet vs. dry summers or years, and over the long term as climate
changes).
EPA Response
We lack the data to demonstrate this interconnectedness in a spatially and temporally uniform
manner, but do include examples of known points of high connectivity (Chapter 7) and
qualitatively discuss the potential role of climate change (Chapter 3).
Reviewer Response
Section 7.3.2.1 of the revised Assessment (page 7-51) discusses the complex groundwater-
surface water connectivity in the deposit area and the lack of information to describe linkages
and possible changes with mine development. This chapter also discusses using the
sustainability boundary approach (page 7-51). I believe the initial comment was written, in part,
to request that the Assessment recognize the importance of these interconnected systems. The
discussion of the sustainability boundary approach includes possible risks to aquatic habitats. I
recognize the paucity of information to model these linkages and believe the general discussion
presented in this section of the Assessment addresses the initial comment.
11
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
Provide information on all rivers, including ephemeral and intermittent streams, and first-order to
main-stem streams that could be potentially influenced by the proposed mine, its ancillary
facilities, and the transportation corridor.
EPA Response
Due to lack of consistent coverage, we rely on the NHD hydrography layer in this analysis, and
can only address ephemeral and intermittent streams qualitatively (Chapter 7).
Reviewer Response
This comment highlights the need for on-the-ground surveys of potentially affected areas should
mine development proceed. The need for more information was discussed in various sections of
the revised Assessment.
Initial Comment from Review Team
Emphasize the importance of a thorough characterization of the leaching potential of acid-
generating and non-acid generating waste rock and tailings, given the low buffering capacity and
mineral content in the streams and wetlands that could receive runoff and treated water from the
proposed mine. Recognize that collection and treatment of runoff and leachate generated will be
critical to maintain baseline water chemistry in these streams and wetlands.
EPA Response
We agree that these are important issues, and the discussion of leachate from waste rocks and
tailings has been expanded in the revised assessment (Chapter 8).
Reviewer Response
The reviewer is satisfied with the additional information.
Initial Comment from Review Team
Reference the most current geochemistry data on potentially acid-generating, non-acid
generating, and metal leaching so as to describe any potential effects of seepage and changes to
surface- and ground-water quality via non-catastrophic failure.
EPA Response
We used the geochemistry data in PLP's Environmental Baseline Document, as summarized by
the USGS in Appendix H. The effects of seepage on water quality are analyzed in Chapter 8 of
the revised assessment.
12
-------
Dr. Phyllis Weber Scannell
Reviewer Response
No further comment.
Initial Comment from Review Team
Explain how contaminants/metals were selected (and others ignored) by EPA as causes for
concern. Information should be included on additional metals and their toxicity so as to assess
impacts of potential leachates. The Pebble Limited Partnership baseline document presented
additional metals that might be useful to include in the assessment.
EPA Response
The revised assessment describes the selection of contaminants and other stressors of concern in
Section 6.4.2. Additional metals, process chemicals and dissolved solids are now included.
Reviewer Response
The additional discussion of other elements, process chemicals and dissolved solids addresses
the initial comment.
Mitigation Measures
Initial Comment from Review Team
Incorporate the critical mitigation information from Appendix I into the main report's mine
scenarios. Include standard mitigation measures that could provide insight into how well they
might work in this context. If this information is not included in the main report, then justify its
absence.
EPA Response
Mitigation measures incorporated into design and operation to minimize potential impacts were
included in the assessment, as were some reclamation measures for closure; these measures are
made clearer in the revised assessment. These mitigation measures were a sub-set of those
presented in Appendix I. The assessment assumes that measures chosen for the scenarios would
be effective. Mitigation to compensate for effects on aquatic resources that cannot be avoided or
minimized by mine design and operation would be addressed through a regulatory process that is
beyond the scope of this assessment. Nevertheless, in response to public and peer comments we
have included a discussion of compensatory mitigation in Appendix J of the revised assessment.
Reviewer Response
The revised Assessment now includes more information on mitigation measures; the revised
Appendix I is a valuable addition to the Assessment. The reviewer is satisfied with these
changes.
13
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in the main
report, as they ultimately influence the range of mining impacts and consider time frames of
mitigation or reclamation measures (e.g., immediate response, long-term reclamation).
EPA Response
See response to previous comment. Mitigation measures are discussed at greater length in the
revised assessment report (e.g., Chapter 4 and Appendix J).
Reviewer Response
The reviewer is satisfied. Further note: Appendix J presents a realistic summary of possible
mitigation measures and their value in the subject area.
Uncertainties and Limitations
Initial Comment from Review Team
Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of uncertainty and (2)
assessing the certainty of some mine impacts. Discuss data limitations in the context of
uncertainty.
EPA Response
The individual analysis chapters and the revised Integrated Risk Characterization (Chapter 14)
discuss certainties and data limitations to a greater extent, as suggested.
Reviewer Response
The reviewer is satisfied with the changes.
Initial Comment from Review Team
Articulate early in the document how much uncertainty is acceptable. The assessment
provides little insight with respect to the decisions the document is intended to support.
EPA Response
Acceptable levels of uncertainty can be defined prior to an assessment if a decision and a
decision maker are identified and if data will be collected by a specified design to implement a
specified model, as described in the EPA's Data Quality Objectives process, However, because
this assessment is based on available data and is intended as a background scientific document
rather than a decision document, it is not possible to specify the amount of uncertainty that is
14
-------
Dr. Phyllis Weber Scannell
acceptable. Rather, the available data determine the uncertainty and if the assessment is
subsequently used to inform a decision, the decision maker must determine whether the level of
uncertainty is acceptable.
Reviewer Response
The information presented in the Executive Summary and Chapter 1 clarify this. The reviewer
accepts that this is an appropriate approach, given the amount of available data.
Editorial Suggestions
Initial Comment from Review Team
The title of the document leads one to believe that the assessment addresses the entire Bristol
Bay watershed; rather, the report deals with two major rivers and their watersheds, the Nushagak
and Kvichak. Thus, the title should be changed to reflect the emphasis on these two rivers and
their watersheds. A possible title may be "An Examination (or identification) of the Potential
Impacts of Mining and Mining Associated Activities on Salmon Ecosystems in the Nushagak
River and Kvichak River watersheds, Bristol Bay."
EPA Response
The assessment addresses multiple scales: the Bristol Bay watershed, the Nushagak and Kvichak
River watersheds, the watersheds of the three streams draining the Pebble deposit, and the
watersheds crossed by the transportation corridor. These multiple scales, and how they are used
throughout the assessment, are described more clearly in the revision (Chapter 2).
Reviewer Response
The reviewer has no further comment and accepts the title as is.
Initial Comment from Review Team
Revise the Executive Summary to more precisely reflect the findings in the document.
EPA Response
The Executive Summary has been rewritten to reflect the revised assessment findings.
Reviewer Response
The reviewer is satisfied with the expanded Executive Summary.
15
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
The appendices contain detailed and useful information that should be summarized and included
in the main document (e.g., Appendix E: Economics, Appendix G: Road and Pipelines, and
Appendix I: Mitigation). Additionally, consider expanding the preface to include information on
the use of the appendices. If the information is not included in the main report, then justify its
absence.
EPA Response
More information from the appendices was brought forward into appropriate chapters of the
revised report. The purpose of the appendices—to provide the detailed background
characterization necessary for the ecological risk assessment—has also been clarified in that
material has been incorporated into Chapters 1 and 2.Chapter 2. The document no longer
contains a preface because that material has been incorporated into Chapters 1 and 2.
Reviewer Response
In the response to charge questions, the reviewer noted the additional information in the main
body of the document. The changes enhance the discussions and provide valuable information.
The concerns of this reviewer have been satisfied.
Initial Comment from Review Team
Discuss in more detail the instructive and well-thought-out conceptual models (pages 3-7 to 3-
11) illustrating the impacts of mining on Bristol Bay ecosystem processes. Also, consider
expanding the conceptual models to include wildlife, fish-wildlife interactions,
vegetation/terrestrial habitat, and hydrologic processes. Allow them to guide the text because they
appear detailed and complete.
EPA Response
Additional information on the use of conceptual models throughout the assessment has been
incorporated into Chapter 2. The more comprehensive conceptual models presented in Chapter 6
(Chapter 3 in the first draft) have been broken into their relevant component parts throughout the
risk analysis and characterization chapters, to better frame the specific pathways addressed in
each chapter. Additional conceptual models considering impacts on wildlife, Alaska Native
populations, and cumulative effects of multiple mines have been added to Chapters 12 and 13.
Reviewer Response
Chapter 6 contains an expanded and clarified discussion of conceptual models used in the
Assessment. Although Figure 3.2A-D in the earlier draft presented figures representing the
conceptual models used in the Assessment, these diagrams were difficult to interpret without
sufficient accompanying information. Discussions on conceptual models in the current revision
have been augmented by the list of stressors (Table 6-9) and following text. The discussions in
16
-------
Dr. Phyllis Weber Scannell
Chapters 12 and 13 also help clarify the potential stressors and endpoints. The reviewer is
satisfied that the initial comments have been addressed.
Initial Comment from Review Team
Incorporate the information contained in the conceptual models into a formal framework,
such as a Bayesian or other decision-analysis models.
EPA Response
This is an excellent suggestion for future efforts, but is beyond the scope of the current
assessment. Creating a Bayesian Belief Network would require that the Agency convene experts
to subjectively estimate the probabilities of each transition in the conceptual models. In contrast,
this assessment is intended to elucidate the risks from potential mining based on available data and
analyses of those data.
A Decision Analysis would require that alternative outcomes be specified, the utility of each
outcome for a decision maker be defined and the probabilities of each outcome be estimated for
each possible decision so that the expected utilities of each outcome can be calculated. Because
this assessment is not a decision document, these requirements are not feasible or appropriate.
Reviewer Response
The reviewer accepts the EPA response.
Initial Comment from Review Team
Generate a standard operating protocol for significant figures and use it throughout the
document.
EPA Response
The authors have carefully addressed this issue. Numbers from the literature or from the PLP EBD
retain the number of significant figures in the original. Numbers derived for this assessment have
the appropriate number of significant figures given the precision of the input data and
uncertainties due to modeling and extrapolation.
Reviewer Response
Perhaps a statement in the text could be added to explain this (there may already be such
clarification, but I did not find it.)
17
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
Remove all references to Mount St. Helens as a surrogate for a TSF failure. Using a nonhuman-
caused release of material into the ecosystem as an analogue for a mine failure is not comparable
in terms of likelihood or risk for a human-caused release. It would be more appropriate to
extrapolate from the impacts of known mine failures.
EPA Response
We are puzzled by this comment. The Mount Saint Helens data were used strictly to address the
rate of benthic habitat recovery from a massive deposition of fine mineral particles. The
hydrological processes that determine the recovery of substrate texture and the requirements of
fish or aquatic invertebrates are not known to depend on whether mineral particles were from a
natural event or an anthropogenic event. We have reviewed the literature on known mine
failures. They studied tailings spills in terms of toxicity but not in terms of physical habitat
effects, which is why we used Mount Saint Helens data. Nevertheless, we have removed
references to Mount St. Helens in the revised assessment to eliminate concern.
Reviewer Response
The reviewer is satisfied with this change; removing reference to Mt. St. Helens eliminates
potential confusions.
Initial Comment from Review Team
Ensure that the draft assessment remains part of the public record, allowing the document
history to remain intact.
EPA Response
All drafts of the watershed assessment will remain part of the public record.
Reviewer Response
The reviewer appreciates that this is being done.
Research Needs
Initial Comment from Review Team
What are the acute and chronic impacts of mixtures of contaminants, including metals, acid
mine drainage, etc., on the fauna and flora of the Nushagak River and Kvichak River
watersheds? What species are most sensitive and might surrogate species exist for those for
which we do not have data? Review the European literature and regulatory requirements for
18
-------
Dr. Phyllis Weber Scannell
additional data.
EPA Response
The acute and chronic impacts of contaminant mixtures, including metals and acid mine drainage
(i.e., metals in low pH-waters) were addressed using concentration additivity models in the
leachate chemistry tables in Chapters 5 and 6 (now Chapters 8 and 11). Additional toxicity data
were obtained by searches of the EU and OECD database eChem, the EPA's ECOTOX and the
Environment Canada site. More metals are now included. In general, metals are most toxic to
aquatic arthropods rather than fish, as discussed for copper.
Reviewer Response
The reviewer is satisfied, but realizes that should mine development go forward, it may be
necessary for permit applicants to conduct specific toxicity studies. These potential studies are
beyond the scope of the Assessment.
Initial Comment from Review Team
Can an inventory of nutrients, total organic carbon, and dissolved organic carbon inputs to
aquatic environments be developed that demonstrates their relative magnitude and spatial
variation from headwaters to Bristol Bay? What is the relative importance of marine-derived
nutrients relative to other nutrients from watershed and terrestrial sources? What is the current
atmospheric input of nutrients?
EPA Response
These data would be very useful in the risk assessment, but are not currently available for the
Bristol Bay region. We agree this is a research need.
Reviewer Response
The reviewer agrees with the EPA response; this is a research need that should be fulfilled at the
permitting stage.
Initial Comment from Review Team
What are the locations of subsistence areas and can these areas be characterized and
differentiated by collecting local environmental and ecological knowledge (e.g., fish overwintering
areas, climate change, ecological shifts, etc.)?
EPA Response
The revised assessment incorporated current data on subsistence use areas available from
ADF&G. EPA acknowledges that these data are incomplete and would encourage additional
collection of subsistence data and Traditional Ecological Knowledge.
19
-------
Dr. Phyllis Weber Scannell
Reviewer Response
This comment is outside my area of expertise. I defer to the other reviewers who are more
qualified to comment on subsistence issues.
Initial Comment from Review Team
What impact might mining have on other important wildlife species in the basin (e.g.,
freshwater seals in Iliamna Lake)?
EPA Response
The scope of the assessment is focused on potential risks to salmon from large-scale mining and
salmon-mediated effects to indigenous culture and wildlife. Direct effects on wildlife from large-
scale mining are likely to be important and Appendix C (now a stand-alone US Fish and Wildlife
report) provides useful information for a future evaluation of direct effects on wildlife from
large-scale mining. We agree that this is an important area for future research.
Reviewer Response
The reviewer is satisfied.
Initial Comment from Review Team
What is the comprehensive hydrologic regime of the specific project mining area, and the
broader watershed system as characterized by baseline monitoring, spatial distribution, and
quantitative flow of surface- and ground-waters?
EPA Response
Comprehensive spatial estimates of mean annual flow are now presented in Chapter 3.
Quantification of spatial and temporal patterns of groundwater flows is an acknowledged highly
desirable product, but it not feasible within the scope of this assessment. Results of an
independent groundwater-surface water modeling effort are described in Chapter 7.
Reviewer Response
This reviewer acknowledges that in-depth discussions of groundwater and surface water are
beyond the scope of the Assessment and beyond the availability of information at this time. In-
depth data on flows will be required should the project proceed to the permitting stage. The
reviewer is satisfied that sufficient available data have been used in the Assessment.
20
-------
Dr. Phyllis Weber Scannell
Initial Comment from Review Team
What is the cumulative impact of commercial fisheries on the Bristol Bay watershed,
especially in an ecosystem context as related to marine-derived nutrient and energy flow?
Acknowledge that commercial fishing has had an impact on the amount of marine-derived
nutrients returned to the watersheds.
EPA Response
The impact of commercial fisheries on the watershed is not within the scope of this assessment.
Information on commercial fisheries management has been added in Box 5-2. However, the
purpose of this assessment is not to assess the relative effects of potential mining and
commercial fishing—it is to evaluate potential effects on endpoints if a mine were to be
developed, given existing conditions and activities in the region.
Reviewer Response
One of the intents of the review committee in making this comment was to state that this is not a
completely unaltered watershed; that there have been effects to the ecosystem by fishing and
subsequent removal of nutrients. However, the EPA statement "given existing conditions"
serves to separate pristine from existing conditions. Perhaps a statement in the Executive
Summary that the Assessment examines impacts of mining activities in the existing conditions of
the Bristol Bay region (for example, Page ES-4, paragraph 3: "This is not an in-depth
assessment. . .") would clarify this issue.
21
-------
Dr. Phyllis Weber Scannell
RESPONSES TO CHARGE QUESTIONS
Qiiestlf
The EPA's assessment focused on identifying the impacts of potential future large-scale mining
to the fish habitat and populations in these watersheds. The assessment brought together
information to characterize the ecological, geological, and cultural resources of the Nushagak
and Kvichak watersheds. Did this characterization provide appropriate background information
for the assessment? Was this characterization accurate? Were any significant literature, reports,
or data missed that would be useful to complete this characterization, and if so what are they?
Initial Comment by Reviewer
The Environmental Assessment presents a well-documented discussion of the fish and wildlife
resources of the Nushagak River and Kvichak River Watersheds, with more limited discussions
of the remainder of the Bristol Bay Watershed. The document discusses interactions among
species, including nutrient flows and the importance of groundwater systems; however,
information on contributions of marine-derived nutrients and existing pressures on the
environment are not as complete, or lacking. The information is general in nature. Should mine
development go forward, it will be necessary to obtain ecological information specific to the
potentially affected areas. The information should include timing of fish spawning, egg hatch, in-
migration and out-migration, and similar specific life-history information for important wildlife
species.
Response by EPA:
We have clarified the use of information at different scales (Bristol Bay watershed and Nushagak
and Kvichak River watersheds in the problem formulation chapters, smaller spatial scales in the
risk analysis and characterization chapters). General information on assessment endpoints is
included in Chapter 5, with more detailed information included in the appendices.
Reviewer Response
Reviewer is satisfied with this approach.
Reviewer Response to Revised Bristol Bay Assessment
The revised draft contains new sections 2.2.1 Topical Scope and 2.2.2 Spatial Scales. The
section 2.2.1 defines the limits of the assessment ("we do not consider all potential sources of
risk . . . ") and the focus ("potential effects on freshwater habitats").
Section 2.2.2 defines the five spatial scales used in the Assessment and the use of broader special
scales for describing the physical, chemical and biological environments and effects of multiple
mines.
These two sections provide needed clarity to the Assessment and satisfy the initial concerns
expressed by this reviewer.
22
-------
Dr. Phyllis Weber Scannell
Questi(
A formal mine plan or application is not available for the porphyry copper deposits in the Bristol
Bay watershed. EPA developed a hypothetical mine scenario for its risk assessment, based
largely on a plan published by Northern Dynasty Minerals. Given the type and location of copper
deposits in the watershed, was this hypothetical mine scenario realistic and sufficient for the
assessment? Has EPA appropriately bounded the magnitude of potential mine activities with the
minimum and maximum mine sizes used in the scenario? Are there significant literature, reports,
or data not referenced that would be useful to refine the mine scenario, and if so what are they?
Initial Comment by Reviewer, Question 2(1)
The Environmental Assessment discusses a hypothetical mine (given that mine plans have not
been developed). Page 4-5 of the document states that "rocks associated with porphyry copper
deposits tend to straddle the boundary between net acidic and net alkaline . . " The Pebble
Project Environmental Baseline Report (SRK 2011, Chapter 11) summarizes testing on the
samples from the pre-Tertiary porphyry mineralized rock in Pebble East Zone (PEZ) and Pebble
West Zone (PWZ). The metals leaching/acid rock drainage study showed acidic conditions
occurring immediately in core with low NP, but the average delay to onset of acidic conditions
was estimated to be about 20 years. Copper was leached in the highest concentrations, but Co,
Cd, Ni, and Zn also leached from samples from PEZ. Wacke (sedimentary rock) samples from
PEZ and PWZ leached As, Sb, and Mo, in addition to Cu. (SRK, page 58). The available
information on acid generation and metals leaching appears to be preliminary. Development and
permitting of a viable mine plan will require extensive sampling and data analysis of ore
samples, plans for classifying waste rock (as PAG and NAG), and, possibly, plans for collecting
and treating runoff and seepage waters.
Response by EPA:
EPA agrees that developing and permitting a viable mine would require extensive information.
The assessment is not a mining plan. The scenario presents a suggested treatment option for
mining influenced water and settling ponds for water that is simply storm water runoff. No change
required.
Reviewer Response
Reviewer agrees that no change is required; however, reviewer still believes that the available
information on metals leaching was not adequately addressed in the initial draft Environmental
Assessment. Perhaps some statement in the Assessment that addresses the need for in-depth
sampling for acid generation and metals leaching would suffice. The primary concern of this
reviewer is that the Assessment did not adequately discuss potential leaching of metals other than
Cu and that much of the information contained in the SRK report was not used.
23
-------
Dr. Phyllis Weber Scannell
Reviewer Response to Revised Bristol Bay Assessment
The reviewer notes that Section 8.2.2.2 Other Metals has been added to the revised draft
Assessment. This new section provides much needed information on the toxicity of elements
other than Cu and recognition that these elements may be of concern. New Section 8.2.2.3
addresses concerns of total dissolved solids, a potentially significant stressor to aquatic
communities. Inclusion of these two sections greatly strengthens the revised Assessment.
Initial Comment by Reviewer, Question 2(2)
The Environmental Assessment seems a bit premature in making an assessment of the potential
for acid rock drainage (ARD) or metals leaching (ML). Data on metals other than Cu are
insufficient and possible toxicities to fish are not addressed. Further, the description of the
potential mine may not reflect a likely mine scenario. It is difficult to calculate potential risks to
the environment without a specific mine plan. The section of the Environmental Assessment
should be revised as more data on ARD and ML become available.
Response by EPA:
The assessment uses the geochemistry data that are available from the Pebble Limited
Partnership. Copper was emphasized in the review draft because the EPA believed, and still
believes, that it is the contaminant of greatest concern. Toxicities to fish of the other metals were
not discussed because they had been screened out. However, the revised assessment explains the
screening process and the selection of copper in more detail in a new section on the identification
of stressors of concern (Section 6.4.2) and more metals have been added to the screening
assessment. The toxicities of all metals reported in the leachate are now addressed either as
individual elements or, in the case of major ions, as contributors to total dissolved solids. The
mine scenario is based on the most recent preliminary plan released by Northern Dynasty
Minerals (Ghafari et al. 2011).
Reviewer Response
The reviewer is pleased that this additional information has been added.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer notes the additional information on different elements and total dissolved solids, as
noted above (Comment 2(1)).
Questit
EPA assumed two potential modes for mining operations: a no-failure mode of operation and a
mode involving one or more types of failures. Is the no-failure mode of operation adequately
described? Are engineering and mitigation practices sufficiently detailed, reasonable, and
24
-------
Dr. Phyllis Weber Scannell
consistent? Are significant literature, reports, or data not referenced that would be useful to
refine these scenarios, and if so what are they?
Initial Comment by Reviewer, Question 3(1)
Chapter 4 provides a detailed description of a hypothetical mine design for a porphyry copper
deposit in the Bristol Bay watershed. Some of the assumptions appear to be somewhat
inconsistent with mines in Alaska. In particular, the descriptions of effects on stream flows from
dewatering and water use do not account for recycling process water, bypassing clean water
around the project, or treating and discharging collected water.
Response by EPA:
The issues mentioned were discussed in Sections 4.2.3 (tailings storage) and 4.3.7 (water
management) in the original draft document. They are now addressed in Chapter 6 of the revised
document, which describes the mine scenarios, and in Chapter 4, which provides generic
background on porphyry copper deposits and mining. Stream flow effects presented in Chapter 7
now reflect a complete water balance, including water capture and re-use, bypass, and discharge
from the wastewater treatment facility, as suggested by the treatment.
Reviewer Response
The reviewer is pleased that this additional information has been added.
Reviewer Response to Revised Bristol Bay Assessment
Chapter 4 of the revised Assessment contains an expanded description of the chemistry and
associated risks of porphyry Cu deposits (Section 4.2.2) and an expanded overview of the mining
process (Section 4.2.3). Chapter 6 of the revised Assessment presents descriptions of potential
mine development alternatives. The revised document has been reorganized to provide in-depth
descriptions of plausible mine scenarios (Chapter 6). The descriptions of mine scenarios include
expanded discussions of mining processes, ore processing, waste rock management, tailings
storage facilities, and water management. Chapter 6 of the revised document provides an in-
depth and needed background to the document. The reviewer also is pleased that there is
recognition of the possibility of activities extending in perpetuity (page 6-16).
Much of the discussion of fish distribution has been moved to Chapter 7 of the revised
Assessment. This discussion has been revised and now includes more in-depth discussions
relating potential habitat loss and stressors to the mine scenarios.
The revisions to chapters 4, 6 and 7 satisfy the initial concerns raised by the reviewer.
25
-------
Dr. Phyllis Weber Scannell
Initial Comment by Reviewer, Question 3(2)
Section 4.3.8, Post-closure Site Management, raises critically important issues - can a mine in
this area be designed for closure? Is it acceptable to develop and operate a mine that will require
essentially perpetual treatment? It is my belief that these are the essential questions that should
be addressed during any mine permitting process.
Response by EPA:
EPA agrees that these are important questions to be addressed, but they are risk management, not
risk assessment, questions. The purpose of the assessment is to evaluate risks to the salmon
fishery from large-scale mining. Risk management decisions will be made during the permitting
process. Thus, no changes to the assessment were made in response to this comment.
Reviewer Response
The reviewer still believes that there is a potential risk to the salmon fishery from large-scale
disturbance of a mineralized area and that the sources of risk may change over time as
weathering of the exposed minerals occurs.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer notes some references to the possibility of perpetual treatment in the revised
document (e.g. page 6-16). I realize that the concerns raised in this comment are issues of risk
management; however, I believe it was necessary to suggest the possibility for long-term
management of the site.
Initial Comment by Reviewer, Question 3(3)
Section 4.3.8.1 raises concerns about long term water quality and quantity from the mine pit.
These concerns need to be addressed during a mine permitting process. Pit water quality depends
on how the pit is developed, what reclamation will occur, if reclamation will be concurrent with
mining, and what kinds of water treatment will be used. Tailings storage facility (TSF) water
quality depends on how the mine tailings are managed; it may be possible to use dry stack
tailings with sulfide removal rather than submerged tailings.
Response by EPA:
EPA agrees that water quality can be influenced by design and reclamation, but when the latter
entails creating a pit lake there is little flexibility for reclamation concurrent with mining. How
tailings are managed within the impoundment can affect water chemistry, and a dry stack with
sulfides removed may produce the best water quality results after reclamation if the fate of the
26
-------
Dr. Phyllis Weber Scannell
sulfide tailings is never considered. According to Ghaffari et al. (2011), 14% of the tailings
produced will be pyritic, which equates to an average of 28,000 tpd in a 200,000 tpd mining
operation (over 255 million tons during a 25 year mine life). These tailings need to be managed
in such a manner that oxidation does not lead to acidic drainage, so the most effective way is to
deposit them subaqueously. Some suggested common mitigation measures for management of the
pit at and post closure are included in the revised assessment (Chapter 6).
Reviewer Response
The reviewer is pleased that a discussion of the need for mitigation and management of the pit at
post closure has been included.
Reviewer Response to Revised Bristol Bay Assessment
The discussion of the mine pit (Section 6.3.1) satisfies the concerns raised in the initial comment.
Qiiestlf
Are the potential risks to salmonid fish due to habitat loss and modification and changes in
hydrology and water quality appropriately characterized and described for the no-failure mode of
operation? Does the assessment appropriately describe the scale and extent of risks to salmonid
fish due to operation of a transportation corridor under the no-failure mode of operation?
Initial Comment by Reviewer, Question 4(1)
The no-failure model makes a number of assumptions about how the mine will be developed -
some may be accurate, some may be considerably different. It is important to take under
consideration that Pebble is currently a prospect, not a mine. Should this project proceed to mine
development, it will be incumbent on the mining company to develop a rigorous mine plan that
includes detailed information on all aspects of a future project. This mine plan will be reviewed
by state and federal staff with experience in large project development.
Response by EPA:
The EPA agrees with this comment. No changes suggested or required.
Reviewer Response
The reviewer is satisfied.
Reviewer Response to Revised Bristol Bay Assessment
The expanded descriptions of different mine scenarios contained in the revised Assessment
clarify the possible sources of risk to salmonid fish.
27
-------
Dr. Phyllis Weber Scannell
Initial Comment by Reviewer, Question 4(2)
The no-failure model discusses the amount of riverine habitat that will be lost to mining by the
mine pit, tailing storage facility, and waste rock dumps. Anadromous fish habitat is protected
under Alaska Statute 16.05.840-870. The statute requires review of a project potentially
affecting fish habitat and, where necessary, avoidance, mitigation, or compensation. A project
must provide free passage of fish; the project cannot be placed in such a way that fish are
prohibited from moving into the upstream reaches. Estimates of habitat loss from the mine
footprint are not possible without a more detailed plan of operations for the mine.
Response by EPA:
The scenarios presented are meant to represent those expected as typical for mining of porphyry
copper deposits of this type, and are based on preliminary mine plans from NDM (Ghafari et al.
2011). Although layout of mining components at a site may differ somewhat from what we
present in the scenarios, the main components of mining will remain the same for open-pit
mining. Given stream density in the area, direct losses of stream and wetland habitat of a similar
magnitude would be inevitable with projects of the specified magnitudes.
Reviewer Response
The reviewer agrees.
Reviewer Response to Revised Bristol Bay Assessment
The expanded discussions of the possible mine scenarios are important additions and help clarify
possible effects to the aquatic environment.
Initial Comment by Reviewer, Question 4(3)
There are many aspects of the development of a large mine project that need thorough review to
ensure that habitats are protected. These include, but are not limited to: classification and
storage of waste rock, lower grade ore, overburden, and high grade ore; development and
maintenance of tailings storage facilities; development and concurrent reclamation of disturbed
areas, including stripped areas and mine pits; collection and treatment of point and non-point
source water; quantity and timing of discharges of treated water; monitoring of ground water,
seepage water and surface water; and biomonitoring. The transportation corridor will require
review and permitting of every stream crossing of fish-bearing waters. In addition, plans should
be developed for truck wheel-washing to minimize transport of contaminated materials.
28
-------
Dr. Phyllis Weber Scannell
Response by EPA:
The EPA agrees that these aspects would need to be subject to a thorough review during the
development and approval of a detailed mining plan. No changes suggested or required.
Reviewer Response
The reviewer is satisfied.
Reviewer Response to Revised Bristol Bay Assessment
The revised Assessment provides some discussion of the in-depth reviews that would be required
during the permitting process. This information addresses the concerns raised by the reviewer.
Questi(
Do the failures outlined in the assessment reasonably represent potential system failures that
could occur at a mine of the type and size outlined in the mine scenario? Is there a significant
type of failure that is not described? Are the probabilities and risks of failures estimated
appropriately? Is appropriate information from existing mines used to identify and estimate types
and specific failure risks? If not, which existing mines might be relevant for estimating potential
mining activities in the Bristol Bay Watershed?
Initial Comment by Reviewer, Question 5(1)
This section focuses on catastrophic failures; however, there are a number of non-catastrophic
failures that can occur at a mine site. Non-catastrophic failures include leakage of contaminated
water to ground or surface waters from PAG waste rock, the tailing storage facility, and exposed
ore surfaces, and from emergency discharge of untreated water from the TSF and ore spills from
trucking accidents. Such failures can be minimized or prevented with good site planning and
monitoring. An additional "failure" has been experienced at a mine in Alaska when the water
elevation of the tailing pond was sufficiently high to cause groundwater flow across a natural
divide into an opposite drainage.
Response by EPA:
Because the number of potential failures is extremely large, it is necessary to choose a
representative set of failure scenarios. The review draft contained a scenario in which tailings
leachate was not fully contained and reached a stream (Section 6.3 in the May 2012 draft). The
revised assessment document includes more failure scenarios (e.g., diesel pipeline failure, truck
accidents, quantitative water treatment failure, and refined seepage scenarios) in Chapters 8, 10
and 11 and explains why these particular failure scenarios were chosen.
Reviewer Response
The reviewer is pleased that this section has been expanded.
29
-------
Dr. Phyllis Weber Scannell
Reviewer Response to Revised Bristol Bay Assessment
The sections of the revised draft that address potential failures have been substantially revised
and expanded. The current draft Assessment is much improved and satisfies the concerns
initially expressed by this reviewer. In particular, I note the following sections as providing
important information:
Chapter 8, Section 8.1.3 provides an in-depth discussion of post-closure sources of
contamination, primarily from the pit lake. This section concludes: "In sum, failure to collect
and treat waters from the waste rock piles, TSFs, or mine pit could expose biota in the streams
draining the post-closure mine site to contaminated water." It is important to recognize that
these risks may extend in perpetuity.
Chapter 10 provides an expanded description of the transportation corridor. The explanation of
bridges and culverts has been clarified (Section 5.4.1 in the previous draft; section 10.3.2.1 in the
current draft).
Chapter 10, Chemical Contaminants in Stormwater Runoff (Section 10.3.3 in the current draft;
Section 5.4.2 in the previous draft) has been revised. The current revision provides more
thorough explanations of possible risks from storm water.
Questi(
Does the assessment appropriately characterize risks to salmonid fish due to a potential failure of
water and leachate collection and treatment from the mine site? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
Initial Comment by Reviewer, Question 6(1)
This section of the report provides an in-depth discussion of possible sources and fates of
contaminated water. Chapter 6.3 discusses possible adverse effects from early mine closure or
prematurely shutting down a water treatment system. These issues highlight the need for a mine
plan that includes concurrent reclamation, sufficient bonding to conduct reclamation in the event of
an early shut down, and plans and specifications for collection and bypass of clean water and
collection and diversion to a water treatment system of contaminated water.
Response by EPA:
The EPA agrees with this comment. No change suggested or required.
Reviewer Response
The reviewer is satisfied.
30
-------
Dr. Phyllis Weber Scannell
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that no changes are needed to the Assessment. The reviewer notes
information in the revised assessment that supports the need for a rigorous review and permitting
process (Box 4.2) and financial assurance (Box 4.3) and references to the Red Dog Mine as
requiring perpetual treatment. This additional information addresses the concerns raised by the
reviewer in her initial comments.
Initial Comment by Reviewer, Question 6(2)
The Risk Characterization (Section 6.3.3) discusses possible contaminant loads to downstream
waters. As stated in this section, it "serves to indicate the large potential risk from improperly
managed waste rock leachate." This statement highlights the need for an in-depth mine plan with
sufficient monitoring and fail-safe provisions. An emergency discharge of untreated waters from
a tailings storage facility could be made to a collection pond for later treatment or the tailings
pond could be engineered to accommodate a higher flood event so the likelihood of overtopping
is minimized.
Response by EPA:
The waste rock leachate scenario referenced by the commenter has been eliminated and replaced
with a more detailed and realistic scenario for waste rock leachate collection and treatment
(Chapter 8).
Reviewer Response
The reviewer is satisfied that this issue has been addressed.
Reviewer Response to Revised Bristol Bay Assessment
The additional information in Chapter 8 satisfies the concerns initially raised by the reviewer.
Initial Comment by Reviewer, Question 6(3)
Section 6.3.3 (Risk Characterization) states "Alternatively, water collection and treatment failure
could be a result of an inadequately designed water treatment system which could result in the
release of inadequately treated water as at the Red Dog Mine, Alaska (Ott and Scannell 1994,
USEPA 1998, 2008). In that case, the failure could continue for years until a new or upgraded
treatment system is designed and constructed." This statement is misleading and overly
simplistic; the water treatment system at the Red Dog Mine was designed to treat the predicted
flows. However, the stream bypass and collection systems were constructed in 1991 to intercept
seepage waters. The additional water that was collected and treated dictated construction of a
second water treatment system in 1992. Sand filters were added in 1993 to remove fine
particulate Zn. The issue was not that the water treatment was inadequate, but that the pre-
31
-------
Dr. Phyllis Weber Scannell
mining hydrologic data was insufficient and that state, federal, and mine officials lacked
experience in mine construction on permafrost soils.
Response by EPA:
The water treatment failure at Red Dog was, as the commenter describes, the result of an
unintended failure to design a plant that was adequate for the mine and site. The passage has
been expanded to clarify the nature of the failure.
Reviewer Response
The reviewer is satisfied that this issue has been addressed.
Reviewer Response to Revised Bristol Bay Assessment
The description of water treatment failure at Red Dog (Chapter 8) is sufficiently detailed and
accurate for the goals of this Assessment. The reviewer is satisfied that the initial comment has
been addressed.
Initial Comment by Reviewer, Question 6(4)
Overall, the discussions of risks to salmonid fish due to a potential failure of water and leachate
collection and treatment from the mine site highlight the need for more comprehensive
information on groundwater, including delineating flow pathways, depth to surface, and water
volumes. Additional information is needed on water collection, storage, and treatment at future
mine facilities.
Response by EPA:
Discussions of wastewater collection and treatment have been considerably expanded in Chapter
8.
Reviewer Response
The reviewer is satisfied that this issue has been addressed.
Reviewer Response to Revised Bristol Bay Assessment
The discussions of wastewater collection and treatment contained in Chapter 8 of the revised
Assessment are a valuable addition to the document. The initial concerns expressed by the
reviewer have been addressed.
Questi(
Does the assessment appropriately characterize risks to salmonid fish due to culvert failures
along the transportation corridor? If not, what suggestions do you have for improving this part of
32
-------
Dr. Phyllis Weber Scannell
the assessment? Are significant literature, reports, or data not referenced that would be useful to
characterize these risks, and if so what are they?
Initial Comment by Reviewer
The risks to salmonid fish due to culvert failures would be minimized by implementation of
permits by Alaska Department of Fish and Game (ADF&G), Habitat Division. Under A.S.
16.05.840-870, Alaska has some of the most protective laws for fish and fish habitat in the
United States. Further, given the lack of specific information on road alignments, construction
methods and stream crossings, it is not possible to calculate lengths of affected streams, quantify
loss of fish habitats, or predict failures of culverts, side slopes, etc. The document would be
strengthened if it included specific information on locations of spawning and rearing habitats and
estimated the contribution of fish habitats in the Nushagak River and Kvichak River Watersheds
to the Bristol Bay fishery.
Response by EPA:
Best management practices (BMPs) or mitigation measures that would be used to minimize
potential impacts to salmon ecosystems from construction and operation of the proposed
transportation corridor are now discussed in text boxes throughout Chapter 10. Box 10-2
specifically refers to fish habitat regulations under Title 16.
As noted in the revised assessment, uncertainty exists in the characterization of streams and
wetlands affected by the proposed transportation corridor. Based on the chosen road alignment
scenario (which agrees with that proposed in Ghaffari et al. 2011) we feel that we are justified in
estimating the potential footprint of the proposed corridor and its potential impact on fish habitats
and populations. We note in the revised assessment that "Although this route (the one proposed in
the EPA scenario) is not necessarily the only option for corridor placement, the assessment of
potential environmental risks would not be expected to change substantially with minor shifts in
road alignment. Along any feasible route, the proposed transportation corridor would cross many
streams, rivers, wetlands, and extensive areas with shallow groundwater, including numerous
mapped (and likely more unmapped) tributary streams to Hiamna Lake (Figures 10-1 and 10-2)."
Specific information on locations of spawning and rearing habitats along the proposed
transportation corridor is difficult to obtain; as noted in the revised assessment, the Alaska
Anadromous Waters Catalog and Alaska Freshwater Fish Inventory do not necessarily
characterize all potential fish-bearing streams because of limited sampling along the corridor.
Nonetheless, the revised assessment summarizes the species, abundances, and distributions that
would potentially be affected, and places the streams along the transportation corridor into the
context of the entire Nushagak and Kvichak River watersheds with respect to important
watershed attributes such as discharge, channel gradient, and floodplain potential. As far as
placing potential mining impacts in the context of the entire Bristol Bay watershed, we are
33
-------
Dr. Phyllis Weber Scannell
unable to build a complete IP model, as this would require validation and more elaborate
construction of metrics appropriate to this region. However, our preliminary characterization
provides the building blocks for assessing the distribution of key habitat-forming and
constraining features across these watersheds.
Reviewer Response
The reviewer agrees with the EPA response; however, it is important to note that extensive fish
sampling before project development will be necessary to provide adequate protection to the
fishery.
Reviewer Response to Revised Bristol Bay Assessment
The discussions of fish and fish habitats have been revised and expanded in the 2013 draft
Assessment. Chapter 10 presents a more thorough discussion of fish presence, what sampling
has been done and where there is a lack of information. This revised chapter relates the
information on fish presence, abundance (where known) and habitat use to different scales of the
Nushagak and Kvichak River watersheds. As presented, the reader is better able to understand
the potential loss of habitat, especially with development of a transportation corridor. EPA is to
be commended for this revision; Chapter 10 strengthens the entire Assessment.
Question 8
Does the assessment appropriately characterize risks to salmonid fish due to pipeline failures? If
not, what suggestions do you have for improving this part of the assessment? Are significant
literature, reports, or data not referenced that would be useful to characterize these risks, and if
so what are they?
Initial Comment by Reviewer, Question 8(1)
This section of the document focuses on effects of pipeline failures; however, without a viable
mine plan, descriptions of pipelines and estimates of possible effects are speculative. The
resource developer may opt to build a pipeline to transport fuel from the coast to the mine site or
slurry concentrate to the port. Construction of any pipelines would require review and approval
by state and federal agencies, such approvals would likely contain monitoring plans to ensure
pipeline integrity. However, the risks of pipeline failures should not be minimized; the Fort
Knox Mine near Fairbanks recently experienced a 45,000 gallon spill of cyanide solution after a
bulldozer struck a supply line (Fairbanks Daily News Miner, August 24, 2012).
Response by EPA:
The EPA agrees that the specific locations of pipelines and requirements for monitoring may
differ from our scenario, which is based on preliminary mine plans (Ghaffari et al. 2011).
However, we believe our scenario is plausible and allows an evaluation of potential impacts
34
-------
Dr. Phyllis Weber Scannell
from pipeline failures. The Fort Knox spill has been added to the assessment as an example of
spills due to human error in Section 11.1.
Reviewer Response
The reviewer is pleased that the Fort Knox spill was added as an example of human error;
including the Fort Knox example demonstrates that pipeline failures can and do happen.
Reviewer Response to Revised Bristol Bay Assessment
No further comment; my concerns have been addressed.
Initial Comment by Reviewer, Question 8(2)
The risks of a pipeline failure to salmonid fish depend on the duration of the spill, the type of
material spilled (return water or concentrate), the location of the spill (in the uplands or in a
waterway), and the timing. The effects of a pipeline failure in a waterway when juvenile
salmon are present would be far more severe than a pipeline failure in an upland area.
Response by EPA:
EPA agrees with this comment. Discussion of spill location and life stage exposed has been
expanded in Section 11.3, and the other issues have been carried over from the May 2012 draft.
Reviewer Response
The reviewer is pleased that this discussion has been expanded.
Reviewer Response to Revised Bristol Bay Assessment
The revised draft Assessment has expanded the information on pipeline failures to include
clearer explanation of different types of failures, probability of failure and possible effects to fish
at different life stages. The arguments presented in this section (Chapter 11) are made stronger
by discussing the role of human failure (page 11-6). The sections on exposure (11.3.2) and
exposure-response (section 11.3.3) are well-documented and a valuable contribution to this
chapter. This reviewer is pleased with the revisions made to this section; the concerns expressed
in the initial comments have been satisfied.
Initial Comment by Reviewer, Question 8(3)
Given that there currently is no information on road alignments or locations of future pipelines, it
is not possible to estimate the number of stream crossings (70, page 6-30) or an exact length (269
km, page 6-30) of potentially affected waterways. The risks from pipeline failures outlined in the
draft document should be revised when more specific information on the mine plan of operations
becomes available.
35
-------
Dr. Phyllis Weber Scannell
Response by EPA:
The road alignment and length in Northern Dynasty Mineral's preliminary mine plan (Ghafari et
al. 2011) were used in this assessment. The number of stream crossings was also detailed in that
plan, and was checked by the EPA using USGS data. We would expect that risks would be re-
evaluated as part of a future specific transportation corridor plan. No change required.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that this comment has been adequately addressed.
Question 9
Does the assessment appropriately characterize risks to salmonid fish due to a potential tailings
dam failure? If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to characterize these
risks, and if so what are they?
Initial Comment by Reviewer, Question 9(1)
The assessment considers two possible failures of the tailings dam: a partial-volume failure
occurring during mine operations and a catastrophic failure occurring during or after mine
operations. The partial-volume failure (as modeled in the assessment) would result in a greater
than 1,000-fold increase in discharge and the catastrophic failure in a greater than 6,500-fold
discharge.
Response by EPA:
No change suggested or required.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that this comment has been adequately addressed.
36
-------
Dr. Phyllis Weber Scannell
Initial Comment by Reviewer, Question 9(2
The discussion of tailings dam failures describes possible changes in channel and floodplain
morphology and briefly mentions that the tailings deposition would be a source of easily
transportable, potentially toxic material.
Response by EPA:
No change suggested or required.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that this comment has been adequately addressed.
Initial Comment by Reviewer, Question 9(3)
The potential for increased metals loadings to river and lake systems is understated. Although
there are no current predictions of tailings water quality, the water quality of tailings water from
similar mines could be used to model increases in metals loading from dam failures.
Response by EPA:
The original draft and the revised assessment use tailings leachate data from the PLP EBD (see
Chapter 8). We believe that is more defensible than use of tailings leachates from other mines.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that this comment has been adequately addressed.
Initial Comment by Reviewer, Question 9(4)
In addition to the partial-volume failure and the catastrophic failure, there are other possible
sources of metals loadings from the tailings pond. Examples are emergency releases of
untreated tailings water, seepage of tailings water into the groundwater, and flow from the
tailings pond to groundwater in an adjacent drainage as the head (i.e. hydrostatic pressure) is
37
-------
Dr. Phyllis Weber Scannell
increased as the tailings pond is filled. The last example was experienced at the Red Dog Mine
when the increased elevation of the tailings pond caused water to flow underground into the
Bonns Creek drainage instead of the Red Dog Creek drainage. Interception ditches were
installed after the increases in metals loading to Bonns Creek were detected.
Response by EPA:
Because the number of potential failures is extremely large, it is necessary to choose a
representative set of failure scenarios. The original and revised drafts of the assessment include
"seepage of tailings water into the groundwater, and flow from the tailings pond to groundwater
in an adjacent drainage" but not emergency releases of tailings water.
Reviewer Response
The reviewer is satisfied.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer agrees that this comment has been adequately addressed. The reviewer notes that
sections on potential failures have been expanded. The current draft contains much valuable
information that is presented in a clear and defensible way.
Qiiestii
Does the assessment appropriately characterize risks to wildlife and human cultures due to risks
to fish? If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that would be useful to characterize these
risks, and if so what are they?
Initial Comment by Reviewer
The document focuses on effects to wildlife that would occur from failures - tailings dam failure,
pipeline failure, etc. There are other sources of disturbance to wildlife that should be addressed
in a future mine plan and agency review and permitting. Other mines in Alaska limit truck traffic
on the haul road during caribou migrations, incinerate all kitchen waste, educate workers on bear
safety, and prohibit inappropriate disposals of food containers or other wildlife attractants. Other
factors that might need to be addressed to protect wildlife are limiting air traffic and noise during
certain times of the year. Unless addressed, these issues are more likely to cause detrimental
effects to wildlife than dam or pipeline failures.
Response by EPA:
No change suggested or required. The EPA agrees with the reviewer that any future mine plan
would require an evaluation of, and mitigation for, direct effects on wildlife.
38
-------
Dr. Phyllis Weber Scannell
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied with the revised Assessment.
Qiiestlf
Does the assessment appropriately describe the potential for cumulative risks from multiple
mines? If not, what suggestions do you have for improving this part of the assessment?
Initial Comment by Reviewer, Question 11(1)
There are two issues that should be considered: cumulative effects from a single mine and
cumulative effects from multiple mines. Cumulative effects from a single mine might include
aquatic habitat degradation from non-point sources, including run-off from exposed mineralized
rock, seepage from the tailings impoundment, contaminated dust, noise, and other forms of
disturbance.
Response by EPA:
We added a discussion on the cumulative effects from a single mine (Box 6-1), as well as
discussion of multiple transportation corridors, induced development and increased access
(Chapter 13).
Reviewer Response
The reviewer is satisfied that these items have been added.
Reviewer Response to Revised Bristol Bay Assessment
The additional information on cumulative effects contained in the revised Assessment satisfies
the initial concerns of the reviewer.
Initial Comment by Reviewer, Question 11(1)
Cumulative effects from multiple mines are difficult to predict because there are too many
unknowns. It is frequently to the advantage of a mining company to take advantage of existing
infrastructure, without building new camps, new mills, etc. It is also possible to use an old mine
pit for tailings or waste rock disposal from a new site; however, none of these features can be
determined until there is sufficient exploration to determine if mining is feasible, to characterize
the deposit, and to develop a detailed mine plan. To date, there is not sufficient information to
predict cumulative effects from multiple mines.
39
-------
Dr. Phyllis Weber Scannell
Response by EPA:
We have expanded the discussion of cumulative impacts from multiple mines in Chapter 13. The
analysis of cumulative effects is largely qualitative rather than predictive. It is a plausible example
of future cumulative effects that can shed light on whether cumulative effects is an important
topic for consideration as plans for mining in the Nushagak and Kvichak River watersheds move
forward.
Reviewer Response
The reviewer is pleased that this discussion has been expanded.
Reviewer Response to Revised Bristol Bay Assessment
A number of changes have been made to the discussion of cumulative impacts from multiple
mines. I note the following changes that are particularly valuable:
The revised assessment has expanded table 7.1 (now Table 13.1), a list of known prospects in the
region. Each of these prospects is described in detail in the following text.
Tables 13.2 through 13.7 summarize the potential effects to different fish species and subsistence
uses by development of different prospects in the region. These tables increase the
understanding of possible cumulative effects.
The final section of Chapter 7 (Section 7.4.7 of the original draft) has been expanded to include a
more in-depth discussion of the cumulative impacts of multiple mines. This discussion
now includes a more thorough description of possible habitat loss (Section 13.2.7.1) and
water quality degradation (13.2.7.3). The section on potential effects on assessment
endpoints (Section 13.4.1) is an especially valuable contribution to the Cumulative
Impacts chapter.
The changes to the Cumulative Impacts discussion address the concerns I raised in my initial
comments.
Qiiestlf
Are there reasonable mitigation measures that would reduce or minimize the mining risks and
impacts beyond those already described in the assessment? What are those measures and how
should they be integrated into the assessment? Realizing that there are practical issues associated
with implementation, what is the likelihood of success of those measures?
Initial Comment by Reviewer, Question 12(1)
There are many avoidance or mitigation measures that would be implemented to reduce or
minimize mining risks. I have described some possible approaches when answering the previous
questions. To summarize:
40
-------
Dr. Phyllis Weber Scannell
The two most important questions for reducing or minimizing mining risks are:
Can a mine in this area be designed for closure?
Is it acceptable to develop and operate a mine that will require essentially perpetual
treatment?
Response by EPA:
EPA agrees that these are key questions that must be addressed in the regulatory process. Our
purpose in the assessment is to evaluate the risks resulting from a mine operated with modern
conventional mitigation measures for design, operation, monitoring and maintenance, and
closure. The regulatory process addresses significant and unacceptable risks. Chapter 4 of the
revised assessment discusses regulatory and financial assurance requirements for mining in
Alaska.
Reviewer Response
This comment by the reviewer was intended to highlight the importance of a rigorous regulatory
process and the potential for long term risks to the environment; she understands and agrees with
the intent of EPA's risk evaluation.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is pleased that Chapter 4 now includes a discussion of regulatory and financial
assurance requirements.
Initial Comment by Reviewer, Question 12(2)
Specific Measures that can be taken to minimize risk include:
Limiting metals contamination and acid drainage:
• Design the mine pit to limit oxidation on pit walls. Where feasible, conduct concurrent
reclamation.
• Develop plans for classification and storage of waste rock, lower grade ore, overburden, and
high grade ore.
• Develop and maintain tailings storage facilities with fail-safe provisions. An emergency
discharge of untreated waters from a tailings storage facility could be made to a collection
pond for later treatment or the tailings pond could be engineered to accommodate a higher
flood event so the likelihood of overtopping is minimized. Consider alternate methods for
tailings disposal (dry stack following sulfide removal, etc.).
• Implement concurrent reclamation of disturbed areas, including stripped areas and mine
pits.
• Collect and treat point and non-point source water.
41
-------
Dr. Phyllis Weber Scannell
• Design and implement plans for the quantity and timing of discharges of treated water;
especially if the treated water is high in total dissolved solids. Monitor ground water, seepage
water, and surface water.
• Design system for collection and bypass of clean water and collection and diversion of
contaminated water to a water treatment system.
• Require stations for truck wheel washing.
Response by EPA:
Many of the measures presented here were included in the scenarios and in Appendix I and are
mitigation measures commonly included for mining of this type. Other bullets noted here are
good suggestions for things to address during the regulatory process, should a permit application
be submitted.
Reviewer Response
The reviewer agrees that many of the issues listed above should be addressed during the
regulatory process. The comment was initially made to highlight the importance of a rigorous
regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
Initial Comment by Reviewer, Question 12(3)
Protection of Fish Habitat:
• Review all in-stream activities in waters important to the spawning, rearing or migration of
anadromous and resident fish.
• Design and implement a biomonitoring program.
• Review every road crossing of fish bearing waters to ensure free passage of fish.
• Design and implement a biomonitoring program
Response by EPA:
Suggestions noted for consideration during any future regulatory permitting process.
Reviewer Response
The reviewer agrees that many of the issues listed above should be addressed during the
regulatory process. The comment was initially made to highlight the importance of a rigorous
regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
42
-------
Dr. Phyllis Weber Scannell
Initial Comment by Reviewer, Question 12(4)
Possible Measures to Limit Effects to Wildlife:
• At the planning stages, design aspects of the project to create or enhance wetland and
aquatic habitats for fish, bird, and wildlife species.
• Limit truck traffic on the haul road during migrations.
• Incinerate all kitchen waste
• Educate workers on bear (or other wildlife) safety.
• Limit air traffic and noise during critical times of the year.
Response by EPA:
Suggestions noted for consideration during any future regulatory permitting process.
Reviewer Response
The reviewer agrees that many of the issues listed above should be addressed during the
regulatory process. The comment was initially made to highlight the importance of a rigorous
regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
Questi(
Does the assessment identify and evaluate the uncertainties associated with the identified risks?
Initial Comment by Reviewer, Question 13(1)
The important features of the Environmental Assessment are to describe the fish, wildlife, and
human use of the subject area and to define possible risks from development of a large porphyry
copper mine. There are many uncertainties associated with the identified risks and most were
identified in the document. The document could be strengthened by putting a greater emphasis
on sources of contamination (such as mine seepage, poorly designed collection systems, exposed
pit walls, etc.) in relation to the permeability of the soils.
43
-------
Dr. Phyllis Weber Scannell
Response by EPA:
The revised assessment includes more emphasis on sources of contamination (e.g., diesel
pipeline failure, quantitative wastewater treatment plant failure, and a refined seepage scenario)
and their potential hydrologic transport including through permeable soil and rock.
Reviewer Response
The reviewer is pleased that the revised Assessment included more in-depth discussions of
sources of contamination.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
Initial Comment by Reviewer, Question 13(2)
The 5thbullet on page 8-11 outlines important uncertainties for protecting fish species. These
uncertainties include life-stage-specific sensitivities to temperature, habitat structure, prey
availability, and sublethal toxicities. These factors must be considered should a mining project
go forward.
Response by EPA:
The EPA agrees with this comment. No change suggested or required.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
Initial Comment by Reviewer, Question 13(3)
The 6th bullet on this page discusses the preliminary nature of leaching test data. These tests
must be sufficiently comprehensive to predict both short term and long term water quality from
all sources, including PAG and NAG waste rock, pit walls, and pyritic tailings.
44
-------
Dr. Phyllis Weber Scannell
Response by EPA:
The EPA used the available leaching test data and agrees with the comments about future tests.
Reviewer Response
The reviewer agrees that no change is required.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied and finds that the concerns have been addressed by additional
information in the revised Assessment.
Questi(
Are there any other comments concerning the assessment, which have not yet been addressed by
the charge questions, which panel members would like to provide?
Initial Comment by Reviewer, Question 14(1)
At present, Pebble remains a prospect and there is no plan of operations for the mine. Should the
project move forward to development of a mine, it will be necessary to develop an in-depth
mining plan of operations. The mining plan should include the following:
Transportation - of equipment and personnel and for shipping ore. Transportation of
ore, including loading facilities, wheel washing, and other measures to prevent ore
spillage and contamination.
Siting of mine facilities, including tailings ponds, waste rock storage areas,
concentrate storage area, bypass systems for clean water, and collection systems for
contaminated water.
Mill operations, including a description of the process for concentrating ore.
Chemical and fuel storage and Spill Prevention and Contingency Plans.
Personnel housing, including handling of domestic waste (sewage, garbage).
Water treatment plant. Processes that will be used, anticipated concentrations of metals
and TDS, anticipated discharge volumes, and predicted mass loadings.
Monitoring plans for seepage from tailings ponds, waste rock storage areas, etc.
Monitoring likely will include a series of wells and possibly, a pump-back system. •
Predictions for acid rock generation and measures that will be put in place during
mining to minimize future seepage from the mine site.
Plans for concurrent reclamation and future closure of the mine.
Specifications for sufficient bonding to provide site stabilization and water treatment in
the event of a premature or temporary shut-down and reclamation at closure.
Response by EPA:
The commenter is correct that these are important points that should be considered in evaluation
of a mining plan once submitted. No change suggested or required.
45
-------
Dr. Phyllis Weber Scannell
Reviewer Response
The reviewer agrees that many of the issues listed above should be addressed during the
regulatory process. The comment was initially made to highlight the importance of an in-depth
mining plan and rigorous regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The concerns raised by the reviewer in Question 14 have been addressed by the additional
information contained in the revised Assessment. The reviewer especially notes the expanded
descriptions of waste water management and potential failures, the descriptions of permitting for
large mines in Alaska (Box 4.2), the need for more in-depth sampling for PAG and NAG and the
expanded descriptions of potential sources of toxicity.
Initial Comment by Reviewer, Question 14(1)
After the Mine Plan of Operations is developed, an environmental assessment plan should be
developed that identifies potential effects to fish and wildlife and their habitats from specific
components of the mine (as listed above). In addition, the assessment should include cumulative
effects of nearby mines (if appropriate) on fish and wildlife habitats and water quality.
Response by EPA:
The commenter is correct and we would expect these things to be considered during the
regulatory permitting process. No change suggested or required.
Reviewer Response
The reviewer agrees that many of the issues listed above should be addressed during the
regulatory process. The comment was initially made to highlight the importance of an in-depth
mining plan and a rigorous regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The concerns raised by the reviewer in Question 14(1) have been addressed by the additional
information contained in the revised Assessment. The reviewer especially notes the expanded
descriptions of waste water management and potential failures, the descriptions of permitting for
large mines in Alaska (Box 4.2), the need for more in-depth sampling for PAG and NAG and the
expanded descriptions of potential sources of toxicity. In addition, Box 4.2 describes permit
requirements.
46
-------
Dr. Phyllis Weber Scannell
Initial Comment by Reviewer, Question 14(2)
Among the most important issues that must be addressed are transportation, potential for acid
mine drainage and metals leachate, control of point and non-point pollution, and developing the
mine for future closure.
Response by EPA:
The commenter is correct and we would expect these things to be considered in more detail during
the regulatory permitting process. No change suggested or required.
Reviewer Response
The reviewer agrees with the EPA response. The comment was initially made to highlight the
importance of an in-depth mining plan and a rigorous regulatory review.
Reviewer Response to Revised Bristol Bay Assessment
The concerns raised by the reviewer in Question 14(2) have been addressed by the additional
information contained in the revised Assessment. The reviewer especially notes the expanded
and clarified descriptions of possible effects from transportation (including roads and culverts)
and metals toxicity to fish.
¦ i'EC i i '
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
Page 4-11:
Initial Reviewer Comment
The document states "geomembranes are generally estimated by manufactures to last 20 to 30
years when covered by tailings (North pers. comm.)." Unless North is a P.E. with experience in
geomembranes, the statement needs a stronger reference. For example, Erickson et al (2008)*
discuss the quality issues with geomembranes related to manufacture, installation and
application of a soil-based cover (such as bentonite).
EPA RESPONSE:
The reference is appreciated, but others were chosen for use related to time of life for liners. The
discussion of liners has been expanded and moved to Chapter 4 (Section 4.2.3.4) in the revision,
and includes additional material and references on lifetime. The personal communication
reference has been removed.
47
-------
Dr. Phyllis Weber Scannell
Reviewer Response
Reviewer agrees with this change, provided the appropriate references were given. North pers.
comm. is not an appropriate reference, unless as stated in the initial comment, North is a PE with
recognized experience pertaining to geomembranes.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer notes that more appropriate citations were used in revised Section 4.2.3.4.
Page
Initial Comment
This first sentence is confusing and implies that oxygen has low solubility because it is in the
tailings pond. Suggested change: eliminate the first phrase "In a TSF".
EPA Response
The change has been made in the revised assessment.
Reviewer Response
Satisfied with change.
Reviewer Response to Revised Bristol Bay Assessment
The revised Assessment has been corrected.
Page
Initial Comment
Last sentence states: [water] treatment would continue until institutional failures ultimately
resulted in abandonment of the system, at which time untreated leachate discharges would occur.
This statement is not supported by any documentation and is not clear what is being implied.
Failure of governments? As stated in my response to questions, any mine plan must include
sufficient bonding and plans for reclamation, including necessary water treatment.
EPA RESPONSE:
Bonding and a perpetual trust could result in treatment for some period, but it possible that long-
term funding will not last to the end of time.
48
-------
Dr. Phyllis Weber Scannell
Reviewer Response
Satisfied with recognition that this is a potential problem.
Reviewer Response to Revised Bristol Bay Assessment
Reviewer notes that the issue of long-term or perpetual treatment is mentioned in the revised
Assessment. The initial comment has been addressed.
Page
Initial Comment
End of paragraph states "premature closure could leave waste rock piles in place." Again, there
is a need for plans for mine closure, concurrent reclamation and sufficient bonding.
EPA RESPONSE:
The premature closure scenario is intended to address the possibility that the closure plan is not
followed. It highlights the need for adequate bonding and for plans that include closure by some
agency.
Reviewer Response
Satisfied with recognition that this is a potential problem. Adequate bonding and plans for
unexpected (temporary or permanent) closure are critical.
Reviewer Response to Revised Bristol Bay Assessment
The reviewer is satisfied with the approach used in the revised Assessment.
App G, Paf 1 . :t II " ¦
Initial Comment
Document states: ".. .other short road segments connect Dillingham to Aleknagik and Naknek to
King (Figure 1)." Shouldn't King be King Salmon?
EPA RESPONSE:
Corrected.
Reviewer Response
Satisfied with correction.
Reviewer Response to Revised Bristol Bay Assessment
The revised Assessment has been corrected.
49
-------
Dr. Phyllis Weber Scannell
App G, Paf
Initial Comment
Dolly Varden should be capitalized throughout the document. The list of fish species should
contain scientific names or reference a table of common and scientific names.
EPA RESPONSE:
Corrected.
Reviewer Response
Satisfied with correction.
Reviewer Response to Revised Bristol Bay Assessment
The revised Assessment has been corrected.
Initial Comment
The document states "In the most comprehensive published field inventory, Woody and
O'Neal (2010) reported." Because the authors have not reviewed other documents on field
inventories, the phrase "most comprehensive" should be changed to "a comprehensive"
EPA RESPONSE:
Corrected.
Reviewer Response
Satisfied with correction.
Reviewer Response to Revised Bristol Bay Assessment
The revised Assessment has been corrected.
Additional comment on revised Assessment
The revised Assessment has been substantially re-organized and includes important additional
information. EPA is to be commended for an excellent job. The revised Assessment is
50
-------
Dr. Phyllis Weber Scannell
thoroughly documented, contains better explanations and incorporates the concerns of this
reviewer.
51
-------
Dr. Phyllis Weber Scannell
Comments on Revised Appendix I
Revised Appendix I: Conventional Water Quality Mitigation Practices for Mine Design,
Construction, Operation and Closure
Many of the comments I raised in the initial review focused on the need for sufficient planning,
design, construction, management and closure of waste and water containment and for testing
materials for acid rock drainage (ARD) and metals leaching (ML). I stated these concerns in the
context of a need for an in-depth mining plan of operations. Appendix I addresses these
concerns; Page 1, paragraph 2 states: "The most important aspects of mitigation for any mining
site are proper planning, design, construction, operation, management, and closure of waste and
water containment and treatment facilities, and monitoring and maintenance over all mine-life
phases, including following closure." On page 3, the author addresses the issue of progressive
reclamation, an additional issue I raised in my initial comments. The information contained in
this Appendix satisfies many of the concerns I had with the initial Assessment. I have a few
minor comments, as listed below.
Page 5, Section 1.2 Accidents and Failures.
The author states "If waste rock piles are designed properly with appropriate mitigation
measures, monitored and maintained, release of contaminants is possible, but unlikely ..." I
would like to add "and the waste rock is adequately tested to classify as non-acid producing and
non-metals leaching." There are examples of mines in Alaska where supposedly benign waste
rock was used to construct berms, etc and later proved to be both metals leaching and acid
producing because the testing to classify the rock was inadequate to predict long-term chemistry.
Section 2.1.1 Operational Phase, page 8, paragraphs 3 and 4.
The author discusses different methods for constructing tailings impoundments and their relative
stabilities and the design criterion to potential earthquakes. This discussion addresses concerns
raised by the review committee in our initial review.
Section 2.1.2 Closure and Post-Closure
The author states "Sufficient capital is required to finance inspections, maintenance, and repairs
in post-closure for as long as the tailings exist." This section of Appendix I recognizes many of
my initial comments about treatment into perpetuity and a need for sufficient bonding. I am
pleased that the discussions of closure and post-closure have been included.
Overall Comments
Appendix I contains a well-documented, thorough and informative discussion of possible
mitigation practices for mine design, construction, operation and closure along with possible
52
-------
Dr. Phyllis Weber Scannell
sources of failure. The appendix is an important contribution to the Assessment and addresses
many of the comments and concerns I raised in the initial review.
Comments on New Appendix J
Appendix J: Compensatory Mitigation and Large-Scale Hardrock Mining in the Bristol Bay
Watershed
Appendix J provides the legal framework for compensatory mitigation (Section 1) followed by a
discussion of different approaches. Section 2 relates more directly to the Bristol Bay area.
Section 2 contains a summary of the important ecological functions of the headwater streams and
associated wetlands in this region. These sections (1 and 2) provide background and authority
for compensatory mitigation and are important additions to the revised Assessment. The
Appendix next presents different opportunities for mitigation, most of which would not be
beneficial. The arguments presented in this section are well-documented and clearly describe the
limitations. The author concludes "There are significant challenges regarding the potential
efficacy of compensation measures proposed by commenters for use in the Bristol Bay region,
raising questions as to whether sufficient compensation measures exist that could address
impacts of this type and magnitude.
I found this discussion to be realistic, well-documented and defensible. There are no changes I
would recommend. In fact, it is well-written and provides important supporting information to
the Assessment.
Task 2(1): 1
Task 2(2): 1
Task 2(3): 1
Initial Comments by Phyllis Weber Scannell (August, 2012) Responses by US EPA follow on
comments by Phyllis Weber Scannell 1
SUMMARY OF KEY RECOMMENDATIONS FROM PEER REVIEWERS 1
Scope of the Document: 1
Risks to Salmonid Fish 6
Wildlife 8
Human Cultures 9
Water Balance/Hydrology 10
Mitigation Measures 13
Uncertainties and Limitations 14
53
-------
Dr. Phyllis Weber Scannell
Editorial Suggestions 15
Research Needs 18
RESPONSES TO CHARGE QUESTIONS 22
Question 1 22
Question 2 23
Question 3 24
Question 4 27
Question 5 29
Question 6 30
Question 7 32
Question 8 34
Question 10 38
Question 11 39
Question 12 40
Question 13 43
Question 14 45
3. SPECIFIC OBSERVATIONS 47
Page 4-11: 47
Page 4-23 (LI): 48
Page 6-36 (P3): 48
Page 6-37 (P4): 49
App G, Page 5 (P 1, last line): 49
App G, Page 5 (P3): 50
Additional comment on revised Assessment 50
Comments on Revised Appendix 1 52
Comments on New Appendix J 53
54
-------
Dr. Paul Whitney
Peer Review Follow-On Comments
on the April 2013 Draft of
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
-------
From: Paul Whitney
Sent: Wednesday, July 10, 2013 9:16 AM
To: Thomas, Jenny
Subject: Bristol Bay Peer Review
Jenny,
Here are my responses incorporated into the Peer Review document. As
I have mentioned before, word processing is not a strength. You might
want to have a word processor touch up my responses? Let me know if
you or others have any questions.
Do I need to submit an invoice or is this submittal adequate to get the
compensation going?
Paul
-------
Dr. Paul Whitney
DRAFT
EPA Response to Peer Review Report
External Peer Review of EPA's Draft Document
An Assessment of Potential Mining Impacts on
Salmon Ecosystems of Bristol Bay, Alaska
11 June 2013
U.S. Environmental Protection Agency
Seattle, WA
DRAFT - EPA Use Only - Do Not Distribute, Cite or Quote - DRAFT
-------
Dr. Paul Whitney
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS
An Assessment of Potential Mining Impacts on Salmon Ecosystems
of Bristol Bay, Alaska
OVERVIEW
In May 2012, the U.S. Environmental Protection Agency (EPA) released the draft assessment
entitled An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay,
Alaska (hereafter, the assessment). Because this document is considered a highly influential
scientific assessment, it has strict requirements for peer review as laid out in the EPA's Peer
Review Handbook (3rd Edition). These requirements included peer review by external expert
reviewers and documentation of how peer review comments are incorporated into the revised
work product.
From May to August 2012, this draft assessment was evaluated by twelve external expert
reviewers; comments from these reviewers were incorporated into a final peer report, which was
submitted to the EPA in September 2012. This document details the EPA's draft responses to the
peer review comments received on the May 2012 draft of the report, as provided in the final peer
review report. Please note that this is a draft document, and that it should not be distributed
beyond the EPA and the peer reviewers. It is considered deliberative material and intended for
internal use only, as the assessment continues to undergo revision. Once the assessment has been
finalized, a final draft of this document will be completed and released to the public.
STRUCTURE OF THE DRAFT RESPONSI. TO COMMENTS DOCUMENT
This draft response to comments document follows the structure of the final peer review report
for the assessment. It is organized into two main sections:
(11 Summary (if Key Recommendations from Peer Reviewers. This section details
general big-picture issues raised by the peer reviewers in the August 2012 peer review
meeting.
(2) Written Peer Review Comments. This section details each reviewer's individual
comments on the assessment, organized according to general impressions, responses to
charge questions, and specific observations.
Within each section, the original text from the final peer review report is included here, followed
by EPA responses to the comments contained in that text (in bold italics). The EPA considered
all comments provided in the final peer review report, although not all comments resulted in
changes to the draft assessment. The EPA responses provided in this document explain either
how the assessment was revised to address the comment, or why the assessment was not changed
in response to the comment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 2
-------
Dr. Paul Whitney
ORGANIZATION OF THE APRIL 2013 DRAFT ASSESSMENT
In response to several comments, the revised April 2013 draft of the assessment has been
reorganized to highlight its foundation as an ecological risk assessment, improve overall clarity,
and reduce redundancy. A cross-walk between chapters of the May 2012 andApril2013 drafts is
provided below to help illustrate where material from the May 2012 can be found in the revised
assessment.
May 2012 Draft
April 2013 Draft
1. Introduction
Introduction
Characterization of Current Condition
Overview of Assessment
Region
Problem Formulation
Type of Development
Endpoints
Mine Scenarios
4. Mining Background and Scenario
Mine Footprint
Water Collection, Treatment, and Discharge
Risk Assessment: No Failure
Tailings Dam Failure
10. Transportation Corridor
11. Pipeline Failures
6. Risk Assessment: Failure
12. Fish-Mediated Effects
Cumulative and Watershed-Scale Effects of Multiple Mines
13. Cumulative Effects of Large-Scale Mining
8. Integrated Risk Characterization
14. Integrated Risk Characterization
Cited References
15. References
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 3
-------
Dr. Paul Whitney
SUMMARY OF KEY RECOMMENDATIONS FROM PEER
REVIEWERS
This section summarizes the significant general recommendations put forth by the peer reviewers
regarding EPA's draft assessment. In developing these recommendations, peer reviewers
provided input on three major areas of the assessment: (1) scope, (2) technical content and (3)
editorial suggestions. Reviewers also identified research needs for EPA to consider. Please note
that this summary of peer review comments did not reflect a consensus or group perspective, but
was compiled from a discussion of individual peer reviewer recommendations. Additional
details, including references cited, can be found in the reviewers' individual comments in
Section III.
PHW Response: The revised assessment is responsive to many of the suggestions offered by the
peer review members. As a result the revised assessment is much improved.
My comments on the EPA Draft Assessment were made before I met the other members of the
peer review group and before I was -familiar with their expertise. Once I was able to discuss my
many concerns with other members of the peer review group it became clear that many of my
concerns would be addressed by reviewers with more expertise in a particular area. For example,
I was concerned about the level of detail given to water balance calculations, cone of depression
and water routing. Others with mine engineering and hydrology expertise expressed similar
concerns, outlined specific problems and offered more detailed suggestions for ways to improve
the assessment. In this current review, I mostly defer my responses to EPA responses to those
with more expertise. As a result, for the most part, I limit my responses to issues related to
wildlife. If I do not offer a comment regarding an EPA response, one can assume I agree with the
EPA information offered or have addressed the topic in another place in this document.
Needless to say, I do not agree with the EPA decision not to include direct impacts on wildlife
and wildlife functions in their assessment. The decision to only address fish mediated impacts
on wildlife side steps the facts that wildlife are an important component of the salmon ecosystem
and that impacts on wildlife mediate impacts on fish. I doubt if EPA will change their
position/decision based on my lone voice for a true Bristol Bay ecosystem analysis that includes
direct impacts on fish and wildlife and their habitats. None-the-less, I state my contrary opinions
as I feel so moved based on my scientific training. My contrary opinions are expressed one time
in the following PHW Responses. I save my time and EPA's time by not repeating the same
contrary opinions over and over.
Scope of the Document:
• Articulate the purpose of the document more clearly via a primer on the Ecological Risk
Assessment process. If the purpose of the assessment is to inform EPA as the decision maker,
then the level of detail should correspond to this purpose. The authors should justify and
explain what level of detail is required.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (I JiA) in general has been added to Chapters 1 and 2, as well as the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 4
-------
Dr. Paul Whitney
Executive Summary. Section 1.2 includes information about the use of the assessment.
The assessment has been reorganized into two major sections (problem formulation, risk
analysis and characterization) to clarify where different chapters fall in the typical ERA
process.
• Include a statement upfront about the role of risk managers and other audiences, such as
project managers/engineers, regulators, mine owners/operators. Knowing their role ensures
inclusion of information necessary for any risk assessment by (1) describing the need for a
risk assessment, (2) listing those decisions influenced, and (3) characterizing what risk
managers require from the risk assessment.
RESPONSE: Section 1.2 of the revised assessment discusses the use of the assessment.
• Explain why the scope for human and wildlife impacts was limited to fish-mediated effects,
as well as why fish-mediated effects on humans were limited to Alaska Native cultures.
Reviewing effects beyond fish-mediated ones (e.g., potential for complete loss of the
subsistence way of life) would improve the assessment.
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters 1 and 2. Throughout the assessment we acknowledge that direct
effects of large-scale mining on midlife and Alaska Native cultures may be significant, but
that these direct effects are outside the scope of the current assessment.
PHW Response: The revised assessment does a better job of clarifying that direct effects of
mining on wildlife are outside the scope of the current assessment but does not explain
why direct effects on wildlife are outside the scope of work. The concept that wildlife are
not a component of the salmon ecosystem is not addressed If wildlife are thought not to be
a component of the salmon ecosystem, the assessment should explain why this is the case.
• Be more consistent throughout the document in terms of the level of detail provided for the
different scenarios and stressors. For example, the document has devoted 36 pages to the
discussion of catastrophic Tailings Storage Facility (TSF) failure, while sections on the
pipeline, water treatment, and road/culvert failures are brief. Indeed, the long discussion on
the TSF failure belies a certainty and understanding of dam failure dynamics that is
inaccurate.
RESPONSE: The final document includes more failure scenarios (e.g., diesel pipeline
failure, wastewater treatment plant failure, and refined seepage scenarios) in Chapter 8. It
also explains why these specific failure scenarios were chosen, and discusses these
scenarios in greater detail than the previous draft (i.e., to more closely match the level of
detail originally provided only for the TSFfailure scenario). Also see detailed responses to
comments on Peer Review' Question 5.
Technical Content:
Mine Scenario
• Consider the document to be a screening-level assessment of all potential stressors. Focusing
on failure mode overemphasizes catastrophic events (e.g., TSF failing), rather than
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 5
-------
Dr. Paul Whitney
considering all potential stressors, such as holding mine owners strictly accountable for their
day-to-day activities with regard to best practices.
RESPONSE: Additional information on the purpose and scope of the assessment has been
added to Chapters 1 and 2. A screening of aU. potential stressors, including individual
chemicals, is presented in Section 6.4.2. Also see detailed responses to Peer Review
Question 2 on the use of "best practices " and responses to Peer Review Question 5 on
failure scenarios.
• Reexamine the document's use of historical data and case studies to describe and estimate the
risk of failure for certain mine facilities (including the TS1*. pipeline, water treatment, etc.),
as these examples from extant mines may not be an appropriate analog for a new mine in the
Bristol Bay watershed.
RESPONSE: The TSFfailure range was, and still is, based on design goals, not the
historical data. The historical TSF failure data are provided as background The pipeline
failure rates are based on the most relevant historical data from the petroleum industry.
They are directly relevant to the diesel pipeline, and experiences at the Alumbrera mine
(described in the previous draft) and the Antamina and Bingham Canyon mines (added to
this draft) suggest that they also are relevant to the product concentrate pipeline. Water
treatment failure rates were not quantified. However, recent reviews cited in the revised
draft indicate that, water collection and treatment failures have been reported at nearly all
analogous mines in the U.S. The estimation of culvert failure frequencies has been revised
and is now based on only recent literature (2002 and later). II e believe that these estimates
are appropriate.
• Expand the discussion on the use of "best" management practices, as the document states that
the mine scenario employs "good," but not necessarily "best" practice. For a mine developed
in the Bristol Bay watershed, only "best" practice likely would be appropriate and anything
less may not be permitted. Even so, without a track record of "best" practice (e.g., new
technologies), we cannot assume that technology, by itself without appropriate operational
management controls, can always mitigate risk.
RESPONSE: The term "best management practices " is a term generally applied to
specific measures for managing non-point source runoff from storm water (40 CFR Part
130.2(m)). Measures for minimizing and controlling sources of pollution in other
situations often are referred to as best practices, state of the practice, good practice,
conventional, or simply mitigation measures. We assume that these types of measures
would be applied throughout a mine as it is constructed, operated, closed, and post-closure,
and have used the term "conventional modern" throughout the assessment to refer to these
measures. To remove any ambiguity related to the subjectiveness of terms "good" or
"best", we have removed them in the revision and have provided definitions for relevant
terms used in Box 4-1.
• Adopt a broader range of mine scenarios (not only minimum and maximum) so as to bound
potential impacts, especially at smaller mine sizes (e.g., 50th percentile). Underground mine
development, with its different impacts, also should be considered and included in the
assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 6
-------
Dr. Paul Whitney
RESPONSE: A third mine size scenario (250 million tons) has been added to the
assessment, to represent the worldwide median sized porphyry copper mine (based on
Singer et aL 2008).
• Based on the hypothetical mine scenario, perpetual management of the geotechnical integrity
of the waste rock and tailings storage facilities, as well as perpetual water treatment and
monitoring, will most likely be necessary (i.e., a "walk away" closure scenario after mining
ends may not be possible). Therefore, emphasize how monitoring and management of the
geotechnical integrity of waste rocks and tailing storage facilities should continue "In
Perpetuity" (i.e., for at least tens of thousands of years). Discuss what conditions would need
to be met to allow "walk away" closure in the Bristol Bay environment gaining insight into
these observations from mines where perpetual treatment and monitoring are ongoing (e.g.,
the Equity Silver Mine in British Columbia).
RESPONSE: The conditions for closure and the potential need for perpetual site
management are discussed in general terms in the revised assessment. The primary
condition assumed to be required is water chemistry that meets all criteria and permit
conditions and that is stable or improving. However, even though there are some facilities
with "perpetual treatment" conditions in place, there is obviously no information about
how these facilities perform over very long periods of time.
• Identify, in technical detail, how exploratory effects (e.g., drill holes, blasting, overflight,
etc.) were managed. This includes roads, airstrips, helipads, camps, fuel dumps, and ATV
trails that have already been developed or imposed on the watershed, and what "mitigation"
already has been undertaken on those sites. Assess the consequences/impacts of these
activities in the Cumulative Risks section.
RESPONSE: The effects of exploratory activities are outside of the scope of this
assessment.
PHW Response: If exploratory activities are outside the scope of the assessment it may not be
possible to conduct a Cumulative Assessment that should address past activities. Formatted: Font: Bold
Risks to Salmonid Fish
• Place potential mining impacts in the context of the entire Bristol Bay watershed by
emphasizing the relative magnitude of impacts. For example, of the total salmon habitat,
assess the proportion lost due to mining. Further, reflect on the non-linear nature of the
relationship between habitat and salmon production; 5% of the habitat could be critical and
thus responsible for 20% or more of salmon recruitment. Intrinsic potential, which measures
the ability of particular habitats to support fishes, would lend credibility to this analysis.
RESPONSE: We are unable to build a complete Intrinsic Potential (IP) model, as this
would require validation and more elaborate construction of metrics appropriate to this
region. Our preliminary characterization provides the building blocks for assessing the
distribution of key habitat-forming and constraining features across these watersheds. We
now include a characterization of the major drivers of habitat potential across the
watershed and place the mine-site specific effects in this context (Chapters 3, 7, and 10).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 7
-------
Dr. Paul Whitney
• Include a section on the impact of Global Climate Change with explicit reference to a
monitoring program that will allow scientists, if the mine is built, to distinguish between
effects of climate change and mining effects on the physical and biological components of
this ecosystem.
RESPONSE: Climate change projections and potential impacts are now included in
Chapter 3, and as important external factors in the risk analyses presented in Chapters 7,
9,10, and 14. Development of a monitoring program to distinguish behveen mining and
climate change effects is outside of the scope of the assessment.
• Explicitly recognize that the transportation corridor and all associated ancillary development,
including future resource developments made possible by the initial mining project, will
necessarily and inevitably have impacts (hydrologic, noise, dust, emissions, etc.). These
impacts will vary in duration, intensity, severity, relative importance, spatial dispersion, and
inevitably expand geographically through time with further "development." These impacts
should be incorporated into the Cumulative Risks section.
RESPONSE: The cumulative risk section (Chapter 13) has been expanded to include the
multiple transportation corridors, ancillary mining development and secondary
development associated with multiple mines in a qualitative discussion. The issues
addressed in the assessment of the transportation corridor (Chapter 10) have also been
expanded to include chemical spills, dust, invasive species, and road treatment salts.
PHW Response: The Cumulative Risk Section does not address the impact of commercial
fishing. Merely stating that the commercial fishery is sustainable and closely monitored does
not mean that it has no impact on the salmon ecosystem. I cited HilbonTs (2005 - citation in
original contents) opinion that fishing has an impact in my first set of comments and I
commented that commercial fishing reduces the amount of MDN returned to the salmon
ecosystem. Neither of these points are addressed. 4Formatted: Font: Not Bold, Not italic
• Incorporate current research findings into stream crossing and culvert-design practices (e.g.,
arch culverts, bridges, etc.).
RESPONSE: We describe current culvert design practices in a box titled "Culvert
Mitigation" in Chapter 10.
• Recognize in the assessment that risk and impact are not equivalent. Risk may be low, but the
potential impact could be huge (e.g., in the case of a TSF failure).
RESPONSE: Risk has been defined in many ways, even by risk assessors. The commenter
seems to define risk as probability. To avoid that potential source of confusion, we use the
term "probability"for that concept. Similarly, the commenter seems to use "impact"
where we use "effect" or "magnitude of effect". We use "risk" to refer to both concepts
combined—that is, an event or effect and its probability).
• Recognize and justify chronic behavioral endpoints, such as those potentially affecting
survival and long-term success of fish populations.
RESPONSE: The chronic behavioral effects of copper on salmonids, the primary endpoint
of concern, were described in Chapter 5 and are now described in Chapter 8. Although
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 8
-------
Dr. Paul Whitney
those effects occur at lower levels of copper than conventional survival, growth and
reproduction endpoints for salmonids, they are less sensitive than the conventional
endpoints for aquatic invertebrates.
Wildlife
• Recognize that the draft assessment did not account for all levels of ecology, such as the
individual (e.g., a bald eagle nest), population, community, ecosystem, and landscape levels.
Fold other levels of organization into the stressors assessment where appropriate or justify a
more limited approach.
RESPONSE: As is appropriate for an ecological risk assessment (as opposed to an
environmental impact assessment), this assessment focuses on a specific, limited set of
endpoints as defined in Chapter 5. We have added text in Chapters 2 and 5 to explain both
why these endpoints were selected, and that responses other than those considered in the
assessment, at multiple levels of ecological organization, are likely but are outside the
scope of the assessment.
PHW Response: The revised assessment does a much better job of defining endpoints. Yet
more work needs to be done to define the salmon ecosystem. For example, if multiple levels
of ecological organization, such as "ecosystem" are outside the scope of the assessment as
stated above. EPA should consider changing the current title of the document that focuses the
assessment on "salmon ecosystems".
EPA's assessment discusses a few ecosystem functions and passes over many others. For
example, salmon perform an important ecosystem function bv returning MDNs to the
watershed portion of the Bristol Bay ecosystem. I can't help wonder why this ecosystem
function is discussed in detail but is supposedly "outside the scope of the assessment."
Wildlife provide a variety of ecosystem functions in the salmon ecosystems but are judged to
be "outside the scope of the assessment." So it appears that some ecosystem functions such
as returning MDNs are within the scope of work but some aspects of returning the MDN
(i.e.. movement of MDN to terrestrial components of the watershed are outside the scope of
work. Are wildlife a component of the salmon ecosystem or not? If thev are not, don't
mention wildlife's role in the MDN processes. I'm not clear where the salmon ecosystem
(structure and function) starts and where it stops in the watersheds.
The assessment title states .salmon ecosystems of Bristol Bay..." but has not included the
impact of a major predator in the salmon ecosystems, commercial fishing, in Bristol Bay, as
part of the analysis. Not including the impact of commercial fishing in the assessment and
not comparing the impact of commercial fishing to potential impact(s) of the mine appears to
side step a very important issue that should be explicitly addressed in any direct or
cumulative assessment of mining impacts. Would the mine bring an end to commercial
fishing in Bristol Bay? greatly reduce commercial fishing? slightly reduce commercial
fishing? or have little or no effect on commercial fishing? The answer to these questions
seem critical to a CWA assessment., Formatted: Font: 11 pt. Not Bold, Not Italic )
• Discuss in the document fishes other than salmonids The assessment focuses on risks to
sockeye salmon in the Bristol Bay watershed (and also considers anadromous salmonids,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
Page 9
-------
Dr. Paul Whitney
rainbow trout, and Dolly Varden), but does not account for potential impacts to other
members of the resident fish community. Further, primary and secondary production,
including nutrient flux was not addressed. Expanding the assessment to consider other levels
of organization, including direct as well as indirect effects on wildlife and other fish, would
provide additional context in the assessment of mine-related impacts.
RESPONSE: See response to comment above; we also incorporated additional
information from Appendices A, B, and C into the Chapter 5 text, to provide additional
detail on the area's biota We chose our endpoints for reasons described in Chapters 2 and
5. Other endpoints, including indirect effects on fish and wildlife, are now discussed more
explicitly, but are generally considered outside the scope of the assessment
Human Cultures
• Use case histories to provide insight and anticipate mining impacts on Alaska Natives (e.g.,
those exemplifying the Exxon Valdez oil spill impacts, cumulative effects of oil and gas
development in the North Slope region, and social impacts related to mining development in
Alaska).
RESPONSE: Examples from applicable case studies, including the Exxon Valdez oil spill,
are cited in Chapter 12 of the revised assessment.
• As noted above (Scope of the Document), clarify why the scope was limited to fish-mediated
effects. The potential direct and indirect impacts for human cultures extend far beyond fish-
mediated impacts (e.g., potential complete loss of the subsistence way of life). The rationale
for this narrow Ibcus should be fully explained. In addition, a clear explanation should be
given for why fish-mediated human impacts focused only on Alaska Native cultures.
RESPONSE: The assessment focuses on a specific, limited set of endpoints as defined in
Chapter x JVe have added text to explain both why these endpoints were selected, and that
responses other than those considered in the assessment are likely but are outside the
scope of the assessment. The assessment was expanded (Chapters 5 and 12) to
acknowledge that there are a wide range of potential direct and indirect impacts to
indigenous culture, but they are outside of the scope of this assessment. The discussion of
potential effects to indigenous cultures was expanded to explain that a loss of subsistence
resources would extend beyond a loss of food resources to social, cultural, and spiritual
disruption. The text has been expanded to acknowledge the strong cultural ties of many
non-Alaska Natives to the region, and potential effects on all residents from loss of a
subsistence way of life. However, the focus of the assessment remains on effects on
indigenous cultures resulting from effects on salmon.
Water Balance/Hydrology
• Better characterize water resources and assess the potential effect of mine development on
these resources by (1) generating a diagram similar to the conceptual models beginning on
page 3-7 to illustrate the potential effects of mine construction and operation on surface- and
ground-water hydrology; (2) developing a quantitative water balance and identifying water
gains and losses; (3) identifying seasonality of hydrologic processes, including frozen soils
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 10
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
and their associated values (e.g., mm/yr) for each component of the water balance; (4)
incorporating these processes into a landscape characterization; (5) evaluating how global
climate change will influence these hydrologic processes and rates; and 6) using this
characterization to demonstrate the expected hydrologic modification associated with the
mine scenarios and infrastructure development.
RESPONSE: The original Figure 4-9 (new Figure 6-5) has been revised to more clearly
show water management in the assessment's mine scenarios. In addition, three schematics
illustrating water flows under each of the mine size scenarios (Figures 6-8 through 6-10)
have been added to Chapter 6, as have quantitative water balances for each mine size
scenarios. A qualitative discussion of climate change is included in Chapters 3 (Section
3.8) and 14 (Box 14-2).
• Demonstrate the interconnectedness of groundwater, surface water, hyporheic zone, and its
importance to fish habitat. Address how interconnectedness changes over time - seasonally,
and with varying weather (e.g., wet vs. dry summers or years, and over the long term as
climate changes).
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
(Chapter 7) and qualitatively discuss the potential role oj climate change (Chapter 3).
• Provide information on all rivers, including ephemeral and intermittent streams, and first-
order to main-stem streams that could be potentially influenced by the proposed mine, its
ancillary facilities, and the transportation corridor.
RESPONSE: Due to lack of consistent coverage, we rely on the NHD hydrography layer in
this analysis, and can only address ephemeral and intermittent streams qualitatively
(Chapter 7).
• Emphasize the importance of a thorough characterization of the leaching potential of acid-
generating and non-acid generating waste rock and tailings, given the low buffering capacity
and mineral content in the streams and wetlands that could receive runoff and treated water
from the proposed mine. Recognize that collection and treatment of runoff and leachate
generated will be critical to maintain baseline water chemistry in these streams and wetlands.
RESPONSE: We agree that these are important issues, and the discussion of leachate
from waste rocks and tailings lias been expanded in the revised assessment (Chapter 8).
Geochemistry/Metals
• Reference the most current geochemistry data on potentially acid-generating, non-acid
generating, and metal leaching so as to describe any potential effects of seepage and changes
to surface- and ground-water quality via non-catastrophic failure.
RESPONSE: We used the geochemistry data in PLP's Environmental Baseline Document,
as summarized by the USGS in Appendix H. The effects of seepage on water quality are
analyzed in Chapter 8 of the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 11
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
• Explain how contaminants/metals were selected (and others ignored) by EPA as causes for
concern. Information should be included on additional metals and their toxicity so as to
assess impacts of potential leachates. The Pebble Limited Partnership baseline document
presented additional metals that might be useful to include in the assessment.
RESPONSE: The revised assessment describes the selection of contaminants and other
stressors of concern in Section 6.4.2. Additional metals, process chemicals and dissolved
solids are now included
Mitigation Measures
• Incorporate the critical mitigation information from Appendix I into the main report's mine
scenarios. Include standard mitigation measures that could provide insight into how well they
might work in this context. If this information is not included in the main report, then justify
its absence.
RESPONSE: Mitigation measures incorporated into design and operation to minimize
potential impacts were included in the assessment, as were some reclamation measures for
closure; these measures are made clearer in the revised assessment. These mitigation
measures were a sub-set of those presented in Appendix I. The assessment assumes that
measures chosen for the scenarios would be effective. Mitigation to compensate for effects
on aquatic resources that cannot be avoided or minimized by mine design and operation
would be addressed through a regulatory process that is beyond the scope of this
assessment. Nevertheless, in response to public and peer comments we have included a
discussion of compensatory mitigation in Appendix J of the revised assessment.
PHW Response: Any discussion of CWA Section 404 should at least mention "No Net
Loss" and the likelihood that No Net Loss could be achieved. The top of page 12 of
Appendix J implies that No Net Loss cannot be achieved with in the salmon ecosystems of
Bristol Bay. If this is the case, it should be clearly stated.
I am pleased that Appendix J discounts the proposal to remove beaver dams.
The EPA response to the ("rhetorical) suggestion for discussion of the concept that
commercial fishing could be reduced as a mitigation measure (Appendix J Section 3.3.2.6)
seems to assume that commercial fishing is not at all related to or concerned with the
potential impact(s) of the mine. L on the other hand, consider commercial fishing to be an
important component of the salmon ecosystem of Bristol Bay. EPAs broad jump in logic
from Hilborn's paleontological analysis to the conclusion of no opportunity for mitigation is
VERY wide and appears to obfuscate Hilborn's (2005) statements that fishing, as most
human activities, has a negative impact on biodiversity. As stated above and below, the
revised assessment would be improved by addressing commercial fishing for several reasons:
1. A cumulative impact analysis should consider past impacts and commercial fishing
has had a past impact on the salmon ecosystem.
2. Some commercial fishermen are likely native Americans and a reduction in the
commercial take could be a fish mediated impact.
Formatted: List Paragraph, Numbered +
Level: 1 + Numbering Style: 1, 2, 3,.
. + Start
at: 1 + Alignment: Left + Aligned at:
0.5" +
Indent at: 0.75"
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 12
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
3. Commercial fishing has definitely reduced the number of fish returning to spawn
and returning MDNs to the ecosystem.
4. The assessment certainly addresses the positive importance of MDNs to the salmon
ecosystem. If it addresses the positive importance, it should also address the
negative importance of the past loss of MDNs due to the commercial take^ Formatted: Font:
I was hoping to see some analysis of using diversion ditches as opportunities to mitigate for
both wetland and riparian impacts. What I saw was a general statement that spawning
channels are not good mitigation. I realize that some mitigation measures are not successful
but my experience is that with proper engineering and biological success criteria plus a good
monitoring program it is possible to achieve effective wetland, riparian and fish mitigation. A
good example is the Weyerhaeuser Headquarters Landfill in Washington where Interfluve
Consultants created good habitat for cutthroat trout out of a run-on-diversion channel around
the landfill footprint. There are many more examples of effective fish mitigatin in created
channels that a brief conversation with Interfluve (Hood River. Oregon) could provide. k Formatted: Font: Not Bold, Not italic
• Emphasize mitigation measures (e.g., minimization, compensation, reclamation) in the main
report, as they ultimately influence the range of mining impacts and consider time frames of
mitigation or reclamation measures (e.g., immediate response, long-term reclamation).
RESPONSE: See response to previous continent. Mitigation measures are discussed at
greater length in the revised assessment report (e.g.. Chapter 4 and Appendix J).
Uncertainties and Limitations
• Clarify the uncertainty vs. certainty in Chapter 8 by (1) defining levels of uncertainty and (2)
assessing the certainty of some mine impacts. Discuss data limitations in the context of
uncertainty.
RESPONSE: The individual analysis chapters and the revised Integrated Risk
Characterization (Chapter 14) discuss certainties and data limitations to a greater extent,
as suggested
• Articulate early in the document how much uncertainty is acceptable. The assessment
provides little insight with respect to the decisions the document is intended to support.
RESPONSE: Acceptable levels of uncertainty can be defined prior to an assessment if a
decision and a decision maker are identified and if data will be collected by a specified
design to implement a specified model, as described in the EPA's Data Quality Objectives
process, However, because this assessment is based on available data and is intended as a
background scientific document rather than a decision document, it is not possible to
specify the amount of uncertainty that is acceptable. Rather, the available data determine
the uncertainty and if the assessment is subsequently used to inform a decision, the
decision maker must determine whether the level of uncertainty is acceptable.
Editorial Suggestions:
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 13
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
• The title of the document leads one to believe that the assessment addresses the entire Bristol
Bay watershed; rather, the report deals with two major rivers and their watersheds, the
Nushagak and Kvichak. Thus, the title should be changed to reflect the emphasis on these
two rivers and their watersheds. A possible title may be "An Examination (or identification)
of the Potential Impacts of Mining and Mining Associated Activities on Salmon Ecosystems
in the Nushagak River and Kvichak River watersheds, Bristol Bay."
RESPONSE: The assessment addresses multiple scales: the Bristol Bay watershed, the
Nushagak and Kvichak River watersheds, the watersheds of the three streams draining the
Pebble deposit, and the watersheds crossed by the transportation corridor. These multiple
scales, and how they are used throughout the assessment, are described more clearly in the
revision (Chapter 2).
• Revise the Executive Summary to more precisely reflect the findings in the document.
RESPONSE: The Executive Summary has been rewritten to reflect the revised assessment
findings.
• The appendices contain detailed and useful information that should be summarized and
included in the main document (e.g., Appendix E: Economics, Appendix G: Road and
Pipelines, and Appendix I: Mitigation). Additionally, consider expanding the preface to
include information on the use of the appendices. If the information is not included in the
main report, then justify its absence.
RESPONSE: More information from the appendices was brought forward into
appropriate chapters of the revised report. The purpose of the appendices—to provide the
detailed background characterization necessary for the ecological risk assessment—has
also been clarified in Chapter 2. The document no longer contains a preface because that
material has been incorporated into Chapters 1 and 2.
• Discuss in more detail the instructive and well-thought-out conceptual models (pages 3-7 to
3-11) illustrating the impacts of mining on Bristol Bay ecosystem processes. Also, consider
expanding the conceptual models to include wildlife, fish-wildlife interactions,
vegetation/terrestrial habitat, and hydrologic processes. Allow them to guide the text because
they appear detailed and complete.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual
models presented in Chapter 6 (Chapter 3 in the first draft) have been broken into their
relevant component parts throughout the risk analysis and characterization chapters, to
better frame the specific pathways addressed in each chapter. Additional conceptual
models considering impacts on wildlife, Alaska Native populations, and cumulative effects
of multiple mines have been added to Chapters 12 and 13.
PHW Response: The conceptual wildlife model in Chapter 12 is a good addition and clearly
shows that direct effects on wildlife are outside the scope of the current assessment. See
comments above and below for the need to explain whv direct impacts on wildlife are
I Formatted: Font: Bold
)
outside the scope and whv wildlife species are not a part of the salmon ecosystems of Bristol
Bay., ^
Formatted: Font: Not Bold, Not Italic
)
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 14
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
• Incorporate the information contained in the conceptual models into a formal framework,
such as a Bayesian or other decision-analysis models.
RESPONSE: This is an excellent suggestion for future efforts, but is beyond the scope of
the current assessment
Creating a Bayesian Belief Network would require that the Agency convene experts to
subjectively estimate the probabilities of each transition in the conceptual models. In
contrast, this assessment is intended to elucidate the risks from potential mining based on
available data and analyses of those data
A Decision Analysis would require that alternative outcomes be specified, the utility of
each outcome for a decision maker be defined and the probabilities of each outcome be
estimatedfor each possible decision so that the expected utilities of each outcome can be
calculated Because this assessment is not a decision document, these requirements are not
feasible or appropriate.
• Generate a standard operating protocol for significant figures and use it throughout the
document.
RESPONSE: The authors have caref ully addressed this issue. Numbers from the literature
or from the PLP EBD retain the number of significant figures in the originaL Numbers
derivedfor this assessment have the appropriate number of significant figures given the
precision of the input data and uncertainties due to modeling and extrapolation.
• Remove all references to Mount St. I Ielens as a surrogate for a TSF failure. Using a non-
human-caused release of material into the ecosystem as an analogue for a mine failure is not
comparable in terms of likelihood or risk for a human-caused release. It would be more
appropriate to extrapolate from the impacts of known mine failures.
RESPONSE: We are puzzled by this comment The Mount Saint Helens data were used
strictly to address the rate ofbentliic habitat recovery from a massive deposition offine
mineral particles. The hydrological processes that determine the recovery of substrate
texture and the requirements of fish or aquatic invertebrates are not known to depend on
whether mineral particles were from a natural event or an anthropogenic event. We have
reviewed the literature on known mine failures. They studied tailings spills in terms of
toxicity but not in terms of physical habitat effects, which is why we used Mount Saint
Helens data. Nevertheless, we have removed references to Mount St Helens in the revised
assessment to eliminate concern.
• Ensure that the draft assessment remains part of the public record, allowing the document
history to remain intact.
RESPONSE: All drafts of the watershed assessment will remain part of the public record
Research Needs:
• What are the acute and chronic impacts of mixtures of contaminants, including metals, acid
mine drainage, etc., on the fauna and flora of the Nushagak River and Kvichak River
watersheds? What species are most sensitive and might surrogate species exist for those for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 15
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
which we do not have data? Review the European literature and regulatory requirements for
additional data.
RESPONSE: The acute and chronic impacts of contaminant mixtures, including metals
and acid mine drainage (i e., metals in low pH-waters) were addressed using concentration
additivity models in the leachate chemistry tables in Chapters 5 and 6 (now Chapters 8 and
11). Additional toxicity data were obtained by searches of the EU and OECD database
eChem, the EPA's ECOTOX and the Environment Canada site. More metals are now
included In general, metals are most toxic to aquatic arthropods rather than fish, as
discussedfor copper.
• Can an inventory of nutrients, total organic carbon, and dissolved organic carbon inputs to
aquatic environments be developed that demonstrates their relative magnitude and spatial
variation from headwaters to Bristol Bay? What is the relative importance of marine-derived
nutrients relative to other nutrients from watershed and terrestrial sources? What is the
current atmospheric input of nutrients?
RESPONSE: These data would be very useful in the risk assessment, but are not currently
available for the Bristol Bay region. We agree this is a research need.
• What are the locations of subsistence areas and can these areas be characterized and
differentiated by collecting local environmental and ecological knowledge (e.g., fish
overwintering areas, climate change, ecological shifts, etc.)?
RESPONSE: The re\'ised assessment incorporated current data on subsistence use areas
available from ADE&G. EPA acknowledges that these data are incomplete and would
encourage additional collection of subsistence data and Traditional Ecological Knowledge.
• What impact might mining have on other important wildlife species in the basin (e.g.,
freshwater seals in Iliamna Lake)?
RESPONSE: The scope of the assessment is focused on potential risks to salmon from
large-scale mining and salmon-mediated effects to indigenous culture and wildlife. Direct
effects on wildlife from large-scale mining are likely to be important and Appendix C (now
a stand-alone US Fish and Wildlife report) provides useful information for a future
evaluation of direct effects on wildlife from large-scale mining. We agree that this is an
important area for future research.
• What is the comprehensive hydrologic regime of the specific project mining area, and the
broader watershed system as characterized by baseline monitoring, spatial distribution, and
quantitative flow of surface- and ground-waters?
RESPONSE: Comprehensive spatial estimates of mean annual flow are now presented in
Chapter 3. Quantification of spatial and temporal patterns of groundwater flows is an
acknowledged highly desirable product, but it not feasible within the scope of this
assessment. Results of an independent groundwater-surface water modeling effort are
described in Chapter 7.
• What is the cumulative impact of commercial fisheries on the Bristol Bay watershed,
especially in an ecosystem context as related to marine-derived nutrient and energy flow?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 16
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Acknowledge that commercial fishing has had an impact on the amount of marine-derived
nutrients returned to the watersheds.
RESPONSE: The impact of commercial fisheries on the watershed is not within the scope
of this assessment. Information on commercial fisheries management has been added in
Box 5-2. However, the purpose of this assessment is not to assess the relative effects of
potential mining and commercial fishing—it is to evaluate potential effects on endpoints if
a mine were to be developed, given existing conditions and activities in the region.
PHW Response: The logic used here is not clear. The logic assumes commercial fishing is
sustainable and therefore has no impact on salmon or the salmon ecosystem and therefore
commercial fishing can be ignored without an adverse consequence to the assessment. Such
an assumption is not factual. Commercial fishing kills70% of the sockeve every year. This
has to have an impact on the salmon ecosystems of Bristol Bay. If this is not the case explain
why and explain why Hilborn (2005) as cited in mv original comments is wrong.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 17
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
WRITTEN PEER REVIEW COMMENTS
1. GENERAL IMPRESSIONS
David A. Atkins, M.S.
The Bristol Bay Watershed Assessment (the Assessment) presents a comprehensive overview of
current conditions in the watershed and establishes the uniqueness and global importance of the
area to global salmon ecology (e.g., the report states that nearly 50% of the global sockeye
salmon population comes from Bristol Bay and nearly 50% of the salmon in Bristol Bay come
from the Nushagak and Kvichak Rivers, which encompass nearly half of the watershed area).
The report also describes in detail the importance of the fishery to Native Alaska cultures, the
importance and uniqueness of subsistence activities, and the scale of the commercial fishery.
Furthermore, the report also outlines the reliance ol" the local economy on the salmon fishery.
RESPONSE: No change suggested or required.
There is no question that a mine, especially of the type and magnitude analyzed in the
Assessment, could have significant impacts and that if these impacts are not or cannot be
properly managed and/or mitigated, the consequences could be profound. The Assessment
presents a mining scenario based on preliminary documents prepared for the Pebble Project,
which sets out a conventional approach for development of a very large mine that includes open-
pit and block-cave underground mining methods and conventional waste rock and tailings
management. Development of the mine as proposed would eliminate streams and wetlands in the
project area permanently. The importance of this impact is not put in context of the watershed as
a whole, so it is not possible to determine the magnitude of the risk to salmon. The Assessment
also did not consider whether there are any methods that could effectively minimize, mitigate or
compensate for these impacts.
RESPONSE: . I characterization of the landscape factors influencing salmon habitat potential
is now included to provide context for the stream habitat impacts described in the document
(Chapter 3). The assessment describes the magnitude of risks to salmon habitat. Due to lack
of knowledge of limiting factors, ascribing comprehensive risks to salmon populations is not
feasible in this assessment. Mitigation to compensate for effects on aquatic resources that
cannot be avoided or minimized by mine design and operation would be addressed through a
regulatory process that is beyond the scope of this assessment Nevertheless, in response to
public and peer review comments we have included a discussion of compensatory mitigation in
Appendix J of the revised assessment
The Assessment also focuses on the risk of failure of the tailings storage facility, a low
probability, but high impact scenario. The Assessment further describes the potential for long-
term acid and metals production from waste rock and the necessity for water treatment. Under
the mining scenario as described, perpetual management of the geotechnical integrity of the
waste rock and tailings storage facilities and perpetual water treatment could be necessary. In
addition, failure is always a possibility, albeit a possibility that is difficult to quantify with any
degree of certainty as explained in the Assessment. The Assessment also does not consider
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 18
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
alternative engineering strategies (so called 'best practice' approaches) that could lessen the risk
of failure and possibly the necessity for perpetual management and water treatment. As such, the
report could be considered a screening level assessment that presents the likelihood of
occurrence and corresponding consequences of failures under the presented development
scenario, but does not describe the magnitude of risk to salmon.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed modern
mining technology and operations. Measures for minimizing and controlling sources of
pollution, outside of stormwater requirements, may be referred to as best practices, state of the
practice, good practice, conventional, or simply mitigation measures. We have added a text
box in the revision (Chapter 4) to discuss terms. Mitigation measures consideredfeasible,
appropriate, and 'permittable' (as per Ghaffari et aL 2011) were considered in the assessment,
and these are measures common to other copper porphyry mines. Evaluation of alternative
strategies (e.g., other options presented in Appendix I) is outside the scope of this assessment,
but such evaluation should be part of the permitting process for a specific mining plan. The
assessment describes the magnitude of risks to salmon habitat. Due to lack of knowledge of
limiting factors, production, and demographics, ascribing comprehensive risks to salmon
populations is not feasible for this assessment.
Steve Buckley, M.S., CP(i
The assessment attempts to evaluate the potential impacts of mining development in the
Nushagak and Kvichak watersheds. The main deficiency in the assessment is that it uses only
two hypothetical mine scenarios to bracket the potential impacts of mining activities on the
ecological resources in the watershed. Both of these mine scenarios are larger than the 90th
percentile of all porphyry copper deposits in the world. In order to properly assess the potential
effects of mining activities, in the absence of any specific mining proposal, a minimum mine
scenario on the order of the 50 percentile of worldwide porphyry copper deposits would be
more appropriate. Three or four mine scenarios would allow for a broad range of analysis, and
the reader would be able to put the potential impacts of mining development in wider
perspective.
RESPONSE: A third mine .si:.e scenario (Pebble 0.25, 250 million tons) is included in the
revised assessment, to represent the worldwide median size porphyry copper mine (Singer et aL
2008).
A large part of the assessment provides information related to catastrophic potential system
failures such as tailings dam failures and pipeline ruptures. There is inadequate information on,
and analysis of, potential mitigation measures at the early stages of mine development, which
would attempt to reduce the impacts of mining activities on fish and water quality. The bulk of
the document is dedicated to evaluating the impacts of tailings dam failure on aquatic resources
and yet in Chapter 4, the assessment provides a probability of tailings dam failure at 1 in every
2,000 mine years.
RESPONSE: Mitigation measures were included in the draft assessment and are more clearly
identified in the revision. Analysis of alternative mitigation measures would be part of a
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 19
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
permitting process and is outside the scope of this assessment. A discussion of compensatory
mitigation in the Bristol Bay watershed has been added as Appendix J in the revised
assessment.
While failures of a TSF might be rare, they do happen, their effects may be very damaging,
and they could be devastating to local communities; thus, the assessment evaluates what risks
might be evident should such an event occur. The revised assessment also expands the
evaluation of risks for some lesser magnitude, but higher probability, events.
The assessment identifies the interconnectivity of groundwater, surface water, and fish habitat as
being a major component of the quality of the fishery in the watershed yet puts relatively little
effort into the analysis of the detailed relationships between sroundwater, surface water, water
quality, and fish habitat, even though this is likely the most important factor in assessing the
potential impacts of mining activities on the fisheries in the watershed.
RESPONSE: We lack the data to demonstrate this interconnectedness in a spatially and
temporally uniform manner, but do include examples of known points of high connectivity
and modeled locations of high groundwater-surface water interaction (Chapter 7)
Additional mine scenarios and a more detailed investigation of the geomorphology, surface, and
groundwater hydrology and their relation to fish habitat would provide the reader with a more
accurate and more useful scope of analysis.
RESPONSE: We now describe the broad geomorphic context for stream habitat in the
Nushagak and Kvichak River watersheds by characterizing gradient and watershed terrain
(Chapter 3). The revised scenarios include an additional mine size (representative of the
worldwide median size). The revision includes more failure scenarios (e.g., dieselpipeline
failure, quantitative water treatment failure, and refined seepage scenarios) and explains why
these particular failure scenarios were chosen.
Courtney ('tiwrlurs, Ph. I).
Synopsis: EPA's draft document examines the potential impacts of large-scale mining
development on the quality, quantity, and genetic diversity of salmonid fish species in the
Nushagak River and Kvichak River watersheds of Bristol Bay, Alaska. To the extent that both
wildlife and Alaska Native communities in the region depend upon salmonids, fish-mediated
impacts to these other "endpoinls of interest" are also explored. A hypothetical mining scenario,
informed by current exploration, planning, and study in the Pebble deposit area, is described
using minimum and maximum estimates for mine production and includes the construction of a
transportation corridor to Cook Inlet. Even in the absence of any failures or accidents,
construction and operation of such a mine would have significant impacts to salmonids in stream
systems proximate to the mine footprint with some related impacts to wildlife and human
communities. At least one or more accidents or failures are expected to occur over the long
lifetime of the mine. Immediate and long-term severe impacts to salmonids are expected to occur
with any significant failure, with relatedly pronounced impacts to wildlife and Alaska Native
communities in the region. Multiple mines in the region would amplify these impacts.
RESPONSE: No change suggested or required
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 20
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
General impressions: Overall, the main report is well-written and presents information in
multiple ways, including: narrative, conceptual models, images, figures, and tables. The report
synthesizes a large amount of information, much of which is described in detail in the report's
appendices. The report highlights the unique characteristics of this watershed: incredibly
productive and sustainable salmon fisheries, relatively little large-scale modification of the
natural environment, and active subsistence-based indigenous cultures still occupying their
homelands and many still using their Native language. Making central these features of the
watershed, the tone of the report suggests that some negative impacts to salmonids, wildlife, and
Alaska Native cultures are necessarily expected to accompany any large-scale mining
development and operation in this region.
RESPONSE: No change suggested or required.
The document should provide a clear articulation of the scope of human impacts considered in
this assessment. The main report considers only fish-mediated impacts to Alaska Native cultures.
The restriction of scope to only fish-mediated impacts should be further clarified. A host of
social, cultural, and economic impacts would accompany large-scale mining development in this
region. These direct and indirect human impacts, both positive and negative, were the focus of
many public comments on the EPA draft document, vet they fall outside of the scope of
consideration in this report. If the narrowed scope of fish-mediated impacts is justified, these
other impacts should be clearly identified as outside of the scope of this report. At times in the
report (e.g., p. 5-77), these other impacts are superficially mentioned. Unless a full treatment of
these impacts is included (including a presentation of a large literature explores these impacts
internationally, e.g., Ballard and Banks 2003), this cursory discussion should be removed. If
maintained, the narrow scope should be reiterated throughout the report to remind the reader that
these larger human impacts are not considered.
RESPONSE: The scope of the assessment has been more clearly articulated in Chapter 2,
which also now contains an overview conceptual model diagram demonstrating which
potential sources, stressors, and responses associated with large-scale mining were considered
outside of scope. The fact that direct impacts to Alaska Native cultures are not within the
scope of the assessment does not imply that they will not occur or that they are unimportant,
and this has been clarified in < hapters 2 and 12.
The report should articulate more clearly why Alaska Native cultures are the only human groups
included in the assessment offish-mediated human impacts. The report notes:
"because.. .Alaska Native cultures are intimately connected and dependent upon fish, .. .the
culture and human welfare of indigenous peoples, as affected by changes in the fisheries are
additional endpoints of the assessment" (ES-1-2). This suggests that the limitation of fish-
mediated human considerations to Alaska Native cultures is not due to government-to-
government relationship between tribes and the federal government, nor the special status
afforded by environmental justice concerns, but rather because of their close connections to, and
dependence on fish. Arguably, other human groups also have connections to fish and depend
upon on salmon in this region in various ways, but are excluded from analysis of potential
impact in this report. This comment is not meant to detract from the importance of the focus on
Alaska Native cultures and the primarily indigenous communities in this region for assessing
fish-related impacts. Rather, the comment is made to suggest the inclusion of a clear justification
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 21
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
for this focus, or the broadening of scope to include other human groups who are also connected
to, and dependent upon, salmon in this region (e.g., substantial information on the economic
dimensions of salmon resources in this region is summarized in Appendix E, but little is
presented in the main report). Additionally, the assessment of fish-mediated effects to Alaska
Native cultures is primarily focused on subsistence fisheries. More discussion of the role of
commercial engagements in salmon fisheries (e.g., commercial harvesting, processing,
recreational fishing businesses and employment) in the watershed communities in this region
would be helpful.
RESPONSE: EPA focused on the Alaska Native communities in response to the original
request we receivedfrom nine federally recognized trihal governments. The text (Chapter 12)
has been expanded to acknowledge the strong cultural ties oj many non-Alaska Natives to the
region, and potential effects on all residents from loss oj a subsistence way of life. However,
the focus of the assessment remains on effects on indigenous cultures. The importance of the
commercial fishery to the regional culture has been added to the text.
Dennis D. Dauble, Ph.D.
Overall, the main report and each of the accompanying appendices were well written. I was
unable to identify major inaccuracies or bias in the material as presented. There were
shortcomings in the main report, however. For example, some topics would benefit by being
expanded (Sections 5.6 and 8.7), while others have more detail than appeared necessary (Section
6.1). The assessment effectively addressed three appropriate time periods: (1) operation, (2) post-
closure, and (3) perpetuity. Potential effects are bounded by a minimum and maximum mine
size, which is also appropriate. Inclusion of inference by analogy strengthened the conclusions
reached in the assessment and helped validate results obtained from model predictions.
RESPONSE: Previous Section S. 6 (wildlif e and culture) has now been expanded and treated
as a stand-alone chapter (Chapter 12). The summary of risks from the mine scenarios
(previous Section 8. 7, now Chapter 14) has been expanded to include fish-mediated risks to
wildlife and culture, and more numerical results are included
Most figures and tables were useful. The conceptual models and accompanying illustrations of
potential habitat effects (Figs 3-2A and C) are important because they provide a view of
complicated pathways and relationships among potential activities and environmental attributes.
However, these relationships are not revisited in any detail later in the document. I recommend
discussing the conceptual models in more detail in the main report (Section 3.6) and summary
section in Chapter 8.
RESPONSE: Additional information on the use of conceptual models throughout the
assessment has been incorporated into Chapter 2. The more comprehensive conceptual models
presented in Chapter 6 (previously in Chapter 3) have been broken into their relevant
component parts throughout the risk analysis and characterization chapters, to better frame
the specific pathways addressed in each chapter.
The Integrated Risk Assessment (Chapter 8) did a creditable job of summarizing habitat losses
and risks from mine operations. What is missing, however, are quantitative descriptions of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 22
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
habitat lost relative to total habitat available in the larger watershed and individual systems.
Habitat loss should be further discussed in terms of salmonid life stage and productivity (i.e., not
all stream miles are equal).
RESPONSE: Unfortunately, no salmon habitat characterization is available for the region.
The State of Alaska has not even identified all anadromous streams in the region.
Productivity data are not available, even for the streams studied by the PLP. However, the
revised Chapter 14 contains tables summarizing habitat loss in stream lengths and wetland
areas.
If anything, the conclusions could be strengthened. The summary of uncertainties and
limitations (Section 8.5) dwells on things that "could not be quantified" due to lack of
information, model limitations, or insufficient resources. Thus, this reader was left somewhat in
limbo as to the potential magnitude of effects from mining activities. (Note that this "neutral
voice" is carried throughout the Executive Summary). Many people might interpret such
statements of uncertainty as no proven effect. My point is that probable environmental
consequences of mining activities are much greater than this report alludes to. eiven that
consequences are likely, even if their magnitude is "uncertain."
RESPONSE: We use a neutral voice throughout the document to convey the neutral scientific
perspective of this scientific assessment. We tried to convey the qualitative likelihood of
occurrence when quantitative probabilities were not obtainable, t his section has been edited
in Chapter 14 of the revised version to make the relationship between uncertainty and
probability of occurrence clearer.
Section 8.7 is perhaps the most important section of the report. It should be comprehensive, i.e.,
cover all resources and be more quantitative. Missing from the summary were impacts on
wildlife, human culture, resident fish, and other ecological resources. Essential details from
Appendices A, C, E, F, and L for example, could be synthesized and moved into the main report.
RESPONSE: The summary of risks from the mine scenarios (Chapter 14 in the revised
version) has been expanded to include fish-mediated risks to wildlife and culture and more
numerical results.
Gordon H. Reeves, Ph.D.
The purpose of the report is unclear, which makes it difficult to assess. The report focused on
the potential impact of a hypothetical mine on salmon and salmon habitat in two watersheds in
Bristol Bay, AK. However, it is not clear whether the analysis was intended to be a case study of
the potential impacts of a hypothetical mine under the various scenarios presented or whether the
intent was to develop a framework for assessing mining scenarios. These are two very different
objectives, which makes it critical that the purpose be clearly stated in the beginning of the
document so that reviewers and others understand the purpose of the document. There certainly
was much confusion among members of the review panel and the people who commented on the
report because of this.
RESPONSE: We have clarified the purpose of the assessment in Chapters 1 and 2.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 23
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
I think that the credibility of the report could be improved substantially if the analyses were
formalized and more clearly articulated and defined. The authors could consider using a
decision support process, such as a Bayesian approach (see Marcot, B.G., J.D. Steventon, G.D.
Sutherland, and R.K. McCann. 2006. Guidelines for developing and updating Bayesian belief
networks applied to ecological modeling and conservation. Canadian Journal of Forest Research
36: 3063-3074). This would provide more transparency to any analysis and allow others to
better understand how results and conclusions were derived. Also, it would identify critical
relations that should be considered and provide insight about the consequences of not
considering them. This will undoubtedly take additional time and effort, but I believe it would be
well worthwhile. Examples of where such analysis has been done are in: (1) Armstrup et al.
2008. A Bayesian Network Modeling Approach to Forecasting the 21st Century Worldwide
Status of Polar Bears. Pages 213-268. In E.T. DeWeaver el al., editors. Artie Sea Ice Decline:
Observations, Projections, Mechanisms, and Implications. Geophysical Monograph 180.
American Geophysical Union, Washington, D.C.i and (2) Lee, D.C. et al. 1997. Broadscale
Assessment of Aquatic Species and Habitats. Vol. Ill, Chapter 4. U.S. Forest Service, General
Technical Report PNW-GTR-405. Portland, Oregon.
RESPONSE: Creating a Bayesian Belief Network would require that the Agency convene
experts to subjectively estimate the probabilities of each transition in the conceptual models.
In contrast, this assessment is intended to elucidate the risks frompotential mining based on
available data and analyses of those data.
I thought one of the strongest aspects of the report were the conceptual diagrams of relations
between the various aspects of the development and operation of a mine and the components of
the ecosystem that influence salmon and their habitat (Chapter 3). These diagrams show the
components of the ecosystem, the relation among them, and how mine impacts could potentially
influence given parts of the ecosystem directly or indirectly as a result of cascading effects. They
are a good first step in developing a decision support framework, as suggested in the previous
paragraph. There was, however, little discussion about them in the text and it was not clear if or
how they were used or considered in the analyses. The authors should, at the very least, clearly
identify which parts of the networks were considered and why these particular avenues were
pursued and others were not. This would provide additional insights into potential limitations of
the analyses and results.
RESPONSE: The more comprehensive conceptual models presented in Chapter 6 (previously
in Chapter 3) have been broken into their relevant component parts throughout the risk
analysis and characteri-ation chapters, to better frame the specific pathways addressed in each
chapter.
If this was a case study, the report appeared to have considered available literature and reports on
all aspects of the mine, its operation and the parameters that could be affected by it. I am not
familiar with this literature so it is not possible for me to comment on the adequacy of the
literature and reports considered. Assumptions about the location and operation of the mine
seemed reasonable and the authors clearly articulated limitations of available data and other
information concerning the mine's location and operation. I found the consideration of the mine
during the various phases of development and operation and the discussion about potential
development of other mines in the area particularly insightful. Inclusion of experiences from
other mining operations was also helpful in understanding the conclusions about potential
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 24
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
impacts of the mine and its operation over time. Additionally, the consideration of the potential
development of other mines in the area was particularly insightful and provided a good picture,
albeit not in depth, of potential cumulative effects on aquatic resources in the Bristol Bay area.
RESPONSE: No changes suggested or required.
Parts of the report on the ecology of fish and aquatic ecosystems, road, and culverts - topics that
I am familiar with - were covered very well and the conclusions about potential impacts of the
mine and its operation generally seemed justified. The authors presented available data and
information on fish distribution and abundance relative to the presumed location of the various
components of the mine operation. Their analyses were appropriate but rather cursory, which is
not unexpected given the restrictions of time and available data. However, there are some
additional considerations and analyses that could be done, which I think would improve the
report. I identify these in answers to specific charge questions. Limitations of the results were
readily acknowledged. However, as mentioned above, there are additional limitations that
resulted from only considering selected potential avenues of impacts. These should be discussed
in the revision.
RESPONSE: The discussion of scope (Chapter 2), endpoints (Chapter 5), and uncertainties
(throughout the risk analysis and characterization chapters) has been expanded in the revised
assessment.
The authors do a good job of summarizing the scientific literature on salmon ecosystems, roads,
and culverts. Most of this is from studies in areas outside of Bristol Bay. Interpretations of the
findings were accurate. However, there was no discussion about potential limitations on the
application of the studies to the area being considered. For example, Furniss et al. (1991) deals
with roads in forest and rangeland settings. These are very different environments than Bristol
Bay, which suggests that road impacts will likely differ. Much attention is given to "headwater
streams" and their ecological importance (p. 5-19 - 5-21). Headwater streams for the area of
consideration need to be defined so that appropriateness of the application of the literature can be
better judged.
RESPONSE: Headwater streams in the study area are now more fully described (Chapter 7).
Because the potential mining described in the assessment would take place in an undeveloped
area, much of the literature is necessarily from areas outside of Bristol Bay. However, to the
extent possible we used examples from representative environments. With respect to Furniss et
aL (1991), though it focuses on forest and rangeland roads, it is a seminal publication on the
potential effects of roads, particularly as they relate to salmon. The general conclusions of that
paper should be applicable to the transportation corridor described in the assessment
A major component that is missing from the report is consideration of the potential impacts of
climate change. Climate change is identified as a factor in the conceptual model of potential
habitat and water quality effects associated with mine accidents and catastrophic failures (Fig. 3-
2D). However, I believe that it is a key factor that will have influence in all aspects of the
assessment, not just failures and natural disturbance events (Fig. 3-2C). It needs to be considered
in other aspects, such as water quality and availability. Climate change should also be included
in any analysis because it will be critical to build it into any monitoring program that is
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 25
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
developed in order to be able to differentiate its impact on salmon and their habitat from
potential impacts of the mine.
RESPONSE: Climate change projections and potential impacts are now included in Chapter 3
(Section 3.8). It is mentioned as an important external factor in the risk analyses presented in
Chapters 7, 9, and 10, and the issue is summarized in Box 14-2.
Charles Wesley Slaughter, Ph.D.
The Assessment (Volume 1 - Main Report) provides a fairly comprehensive review of fisheries-
driven issues, from the perspective of salmonids. Appendices (Volumes 2 and 3) are very
informative. The high significance of the Bristol Bay watershed, specifically of the Nushagak
and Kvichak river systems, for commercial fisheries on the global scale and for sport and
subsistence fisheries at the regional and local scales, was appropriately described.
RESPONSE: No changes suggested or required.
The potential risks and impacts are fairly and succinctly staled. Given the extremely long-term
nature of the projected Pebble project, and the irreversible changes which would be imposed to
the region, the risks seem, if anything, understated. 1 attribute this to the decision to focus this
Assessment on salmon and anadromous fisheries, with less attention on "salmon-mediated"
impacts - i.e., effects on indigenous culture, on wildlife other than salmon, etc.
RESPONSE: No changes suggested or required.
Chapter 2 (Characterization of Current Condition) provides only a superficial overview of the
landscape of the Bristol Bay watersheds; a reader would preferably have access to Wahrhaftig
(1965) or Selkregg (1976), as only two (relatively dated) suggestions, to gain a more
comprehensive understanding of the region.
RESPONSE: Additional, information on the physical environment of the region (e.g., geology,
vegetation, etc.), along with an expanded treatment of the regional landscape, has been
incorporated into Chapter 3 (e.g., Figures 3-4 through 3-7). Chapter 3 also includes the
suggested citations.
The "Water Management" section (4.3.7) seems cursory, highly generalized, and optimistic.
Statements such as "uncontrolled runoff would be eliminated"; "water from these upstream
reaches would be diverted around and downstream of the mine where practicable"; and
"Precipitation.. .would be collected and stored..." do not indicate actual (proposed) practices or
techniques, nor inspire confidence that actual runoff events during "normal" conditions, let alone
during hydrologic extremes (such as a rain-on-snow event with underlying soils still frozen),
would be planned for or actually managed adequately.
RESPONSE: Water management measures are more clearly described and discussed in
Section 6.1.2.5 of the revised draft, and in sub-sections for the mine components in the
scenarios. The assessment no longer contains a no failure scenario, so complete water
collection is not longer assumed. Rather, standard and common practices are incorporated.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 26
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Perhaps I missed it, but I found no acknowledgment of the potential presence of or consequences
of perennially frozen soils - permafrost - in the Bristol Bay watershed, or more specifically in
the Pebble ore deposit locale or the proposed transportation corridor. Selkregg (1976), Fig. 136,
shows soils of the Pebble locale as INT/2g, INT/lg - HYP, or SOU/2g-HYP - that is, well-
drained gravelly soils (INT) or well-drained acidic soils (SOU) with interspersed peaty, poorly-
drained shallow discontinuous permafrost. There is abundant literature on the influence of
permafrost on engineered structures, roads, hydrology, etc. Even if the bulk of the terrain
involved in the proposed Pebble mine, road and infrastructure project is founded on well-drained
gravelly soils, any interspersed permafrost-underlain terrain can prove problematic in terms of
landscape stability, potential erosion, and consequent structural, engineering, hydrologic and
water quality issues. See Specific Observations for a few suggested references in.
RESPONSE: We have expanded our characterization of the soils and permafrost distribution
in the Bristol Bay watershed in Chapter 3 of the revised assessment. As part of this expansion,
we summarize the nature and distribution of permafrost by physiographic region.
While there is extensive discussion of a proposed transportation corridor, there was no mention
of construction of a major airfield. A project of this magnitude would undoubtedly require
development of a facility in close proximity to the mine(s) capable of handling CI 30 and
commercial jet passenger and cargo traffic, at least to the 737 class, if not 747. I don't know
what the footprint for such an airfield would be, but it would be substantial, and with requisite
roads, fuel handling, etc., would be a major project in itself. This would seem to be a logical
component of a comprehensive assessment of the potential Pebble project.
RESPOSSE: The scope of the assessment has been clarified in Chapter 2, and construction
and operation of a new airport is considered outside the scope of the assessment We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy \ct Environmental Impact Statements would consider these
effects if a new airport is proposed.
As noted in the Executive Summary, the Assessment does NOT address several major
components of the (hypothetical) Pebble project, including electrical generation and
transmission, a deep-water port, or "secondary development" and associated infrastructure which
would follow an initial mining project. A truly comprehensive analysis should incorporate full
analysis of these aspects. This Assessment is thus inadequate in terms of considering potential
broader consequences for the Bristol Bay watershed system.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and we have stated
throughout the text that areas outside of scope may also be important factors.
John D. Stednick, Ph.D.
The purpose of the document is not clearly stated in either the Executive Summary or the
Introduction. Need to specifically identify the document as an environmental risk assessment.
There is a misconception that it is a CWA Section 404(c) review, rather than an environmental
risk assessment. The document should have the utility to inform future users of the risk to the
watershed resources from mining activities in the watershed. The assessment can be used by
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 27
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
others for decision making purposes, and includes current and appropriate methodologies for all
identified stressors, such that study results can be duplicated. And all stressors are evaluated to a
similar level of detail.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (ERA) in general has been added to Chapters 1 and 2, as well as the Executive
Summary. Section 1.2 includes information about the use of the assessment. The assessment
has been reorganized into two major sections (problem formulation, risk analysis and
characterization) to clarify where different chapters faU. in the typical ERA process.
The document characterizes the potential environmental effects of an open pit mine over a
copper porphyry complex in southwest Alaska using a hypothetical mine design based on similar
ore deposits and mine complexes elsewhere. Proposed mine activity has been identified by the
Pebble Limited Partnership though Northern Minerals Dynasty and should be cited to improve
applicability of the risk assessment. Furthermore, a wider range of mining scenarios should be
developed and analyzed for environmental risk assessment. Environmental consequences were
estimated by the environmental risk assessment model approach for both 'no-failure' and
'failure' scenarios. The Executive Summary concluded that the effects of mine development
resulted in significant salmon habitat losses. Potential effects on other aquatic species were not
identified. The assessment evaluated environmental risks under the development and closure
scenarios using large catastrophic events and did not include smaller, yet more frequent
excursions or system failures. Nor did the assessment look at the full range of mine development
scenarios, specifically what are the risks associated with a smaller underground operation?
RESPONSE: The assessment used the Pebble deposit and its characteristics, as described by
Northern Dynasty Minerals in the (ihaffari et a I. (2011) report. That report is cited extensively
in both the original review' draft and the revision. .1 median-sized mine (based on worldwide
mine sizes) has been added to the scenarios in the revised assessment. Because the number of
potential failures is extremely large, it is necessary to choose a representative set offailure
scenarios. The final document includes more failure scenarios (e.g., diesel pipeline failure,
quantitative water treatment failure, and refined seepage scenarios) and explains why the
particular failure scenarios were chosen. Underground mining is a potential for any mining
site that has high-quality ore located at depth, but sources of potential impact considered in
scope for the assessment would be common for either a surface or an underground mine (e.g.,
water withdrawal, tailings dam failure, water treatment failure, seepage, etc).
The conclusions of the Executive Summary are strongly worded (e.g., pages ES 13 to 24), yet the
uncertainties presented later in the report make the strong conclusions tenuous. An expanded
discussion of uncertainties and limitations may temper those 'conclusions.'
RESPONSE: Each risk analysis chapter of the revised assessment now includes an
uncertainty section The Executive Summary has been rewritten to reflect the revised
assessment text.
Site characterization/description of current conditions is too brief. More information is needed
for a full site characterization. Any reader unfamiliar with the setting would not fully understand
the physical, biological, or ecological inventories and linkages in the study area. The risk
assessment of failure and no failure are covered in Chapters 5 and 6 with varying levels of detail
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 28
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
and substantiation of conclusions. Statements like "salmon is important in the human diet, thus a
salmon loss affects human health" seem like a weak argument, especially when additional
information in the appendix suggests a larger effect.
RESPONSE: Additional information on the region's physical environment has been added to
Chapter 3 (e.g.. Figures 3-4 through 3-7), and additional information on the region's
biological communities from Appendices A through C has been incorporated into the main
text The purpose of the appendices—to provide the detailed background characterization
necessary for the ecological risk assessment—has also been clarified in Chapter 2.
The Pebble Limited Partnership has a large environmental baseline database (EBD), but does not
appear to be cited or used. Justification for the inclusion or exclusion of these data should be
made. Reference is often made to various data, but these data were not presented.
RESPONSE: The EBD was used and cited more than ~0 times in the May 2012 review draft
and even more in the revised assessment. Data from the PLP EliD concerning hydrology,
water quality, and biology of the streams on the site and along the transportation corridor
have been extensively incorporated into the assessment in the analyses in Chapters 7 through
11. However, to the extent possible, the assessment relies on peer-reviewed literature.
Review and revise the water balance section, which would include: 1) generating a diagram or
conceptual figure similar to page 3-7 to illustrate the potential effects of mine construction and
operation on surface and groundwater hydrology; 2) developing a quantitative water balance for
surface and groundwater resources; 3) incorporating seasonality (especially assessing the role of
frozen soil); 4) identifying hydrologic processes and their associated values (e.g., mm/yr) for
each component of the water balance in time and space, and then incorporating into a landscape
characterization; 5) demonstrating the interconnectedness of groundwater, surface water, and the
importance to fish habitat and stream productivity; 6) evaluating the influence of global climate
change on these hydrologic processes and rates; and 7) using this characterization demonstrate
the expected hydrologic modification associated with the mine scenarios and infrastructure
development and closure scenarios.
RESPONSE:
1) We included schematics to illustrate potential effects of mine construction and operation of
surface hydrology, including e//ects via groundwater changes (e.g.. Figure 6-5, Figures 6-8
through 6-10).
2) Our water balance focuses on surface water hydrology (including interactions with
groundwater). A comprehensive groundwater hydrology water balance is beyond the scope of
the assessment.
3) The core of our analyses is an annual water balance, but we have maintained the simple
approach to seasonality used in the first draft of the assessment.
4) We have adopted a basic approach to representing the dominant hydrologic processes at the
mine site; a comprehensive representation of all hydrologic processes is beyond the scope of
this assessment
5) Throughout the assessment, we have identified and quantified the interconnectedness of
surface water, groundwater, and their importance to fish habitat and stream productivity.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 29
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
6) A section on potential climate change effects has been added to Chapter 3.
7) We have used our updated water balance and hydrologic modeling approaches to estimate
expected responses of mine scenarios, infrastructure development and closure scenarios.
One common theme that emerged from the public comment session during the peer review
meeting in Anchorage, AK was the questioning of the document timing, from draft release to the
public comment period to the unannounced completion of a final document. These concerns
should be addressed in the new document.
RESPONSE: This type of contextual information is not directly relevant to the ecological risk
assessment, but clarification of the timing and use of the assessment has been included in
Chapter 1.
Roy A. Stein, Ph.D.
Accuracy of Presentation. Overall, I was pleased w ith the accuracy of the presentation.
Typically, peer-reviewed citations to the scientific literature were cited as supportive
documentation for most all of the factual information (though the well-developed appendices,
e.g., Appendix E: Economics; Appendix I: Mitigation, could be used to far better advantage, see
below). Unfortunately, in the main report, many data are missing, especially with regard to
salmonid populations, their diversity (both across species wand within species across
populations), their relative population sizes, their distribution across the watershed, their vital
rates (i.e., recruitment, growth, and survival across life stages), and to what extent the Pebble
Mine and its associated activities will reduce these populations (for there is no question they will
indeed be reduced through both the mine footprint and all allied operations in the drainage), both
through impacts on individual populations and the overall production of salmonids (and other
fishes) in the Bristol Bay watershed.
RESPONSE: We now include figures showing reported salmon species distributions and
salmon diversity by IIUC-I2 watersheds across the Sushagak and Kvichak River watersheds
(Figures 5-3 through 5-8). Information on population sizes and vital rates are limitedfor the
region, but are reported where known. Due to lack of comprehensive estimates of limiting
factors across the impacted watersheds, population-level eff ects could not be quantitatively
estimated except for the most severe cases, where total losses of runs could be reasonably
assumed
Whereas I am relatively confident about accuracy of the fisheries information included, I cannot
comment in detail regarding the accuracy of the mining information or impacts on the Native
Alaskan cultures (though the impact of the mine on this culture was confined to fish-mediated
effects). That a Native Alaskan culture 4,000 years old is in jeopardy bothers me greatly; might
this complete subsistence way of life in the Bristol Bay watershed be eliminated with the
exploitation of the copper via open-pit mining? In turn, what impacts might there be on
subsistence users, other than Native Alaskans? Even though these sections seemed reasonably
well presented (with caveats above) and appropriately supported with citations, they do lie
beyond my expertise.
RESPONSE: No changes suggested or required
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 30
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
My concerns about the document revolve around issues that were not considered, i.e., Global
Climate Change, "In Perpetuity" issues, groundwater-surface water exchange issues (owing to
missing information), impacts of Routine Mine Operations in a more realistic setting, the
seemingly undue influence on a failure of the Tailings Storage Facility, and other somewhat
more minor issues (see comments below). With any revision, the authors should include this
information by eliminating redundancy (see below), thereby not increasing document length.
RESPONSE: We have thoroughly revised our approach to quantifying hydrologic responses
to the mine scenarios. We explicitly include groundwater-mitigated effects on surface waters.
Climate change projections and potential impacts are nmr included in Chapter 3 (Section 3.8),
and are considered as important external factors in the lis It analyses as summarized in Box
14-2. See responses to the commenter's specific comments and to Dr. Stednick's hydrologic
comments below.
Clarity of Presentation. Generally speaking, I believe that the writing was intelligent, reasonably
insightful, and, more specifically, on task. One significant criticism with regard to the
presentation revolves around the organization of the document. As detailed below, the
organizational scheme lent itself to redundancy, from the Introduction through the various
chapters to the Integrated Risks Characterization chapter. Owing to this redundancy, the report
is likely too long by about 20% and any revision and shortening should serve to improve its
impact on readers.
RESPONSE: The assessment has been reorganized to eliminate redundancy and help clarify
the structure of the document
The conceptual block and arrow diagrams (pages 3-7 to 3-11) were quite instructive. They
nicely demonstrate the interactions that occur within this mining scenario. The main report
would be much improved if text were to review this set of interactions. Clearly, a tremendous
amount of time, effort, and thought went into generating these diagrams and it is indeed a true
shortcoming of the main report that essentially no text was spent stepping through these
diagrams.
RESPONSE: Additional information on the use of conceptual models in the assessment has
been incorporated into Chapter 2. The more comprehensive conceptual models presented in
Chapter 6 (previously in Chapter 3) have been broken into their relevant component parts
throughout the risk analysis and characterization chapters, to better frame the specific
pathways addressed in each chapter. Additional conceptual models considering impacts on
wildlife, Alaska Native populations, and cumulative effects of multiple mines have been added
to Chapters 12 and 13.
Soundness of Conclusions. The conclusions were well supported, where there were published
data to support them. Many statements that could be interpreted as conclusions were often more
qualitative than desirable in a review document such as this one, owing to the lack of information
(percent of salmonids lost owing to routine mine operations, impacts of mining and the
transportation corridor on wetlands, extent of groundwater-surface water disruptions, just to
name a few). Consequently, the soundness of the conclusions are somewhat compromised by a
lack of information.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 31
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: No changes suggested or required
In addition, what would aid readers is a succinct statement of the purpose (risk assessment?,
impact on water quality and then through to fishes and beyond?, etc.) and scope (relatively
narrow impact of the mine on salmonids and ripple effects out from there) of the document early
in the initial chapter. In so doing, both reviewers and readers will be informed as to the direction
of the document and thus better informed as they move through the document.
RESPONSE: Additional information on both the purpose of the assessment and ecological
risk assessment (ERA) in general has been added to Chapters 1 and 2, as well as the Executive
Summary.
Finally, a portion of the public testimony complained about the process, specifically about the
time allowed for document review, the data reviewed, the validity of the hypothetical mine, etc.
Though I found most all comments to be somewhat disingenuous, I still would offer the
following advice: Provide a section upfront that deals with process issues surrounding the
review, i.e., explaining the constraints under which EPA was operating; without a section like
this, complaints, such as those described above (coming from just one segment of the public),
will go unanswered.
RESPONSE: Chapters 1 and 2 no ir clari/y the purpose of the assessment and document how
public participation was incorporated into tlic process (e.g.. Ilox I-1).
William A. Stubblefield, Ph.D.
The document, "An Assessment Of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska," is a well-written, comprehensive document that employs a risk assessment-type
approach to an a priori evaluation of potential environmental effects on the ecosystem and
potential receptor species (e.g., salmon) that may be affected by a potential copper mine located
in the Bristol Bay area of Alaska. This document is somewhat unique, in that no actual mine has
been proposed at the location and few site- or project-specific data are available. Therefore, no
specific information about development plans and potential operational and closure activities
associated with the mine are available. Rather, the authors have attempted to develop a
hypothetical mine and attempted to assess possible environmental effects associated with mine
development, operation, and closure. Although interesting, the potential reality of the assessment
is somewhat questionable. It is also unclear why EPA undertook this evaluation, given that a
more realistic assessment could probably have been conducted once an actual mine was
proposed and greater detail about operational parameters available. The approach taken in the
document attempted to be comprehensive and evaluated a variety of scenarios that may affect
aquatic resources in the Bristol Bay region. Given the importance of salmon populations in the
area, both from a financial and societal perspective, it is important that a comprehensive
evaluation of potential environmental effects associated with mine development and operations
be conducted. The authors have attempted to conduct such a comprehensive evaluation and have
attempted to quantify (to the extent possible) the probability of adverse effects occurring.
Implementation of this approach is proper, and with the correct data, can provide a
comprehensive evaluation of potential environmental effects. Unfortunately, because of the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 32
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
hypothetical nature of the approach employed, the uncertainty associated with the assessment,
and therefore the utility of the assessment, is questionable.
RESPONSE: The EPA respectfully disagrees that the hypothetical nature of the approach
compromises the utility of the assessment. All mining plans are hypothetical They change in
response to the results of assessments, regulatory requirements, public input, and unforeseen
conditions and events. They cease to be hypothetical only after the mine is closed At every step
in the process, assessments of the current plan are useful even though plans will change. This
assessment is based largely on a preliminary plan, published by Northern Dynasty Minerals
(Ghaffari et a I. 2011). Although layout of mining components in a future mine plan may differ
somewhat from the preliminary plan or the EPA scenarios, the main components of mining
would remain the same for open-pit mining (and underground mining wouldface the same
waste issues).
A variety of uncertainties and data needs were identified as a result of this effort and this alone
may provide sufficient value to justify the document and approach. For example, the authors note
that there is not an abundance of chronic toxicity data considered in deriving the EPA's ambient
water quality criteria for copper and that there is an uncertainty associated with whether the
biotic ligand model (BLM) adequately protects species of concern in Bristol Bay. It would seem
appropriate for EPA (perhaps in concert with industry) to develop the data to improve our
understanding of copper toxicity and to ensure that regulatory standards are, in fact, appropriate
for their intended use. A substantial body of data evaluating copper chronic toxicity has been
developed by the copper industry as a result of regulatory requirements driven by the European
REACH regulations. It may be beneficial for EPA to examine these data, thus resulting in a
reduction in any uncertainty associated with the evaluation of environmentally acceptable metals
concentrations. It should also be noted that similar datasets and biotic ligand models exist for
number of other metals that may be of concern at the Bristol Bay site.
RESPONSE: The EPA has examined the Eli's 2IIIIX Voluntary Risk Assessment of Copper
(the relevant REACH document). Although tluy do derive a chronic species sensitivity
distribution, it is because of the way they include and aggregate data, rather than the
generation of new data. In particular, they have no data for sensitive aquatic insects, so the
EU does not resolve that problem. The HIM was usedfor copper because copper is the
contaminant of greatest concern and because the copper BLM has been approved by the EPA
Office of Water. Other metals with BLMs, such as zinc and nickel, occur at much lower levels
in leachates.
One suggestion that would improve the document is that EPA should include a basic description
of the risk assessment process and the relationship between the risk assessor and the risk
manager, i.e., the decision maker. They must include a discussion of why the assessment is
being conducted, the decisions that will be informed, and what information they need from the
risk assessor.
RESPONSE: Additional contextual information for the assessment has been included in
Chapter 1, and additional information on ecological risk assessment has been incorporated
into Chapters 1 and 2. The assessment has also been restructured into problem formulation
and risk analysis and characterization sections, to make the assessment's structure as an
ecological risk assessment clearer.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 33
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Taken from the USEPA's Guidelines for Ecological Risk Assessment (EPA630/R-95/002F; April
1998). Note 2nd sentence re: the role of the risk manager.
"2.1. THE ROLES OF RISK MANAGERS, RISK ASSESSORS, AND INTERESTED
PARTIES IN PLANNING
During the planning dialogue, risk managers and risk assessors each bring important
perspective to the table. Risk managers, charged with protecting human health and the
environment, help ensure that risk assessments provide information relevant to their
decisions by describing why the risk assessment is needed, what decisions it will
influence, and what they want to receive from the risk assessor. It is also helpful for
managers to consider and communicate problems thev have encountered in the past when
trying to use risk assessments for decision making.
In turn, risk assessors ensure that scientific information is effectively used to address ecological
and management concerns. Risk assessors describe what they can provide to the risk manager,
where problems are likely to occur, and where uncertainty may be problematic. In addition, risk
assessors may provide insights to risk managers about alternative management options likely to
achieve stated goals because the options are ecologically grounded. "
RESPONSE: Section 1.2 in the revised assessment discusses uses of the assessment.
Dirk van Zyi, Ph.D., P.E.
Planning and designing a large mine, and especially one in a sensitive environmental setting such
as Bristol Bay, involves many iterations before a design evolves that is provided for further
public considerations. The EPA elected to use a design, developed for Northern Dynasty
Minerals Ltd. in a preliminary assessment prepared following the guidance of National
Instrument (NT) 43-101, as the basis for extensive evaluations in their risk assessment. The
resulting risk assessment can be at best characterized as preliminary, screening level, or
conceptual. There are both technical and process issues that must be addressed before this risk
assessment can be considered complete or of sufficient credibility to be the basis for a better
understanding of the impacts of mining in the Bristol Bay watershed.
RESPONSE: The EPA respectfully disagrees that the hypothetical nature of the approach
compromises the utility of the assessment. All mining plans are hypothetical They change in
response to the results of assessments, regulatory requirements, public input, and unforeseen
conditions and events. They cease to be hypothetical only after the mine is closed At every step
in the process, assessments of the current plan are useful even though plans will change. This
assessment is based largely on a preliminary plan, published by Northern Dynasty Minerals
(Ghaffari et al 2011). Although layout of mining components in a future mine plan may differ
somewhat from the preliminary plan or the EPA scenarios, the main components of mining
would remain the same for open-pit mining (and underground mining would have the same
waste issues).
With respect to the proposed transportation corridor, we note in the assessment that
"Although this route (the one proposed in the EPA scenario) is not necessarily the only option
for corridor placement, the assessment of potential environmental risks would not be expected
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 34
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
to change substantially with minor shifts in road alignment Along any feasible route, the
proposed transportation corridor would cross many streams, rivers, wetlands, and extensive
areas with shallow groundwater, including numerous mapped (and likely more unmapped)
tributary streams to Iliamna Lake (Figures 10-1 and 10-2)."
There are a number of items that require specific attention prior to finalizing the report. While
my comments below provide further details, from a global perspective the following aspects
must be addressed:
• A better sense about the range of impacts from a mining project that use not only
different technologies but also different lay-out options in its development than that
assumed in the EPA Assessment;
• More attention to the use of appropriate order of magnitude numbers reflective of the
quality of data, e.g. less accuracy is obtained when 1:62.500 scale vs. 1:12,500 scale
maps are used;
• Correction of errors associated with misquoting and incorrect use of information in the
literature; and
• A critical review and rewrite of the Executive Summary to reflect the tone, terminology,
information sources and results of the main body of the report. One example of an error
and one of inconsistent terminology are:
o Page ES-10: "Thus, the mine draws on plans published by the Pebble Limited
Partnership (PLP)", this is incorrect as the plans that were used were prepared for
Northern Dynasty Minerals Ltd.
o Page ES-10: . .our scenario reflects the general characteristics of mineral
deposits in the watershed, contemporary mining technologies and best
practices..." The main body of the report emphasizes on a number of occasions
(such as Page 4-1, 4-17) that "Our mine scenario represents current good, but not
necessarily best, mining practices".
My comments contained above and below are based on a single review of the report, i.e.
contractual lime constraints were such that I could not afford a second review of the report. It is
therefore possible that there are other errors remaining in the report that I did not observe in my
review. It is therefore recommended that after making these corrections and edits that EPA
subject the report again to a rigorous independent review.
RESPONSE: The scenarios evaluated are meant to represent those expected to be present as
typical for mining porphyry copper deposits of this type. Although layout of mining
components at a site may differ somewhat from what we present in the scenarios, the main
components of mining would remain the same for open-pit mining (and underground mining
would have the same waste issues). Theref ore, no change is required for technologies
presented in the original assessment, and we have noted in the assessment that there could be
different layouts than what we have presented
Errors and inconsistencies in sections of the document are noted and have been corrected in
the revised assessment. With regard to the terminology of "best", "good", or other terms for
the practices used, what was intended to be conveyed is that we have assumed modern mining
technology and operations. The terms are qualitative when generally interpreted, or have a
regulatory meaning (for example, "best management practices" applies to the setting of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 35
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
stormwater control, but not specific to mining sites), and thus we have eliminated their use in
the revised assessment. The assessment is being re-reviewed by the external expert reviewers.
Phyllis K. Weber Scannell, Ph.D.
My comments on EPA's draft document, An Assessment of Potential Mining Impacts on Salmon
Ecosystems of Bristol Bay, Alaska, follow a three-day peer review meeting in Anchorage, AK.
On the first day of the meeting, the Peer Review Team heard testimony on the importance of the
resources in the potentially affected area and on possible effects of mineral development on the
fish and wildlife resources and on local residents. The issues of mineral development are
complex, particularly with respect to protecting the environment and the interests of local
residents. I understand and appreciate the complexity of these issues; however, the charge of the
Peer Review Team is to review EPA's draft document, An Assessment of Potential Mining
Impacts on Salmon Ecosystems of Bristol Bay, Alaska, and offer suggestions to strengthen the
report. My comments, included below, are focused on the accuracy and thoroughness of the
draft document.
The document "An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol
Bay, Alaska" and the accompanying appendices provide an in-depth and thoroughly documented
description of the environment and resources of the areas under consideration for mineral
development, although not in the entire Bristol Bay region. Appendices A and B are particularly
thorough in describing the salmon and non-salmon fishes in the region; the discussion of species
specific fish sensitivities to certain toxicants adds important information for future consideration
of project development.
RESPONSE: No change suggested or required.
The assumptions for developing and operating large porphyry copper mine may not be aligned
with features of a future mining project. Too much emphasis was placed on effects of
catastrophic failures, such as failure of a tailings dam or pipeline, and too little emphasis on the
need to identify and control seepage water, run-off from PAG (potentially acid generating) and
NAG (not acid generating) waste rock areas, and water treatment.
RESPONSE: Because the number of potential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
seepage scenarios) and explains why the particular failure scenarios were chosen.
The document discussed effects of dewatering on suppressing stream flows and groundwater
inputs but did not consider effects of the discharge of treated wastewater. The section on
hydrology illustrates the need for more complete hydrologic information before any project
development. The need for bypassing all clean water sources around a development site should
be addressed.
RESPONSE: The revised assessment more clearly presents that clean water would be diverted
around the site, retained in settling ponds, and released following settling and/or treatment, if
required Discharge of treated wastewater is analyzed and discussed in greater depth in the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 36
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
new Chapter 8, Water Collection, Treatment, and Discharge. The assessment is based on the
best hydrologic information available for the site; however, additional hydrologic information
may be available and/or acquiredfor any future mine plan in this watershed We agree that
detailed hydrologic information is critically important for responsible project development. We
have updated our hydrologic analyses to represent the probable influence of mine scenarios
on surface water/groundwater interaction.
As stated in my response to charge questions, I believe that the two most important questions for
mineral development in this region are: can a mine be designed and operated for future closure?
and, if not, is it acceptable to develop a large porphyry copper mine in a region of high value
salmon habitat that will essentially require perpetual treatment? These two questions must be
addressed when considering protection of the fish, wildlife, and human resources of the region.
RESPONSE: We agree that these are important questions to be addressed but they are risk
management, not risk assessment, questions. The purpose of the assessment is to evaluate
risks to the salmon fishery from large-scale mining. Risk management decisions will be made
during the permitting process. No changes to the assessment were made in response to this
comment
Paul Whitney, Ph.D.
Response (with a wildlife perspective) - The main document is fish centric and it should be,
given the importance of salmon in the Bristol Bay ecosystem. Wildlife (aquatic, wetland and
upland species) and terrestrial resources related to potential mine and haul road impacts are
glossed over. The summary write ups for several species of wildlife (Appendix C) are very good
regarding natural history and some potential impacts. Information in Appendix C tends to focus
on the proposed mine site and less on the proposed haul road and game management units in the
Kenai Mountains.
RESPONSE: Direct effects on wildlife and terrestrial resources are outside the scope of the
assessment, as clarified in Chapter 2. Eff ects on wildlif e are now treated in Chapter 12.
PHWResponse: The clarification provided in Chapter 2 is not tight. For example, the statement
"We focus on freshwater habitats, because the most exceptional ecological feature of the Bristol
Bay watershed is its fish populations" (p 2-5. para 3. lines 2 and 31 is subjective at best and is
in the eve of the beholder. What is clear to me is that the decision not to consider direct impacts
to wildlife is political and relates to the Clean Water Act. Asserting that "exceptional" has an
ecological or scientific meaning and explains why direct impacts on wildlife are not considered
confuses rather than clarifies. .,
USFWS RESPONSE: We acknowledge the comment regarding quality of Appendix C.
Information in Appendix C is intended to focus on the entire Nushagak and Kvichak River
watersheds to the extent that data exist. To the extent that a potential mining-related road is
within the Nushagak and Kvichak River watersheds, information about selected wildlife
species is included in Appendix C. Information about selected wildlife species on the Cook
Inlet side of the Chigmit Mountains is not included in the wildlife report. The Kenai
Mountains are not in the Nushagak and Kvichak River watersheds.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 37
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
PHWResponse: Appendix C is what it is but the above response is circular. Direct impacts on
wildlife are outside the scope of the assessment so it really doesn't matter that wildlife resources
along the entire road are not discussed. Is one to assume that impacts along the road segments
outside the Bristol Bay watersheds are not significant? The circular reasoning in the above
responses is of little consequence here but circular reasoning in other responses is more of a
concern., —-{ Formatted: Font: Not Bold, Not Italic
A variety of authors have obviously contributed to the documents and it appears that the
direction given to them or their interpretation of goal statements varies. For example, if one of
the goals of the assessment is to evaluate the risk to wildlife due to risk to fish (Executive
Summary, page 1, last para) it's not clear why so much verbiage in Appendix C (wildlife) is
devoted to species such as caribou that are not closely associated with fish. Information in
Appendix C could be used to assess direct impacts if the scope of the assessment is expanded.
For example, if the goal is to assess the impact of potential mining on the ecosystem (see
Executive Summary page 1, para 1), the information on caribou in Appendix C is more relevant.
The apparent diversity of goal statements cited in the main assessment gives mixed messages
regarding the clarity of the presentation (see more detailed discussion below).
RESPONSE: As the commenter notes, the scope of the assessment is focused on potential
risks to salmon from large-scale mining and salmon-mediated effects to indigenous culture
and wildlife. EPA agrees with the comnienter that direct effects on midlife are likely to be
important and that Appendix C (now a stand-alone USFWS document) provides useful
information for an evaluation of direct effects on wildlife from large-scale nutting. We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
direct effects. The revised assessment acknowledges the potential for direct effects on midlife
as well as risks due to fish, but states that these effects are outside the scope of the assessment.
PHW Response: Please reread the last sentence in the above Response. What does it mean that
risks due to fish are outside the scope?, Formatted: Font: Not Bold, Not Italic
USFWS RESPONSE: The scope of Appendix C is broader than that of EPA's assessment
because it is a USFWS document prepared to serve various purposes, including statewide or
regional land use planning, completion of environmental documentation for permitting of
development projects, and activities related to Landscape Conservation Cooperatives in
Alaska. The former Appendix C is nmv a separate USFWS report which is cited by the
assessment but is no longer an appendix of the assessment. However, information in the
USFWS report document has been used by EPA to provide a more complete assessment of
overall watershed resources at risk due to potential nuning, and to strengthen the assessment
of risks to wildlife from fish-mediated effects of the mine in the revised assessment.
The charge question related to wildlife asks for an evaluation of the risk to wildlife due to the
risk to fish. If the risk to fish cannot be quantified because there is little or no demographic
information, then any evaluation of risk to wildlife can't be quantified and must be qualitative.
Merely stating that a qualitative increased risk for fish will also result in a qualitative increased
risk for wildlife is not adequate. I am not satisfied with such an obvious and general conclusion. I
do not understand why the scope of the main document is limited to an indirect evaluation of
fish-caused risk to wildlife. The following responses to charge questions leans more toward an
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 38
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
ecosystem evaluation that includes, not only risk of fish to wildlife, but also risk of direct
wildlife and vegetation loss to fish and other direct risks to wildlife, such as noise and human
presence.
RESPONSE: EPA acknowledges that there are numerous potential direct risks to wildlife
from large-scale mining. However, this evaluation is outside of the scope of the assessment
The revised assessment provides a clearer explanation of the reasons for its defined scope.
PHWResponse: See other responses in this document., Formatted: Font: Not Bold, Not italic
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 39
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
2. RESPONSES TO CHARGE QUESTIONS
Question 1. The EPA's assessment focused on identifying the impacts ofpotential
future large-scale mining to the fish habitat and populations in these watersheds. The
assessment brought together information to characterize the ecological, geological, and
cultural resources of the Nushagak and Kvichak watersheds. Did this characterization
provide appropriate background information for the assessment? Was this
characterization accurate? Were any significant literature, reports, or data missed that
would be useful to complete this characterization, and if so what are they?
David A. Atkins, M.S.
Based on my general understanding of the watersheds, I consider the general background
information presented in the Assessment accurate and sufficiently complete for the endpoints of
this watershed assessment in the following areas:
• General view of Pacific salmon populations
• General view of resident (non-anadromous) fish
• Wildlife populations
• Native cultures
RESPONSE: No change suggested or required.
The Assessment also describes the current economics of the watershed, including commercial
and sport fishing and subsistence activities.
RESPONSE: No change suggested or required.
Additionally, the report highlights several general aspects of the area that make the fishery
unique in both its abundance and diversity:
• The unique hydrology of the area (strong groundwater and surface water interaction) that
contributes to stable Hows and temperatures favorable for salmon reproduction.
• The importance of anadromous fish in transferring marine-derived nutrients to upland
areas and thus providing nutrients to areas that would naturally be nutrient poor.
• The lack of roads and infrastructure that make the area unique as one of the few intact
ecosystems remaining in the world, and possibly unique for this type of fishery.
RESPONSE: No change suggested or required.
It would be helpful in the background section to better describe the uniqueness of the Bristol Bay
watershed ecosystem in the Pacific Northwest. This could include a description of other similar
ecosystems in the region that have undergone development and documentation of any changes in
fish populations associated with this development. The Assessment does mention the Fraser
River as an analogue, but the scale of development in this watershed, and even the success of the
salmon fishery, seems to be a point of contention, with some saying mining and fish coexist, and
other saying the impacts are severe.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 40
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The unique conservation value of Bristol Bay fisheries is now discussed in
Chapter 5.
It would also be helpful to better explain fish resources in the proposed project area in
comparison to other areas within the watershed. I understand some of the necessary data may not
be available for the project area. It would be helpful to know, however, if the habitat in the
project area is typical, exceptional, or inferior to that in other areas of the watershed.
RESPONSE: We now include figures showing reported salmon species distributions and
salmon diversity by HUC-12 watershed, across the Nushagak and Kvichak River watersheds
(Figure 5-3; Figures 5-4 through 5-8). It is informative to note that sahnonid diversity is
relatively high in the project area. Information on population sizes and vital rates are limited
for the region, but are reported where known. In addition, iiv include summary statistics and
figures of stream and valley characteristics across the assessment area (Section 3.4), and
compare stream attributes in the project area to those of the larger watersheds (Section 7.2.1).
These results generally illustrate that the project area contains streams of a size and gradient
well within the range of suitability for salmon, as amply demonstrated by the distribution of
spawning and rearing salmon within the project area streams (Figures 5-4 through 5-8).
Regarding geological resources, the report describes the Pebble deposit and five other mineral
deposits in the Nushagak and Kvichak watersheds. It would be helpful to know if there are other
mineral resources or oil and gas resources in the Bristol Bay watershed as a whole that could also
be exploited. It would also be helpful to describe the portion of the watershed that is off-limits to
development due to park and protected area status vs. those lands that are open to mineral
development.
RESPO.XSIC: The scope of the assessment was to evaluate the potential impacts from large-
scale mining on salmon resources; thus, consideration of prospective oil or gas development
in the area was outside the scope. The mineral resources identified in the assessment are those
in the Bristol Bay watershed that have had some level of identification or exploration at this
time. Mine claims within the Nushagak and Kvichak River watersheds are shown in Figure
13-1 and discussed in greater detail throughout Chapter 13. The assessment assumes that
mining would occur on lands open to mineral development.
Protected areas within the Bristol Bay watershed and the Nushagak and Kvichak River
watersheds are shown in Figures 2-3 and 2-4. We have clarified in the text that the Nushagak
and Kvichak River watersheds represent the least-protected area of the Bristol Bay watershed
Other state documents exist that map out areas in the Bristol Bay watershed off-limits to
development.
Steve Buckley, M.S., CPG
The background information presented in the characterization of the ecologic, hydrologic, and
geologic resources is overly broad in scope. Specifically, the descriptions of the relationship
between landforms, streams, and surface water and the interaction with groundwater are
mentioned as very important to fish in the watersheds, yet there is insufficient detail to assess
these interactions and consequently, the characterization of these resources is weak. There is
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 41
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
more detailed information available in the Environmental Baseline Document (EBD) regarding
the relation between landforms, streams, groundwater, and fish habitat in the watershed.
RESPONSE: Descriptions of the region's physical environment have been expanded in
Chapter 3. We provide additional detail on the broad-scale habitat characteristics of the
watersheds, but providing the detail necessary to assess groundwater interactions
comprehensively is beyond the scope of this document, and data are not available to do so.
Courtney Carothers, Ph.D.
The background information presented on the ecological and geological resources of the
Nushagak and Kvichak watersheds appears to be appropriate and accurate. The report notes that
there is a lack of quantitative data on salmonid populations in this region, a lack of a full
identification and characterization of salmon presence, spawning, and rearing areas, and a lack of
detailed understanding of how local stream and river system features (e.g., temperature, habitat
structure, predator-prey relationships, limiting factors) affect salmonid production in the region.
Further, climate change is noted to be affecting local conditions. These unknowns are important
to stress throughout the report.
RESPONSE: Each risk analysis chapter of the revised assessment now includes an
uncertainty section. Climate change is now incorporated more explicitly as an important
external factor that could interact with mining impacts (Box 14-2).
The cultural characterization presented in Appendix D presents detailed information on historical
and contemporary Yup'ik and Dena'ina communities of this region, stressing the centrality of
salmon and subsistence in these cultures. This assessment benefits from the time-depth of
relationships developed by Boraas and Knott. Overall, this section of the report is based on
standard ethnographic methods, although the research design and analysis could be explained in
more detail (and described in a separate methods section). The "voices of the people" sections
are helpful to present directly the perspectives given by local people. These quotes reveal the
complexity of subsistence and contemporary village concerns in this region. At times, the
cultural assessment can minimize this complexity.
RESPONSE: Additional detail was added to the methodology section of Appendix D.
As detailed in the specific comments below, potential risks and impacts to subsistence are
underestimated and at times framed in the report as primarily ones of physical health and
economic factors. As described in Appendix D, harvesting, processing, sharing, and consuming
wild foods are central to social, cultural, spiritual, psychological, and emotional well-being in
Yup'ik and Dena'ina cultures. The subsistence lifestyle is considered central to the health of the
people and communities of this region. This is particularly important to note for indigenous
communities who continue to cope with the legacies of colonialism. This point is made in
Appendix D (but at times could also be strengthened there, as suggested below), and is
articulated in some of the quoted interview material.
RESPONSE: The assessment text regarding the importance of the subsistence way of life has
been expanded to recognize the centrality of subsistence to the social, cultural, and spiritual
well-being of the indigenous cultures.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 42
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Recent data on subsistence harvests, use areas, and local context collected for the PLP
Environmental Baseline Document (as well as evaluation and discussion of such data, e.g.,
Langdon et al. 2006) and by the Alaska Department of Fish and Game (e.g., Fall et al. 2012)
would be a useful addition to the cultural characterization. Other studies of local traditional
ecological knowledge (e.g., Kenner 2005) may help to supplement the assessment of the
abundance and distribution of fish species in this region, or to supply information on other less-
studied freshwater fishes. Recent research on the contemporary salmon-based livelihoods of the
region (e.g., Holen 2011, 2009a, and 2009b; Hebert 2008; Donkersloot 2005) would also be
helpful to include. An inclusion of case studies of salmon-based cultures that have suffered
depletions of their resource base would add to the presentation of likely fish-mediated impacts to
culture (e.g., Colombi and Brooks 2012).
RESPONSE: The suggested references were consulted during the revision of the report and
the discussion of subsistence has been expanded. In addition, case studies have been cited
where applicable in the discussion of potential effects to indigenous cultures in Chapter 12.
Appendix E also characterized the economic baseline of the region. Why is this dimension not
asked about here?
RESPONSE: The focus of the assessment is potential effects on salmon from large-scale
mining. There are two secondary eiulpoints: salmon-mediated effects on wildlife and Alaska
Native culture. The economics related to potential salmon-mediated effects are not evaluated
because they are outside the scope of the ecological risk assessment Appendix E presents
information regarding the economic value of salmon is presented as backgroundfor the
descriptive material in Chapter 5, and could he used as a basis for future analyses. However,
this assessment does not include an economic emlpoint
Dennis I). Double, Ph.D.
As noted in the approach, characterization of and risk to ecological resources emphasized salmon
and other important sport and commercial fish species. Consequently, the description of non-
salmonid species generally lacked estimates of population size, except for sport and subsistence
catch statistics. There was a long list of other resident fish in Appendix A, but their role in the
Bristol Bay watershed (including the Nushagak River and Kvichak River watersheds) is not
described in any detail there or in the main report. Available data on known or perceived
ecological interactions among salmonid and resident fish should be included in the assessment.
RESPONSE: The assessment endpoints—salmonidfishes and their effects on wildlife and
Alaska Native cultures—have been clarified in Chapters 2 and 5; other fish species are thus
outside the scope of the assessment. However, we recognize in the text that other fishes (as
well as other biota) are important components of the ecosystem, and have included a table of
aU. documentedfish species in the region in Chapter 5 to better reflect the fish fauna in the
region.
Another limitation to the salmon-centric assessment is that risk assessment endpoints, described
in Chapter 3 of the main report, do not address other aquatic ecological resources. Consequently,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 43
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
while there was acknowledgment of ecological dependencies among salmon, other fishes, and
land mammals, very little information was provided on primary and secondary production
processes of aquatic communities. For example, the relative importance of marine-derived
nutrients (MDN) in the form of salmon eggs and carcasses is discussed, but there is only brief
mention of aquatic insects in the diet salmonid species. What nutrient levels occur in these
stream systems with and without MDN?
RESPONSE: We recognize that nutrient status, and more important prey availability, is a
critical component of habitat capacity for fish in these systems, and may be strongly driven by
salmon derived nutrients. We concur that more information is needed regarding potential
limiting factors for salmon productivity and capacity, and that food availability may be one
such factor. The role of aquatic invertebrates in the diet ofsalmonids receives more attention
in the revised draft, and is an essential part of the risk assessment for water treatment and
discharge, given the relatively high sensitivities of aquatic invertebrate taxa to metals.
However, because water chemistry data may not provide a complete picture of trophic status,
particularly where direct consumption of salmon Jlesli. eggs, and fry is of such high
importance as it is in many of the area streams, nv determined that nutrient status of area
streams is outside the scope of this assessment.
A description of major groups of aquatic invertebrates in terms of biomass and seasonal
abundance should be included in the main report. Further, aquatic and terrestrial food webs and
linkages need more embellishment. One approach might be to add narrative text with the
conceptual model discussion, including descriptions of community structure, function, and
biomass.
RESPOXSE: Additional detail on food webs is beyond the scope of this assessment (as
detailed in Chapters 2 and 5). Further, available data are inadequate to assess risks at that
level of specificity. For example, there are no acute copper toxicity data for any aquatic insects
and only one old chronic value for a caddis fly.
More detail on river and lake limnology would be helpful. For example, the hydrology of the
watershed is mainly limited to a brief discussion of salmonid habitats. The geology of the basin
emphasizes geology of mining areas and mineral processes. A more landscape-based description
is warranted given the importance of geology to surface water processes and groundwater
movement. The report would benefit from having a summary table listing lake size/volume and
river length/discharge for watersheds potentially affected (and not affected) by mining activities.
RESPONSE: We now include maps of geology and estimated mean annual flow for the study
region (Chapter 3).
Also missing were specific habitat requirements for rearing of juvenile salmon. A brief
description of where pink and chum salmon spawn and rear in the Bristol Bay watershed relative
to other salmon species should be included in the main report. There was nothing in Appendix A
on where coho, pink, and chum salmon reside within the Bristol Bay watershed.
RESPONSE: Identified spawning and rearing habitats for the five Pacific salmon species are
reflected in Figures 5-3 through 5-8, and additional text on salmon life histories has been
included in Chapter 5.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 44
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Each appendix has a wealth of supporting information and could serve as a stand-alone
document. However, having to work back-and-forth between the main report and appendices to
interpret critical aspects of the assessment presents a challenge. Don't assume the average reader
will read (and interpret) these appendices. To help remedy, the authors of the main report should
strive to directly cite relevant information (and/or a specific appendix) that supports their
conclusions.
RESPONSE: Additional information from Appendices A andB has been pulled into Chapter
5 of the main assessment. In addition, the purpose of the appendices has been clarified in
Chapter 2.
Gordon H. Reeves, Ph.D.
The assessment, which included the report and appendices, was comprehensive and thorough
regarding the ecological resources of the Nushagak and Kvichak watersheds. The best available
data on fish numbers and distribution (Alaska Dept. of Fish and Game's aerial escapement
counts, records from the Anadromous Waters uatalog and Alaska Freshwater Fish Inventory,
and the Environmental Baseline Document of the Pebble Limited Partnership (2011)) were used
for the assessment. These data formed the foundation for much of the assessment on potential
impacts to anadromous salmonids and their freshwater habitat in these watersheds and their
characterization appeared to be accurate. The authors also appeared to have thoroughly
identified and considered all of the appropriate literature.
RESPONSE: No change suggested or required.
Charles Wesley Slaughter. Ph./>.
If only Volume 1 (the Main Report) is considered, the characterization of some aspects of the
Nushagak and Kvichak watersheds would have to be termed cursory. Chapter 2, Volume 1
(Characterization of Current Condition) provides only a superficial overview of the landscape of
the Bristol Bay watersheds; a reader would preferably have access to Wahrhaftig (1965) or
Selkregg (1976), as only two (relatively dated) suggestions, to gain a more comprehensive
understanding of the region. Similarly, Volume 1 provides a relatively superficial discussion of
non-fish wildlife concerns, or human/cultural concerns
RESPONSE: Additional injormation on the region's physical environment from Selkregg
(1974) has been included in Chapter 3. We have also clarified that our discussion of biological
communities focuses on the assessment endpoints, as defined in Chapters 2 and 5.
By contrast, the information provided in Appendices A-H appears to be comprehensive and
complete for each subject field. (Appendix I appears to be a general "template" summary, not
tailored to the Bristol Bay watershed environment).
RESPONSE: The purpose of the appendices is. the main assessment document has been
clarified in Chapter 2. Appendix I is not meant to be specific to any given region, but discusses
options that are possible and notes that their applicability is dependent on site-specific
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 45
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
constraints. What would be chosen for the Bristol Bay watershed environment, given a mining
plan and permit application, also would be dependent on regulatory decisions.
As noted in the Executive Summary, the Assessment does NOT address several major
components of the (hypothetical) Pebble project, including electrical generation and
transmission, a deep-water port, or "secondary development" and associated infrastructure,
which would follow an initial mining project. A truly comprehensive analysis should
incorporate a full analysis of these aspects.
RESPONSE: The scope of this assessment was tailored to its purpose, as clarified in the first
two chapters. We would expect that a full evaluation of any future mining permit applications
and subsequent National Environmental Policy Act I 'm vironmental Impact Statements would
consider these components.
John D. Stednick, Ph.D.
The site characterization needs to be expanded. The report needs to better characterize the
physical setting. There are a variety of data sources that can be used to better describe the
physical setting. It would be useful to see geology, gcomorphology, soils, vegetation, digital
elevation maps, hypsometric curves of the watersheds in question, streamflow data, and
precipitation data—especially storm events and water quality data lor surface and groundwater
over time and space. Various geographical information system maps would be useful here.
RESPONSE: Maps displaying information about the physical setting have been added in
Chapter 3.
The salmon populations and habitat linkage needs to be better documented since many of the
mine impacts are resulted from hydrologic modification. Figures 3-2A to 3-2E represent good
thinking and an understanding of the linkages and potential effects of mining on these resources.
The linkages to indigenous peoples is illustrated in Figure 3-2E, but little text is presented,
referring the reader to the Appendix. The other conceptual models are not adequately addressed
in the text. These flow charts provide an opportunity to present processes and linkages as related
to potential effects of mine development activity and need to be developed within the text.
Indeed, they seem to stand alone with little discussion of potential effects. Additionally, not all
charts have adequate materials in the appendix for coverage, thus the variability in resource
coverage is inconsistent and infers either a writing bias or data (lack of) bias.
RESPONSE: Conceptual models are now linked with relevant text, and are included in each
of the risk analysis and characterization chapters.
The assessment concludes that a hydrologic modification will have detrimental salmon habitat
consequences. The groundwater contributions to streamflows are important, both hydrologically
and ecologically. Additional streamflow and groundwater data are needed to represent this
linkage. Similarly, additional water quality data over time and space are needed and should
include water hardness for metal standards. Depth to groundwater as related to streamflow, age
dating of waters, and streamflow modeling would all be useful to illustrate the groundwater
upwelling and hyporheic exchanges.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 46
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We have incorporated a figure illustrating modeled and observed groundwater
upwelling zones (Chapter 7).
Site disturbance will be significant, yet there is no discussion of soil erosion. Soil erosion and
subsequent suspended sediment transport would have the potential to have significant effects on
water quality, channel delivery efficiency, salmon, salmon habitat, and metal transport. There is
a generic discussion of road construction related to erosion, but road standards, road location,
road usage, road maintenance (salting, grading, or watering), and length of roads would help in
the risk assessment.
RESPONSE: Soil erosion on the mine site is not assessed because the scenario prescribes that
runoff will be directed to retention basins. Salts used to reduce dust and improve winter
traction on roads are discussed in Section 10.3.3 (Chemical. Contaminants in Stormwater
Runoff). Road usage and length are also factored into the risk assessment (e.g., in the
assessment of chemical spills (Section 10.3.3) and potential impacts from dust (Section
10.3.5)). Potential mitigation measures for stormwater runoff erosion, and sedimentation are
discussed in Box 10-3.
Are any endangered or threatened species present, either state or federally listed?
RESPONSE: Text has been added to Chapter S stating that there are no state or federal
endangered or threatened species in the region.
Roy A. Stein. Ph.D.
Overall Characterization. The characterization of the resources of the Nushagak and Kvichak
watersheds was appropriate and accurate in the ecological arena save for the issues discussed
below. Geological and cultural resources seemed adequately characterized, but they are not
within my expertise. Finally, given the emphasis on these two watersheds (not the entire Bristol
Bay watershed), might there be some consideration of a more circumscribed document title?
RESPONSE: The assessment deals with multiple spatial scales (now clarified in Chapter 2).
The Bristol Hay watershed is the largest spatial scale considered in the assessment, and it is
the only one that encompasses all of the issues discussed in the document.
Broad Scale Comuicnls:
Global Climate Change I. Risks to salmonids seem far greater than what is reviewed
throughout this portion of the document. Missing, in my view, is any consideration of Global
Climate Change, especially in light of the expected life of the mine (25-78 years), applied
directly to the Bristol Bay Watershed (save for a brief mention on page 5-28,2nd full paragraph).
Given our current understanding, general changes likely include more intense precipitation
events and increased temperature (and then of course, all that follows from these two changes
and as models become more sophisticated, more specific geographically localized impacts could
be assessed). With more intense storms come a greater likelihood of a failure of Tailings Storage
Facilities (i.e., commensurate with more frequent and more intense flooding), more acidity from
Pre-Tertiary waste rock (which will enter quite vulnerable, poorly buffered streams), and greater
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 47
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
sediment influx into streams (and increasing fines in the gravel by as little as 5%, quite a small
proportion, .causes unacceptable effects on salmonid reproduction" (page 8-6; also see
Chapter 7), which could occur during "routine operations", especially in light of the fact that
sediment influx into streams is a cumulative process). Increased stream temperatures, depending
on the absolute increase over a period of 78 years (and beyond, see "in perpetuity" comments
below), could lead to reductions in salmon spawning success, as extant populations are
specifically adapted to the current temperature regime. As is apparent, both increasing intensity
of storms and increasing temperature will likely compromise salmon spawning success, and
growth and survival of their offspring in the freshwater environment of Nushagak and Kvichak
rivers.
RESPONSE: Climate change projections and potential imparts are now included and
discussed in Chapter 3 and are included as important external factors in the risk analyses
presented in Chapters 7, 9,10 and 14.
What this would entail, at the very least, is a discussion of a monitoring system to quantify the
impacts of Global Climate Change whose impacts on the ecosystem can then be differentiated
from mine impacts. My concern is that if the mine is built, all negative impacts of the mine on
salmonids, etc., could be attributed to Global Climate Change rather than the true culprit which
would be the mining activities.
RESPONSE: Climate change projections and potential impacts are now included in Chapter
3, and Box 14-2 includes a discussion of the need for future monitoring to differentiate
climate change effects from large-scale mining effects.
Global Climate Change II. Indeed, climate change is affecting Alaskan salmon as
demonstrated (in a paper that just appeared online July 11, 2012) by a loss of a late-migrating
population of pink salmon in a small stream near Juneau, in favor of an early-migrating one.
Genetic evidence supports this explanation for Kovach et al. (2012) had 17 generations of data
(since 1979) showing the reduction of the September spawners in favor of the late-August ones
in response to increasing stream temperatures. As Kovach et al. (2012) write in their concluding
paragraph:
"We no longer observe the clear phenotypic distinction between early- and late-
migrating individuals that was once present in the system. Apparently, the very-
late-migrating phenotvpe has been greatly reduced or potentially lost. Although
microevolution may have allowed this population to successfully track
environmental change, it may have come at the cost of a decrease of within-
population biocomplexily - the loss of the late run. This is not a surprising result;
by definition, directional selection will decrease genetic variation. However, it
does highlight the importance of maintaining sufficient genetic and phenotypic
variation within populations in order for them to have the ability to respond to
environmental change."
The ramifications of this work are obvious. As pointed out in the report (pages ES-8, 2-22, 5-28
as just a few examples), the exceptional quality of the Bristol Bay salmon stocks depend on the
pristine quality of a set of quite diverse aquatic habitats, which has led to the development of
genetically diverse stocks of salmon within species, each uniquely adapted to particular habitats.
Reducing this variability by mining on top of the rivers that produce >50% of the wild sockeye
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 48
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
salmon in Bristol Bay serves to reduce the flexibility with which these stocks respond to any
environmental change (most notably Global Climate Change), and most notably during the time
course of the Pebble Mine.
RESPONSE: Climate change projections and potential impacts are now discussed in Chapter
3 and include these points.
Groundwater Exchange. One of the key aspects of this system is the importance of
groundwater exchange with surface streams and this groundwater contributes mightily to
salmonid egg incubation success and survival (page 2-21). Simultaneous with this is the fact that
the water demands of the proposed mine will require more than just surface waters available to
it, but rather the mine will have to exploit groundwater resources to support its operations. This
is yet another risk to salmonid success for reduction in the availability of groundwater will lead
to increased temperatures in summer (see pages 3-7, 5-28, 5-29) and less inviting overwinter
habitats (pages 5-20, 5-29), further exacerbating both mining and climate change effects.
RESPONSE: We have updated our hydrologic analyses to represent the probable influence of
mine scenarios on surface water/groundwater interaction. Climate change projections and
potential impacts are now discussed in Chapter
Exploration Effects. During the public testimony segment, several Alaskan Natives argued that
impacts owing to exploration have already occurred. A series of points were made: 1)
exploration equipment was left behind, despoiling the landscape, 2) noise from helicopters
frightened moose making them less vulnerable to exploitation, and 3) habitat change has already
begun just due to exploration activities.
RESPONSE: EPA acknowledges this testimony, but potential or actual impacts of exploration
activities are outside the scope of the assessment as defined in Chapter 2.
"In Perpetuity." Following up on the idea of increased risk (see previous points) to salmon, I
struggled with the idea of this mine being monitored and maintained "in perpetuity" (e.g., pages
ES-2, 4-32, 4-34). First, this relates directly to the Global Climate Change issues, in that these
changes likely will continue to build through time, further exacerbating negative impacts on
salmon. Even without climate change, salmon are in peril from mining operations in the
Nushagak and Kvichak rivers; with climate change, the cards are stacked against them.
RESPONSE: The post-closure phase of mining begins when reclanwtion is completed and
monitoring and maintenance commences using the controls put into place during closure;
exactly how long the site would require monitoring and maintenance is unknown, and thus
may be 'in perpetuity'. There are no existing examples from which to evaluate success of
treatment in perpetuity. No mine in Alaska has maintained a tailings pond into post-closure,
although one small mine did maintain a pond during a many year hiatus from operations.
Under AS 72.90.040, financial assurance is required to be sufficient to cover expenses for as
long as treatment need is predicted, even into perpetuity (e.g.. Red Dog Mine). Maintaining a
water cover over the tailings is a part of the reclamation and closure plan for the Red Dog
Mine. The comment is noted and understood. No changes suggested or required.
Second, what regulatory or institutional mechanisms currently available place the responsibility
of these efforts on the corporation "in perpetuity"? Because mining companies come and go,
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 49
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
might there be mechanisms that come into play if this particular company goes bankrupt? Might
there be some sort of bonding process that protects the environment from the mine's remains into
the long-term future? If not, should new legislation be pursued? Acknowledgement of this
important issue should be front and center in the document, in my view.
RESPONSE: There are many requirements that have to be met including compliance with the
CWA § 404(b)(1) Guidelines and adequate financial assurance under AS 72.90.040. The
former would lead to the least environmentally damaging, preferred alternative and the latter
to having adequate financial resources to cover the cost ofperpetual treatment. These issues
would be addressed in a permit process. Our purpose in the assessment is to evaluate the
potential effects of the primary features of a mine, assuming conventional modern mitigation
measures. Additional information on the regulations and financial, assurance issues
associated with mining has been added in Chapter 4 (Hows 4-2 and 4-3). The comment is
noted and understood
Third, I began the review process with idea thai the mine would be built, would capture its
resources, and then would end by restoring the site. The scenario that includes monitoring and
maintenance 1,000 years into the future continues to bother me. One solution that comes to mind
is that Federal or state government would be charged with these monitoring and long-term
maintenance activities, paid for by a hefty tax on the minerals removed from this site.
RESPONSE: Currently, the solution is the requirement and provision of adequate financial
assurance (under AS 72.90.040, when speaking specifically about Alaska) by the company.
The comment is noted and understood. No change suggested or required
Finally, I am not encouraged by any of the text surrounding this issue, the two most relevant
quotes (pages 4-31 and 5-45, respectively) being:
"There are no examples of such successful, long-term collection and treatment
systems lor mines, because lliese time periods (100's to 1000's of year) exceed
the lifespan of inosi pasl large-scale mining activities, as well as most human
inslilulions."
"We know of no precedent for the long-term management of water quality and
quantity on this scale at an inactive mine."
RESPONSE: The post-closure phase of mining begins when reclamation is completed and
monitoring and maintenance commences using the controls put into place during closure;
exactly how long the site would require monitoring and maintenance is unknown, and thus
may be 'in perpetuity'. There are no existing examples from which to evaluate success of
treatment in perpetuity. No mine in Alaska has maintained a tailings pond into post-closure
although one small mine did maintain a pond during a many year hiatus from operations.
Under AS 72.90.040, financial assurance is required to be sufficient to cover expense for as
long as treatment need is predicted, even into perpetuity (e.g.. Red Dog Mine). Maintaining a
water cover over the tailings is a part of the reclamation and closure plan for the Red Dog
Mine. The comment is noted and understood No changes suggested or required
And, finally, a quote from Chapter 8 on page 8-13:
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 50
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
"The promises of today's mine developers may not be carried through by future
generations of operators whose sole obligation is to the shareholders of their time
(Blight 2010)."
RESPONSE: The comment is noted and understood No changes suggested or required
William A. Stubblefield, Ph.D.
The EPA's assessment document presents a seemingly comprehensive compilation of the data
associated with the ecological, geological, economic, and cultural resources of the Bristol Bay
area. The characterization as presented seems to provide appropriate background information for
the assessment considering the hypothetical nature of the evaluation. Without having specific
knowledge of the area in question, it is not possible to provide an assessment as to whether the
characterization was accurate. I'm unaware of significant literature, reports, or data that were
specific to the site and would be useful for consideration. The assessment should be expanded to
include greater detail regarding the environmental aspects of the site.
RESPONSE: Additional information on the physical environment of the region and
assessment endpoints has been incorporated into ('hapten .> and x
Dirk van Zyl, Ph.D., P.E.
The geological information was taken from documents prepared to conform to and in compliance
with the standards set by National Instrument 43-101 (NI 43-101) (Ghaffari et al., 2011). This
regulatory instrument emphasizes resource information for projects. While I cannot comment on
the accuracy of the regional geological information, the document should reflect accurate
geological information of the Pebble District as known at the time when the report was prepared.
RESPONSE: The assessment uses geological information available for the Pebble site area.
Geological, information from Selkregg (1974) has been incorporated into Chapter 3.
My review did not include the Environmental Baseline Document (EBD) of the PLP. However,
in scanning that document, it seems that more site-specific information on site hydrogeology
may be available than was described in the EPA Assessment. While the latter refers to the EBD
extensively in terms of fish populations, etc., it does not refer to it for much of the site physical
characterization. EPA should address this in edits to the Draft Assessment.
RESPONSE: Additional site-specific hydrogeology information has been incorporated in
Chapters 3and 7 and in the calculation of water quality values in Chapter 8. EBD data were
used along with USGS data for hydrologic analysis in both drafts of the assessment, but the
sources of data were not discussed as extensively in the previous draft.
Phyllis K. Weber Scannell, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 51
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The Environmental Assessment presents a well-documented discussion of the fish and wildlife
resources of the Nushagak River and Kvichak River Watersheds, with more limited discussions
of the remainder of the Bristol Bay Watershed. The document discusses interactions among
species, including nutrient flows and the importance of groundwater systems; however,
information on contributions of marine-derived nutrients and existing pressures on the
environment are not as complete, or lacking. The information is general in nature. Should mine
development go forward, it will be necessary to obtain ecological information specific to the
potentially affected areas. The information should include timing of fish spawning, egg hatch,
in-migration and out-migration, and similar specific life-history information for important
wildlife species.
RESPONSE: We have clarified the use of information at di//erent scales (Bristol Bay
watershed and Nushagak and Kvichak River watersheds in the problem formulation chapters,
smaller spatial scales in the risk analysis and characteri:.atiou chapters). General information
on assessment endpoints is included in Chapter >. with more detailed information included in
the appendices.
Paul Whitney, Ph.D.
Fish Population Estimates. There are several places in the text where impacts of the loss and
degradation of habitat on fish populations was not quantified because of the lack of demographic
data for salmonids (e.g., page ES-26, third bullet). These statements are only partially accurate. It
is true that population models such as life tables or Leslie matrices require population age class
data to estimate population numbers. However, even if demographic data are available, these
population models do not relate population estimates to habitat quality. Incomplete data and
relating fish population estimates to habitat quality are not an uncommon problem in ecology and
there are many approaches for dealing with this issue. Approaches such as Ecosystem Diagnosis
and Treatment (McElhany et al. 2010), Expert Panels (Marcot et al. 2012), Bayesian nets (Lee
and Reiman 1997), Discussion with experts (Appendix G), or Weighing Lines of Evidence
(Section 6.1.5) are just some of the methods for relating habitat quality to fish abundance.
Models and expert opinions, of course, bring their own uncertainties but it seems better to have
quantitative estimates (and discussion of the estimates) of all the potential fish losses due to
habitat loss than no estimate at all.
RESPONSE: Approaches such as EDT, mentioned above, were considered, but rejected due to
lack of stream-reach specific information needed to provide the sort of quantitative estimates
desired Expert panels and Bayesian Belief Networks are recognized as potentially providing
useful guidance for identifying key uncertainties and directing future research and
monitoring efforts. However, this was deemed outside the scope of this assessment, and we
insteadfocus on the risks associated with the types of habitat change that would be expected
under the mining scenarios outlined We restrict quantitative estimates of population level
effects to the most severe cases where total losses of runs could be reasonably assumed
Even though the Executive Summary indicates that the impacts of loss and degradation of habitat
on fish populations could not be quantified, the text does provide some estimates. For example,
the assessment (page 6-11, first full para) estimates "that the combined effects of direct losses of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 52
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
habitat in the North Fork Koktuli, down stream in the mainstem Koktuli and beyond, and
impacts on macroinvertebrate prey for salmon could adversely affect 30 to 50% of Chinook
salmon returning to spawn in the Nushagak River watershed." This type of statement, and the
basis for the statement followed by a discussion of uncertainty, is a good example of the
estimates that would better describe possible impacts of the example mine on salmonids. Another
example estimate appears on page 6-39 for four species of salmon.
RESPONSE: We restrict quantitative estimates ofpopulation level effects to the most severe
cases where total losses of runs could be reasonably assumed, such as the example given
above. The text of the revised assessment has been clarified for consistency regarding
feasibility of estimates.
Question 2. A formal mine plan or application is not available for the porphyry copper
deposits in the Bristol Bay watershed. EPA developed a hypothetical mine scenario for
its risk assessment, based largely on a plan published by Northern Dynasty Minerals.
Given the type and location of copper deposits in the watershed, was this hypothetical
mine scenario realistic and sufficient for the assessment? Has EPA appropriately
bounded the magnitude of potential, mine activities with the minimum and maximum
mine sizes used in the scenario? Are there significant literature, reports, or data not
referenced that would be useful to refine the mine scenario, and if so what are they?
David A. Atkins, M.S.
The hypothetical mining scenario presented in the Assessment is based on a "Preliminary
Assessment Technical Report" of the Pebble deposit prepared for Northern Dynasty Minerals by
Wardrop (referred to as Ghaffari et al. 2011), in conformance with Canadian National Instrument
43-101 (NI 43-101) which is used to set standards for public disclosure of scientific and
technical information about mineral projects of companies on bourses supervised by the
Canadian Securities Administrators. By most accounts, the Pebble deposit is a world-class
deposit and the Wardrop report counts nearly 11 billion tonnes of total resource. It is unlikely
that all the ore currently identified would be mined, so 11 billion tonnes would be an upper
bound for this particular deposit. It is also certain that exploiting the Pebble deposit would have
to be at a scale large enough to justify the capital investment to build an infrastructure in such a
remote area. Although the Assessment is ostensibly about any mining development in the Bristol
Bay watershed, the use of the Wardrop scenario for Pebble effectively makes the report an
assessment of mining the Pebble deposit.
RESPONSE: The purpose of the assessment is to estimate potential impacts of large-scale
surface porphyry copper mining on salmon ecosystems in the Bristol Bay watershed The
preliminary plan for mining the Pebble deposit was used as the basis for the assessment
because that deposit is the most likely to advance in the near term Also, the Agency believes
that mining of other porphyry copper deposits in the watershed would proceed with a similar
approach, since the scenarios used are similar to what has been done at other porphyry copper
deposits. Therefore, it is appropriate to use Northern Dynasty Mineral's 2011 plan for the
Pebble deposit (Ghaffari et aL 2011) as the basis for the scenarios; however, a final mining
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 53
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
plan may differ from what is presented in Ghaffari et ah (2011). Chapter 13 of the revised
assessment also considers the potential cumulative effects of additional smaller copper
porphyry mines in the watershed No change suggested or required
The question then becomes what size mine is feasible from a technical and economic point of
view. The Pebble deposit mine plan, as presented in the Wardrop report, outlines three scenarios:
• An "investment decision case" for a 25-year mine life that would mine 2 billion tonnes of
ore;
• A "reference case" for a 45-year mine life that would mine 3.8 billion tonnes of ore; and
• A "resource case" for a 78-year mine life that would mine 6.5 billion tonnes of ore, or
55% of the total measured, indicated and inferred resource.
The Assessment chose minimum and maximum mine sizes of 2 billion and 6.5 billion tonnes of
ore, respectively. Thus, the resource estimate used for the Assessment is the same as that for the
two end members presented by Wardrop. This would make the mine one of the largest in the
world, exceeding the size of the 10th percentile of global porphyry copper deposits by an order of
magnitude (see Appendix H of the Assessment). Mines that ultimately become this size usually
expand by increments, as exploration discovers new ore zones and expansion permits are
granted.
RESPONSE: Yes, the scenarios represent large-scale mines. The purpose of the assessment is
to estimate potential impacts of large-scale surf ace porphyry copper mining on salmon
ecosystems in the Bristol Hay watershed, so large mine sizes are appropriate. It is quite likely
that large mines would be created in increments, but this would not influence our assessment,
as we have evaluated impacts based on volumes of material released in the event offailures or
accidents and on material processed as proposed in (ihaffiuri et al. (2011) as reasonable for a
deposit of this size, regardless of the time period for mine operation. However, we have
included a third, smaller mine in our revision to represent the median-sized porphyry copper
mine on a worldwide basis (250 million tons).
The Wardrop report further delineates Pebble West as a low-grade deposit near the surface that
would most efficiently be mined using open-pit methods, with Pebble East as a deeper, higher-
grade deposit that would most efficiently be mined using underground methods (specifically
block-caving). Mine facilities, as outlined in the Wardrop report, would include:
• Open-pit mining utilizing conventional drill, blast and truck-haul methods for near-
surface deposits.
• Underground, block-cave methods for deeper deposits.
• A process plant with throughput of 200,000 tonnes/day that utilizes conventional crush-
grid-float technology with secondary gold recovery.
• Other mine-site facilities, including:
o Tailings storage.
o Waste rock storage (the estimated waste/ore strip ratio is 2:1).
o A natural-gas fired power plant.
o Shop, office, and camp buildings.
o Pipelines to ship ore concentrate slurry to the port facility; return water from the
tailings slurry after separation at the port facility; and fuel.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 54
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: No change suggested or required
This mining and ore processing approach is conventional, and the Assessment includes these
elements. A mine developer may present alternative plans that could vary or alter how the mine
is developed, but the fundamental components would most likely remain the same.
RESPONSE: The EPA agrees with this comment No change suggested or required
Because the Assessment is presented as a general assessment of mining risks and impacts in
Bristol Bay and not a specific analysis of the Pebble Project, reliance on the scenario presented
in Wardrop makes the assessment overly specific. Further, Chapter 7 provides more specific
information on "Cumulative and Watershed-Scale Effects of Multiple Mines," which presents
analysis of potential impacts from mining five additional deposits in various stages of
development (presumably from early exploration to pre-leasibility). The information presented
in Chapter 7 seems more like another mining scenario than a cumulative impacts assessment.
Therefore, I would suggest a broader range of potential mining scenarios be organized as
follows, with the detail of assessment necessarily becoming more speculative with each
subsequent scenario in the list (due to the lack of geologic and engineering information on the
other deposits):
• Development of one, average-sized porphyry copper deposit (50th percentile or 250
million tonnes of ore as described in Appendix H) in the location of the Pebble deposit.
• Development of a mega-mine in the location of the Pebble deposit (of the range between
2 and 6.5 billion tons of ore) that may develop after multiple expansion and permitting
cycles.
• Development of a mining district consisting of an average-sized Pebble mine and other
potential mines (i.e., those presented in Chapter 7).
• Maximum development of all identified potential resources to their most likely ultimate
extent.
Considering this broader range of scenarios would help the reader to better understand the range
of potential risks and impacts.
RESPONSE: The Pebble deposit is located in the watershed of interest, the deposit is similar
to other copper porphyry deposits in the world, and components of the scenarios are common
and anticipated for any such deposit of this type; thus, we feel that use of the Pebble deposit
characteristics and location is appropriate. The revised assessment includes an additional
mine size scenario (Pebble ft 25), representing the worldwide median size porphyry copper
mine (Singer et a I. 2008). The revised assessment expands the cumulative impacts discussion
(Chapter 13) f urther by including transportation corridors and secondary impacts.
Steve Buckley, M.S., CPG
Additional mine scenarios are necessary to appropriately bound the magnitude of potential mine
activities. The maximum mine size in the mine scenario seems appropriate given the existing
public information on the Pebble deposit. The minimum mine size of 2 billion tons exceeds the
90th percentile of global porphyry copper deposits. Using a minimum mine scenario in the range
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 55
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
of 250 million tons or in the 50th percentile range of global porphyry copper deposits would be
more appropriate to bound the lower end of the magnitude of potential mine activities. It would
also be useful to include some variation in mining methods. This could include incremental
development of a smaller open pit in the lower grade zones of a deposit, along with a portion of
the higher grade deposit being mined by underground block caving methods to further assess the
minimum potential impact of the mine scenario.
RESPONSE: The revised assessment includes a mine size scenario (Pebble 0.25) representing
the worldwide median-sized porphyry copper mine as presented in Singer et al (2008). The
revision does not evaluate risks from hazards for underground mining, but a brief discussion
of underground mining is included in Chapter 4. The failures assessed would apply whether
the mining technique were underground or surface.
Courtney Carothers, Ph.D.
The hypothetical mine scenario was closely based on a probable mine prospect under
development. As such, it appears to be realistic and sufficient, if challenging to conceptualize as
fully hypothetical given this association.
RESPONSE: No change suggested or required
The report notes that the Pebble deposit may exceed 11 billion metric tons (4-17). The rationale
for choosing 6.5 billion metric tons as a maximum si/e is based "most likely mine to be
developed (4-19)." The rationale for not choosing a higher potential maximum could be
explained.
RESPONSE: liotli the 2 and the 6.5 billion ton scenarios were presented in Ghaffari et al
(2011) as economically viable, technically feasible, and permittable. The purpose of the
assessment was to estimate potential impacts of large-scale surf ace porphyry copper mining on
salmon ecosystems in the Bristol Hay watershed: the 6.5 billion ton mine is a large mine.
Because this size mine is on the lower bound of a maximum size, it is a conservative
assumption for the risk assessment. Thus, for the purposes of the assessment, it is not
necessary to hypothesize an even larger mine.
Dennis D. Dauble, Ph. I).
The hypothetical mine scenario initially appeared realistic and useful in terms of potential project
scope. However, it was apparent during the public hearing, and upon further discussion between
members of the panel, that assumptions on mine size should be revisited based on deposit
characteristics and extraction potential. Also, assumed practices and operations should be
verified against current best-practice and State of Alaska permitting guidelines.
RESPONSE: The revised assessment includes a smaller sized mine that is based on the
median-sized porphyry copper mine on a worldwide basis. The State of Alaska does not have
permitting guidelines that address the size of a mining operation. Land use activities were
previously subject to stipulations meant to minimize surface damage or disturbance under 11
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 56
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Alaska Administrative Code (AAC) 96.140, but this regulation was repealed in December
2002. The State does have statutory and/or regulatory requirements for an approved Plan of
Operations (11 AAC 86.800), a Reclamation Plan (Alaska Statute (AS) 27.19.30) and
appropriate Financial Assurance (AS 27.19.040).
Referenced literature provides appropriate context, however, I cannot help believe that
information on environmental impacts from past mining activities conducted in the Rocky
Mountain metal belt would be relevant to this assessment in some cases. It is also possible that
recent published information from Holden Mine in northern Washington State would help
establish context for effects of leachates and model results that predict downstream transport of
tailing material in a wilderness setting, for example.
RESPONSE: Environmental impacts from historic mining are the basis for understanding
that risks from hazards of mining need evaluation. Modeling of tailings transport was based
on the expected characteristics of tailings for the Pebble deposit. There is an expanse of
literature on Superfund sites and interactions oj metals associated with sediments and their
leaching. We included a number of selected sites in our background information, but to
include ail possible sites would get further away from the scope of the assessment, which was
to evaluate potential effects within the Bristol /lay watershed.
Gordon H. Reeves, Ph.D.
No comments on this question.
Charles Wesley Slaughter, Ph. I).
Given the available information base for the ore deposits of the Bristol Bay watershed, and the
publicity which has attended the Pebble planned development over the past several years, the
Assessment's hypothetical mine scenario seems fairly realistic. Further, it is appropriate that the
Assessment consider the probable impacts of other future mineral development projects once an
initial entry (presumably Pebble-Northern Dynasty Minerals) has been accomplished. Such
subsequent development - "cumulative effects over a long time period" - could (and should)
receive more emphasis than is accorded in the Assessment.
RESPONSE: The assessment oj cumulative effects of multiple mines is given more emphasis
in the new Chapter 13.
John D. Stednick, Ph.D.
The document does not adequately bound the range of mine scenarios. The minimum mine
development scenario is not adequately addressed. A frequent criticism during the public
comment session was that mine plans presented in the assessment are not representative of
current standards. A compilation of existing world porphyry mine complexes as well as other
types of mines specific to Alaska would better inform the reader of mining processes and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 57
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
potential risks. The physical setting in Southwest Alaska is not the same as the Bingham Mine in
Salt Lake City. Currently, the document refers to a particular mine in a particular risk
assessment (stressor), e.g., the Fraser River for salmon, Aitika for chemistry, and Altiplano for
pipeline failures.
RESPONSE: The revision includes a smaller mine size that represents the worldwide median
size for a porphyry copper mine (Singer et al 2008), to help with the issue of the range of
scenarios. EPA disagrees that the mine scenarios evaluated are not representative of current
standards. This view apparently stems from use of the term "good" rather than "best"
practices in the draft assessment. The reason for using that term is that the term "best
management practices" is a term generally applied to specific measures for managing non-
point source runofffrom stormwater (40 CFR Part 130.2(m)). Measures for minimizing and
controlling sources ofpollution in other situations are often referred to as best practices, state
of the practice, good practice, conventional practice, or simply mitigation measures. We
assume that these types of measures would be applied throughout a mine as it is constructed,
operated, closed, and maintained post-closure, regardless of the qualifier that one wishes to
place with it. A text box was added to the revised Chapter 4 that discusses terms to help clarify
our intention for descriptors used The Fraser River example was considered because it had
been used as an analogue by others, but was dismissed as not representative of Bristol Bay.
Other mines that are noted in the assessment are illustrative of specific issues only and not
usedfor risk evaluation in the Bristol Hay watershed. II liile physical settings are not the same,
the components and the impacts are .similar and thus included to help a reader understand
where and how these things occur.
The Bureau of Land Management has identified certain lands that will be excluded from
development. This reference needs to be followed up.
RESPONSE: The purpose of the assessment it'u.v to estimate potential impacts of large-scale
surface porphyry copper mining on salmon ecosystems in the Bristol Bay watershed This
presupposes that our mine scenarios are located in areas that are not excluded from
development. The majority of land in the two watersheds is state land that is available for mine
development, and Figures 2-3 and 2-4 now indicate protected areas within the Bristol Bay
watershed
Roy A. Stein, Pli.l).
Hypothetical Mine Scenario. I hough mining does not lie within my area of expertise, I
thought that this scenario helped me understand the potential impact of a mine of this magnitude
in a wilderness, pristine watershed. I find it difficult to comment as to whether this scenario is
realistic and sufficient, though I did use this scenario to guide my comments below. From the
text, it is apparent that this is a realistic scenario, based on documents filed by the company with
the Canadian government. This makes this scenario the most realistic one could expect.
a. Minimum and Maximum Mine Size. For me, as an ecologist, this bounding helped me
to understand the potential impacts of the Pebble Mine, though I did not understand what
the probability of either mine size happening in the near term. Understanding these
probabilities would be helpful to the readers.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 58
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
b. Mine-Size Continuum. Is it more likely that the initial Pebble Mine will be maximum
or minimum in size? Wouldn't it be far better to review a continuum of mine sizes from
the smallest that is economically feasible to one that is intermediate in size and then to
one (or two) that would take to the largest realistic mine size? With this continuum, the
reader begins to understand the overall impact of various mine sizes on the Bristol Bay
ecosystem. Some reflection on these mines sizes and their impacts would have helped
me interpret the Environmental Risk Assessment with some additional insight.
RESPONSE: The State of Alaska does not have permitting guidelines that address the size of
a mining operation. Many identified deposits never become developed mines for various
reasons, and it is unknown how many deposits exist and are economically viable for
exploration of mining feasibility. Once it is decided to develop a site, there are a number of
things that must occur before a mine begins operation. Thus, it is not possible to predict the
probability of either mine size happening in a specific period, at least with any certainty. All
we can say is that there are deposits that have the potential to be mined in the future.
It is more likely that an initial mine would begin at a smaller scale and become larger and
perhaps be permitted in stages of increasing size. However, there are different approaches in
how plans are presented and these depend on multiple factors, including economics and
projected costs/gains in prices of the metal being mined. For example, if it were not
economically viable to mine only a small part of a known large deposit, a larger mine would be
proposed and planned for. The revised assessment includes a size scenario that represents the
worldwide median-sized mine to provide more of a continuum of sizes.
One Watershed. Given the productivity of salmon from these two river systems (50% of the
sockeye salmon in Bristol Bay are produced from these rivers), might there be some thought
given to limiting the mining operations to a single watershed, either the Nushagak or the Kvichak
(page ES-2)? In so doing, in a single stroke, the impact of this mine on salmon is reduced by
50% or more. Could the Pebble Mine be confined to one watershed, such as where the majority
now falls - in the Nushagak River (both the north and south forks of the Koktull River)
watershed? Even so, this suggestion becomes especially pertinent to Chinook salmon spawning
in the Nushagak River, for this run is "near the world's largest" (page ES-5), but yet the
Nushagak watershed is small relative to other watersheds (such as the Kuskokwim and the
Yukon) where Chinook salmon are abundant. As a result, any impacts to the watershed by a
mine of this size are magnified, another concern when considering this location. Without mining
expertise, I cannot judge whether it would be possible to mine in only one of the watersheds,
rather than both. Even so. some consideration should be given to this suggestion.
RESPONSE: Restricting impacts to one watershed would change the risks, and could be a
part offuture mine plans. We chose to represent a suite of realistic mine scenarios based upon
preliminary mining plans and the location of the ore deposit which lies in both watersheds.
William A. Stubblefield, Ph.D.
No comments on this question.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 59
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Dirk van Zvl, Ph.D., P.E.
The hypothetical mine scenario adopted by the EPA relied almost exclusively on the document
prepared for Northern Dynasty Minerals (NDM), one of the partners of the Pebble Limited
Partnership. Developing a mine plan for a specific ore body is a large task and is undertaken by a
large team of engineers and scientists. In the process of developing a mine plan many options are
considered for each facility and its components, including mining methods, process design
options, waste rock management options, tailings management options, shipment of product, etc.
The hypothetical mine scenario was prepared by an independent consulting company for one of
the partners and this plan does not necessarily represent the design and management options that
will be selected for developing this ore body. Because of ore grades and the deposit style, it is
most likely that an open pit mine will be developed as assumed in the report for the western
lower grade ore body and that underground mining will be used for the eastern higher grade ore
body. The size of the ore body and the strip ratio for an open pit mine are completely dependent
on metal prices and production costs at the time of mine development. Metal prices and
production costs will also be a major factor in deciding whether to first develop an underground
mine instead of an open pit mine. While some of the components of the final mine may contain
elements of the conceptual mine, it is impossible to know whether the hypothetical mine scenario
is realistic, as will be further discussed in the comments below.
RESPONSE: It is acknowledged in the assessment that the mine scenarios might not look
exactly like a mine presented in a mining plan. The assessment is not a mining plan and is not
an evaluation of a mining plan; it simply uses current in f ormation for the Pebble deposit
because it is a large ore deposit which has had extensive exploration with potential for
development in the near future. An additional scenario has been included in the revision to
match the worldwide median mine size and show a better continuum of scenarios possible. We
consider our scenarios to be realistic, as thev were presented as possible layouts for the Pebble
deposit and stated in (ihajjari el al. (2011) as being ''economically viable, technically feasible,
and permittable".
To address the issue of sufficiency it is necessary to understand the range of potential outcomes
related to the various options. For the most part, the EPA study used the information from the
NDM document for evaluating impacts to salmonids. Using different options, both technological
as well as site selection, for some or many of the facilities could result in impacts that are
different from those described in the report. I would therefore suggest that using only the present
hypothetical mine scenarios is insufficient. There could be a range of impacts, such as the
surface areas of facilities, which in some cases could be smaller than what was chosen and in
other cases larger. However, this does not mean that the hypothetical mine represents "average
conditions." I therefore consider the mine scenario not sufficient for the assessment.
RESPONSE: It is acknowledged in the assessment that the components in the mine scenarios
might not be exactly what would be proposed in a mining permit application for this location
or for other locations within the Bristol Bay watershed The purpose of the assessment was to
evaluate potential impacts of large-scale surface porphyry copper mining in the Bristol Bay
watershed, so the assessment was not meant to represent "average conditions". However, an
additional mine size scenario has been included in the revision to match the worldwide median
porphyry copper mine size and show a better continuum of scenarios possible.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 60
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The minimum and maximum mine sizes selected by EPA are 2 billion tonnes mined over 25
years and 6.5 billion tonnes mined over 78 years; in both cases, the daily ore processing rate is
200,000 tonnes. As indicated above, the final economic mine size at the time of development
will be determined by metal prices and production costs. Note that production costs, as used
here, include all the considerations related to regulatory, environmental and social aspects of the
mine and its environs. Mining companies typically make investment decisions for periods of 20
to 30 years. It is seldom, if ever, that a new investment will be made based on a 78 year mine
life; however, the upside potential will be taken into account when an investment for a shorter
mine life is made. It is also unlikely that environmental regulatory agencies will consider issuing
a permit, including closure plans, etc. for a 78-year project. Furthermore, even if the mine
ultimately continues for 78 years, it is certain that the operating and environmental control
technologies and societal expectations will change in that period and therefore the elements used
by EPA for the maximum size hypothetical mine will certainly not be valid for such a long mine
life. It is therefore my conclusion that assuming the development of a 2 billion tonne ore body is
realistic, but that assuming development of a 6.8 billion tonne ore body, using static technology
assumptions, is not.
RESPONSE: While it is true that an actual mine likely would be permitted in increments, the
assessment never stated that the ~K-yr scenario would not be done this way. In fiict, the
assessment does not discuss how such scenarios would be permitted at all, and to do so is
outside the scope of the assessment. The purpose of the assessment is to estimate potential
impacts of large-scale surf ace mining of copper porphyry on salmon ecosystems in the Bristol
Bay watershed, and that necessarily assumes that the mine scenarios are permitted It is true
that some technologies would hare advanced over time from 'day one' of mine scenario
development, and we acknowledge this in our assessment It is impossible to predict, however,
how impacts from use of future technologies would differ from those in use today, or how they
would change conditions existing at the time they began being u sed. We can only present and
predict potential, impacts based on use of the most appropriate technologies available at the
current time. No change suggested or required.
The EPA assessment report includes a range of the literature and reports in evaluating the
selected mine scenario. However, I have a number of specific comments about various aspects of
the report as well as the references.
Good practice vs. best practice.< >n p. 4-1 of the report, the EPA states: "Described mining
practices and our mine scenarios reflect the current practice for porphyry copper mining around
the world, and represent current good, but not necessarily best, mining practices". EPA does not
clarify this decision, nor does the report clarify the distinction between "good" and "best"
practices. It can only be concluded that "best" will be better than "good". On the basis of this, it
is inconceivable to me that the Bristol Bay communities, the Alaska regulatory authorities as
well as Federal Regulatory Authorities will not demand that the company follow "best mining
practices", however that is defined at the time. It is also inconceivable to me that the company
will not follow "best mining practices" in the design and development of such a mine. During the
engagement processes, the stakeholders will have to agree what represents "best" practice in the
design of the mining project. It is important to note that most of the failure statistics used as a
basis for the evaluations in the report are derived from data gathered over the last 50 years or so
(e.g. refer to p. 4-45 of report). It may be argued that this information is mostly for mines
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 61
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
following "good" practices and, in many cases, for projects that had a lower standard of care. To
my knowledge, there are no statistics available that compare failure rates of facilities designed
and operated under "good" practice to those designed and operated under "best" practices,
whatever definitions are used for "good" and "best".
RESPONSE: The term "best management practices" is a term generally applied to specific
measures for managing non-point source runofffrom stormwater (40 CFR Part 130.2(m)).
Measures for minimizing and controlling sources of pollution in other situations often are
referred to as best practices, state of the practice, good practice, conventional, or simply
mitigation measures. We assume that these types of measures would be applied throughout a
mine as it is constructed, operated, closed, and post-closure, regardless of the qualifier that
one wishes to place with it. To remove any ambiguity and suhjectiveness of terms "good" or
"best", we have removed them in the revision and have added Box 4-1, which includes
definitions for several terms used
The EPA also is not aware of any statistics available that compare dams designed (and/or
operated) under different standards; however, the probabilities Jor dam failure used in the
assessment were not derived solely from the historical record. Historical, failures were
discussed as supporting background inf ormation and present a defensible upper bound on the
failure probabilities. The failure probabilities used in the assessment are based on Alaska's
dam classification and required safety factors applied to the method ofSilva et aL (2008). The
data presented by Silva et aL (2008) consider only the annual probability offailure from slope
instability, but the methodology is equally applicable to oilier failure modes. The discussion of
failure probabilities in the revision (Chapter 9) is expanded to clarify this issue.
Mine scenarios. The executive summary indicates (p. KS-1 I): "The mine scenario includes
minimum and maximum mine sizes, based on the amount oi ore processed (2 billion metric tons
vs. 6.5 billion metric tons), and approximately corresponding mine life spans of 25 to 78 years,
respectively". This seems to indicate that the mine life cycle in the first case consists of 25 years
of operational life followed by closure and, similarly for the second case, 78 years of operational
life followed by closure. However, a careful review of the water management section (section
4.3.7) indicates that this is not the case. The EPA water balance calculations are simplified to a
set of deterministic values in Tabic 4-5 for four water management stages during the overall
mine life cycle: start-up, operations minimum mine (25 years), operations maximum mine (78
years), and post-closure. For post-closure, only the 78-year mine life numbers are used. It
therefore seems that EPA is not considering that the 25-year mine will close, but that its life will
automatically be extended to 78 years. Does this mean that the EPA really does not evaluate the
minimum mine size completely, i.e. the 25-year mine life followed by closure? It is important
that this be clarified as it would be inconsistent not to evaluate closure of the 25-year mine. It is
possible that additional evaluations, or at least additional explanations, will be required to clarify
this.
RESPONSE: The reviewer is correct. Our water balance calculations only explicitly present
the closure of the Pebble 6.5 scenario. The water balance for the Pebble 0.25 and the Pebble
2.0 scenarios would be similar, but in those scenarios there would be no water captured in
TSF 2 or TSF 3 and the amounts captured in the pit would be proportionally smaller, as they
are in the operating scenarios. While the pit is filling in each of the three scenarios, the total
amount of water captured is slightly less than the amount captured during operations, and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 62
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
about 35% to 45% of the water captured is available for reintroduction to the streams. Once
the pit is full, the amount captured at the mine pit area drops substantially and 100% of the
captured water is reintroduced Post closure flows for the three mine scenarios are presented
in Table 6-8.
Tailings management technologies. Ongoing technology development has resulted in a broader
range of tailings management options than only slurry tailings disposal. Filtered dry stack
tailings can be considered as a realistic option, even for mines with higher production rates.
Flotation of remaining sulfides in the tailings before deposition is also a realistic option for
mines; it has been done successfully at the Thompson Creek Mine in Idaho for the last 18 plus
years. While these technologies are mentioned, they are not selected for reasons such as
technology not being appropriate for the climatic conditions and concerns with disposal of pyrite
waste. Both of these are not insurmountable technical issues and adopting such management
options will reduce failure probabilities and potential impacts following a failure. The failure
mode of a filtered dry stack facility not containing sulfides will be completely different from a
slurry impoundment and the potential environmental impacts of these other tailings management
options will definitely be far smaller than those for the selected mine scenario using slurry
tailings disposal.
RESPONSE: Selective flotation to lower the pyrite content has been added to the discussion of
processing in Chapter 4 and ref erenced in Chapter 6 as being done in the assessment
scenarios. How tailings are managed within the impoundment can affect water chemistry and
a dry stack with sulfides removed may produce the best water quality results after reclamation
if the fate of the sulfide tailings is never considered. According to Ghaffari et aL (2011), 14%>
of the tailings produced would be pyritic (with the selective flotation method used), which
equates to an average of28,000 tons/day in a 200,000 tons/day mining operation, or over 255
million tons during a 25-year estimated lifetime for that scenario. These tailings need to be
managed in such a manner that oxidation does not lead to acidic drainage and the most
effective way is to deposit them subaqueously. Additionally, a "rule of thumb" for design of
dry stack tailings is to allot 25 acres for every thousand dry tons of tailings per day over the
life of a 20-year operation (SME Mining Engineering Handbook 1973). Assuming our
scenario of200,000 tons per day processed and that 99% of this material was waste tailings,
this would amount to 4900 acres (25 *200,000 *. 99/1000) or 19.8 km2 over 20years. This
exceeds the internal, surface area taken up by the TSEfor tailings deposited over 25 years for
the same mass of material processed per day by 5.6 km2; thus, even if there were not a risk of
the PA (1 tailings acidic leaching potential, dry stack tailings disposal at this site would create
additional surf ace area loss versus the scenario \ trailitional dam
Waste rock management. The waste rock management plan on p. 4-13 calls for the potentially
acid generating (PAG) waste rock to be separated from the rest of the waste rock and states that
the "PAG waste rock might be placed in the open pit at closure to minimize oxidation of sulfide
minerals and generation of acid drainage". However, on p. 4-33 it is stated that: "PAG waste
rock will be processed through the flotation mill prior to mine closure, with tailings placed into
the TSF (tailings storage facility) or the mine pit." These two alternatives represent completely
different management, economic and environmental conditions and are not consistent. Milling
the PAG waste rock represents a higher cost than placing the PAG rock in the pit and placing the
PAG waste rock tailings in the TSF will increase the size of the TSF. Placing the PAG tailings in
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 63
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
the pit will set up a completely different management scenario than placing the PAG waste rock
in the pit. The EPA should clarify which option or range of options they select for evaluation and
use that consistently in the assessment.
RESPONSE: These statements have been made consistent The revised assessment scenarios
include processing the PAG waste rock over the course of operations to minimize the length of
time a PAG waste rock pile would be on the surface, and thus minimize its potential for
oxidation and subsequent release of acidic leachate. There is no longer mention of PAG rock
being disposed in the pit at closure (Section 6.3.3).
Water balance and management - waste rock. Mine site water balance and management is a
very complex issue as recognized by the EPA on p. 4-27: "... water balance development is
challenging and requires a number of assumptions". I because of these uncertainties, complex
probabilistic dynamic models are employed at mines where the site details are better defined than
that of the EPA hypothetical mine scenario. The information in Box 4-2 indicates that the
"captured flows include water captured at the mine site and the TSFs (Table 4-5). The total
amount of water captured at the mine site includes net precipitation (precipitation minus
evapotranspiration [footnote: during operations most of these areas will not be covered with vegetation
and the correct terminology here is "evaporation"]) over the areas of the mine pit, the waste rock
piles, and the cone of depression (without double-counting any areas of overlap)". On p. 4-23 it
is stated that: "Monitoring and recovery wells and seepage cut-off walls would be placed
downstream of the piles to manage seepage, with seepage directed either into the mine pit or
collection ponds". Figure 4-9 shows this schematically where leachate from the waste rock enters
the groundwater that then flows to the mine pit or to the monitoring and collection well.
However, if net precipitation only includes the components above (precipitation minus
evapotranspiration), effectively excluding infiltration, and if this net precipitation is captured
from that waste rock pile (as stated in Box 4-2), then there should not be any water available to
infiltrate into the waste rock pile, i.e. there should not be any leachate. All references to seepage
from the waste rock piles are incorrect following the EPA's assumptions of total capture of net
precipitation. In addition, the approach that is used in the water balance is inconsistent with
observed field performance and descriptions in the literature, as is it difficult to imagine a case
where there is zero infiltration into a porous waste rock pile (e.g. Nichol et al., 2005 and Fretz et
al., 2011). The EPA must clarify the whole water balance model and the evaluations. For the
assessment to have any credibility, the water balance and management evaluations should reflect
realistic conditions.
RESPONSE: The term "evapotranspiration" has been corrected to 'evaporation' in the
revised assessment when discussing the operational phase. Total precipitation equals the sum
of evaporation, transpiration (where applicable), runoff and infiltration. Net precipitation
included in the water balance includes all water falling onto the site components minus water
leaving only via evaporation (i. e. it includes both runoff and infiltration). Leachate/seepage
water originates from precipitation which has infiltrated the waste rock piles or tailings onto
which it fell Therefore, the discussion of seepage is not incorrect, and there is no indication
that the scenarios represent "zero infiltration", as we discuss (as noted in the comment made)
how seepage and leachate is managed. However, the water balance section has been revised
for clarity (new Section 6.2.2). As the commenter notes, the assessment discusses seepage
collection systems. However, the assessment does not assume total capture. Our water balance
assumes that 50% of the leachate that is lost from the TSFs andfrom the portion of the waste
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 64
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
rock piles outside the drawdown zone of the pit escapes into the groundwater and eventually
into the streams.
Dam failure - tailings storage facilities. During operations, "water falling within the perimeter
of a TSF would be captured directly in the TSF, but runoff from catchment areas up-gradient of
the TSF would be diverted downstream" (p. 4-27). At closure, water would be removed from the
TSF providing more storage, but also maintaining a small pool to "keep the core of the tailings
hydrated and isolated from oxidation" (p. 4-32). This seems to assume that the diversion systems
will be kept in place and most likely will be upgraded to divert up-gradient surface water around
the tailings impoundment. It is likely that the design criterion for the upgraded diversion system
during the post-closure period will be the probable maximum Hood (PMF) as is done at a number
of mines. Dam failure analyses were done assuming that the Hood leaving the TSF includes the
PMF inflow from the up-gradient catchment, excess water on top of the tailings and 20% of the
tailings volume (Box 4-8). While one can argue that a failure including all these materials may
be a plausible, although a very low likelihood event during operations, it seems less probable that
such a failure will take place for the mine closure period when an upgraded diversion system is
in place. Also, during the closure phase, the tailings will consolidate and be less mobile. Note
that the densification behavior of oil sand tailings referred to on p. 4-32 (i.e. the Wells, 2011
reference) does not apply to copper tailings. The presence of clay minerals and bitumen in the
mature fine tailings portion of the oil sand tailings is the source of the different behavior
(Znidarcic et al., 2011).
RESPONSIC: It is assumed that post closure scenarios will hare drainage facilities in place
that could safely pass storm events. However, the decision was made to assume a failure that
was consistent for both the operational and post operation scenarios. The PMF would overtop
the main dam of the TSF and drain into the North Fork Koktuli as a representative failure,
assuming that drainage facilities were either not yet operational or hadfailed It is important
to note that the PMF peak flow generated only 291 cms, which is small when compared to the
peak flow release at the failing dam (149,263 cms and 11,637 cms for the large and small
failures, respectively). 20",, was selected to represent the volume releasedfrom the TSF
because it fell within a reasonable range when compared to release volumes of historic
failures. The 20% volume included both solids and pore water.
Reclamation slope of waste ruck. On p. 4-32 it is stated that: "We assume that NAG waste rock
would be sloped to a stable angle (less than 15%) (Blight and Fourie, 2003)". I contacted Profs.
Geoff Blight and Andy Fourie about this statement and received the following response from
Prof. Blight: "The only reference to 15 degrees (not 15 %) slopes is the following, talking about
the outer tailings, not waste rock covered, slopes of decommissioned TSFs: "it must be
remembered that the outer slopes will need to be rehabilitated, and that for vegetation to be
stable, and surface erosion minimal, the maximum outer slope should not exceed 15 degrees."
This error in reference must be corrected; it is recommended that more typical closure slopes of
about 30%) (or 3H: 1V, about 18 degrees) for waste rock should be used in the evaluations.
RESPONSE: The text has been changed to correctly read IS degrees in Chapter 4 and
Appendix I.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 65
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Phyllis K. Weber Scannell, Ph.D.
The Environmental Assessment discusses a hypothetical mine (given that mine plans have not
been developed). Page 4-5 of the document states that "rocks associated with porphyry copper
deposits tend to straddle the boundary between net acidic and net alkaline . . The Pebble
Project Environmental Baseline Report (SRK 2011, Chapter 11) summarizes testing on the
samples from the pre-Tertiary porphyry mineralized rock in Pebble East Zone (PEZ) and Pebble
West Zone (PWZ). The metals leaching/acid rock drainage study showed acidic conditions
occurring immediately in core with low NP, but the average delay to onset of acidic conditions
was estimated to be about 20 years. Copper was leached in the highest concentrations, but Co,
Cd, Ni, and Zn also leached from samples from PEZ. Wacke (sedimentary rock) samples from
PEZ and PWZ leached As, Sb, and Mo, in addition to Cn. (SRK, page 58). The available
information on acid generation and metals leaching appears to be preliminary. Development and
permitting of a viable mine plan will require extensive sampling and data analysis of ore
samples, plans for classifying waste rock (as PAG and NAG), and, possibly, plans for collecting
and treating runoff and seepage waters.
RESPONSE: EPA agrees that developing and permitting a viable mine would require
extensive information. The assessment is not a mining plan. The scenario presents a suggested
treatment option for mining influenced water and settling ponds for water that is simply
stormwater runoff No change required.
The Environmental Assessment seems a bit premature in making an assessment of the potential
for acid rock drainage (ARD) or metals leaching (ML). Data on metals other than Cu are
insufficient and possible toxicities to fish are not addressed. Further, the description of the
potential mine may not reflect a likely mine scenario. It is difficult to calculate potential risks to
the environment without a specific mine plan. The section of the Environmental Assessment
should be revised as more data on ARD and ML become available.
RESPONSE: The assessment uses the geochemistry data that are available from the Pebble
Limited Partnership. Copper was emphasized in the review draft because the EPA believed,
and still believes, that it is the contaminant of greatest concern. Toxicities to fish of the other
metals were not discussed because they had been screened out. However, the revised
assessment explains the screening process and the selection of copper in more detail in a new
section on the identification of stressors of concern (Section 6.4.2) and more metals have been
added to the screening assessment. The toxicities of all metals reported in the leachate are now
addressed either as individual, elements or, in the case of major ions, as contributors to total
dissolved solids. The mine scenario is based on the most recent preliminary plan released by
Northern Dynasty Minerals (Ghaffari et aL 2011).
Paul Whitney, Ph.D.
Reclamation Plan. I am not familiar with the Northern Dynasty Minerals mine plan. I wonder if
their mine plan includes a Reclamation Plan. If not, why not? If their mine plan includes a
Reclamation Plan, why isn't it presented as part of the Bristol Bay Assessment? The feasibility
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 66
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
of reclaiming the waste rock and tailings areas and possibility the pit (page 4-23, last para, last
sentence) seems important for evaluating the acceptability of the example mine. I am not aware
of any mine regulating agency that does not require a Reclamation Plan as part of a mine
application. I wonder if a Reclamation Plan that involved placing waste rock and tailings back in
the pit and reducing surface infiltration would greatly reduce the need for water treatment.
RESPONSE: The assessment is not a mine plan. The NDM document relied upon for
information in the assessment (Ghaffari et al 2011) is a preliminary mine plan and does
include conceptual reclamation measures which we have used in our scenarios. The State of
Alaska requires a Reclamation Plan of all mining facilities unless they are very small (AS
27.19.010). EPA \ revised scenarios utilize blending throughout the mine life to eliminate the
needfor long-term storage of PAG waste rock. Reclamation activities for the scenarios are
discussed more clearly in the revision and are activities that are considered feasible and
common at other similar existing mining sites. The scenarios in the assessment assume that
reclamation is properly completed, but concentrates the discussion on impacts that are
expected even with such activities.
Best Mining Practices. The assessment refers to the example mine plan as having both the "best"
mining practices (e.g., page ES-10, five lines from the bottom) and "not necessarily best" mining
practices (e.g., page 4-17, four lines from the top). Both of these statements can't be accurate.
RESPONSE: With regard to the terminology of "best", "good"', or other terms for the
practices used, what was intended to he conveyed is that nv have assumed modern mining
technology and practices. The terms are qualitative when generally interpreted, or have a
regulatory meaning (best management practices), and thus we have eliminated their use in the
revised assessment to avoid confusion.
Noise Levels. The mine plan should provide information on the location, frequency, and size of
blasting, sound level isoplelhs around the mine, and efforts to minimize sound levels as the mine
develops. 1 wonder if a majority of the sound levels will attenuate as mining activities move
deeper into the ground or if will there be a hundred years of blasting at the surface level. The
interviews with the villagers indicate that blasting and helicopter noise is a concern (Appendix
D, Cultural Characterization, page 94). A characterization of current noise levels in relation to
the area and timing of current and past wildlife use would help to determine if the whole or parts
of the watersheds are less than pristine.
RESPONSE: Noise from large-scale mines is outside of the scope of this assessment because it
is not known to affect salmonid populations. EPA agrees that a full evaluation of any future
mining permit applications and subsequent National Environmental Policy Act
Environmental Impact Statements would consider these direct effects on wildlife.
Water treatment during the winter. I wonder if it will be possible to treat water during the winter.
Will such treatment have to occur in a warm building? If so, what are the temperature
consequences of releasing warm treated water into streams?
RESPONSE: Although water quality standards for temperature are directed toward summer
maximum temperatures, we emphasize the importance of the year-round thermal regime to
which salmon are adapted Abnormally warm temperatures could accelerate egg maturation
and reduce survival of incubating salmon. We state that a protective approach would
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 67
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
discharge water that met baseline thermal regime conditions throughout the year, and
describe the risks associated with failure to meet this regime.
Yes, it is possible to treat water during the winter, and treatment facilities would be contained
in structures to prevent freezing of the treatment plant. While the Alaska water quality
standards (18 AAC 70) do contain maximums for water temperature, even the most stringent
of these (13°C) may be too high for winter-time discharge. Very likely, following a permit
application, ADF&G would determine an appropriate temperature for winter discharge and
the facility would have to meet this value, especially for discharges to anadromous streams.
Temperature can be easily lowered during the winter by exposure to ambient air.
Cone of Depression. I have worked on pit mines where hvdroaeologists model the lateral extent
of the cone of depression and have mapped the lateral extent as an area around the pit. The
lateral extent of the cone of depression, illustrated in Figure 4-9. appears to be underestimated
and has no effect on streams or wetlands. The figure has no scale. Is the lateral extent of the cone
of depression in Figure 4-9 based on modeling (see Box 4-2, para 3, last sentence)? If so, how
many NWI wetlands and meters of stream are in the area used for the model? If there are
wetlands or streams in the modeled area, how far down stream will the cone of depression
influence stream flow and wetland hydrology?
The information in Box 4-2 doesn't clearly (at least to me) deal with the proportions of run-on
and run-off water. If the diverted run-on water is supposed to mitigate the cone of depression,
will it be available for down stream resources'.'' Why won't diverted water seep back into the
near-by pit versus mitigating the cone of depression? The answer to these questions is on page 5-
72, but merely indicating there will be a reduction is not very informative.
RESPOXSIC: The cone of depression is now projected to dewaler streams and wetlands with
which it intersects. These losses are incorporated into the water balance calculations used to
estimate changes in downstream stream/low presented in Chapter 7. The lateral extent of the
cone of depression was estimated using the Dupiiit-Forcheimer discharge formula for steady-
state radial flow into a fully penetrating well in a phreatic aquifer with a diameter equal to the
average mine pit diameter. The radius of influence was determined by balancing the net
precipitation falling within the cone of depression with the calculatedflow into the mine pit
Section 6.2.2.1 and Box 6-2 in the revised draft (Section 4.3. 7 and Box 4-2 in the original draft
assessment) contain additional, details on the methodology.
Our estimate of the mine pit inflow agrees closely with the estimate provided in Ghaffari et al
(2011). The estimated cone of depression would extend about 1.2 km beyond the pit rim The
water within the cone of depression that would flow to the mine pit is included in the water
balance. The revised assessment presents estimates of the changes in streamflow at individual
downstream gages.
Figure 4-9 (now Figure 6-5) is merely a schematic of the flows and is not to be interpreted as
an exact representation of the expectedflow regime. The figure is not to scale and is not based
on modeling.
All of the precipitation failing within the cone of depression is considered to flow into the mine
pit These losses are incorporated into the water balance calculations used to estimate changes
in downstream streamflow presented in Chapter 7. Streams within the cone of depression
would dry up. The assessment assumes that water from streams upstream of the cone of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 68
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
depression would be diverted through pipes or channels to locations on streams downstream of
the cone of depression. The diverted water would not seep or flow into the mine pit
Run-on and run-off water terminology. I am used to referring to up gradient or adjacent water
that runs onto the pit or tailings facilities as run-on water and to water from the mine or storage
facilities as run-off water. The assessment doesn't always distinguish these two types of water.
For example, on page 4-13, line 6 refers to precipitation run-off water as up gradient water. On
page 4-26, the first bullet refers to run-off water as water running off mine facilities. The
terminology overlap makes it difficult (at least for me) to understand how the run-on and run-off
water will be captured and diverted around the mine facilities or used for other purposes. In
addition to calculations, diagrams of the diversions would be helpful. Will there be parallel
diversion ditches around the facilities, one for run-on and one for run-off water? Will one or both
of these ditches be lined? How will the water in these ditches be influenced by the cone of
depression? These questions are alluded to in the discussion on page 4-27(second para), but are
not explicitly addressed. I am sure engineers can and have answered these questions for other
mines with water balance analyses. It would be interesting to see an explicit summary of the
water balance for the various facilities. Such analyses would be good for the example mine plan
during operation and once the mine is no longer a net consumer of water (page 5-44, para 2).
Without the water balance analyses, potential impacts are not easily understood or quantifiable.
RESPONSE: A single diversion ditch intercepts lip-gradient surface water and routes it
around the areas disturbed by mining, preventing it from mixing with waters that have
encountered site materials. The revised assessment includes a more detailed water balance
(Chapter 6).
Some ideas for how to manage and separate run-on and run-off water might help determine
which streams might dry up and what type of mitigation measures (i.e., lining ditches) could
minimize the impact. In addition, if run-on water can be maintained in a diversion ditch, what is
the opportunity for developing a reclamation plan for the ditches? Such plans might be able to
minimize and partially compensate for lost reaches of headwater streams.
RESPONSE: The revised assessment does include some possible reclamation activities for
closure, although others could be proposed for an actual mining plan. Water would continue
to be diverted around areas where the water could encounter contaminants. Mitigation to
compensate for lost streams would be addressed through a regulatory process that is beyond
the scope of this assessment. Severtheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment
PHWResponse: See my response to Mitigation Measures on page 12.,
Protective approach. A "protective approach" is mentioned on page 5-30 (para 3, last sentence).
This has something to do with water management and would be good to explain.
RESPONSE: This has been clarified in Chapter 7.
Formatted: Font: Not Bold, Not Italic, Font
color: Red
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 69
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Question 3. EPA assumed two potential modes for mining operations: a no-failure
mode of operation and a mode involving one or more types offailures. Is the no-failure
mode of operation adequately described? Are engineering and mitigation practices
sufficiently detailed, reasonable, and consistent? Are significant literature, reports, or
data not referenced that would be useful to refine these scenarios, and if so what are
they?
David A. Atkins, M.S.
The no-failure scenario attempts to quantify the impacts from developing the footprint of the
project alone. In reality, various failures and accidents inevitably occur, and they may have a
range of impacts from inconsequential to large. So this scenario is presented to describe the
minimum impact that could be expected from project development assuming everything works
as planned.
RESPONSE: This comment is a correct interpretation of our "no-failure" scenario. Because
this distinction was not clear to many other readers, the revision no longer uses the term "no
failure". The no-failure scenario from the draft assessment has been changed to a chapter on
the effects of the footprint of a mining operation, without regard far operational problems
(Chapter 7).
The mine will, by necessity, remove those streams and wetlands that are beneath the pit, waste
rock, tailings and processing plant development areas. There should be some flexibility in siting
facilities other than the pit or underground workings. For the 'no-failure' scenario, the
Assessment presents lengths of stream and areas of wetlands that would be lost due to physical
displacement of the aquatic resources from mine development and reduction in flows from mine
water management. The assessment presents the following resources that would be lost and that
have been shown to be spawning or rearing habitat for coho, Chinook, and sockeye salmon, or
have resident populations of rainbow trout and Dolly Varden:
25-year scenario
78-year scenario
Eliminated or blocked streams (km)
87.5
141.4
Reduced flow (>20%; km)
2
10
Eliminated wetlands (km2)
10.2
17.3
Given the range of uncertainty with the proposed mine plan, presenting stream lengths and
wetland areas to the tenth place implies unrealistic accuracy. Significant figures should be
checked and consistent throughout the document, and ranges should be presented if known (e.g.,
results for the pits could be presented with more accuracy since we know where they will be,
whereas other facilities that could be located in different areas should be presented with an
appropriate range of uncertainty).
RESPONSE: The authors have carefully addressed this issue. Numbers from the literature or
from the PLP EBD retain the number of significant figures in the original. Numbers derived
for this assessment have the appropriate number of significant figures given the precision of
the input data and uncertainties due to modeling and extrapolation.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 70
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The impacts as presented appear substantial, mainly because of the very large nature of the
project. However, it would be helpful to describe the significance of this loss, specifically with
regard to the following questions:
• What impact would the loss to streams and wetlands have on the fishery within the
Nushagak and Kvichak basins?
RESPONSE: Impacts of habitat loss and alteration are very difficult to quantify given the
lack of information on limiting factors, production and capacity estimates. We were unable
to comprehensively evaluate impacts at the population level, except for the most severe
cases where total losses of runs could be reasonably assumed.
• Is this loss significant in comparison to the fishery as a whole'?
RESPONSE: Losses of streams and wetlands under the mine footprint could not be related
to the fishery due to reasons listed above. For the TSF failure scenario that completely
eliminates or blocks access to suitable habitat in the North Fork Koktuli River, we estimate
that the entire Koktuli portion of the run (-28% of Nushagak escapement) could be lost
Higher proportional losses would occur if significant downstream effects occurred due to
transport of toxic tailings fines beyond the Koktuli as modeled under the Pebble 2.0 TSF
failure.
• Are there local communities that could be affected by this specific loss?
RESPONSE: Wildlife and resident fish communities would be affected by reductions in
spawning salmon. Local communities would also be aff ected by the reduction, which is
now discussed in Chapters 12 and 13 (e.g.. Chapter 13 now contains tables which refer to
specific subsistence resources used by individual communities).
• Is fragmentation of the resource from this loss a significant impact (i.e., are there stocks
that are unique to the project area)?
RESPONSE: Stock structure and genetic diversity are not well known at the project scale,
but based on evidence from oilier parts of Bristol Bay watersheds, local adaptation is
highly likely. Discussion of fragmentation effects is now included in Chapters 7 and 10.
There is no discussion of enaineering and mitigation practices in this section. The responsible
regulatory authority would require the project proponent to present a mitigation plan to
compensate for these impacts before permitting. Measures would include minimization of impact
through facility siting, reclamation if possible, and compensation if reclamation were not
feasible. A thorough analysis of possible mitigation approaches and the likelihood of their
success are necessary to fully evaluate impacts from the 'no-failure' scenario.
RESPONSE: Mitigation measures for design and operation are more clearly called out in the
revised assessment. While measures chosen here may differ from what is required during the
regulatory process, the assessment is not a mining plan and not an evaluation of a mining
plan The assessment assumes that measures chosen for the scenarios would be as effective as
possible and examines only accidental failures rather than a failure to choose a proper
mitigation measure. Mitigation to compensate for effects on aquatic resources that cannot be
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 71
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
avoided or minimized by mine design and operation would be addressed through a regulatory
process that is beyond the scope of this assessment. Nevertheless, in response to public and
peer comments we have included a discussion of compensatory mitigation in Appendix J of
the revised assessment.
Steve Buckley, M.S., CPG
The engineering and mitigation designs associated with the no-failure mode of operation are
inadequate. There is no detailed discussion of engineering practices. There is insufficient
discussion of any potential mitigation measures and there is a lack of any detailed research into
applicable engineering and mitigation methods. Appendix I provides some engineering and
mitigation practices along with water quality mitigation and monitoring during closure; however,
these are not discussed or accounted for in the main assessment document.
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included as part of the mine scenarios in the draft
assessment, and this discussion has been expanded in the revised assessment. The mitigation
measures proposed within the mine scenarios are those that could reasonably be expected to
be proposedfor a real mine (they are a subset of options presented in Appendix I), all of which
were presented as appropriate for the I'ehhle deposit in (ihaffari et ah (2011). Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation would be addressed through a regulatory process that is beyond the
scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Xppendix J of the revised assessment
Courtney ( a rot hers. Ph./).
The no-lailure mode of operation appears to be described adequately. The engineering and
mitigation practices appear to be sufficiently detailed, reasonable, and consistent, although I have
no particular expertise with which to evaluate this part of the assessment.
RESPONSE: No change suggested or required.
It would be helpful to have a clear statement about how well the local (geotechnical, hydrologic,
and environmental) conditions in this region have been studied and characterized. How much is
understood about the seasonal variation in these conditions and how those variations would
affect these scenarios? How well are statistics from mines and TSFs constructed in very different
environments likely to apply here?
RESPONSE: TSFfailure probabilities are based on Alaska's dam classification and required
safety factors applied to the method of Silva et aL (2008). The discussion offailure
probabilities is expanded to clarify this issue. Therefore, the failure is not a consequence of
any specific site conditions or seasonal phenomena. However, the discussion of local and
regional conditions has been expanded in the revised assessment It is recognized that some
issues such as hydrology are very complex and additional information will be useful in future
analysis of any mine plans.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 72
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Dennis D. Dauble, Ph.D.
The description of the no-failure mode for mine operation appears adequate in terms of potential
mitigation measures that might be employed. I have limited knowledge of current engineering
practices and subsequent risks to the environment from best practices of modern mines,
including those operating under optimal conditions. However, it would be helpful to include a
short discussion on which mitigation measures would be most applicable to mining activities in
the Bristol Bay watershed.
RESPONSE: Standard design mitigation measures considered feasible, appropriate, and
'permittable' (as per Ghaffari et aL 2011) were considered and are discussed in Chapter 6 and
Appendix I of the revised assessment; these are standard measures common to other copper
porphyry mines. Evaluation of measures that would be proposed for an actual mine would
occur through the regulatory process. Whether these same measures would be appropriate for
all locations within the Bristol Bay watershed would depend on the given site s specific
characteristics.
Gordon H. Reeves, Ph.D.
No comments on this question.
Charles Wesley Slaughter, Ph. I).
Based on the actual history of other major resource extraction projects in Alaska and throughout
the world, a "no failure" assumption seems unrealistic. Rather, the assumption should be that
there will be failures, of varying modes and magnitudes, over the life of the project. This reality
is recognized in several sections of text.
RESPONSE: Our intention for the "no-jailure " scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed The "no
failure" chapter has been eliminated The revised assessment differentiates between potential
effects from the footprint of a mine (Chapter 7), water treatment (Chapter 8), TSFfailures
(Chapter 9), the transportation corridor (Chapter 10), and pipeline failures (Chapter 11).
In some sections in the Assessment, presumed "mitigation practices" are either cursory,
optimistic, or so general as to be un-supported. Examples include Section 4.3.7's cursory,
generalized statements about handling water: "Uncontrolled runoff would be eliminated.. .The
mine operator would capture and collect surface runoff and either direct it to a storage
location... or reuse or release it after testing and any necessary treatment"; "... water from these
upstream reaches would be diverted around and downstream of the mine where practicable";
"precipitation would be collected and stored..."; and "Assuming no water collection and
treatment failures, this excess captured water would be treated to meet existing water quality
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 73
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
standards and discharged to nearby streams, partially mitigating flow lost from eliminated or
blocked upstream reaches." Other examples from Chapter 6: .assuming no water collection
and treatment failures" and "excess captured water would be treated... and discharged to nearby
streams..."
RESPONSE: Water management (mitigation) measures are more clearly described and
discussed in the revised Section 6.1.2.5, and in sub-sections for the mine components in the
scenarios. However, the intent of the assessment is not to specify technologies, beyond those
already identified by the existing preliminary mining plans. Rather, the assessment focuses on
the environmental outcomes of conventional modern mining practices and effluents.
John D. Stednick, Ph.D.
The no-failure mode is not adequately described. Assessment of the effects of the mine is based
on large risk failures of low probability and did not include low risk failures of higher
probability. The report concludes (and emphasizes) that the mine footprint will disrupt/disturb
contributing watershed and wetland areas and result in hydrologic modification. The hydrologic
modification affects salmonid habitats, particularly in low flow conditions. Regulatory oversight
will include the State of Alaska, and certainly mitigation measures would be required. The task
is to address the adequacy of these mitigation measures.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed The revised
assessment no longer uses the term "no-failure", and does address effects from scenarios
having higher probability and lesser magnitude (e.g., f ailure to collect or treat leachate water)
in addition to those with lower probability and higher magnitude (e.g., TSFfailure).
Mitigation to compensate for effects on aquatic resources that cannot be avoided or minimized
by mine design and operation would be addressed through a regulatory process that is beyond
the scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment
Pollutant/toxicity assessment focused on copper. Other metals can be presented to show the
range of metal concentrations for chronic and acute toxicity. Suitability of treatment processes
for all wastewaters can be included to address potential effects on receiving waters.
RESPONSE: Although copper is emphasized due to its dominance of the toxicity ofleachates,
other metals were and are presented (Chapter 8).
The discussion of roads is mostly related to fish blockage and some soil erosion. Information on
current design standards was not included and tended to relay on dated references from logging
roads
RESPONSE: The discussion of roads covers risks related to filling and alteration of wetlands,
stream crossings, fine sediments, dust deposition, runoff contaminants, and invasive species.
Information on current design standards is now included within text boxes throughout
Chapter 10, and relies on recent literature. The failure frequencies cited in the assessment are
from modern roads and not restricted to forest roads. One of the papers usedfor general
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 74
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
information (Furniss et ah 1991) focuses on forest and rangeland roads, but it is a seminal
publication on the potential effects of roads, particularly as they relate to salmon.
There were no engineering or mitigation practices described in this section or in the document.
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included in discussion of the mine scenarios in the
draft assessment as part of the design. This discussion has been expanded in the revised
assessment.
Roy A. Stein, Ph.D.
No Failure Operations and Their Impact. What about the failure of continued monitoring, of
continual inspection, of continual, rigorous oversight? This is more insidious than a catastrophic
failure of some sort, but perhaps just as dangerous (in fact, one research geochemist testified
during public testimony that of 150 hard-rock mines, none operated without leakage of leachate).
How can we be sure that mine operators will be held strictly accountable for their actions with
regard to best mining practices (a point emphasized by those who testified in favor of the Pebble
Mine that indeed best management practices would be used), meeting all the various and sundry
regulations, and communicating all of these activities back to the regulatory organization? Will
there be a force of will on the part of EPA or other regulatory body to be sure that all activities of
the operator are appropriate and within regulatory limits? The down-side of poor monitoring and
lack of rigorous oversight is the loss of salmonid populations. These losses are, in my view, less
important than compromising human health and life. Yet, at the Upper Big Branch Mine in West
Virginia, dust standards have been exceeded for years, leading to a dust explosion that killed 29
miners on April 5, 2010. In turn, even surviving miners were not immune to these dust impacts,
for they suffer from "black lung", a condition that literally shortens their life by decades. In turn,
much of the monitoring of these conditions has historically been the responsibility of the owner
corporation, rather than an independent regulatory body, much like "the fox guarding the
chickens". Here at the Pebble Mine site, where only fish (but, of course, Native Alaskan
subsistence users, plus other human users as well) are at stake, would one expect rigorous
oversight by appropriate regulatory bodies? Skepticism leads to cynicism when contemplating
the Upper Big Branch Mine case history in the context of the Pebble Mine proposal.
RESPONSE: Our intention for the "no-failure " scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed The "no
failure" chapter has been eliminated The revised assessment differentiates between potential
effects from the footprint of a mine (Chapter 7) water treatment (Chapter 8), TSFfailures
(Chapter 9), the transportation corridor (Chapter 10), and pipeline failures (Chapter 11).
Holding the mining company accountable is done through the regulatory process and is
outside scope of this assessment.
Engineering Practices and Mitigation. I did not think that mitigation was well described in
text, but Appendix I is quite well developed and was instructive to me as I moved through the
documents. I would suggest including the ideas in Appendix I in the mitigation section of the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 75
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
main report. Other comments on mitigation issues can be found below associated with Question
12.
RESPONSE: Mitigation measures for design and operation are included in the assessment
(Chapter 6), and are those that reasonably could be expected to be proposed for a real copper
porphyry mine (they are a subset of options presented in Appendix I). Many, if not all, of these
measures were presented as appropriate for the Pebble deposit in Ghaffari et aL (2011).
Mitigation to compensate for effects on aquatic resources that cannot be avoided or minimized
by mine design and operation would be addressed through a regulatory process that is beyond
the scope of this assessment Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in \ppeudix J of the revised assessment
Mitigation measures have been explained more clearly throughout the revised assessment.
William A. Stubblefield, Ph.D.
It is interesting and appropriate that the EPA has included both modes o I" operation in conducting
this assessment. This approach provides some degree of "bounding" for the assessment;
however, the degree of accuracy (i.e., predictability) for either scenario cannot be known at this
time. The document appropriately acknowledges that there are a variety of potential mitigating
factors (e.g., acts of God, accidents, market changes) that may render the assumptions used in
this assessment incorrect.
RESPONSE: Our intention for the "no-failure " scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed There were a
number of comments on this approach so the revised assessment no longer uses the term "no-
failure " and addresses effects from the foot print (i. e., the no failure scenario), failure
scenarios having higher probability and lesser magnitude (e.g., f ailure to collect or treat
leachate water), and failure scenarios having lower probability and higher magnitude (e.g.,
TSFfailure).
No change suggested or required.
Dirk van Zvl, Ph.D.. P.E.
The no-failure mode of operation failures is based on surface disturbances and potential
blockages caused by the various facilities. For example, for the mine pit, TSF and waste rock
facility, the surface areas of these facilities are used as a basis for calculating the streams and
wetlands affected by the mining activities. While the failure mode is adequately described,
engineering and mitigation practices are not adequately described by EPA.
RESPONSE: Mitigation measures associated with the mine components (e.g., waste rock pile,
TSF, etc.) were discussed in the sections presenting those components. They were not repeated
in the section on failures. These measures are included again in the revised assessment
(Chapters 4 and 6), but are described in greater detail However, the emphasis of the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page lb
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
assessment is on the consequences offailures rather than on the details of how inadequate
mitigation may cause the failures.
The EPA Assessment states on p. 8-1 "Routine operations are defined as mine operations
conducted according to conventional practices, including common mitigation measures, and that
meet applicable criteria and standards". The adverse effects listed are: direct impacts as a result
of removal of streams in footprint of mine pit and waste storage areas; reduced streamflow
resulting from water retention; removal of wetlands in the footprint of the mine; indirect impacts
of stream and wetland removal; diminished habitat quality in streams below road crossings; and
inhibition of salmonid movement from culverts that may block or diminish use of full stream
length.
RESPONSE: No change suggested or required.
Any mine in Bristol Bay will have to undergo a rigorous and lengthy regulatory review and
permitting process. I do not know of a process that will exclude consideration of the impact of all
mine facilities on the streams and wetlands in the region. Therefore, I would suggest that the full
implications of "mine operations conducted according to conventional practices, including
common mitigation measures, and that meet applicable criteria and standard" should have been
addressed in the report. The EPA (2003) document 011 Generic Ecological Assessment Endpoints
for Ecological Risk Assessment specifically details the applicability of Section 404 of the CWA
in addressing community and ecosystem-level endpoints. "The CWA provides authority for the
Corps to require permit application to avoid and minimize wetlands impacts and requires EPA to
develop, in coordination with the Corps, the criteria used for Section 404 decisions. When
damages to wetlands are unavoidable, the Corps can require permitees to provide compensatory
mitigation". It is unclear why this was not included in the evaluations.
RESPONSE: This is not a permitting document The purpose of the assessment is to evaluate
the effects of the operation of a mine on salmon ecosystems in the Bristol Bay watershed,
while following conventional practices, including common mitigation measures. Once those
effects are described, then it is appropriate to determine 1) if unacceptable environmental
effects are likely to occur and 2) whether those effects can be offset (made acceptable from a
regulatory standpoint) with compensatory mitigation. In other words, compensatory mitigation
is a next step and not within the scope of this assessment. Nevertheless, we have included a
discussion of compensatory mitigation in Appendix J in the revised assessment
Similarly, one would expect thai the regulatory reviews will require that the impacts resulting
from loss of streams, streamflow and road crossings will be addressed through engineering
designs, proposed mitigation measures, as well as regulatory and community engagement best
mining practices (see discussion above on "good" vs. "best" practices).
RESPONSE: Our scenarios included mitigation measures through engineering design and
operations to reflect standard industry practices. The purpose of the assessment was to
evaluate risks in the presence of these measures. The conunenter is correct that alternatives
for such measures would be evaluated during the permitting (regulatory) process. No change
suggested or required
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 77
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
On p. 4-33, it is stated that "Environmental impacts associated with premature closure may be
more significant than those associated with planned closure, as mine facilities may not be at the
end condition anticipated in the closure plan and there may be uncertainty about future reopening
of the mine". Further text describes potential negative impacts from such a premature closure.
One of the outcomes of the regulatory review and permitting will be the establishment of
financial assurance that will provide State and Federal Regulatory Agencies with the financial
resources to accommodate a closure. These obligations are typically reviewed on a 3 or 5-year
interval to also make sure that they are adequate to cover premature closures. If the mining
company is still managing the site, then they will have responsibilities under all Federal and
State Regulations and the dire picture painted by the EPA Assessment should not come to pass.
RESPONSE: The revision includes language that addresses financial assurance (Box 4-3).
Because of this major oversight of the realities when permitting and operating a mine it is
essential that the scenarios be reviewed by evaluating the effects that regulatory requirements
and resulting mitigation methods would have on the no-failure conditions before completely
reworking the no-failure mode of operations and their impacts. Other significant reports and data
that should be reviewed include typical permitting documents and resulting requirements for
similar mines in the US and Canada to obtain a range of potential outcomes. The results from
such an evaluation will also contribute significantly to the discussions in Alaska when the Pebble
Mine and other mines in Bristol Bay are brought forward to permitting.
RESPONSE: Our intention for the "no-failure" scenario was to identify and evaluate the
unavoidable environmental effects if all systems and mitigation measures operated perfectly,
and to separate those effects from a scenario where systems periodically failed The revised
assessment no longer uses the term "no-failure but simply presents effects from scenarios
having higher probability and lesser magnitude (e.g., f ailure to collect or treat leachate water)
as well as those having lower probability and higher magnitude (e.g., TSFfailure).
Phyllis K. Weber Scannell, Ph.D.
Chapter 4 provides a detailed description of a hypothetical mine design for a porphyry copper
deposit in the Bristol Bay watershed. Some of the assumptions appear to be somewhat
inconsistent with mines in Alaska. In particular, the descriptions of effects on stream flows from
dewatering and water use do not account for recycling process water, bypassing clean water
around the project, or treating and discharging collected water.
RESPONSE: The issues mentioned were discussed in Sections 4.2.3 (tailings storage) and
4.3.7 (water management) in the original draft document They are now addressed in Chapter
6 of the revised document, which describes the mine scenarios, and in Chapter 4, which
provides generic background on porphyry copper deposits and mining. Streamflow effects
presented in Chapter 7 now reflect a complete water balance, including water capture and re-
use, bypass, and discharge from the wastewater treatment facility, as suggested by the
commenter.
Section 4.3.8, Post-closure Site Management, raises critically important issues - can a mine in
this area be designed for closure? Is it acceptable to develop and operate a mine that will require
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 78
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
essentially perpetual treatment? It is my belief that these are the essential questions that should
be addressed during any mine permitting process.
RESPONSE: EPA agrees that these are important questions to be addressed, but they are risk
management, not risk assessment, questions. The purpose of the assessment is to evaluate
risks to the salmon fishery from large-scale mining. Risk management decisions will be made
during the permitting process. Thus, no changes to the assessment were made in response to
this comment.
Section 4.3.8.1 raises concerns about long term water quality and quantity from the mine pit.
These concerns need to be addressed during a mine permitting process. Pit water quality
depends on how the pit is developed, what reclamation will occur, if reclamation will be
concurrent with mining, and what kinds of water treatment will be used. Tailings storage facility
(TSF) water quality depends on how the mine tailings are managed: it may be possible to use dry
stack tailings with sulfide removal rather than submerged tailings.
RESPONSE: EPA agrees that water quality can be influenced by design and reclamation, but
when the latter entails creating a pit lake there is little flexibility for reclamation concurrent
with mining. How tailings are managed within the impoundment can aff ect water chemistry,
and a dry stack with sulfides removed may produce the best water quality results after
reclamation if the fate of the sulfide tailings is never considered. According to Ghaffari et a I.
(2011), 14% of the tailings produced will be pyritic, which equates to an average of28,000 tpd
in a 200,000 tpd mining operation (over 255 million tons during a 25 year mine life). These
tailings need to be managed in such a manner that oxidation does not lead to acidic drainage,
so the most eff ective way is to deposit them subaqueously. Some suggested common mitigation
measures for management of the pit at and post closure are included in the revised assessment
(Chapter 6).
Paul Whitney, Ph.D.
Mitigation Plan. Most mine permit applications I have worked on include both mitigation to
minimize environmental impact and mitigation to compensate for environmental impact. The
assessment outlines a variety of mitigation measures to minimize impact, but no compensatory
mitigation. This is a concern, for I wonder if compensatory mitigation for the example mine is
even possible in the watersheds.
RESPONSE: The purpose of the assessment is to evaluate the effects of the footprint and
operation of a mine that follows conventional practices, including common mitigation
measures. Once those effects are described, then it is appropriate to determine 1) if
unacceptable environmental effects are likely to occur and 2) whether those effects can be
offset (made acceptable) with compensatory mitigation. Determining compensatory mitigation
is a next step and outside the scope of this assessment; however, we have included a discussion
of compensatory mitigation in Appendix J in the revised assessment.
The watersheds are characterized with descriptors such as "pristine" (e.g., page 6-29, last para,
second line), "nearly pristine" (e.g., pages 2-25 and 7-2) and "exceptional quality" (page 2-20). It
is also stated that the return of the salmon "fuel" (i.e., provide energy to) the terrestrial food web.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 79
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
If in fact the watersheds are pristine or nearly pristine, the habitat is high quality and there is
little, if any, opportunity for compensatory mitigation (i.e., improving low quality habitat) in the
terrestrial and fresh water environments. For example, if 55 miles of streams and streamside
wetlands are lost to the mine footprint (page ES 15, first bullet), is it possible to find miles of
very degraded stream to plan for and implement compensatory mitigation? If one assumes a
mitigation ratio of 3:1 for enhancement, one might have to find 165 miles of degraded stream for
compensation. I suspect (but don't know) that there are very few (if any) miles of degraded
stream where compensatory mitigation could occur in the Bristol Bay watershed(s). If this is the
case, it might not be possible to demonstrate no net loss for waters of the US, and this is
something EPA should be interested in.
RESPONSE: The comment is correct in stating that the exceptional, quality of the Bristol Bay
environment leaves little opportunity for compensatory mitigation. Determining compensatory
mitigation is outside the scope of this assessment; however, »iv have included a discussion of
compensatory mitigation in Appendix J in the revised assessment.
I agree that the ecological resources can be ranked as having high quality because the human
footprint on the habitat is small (i.e., few roads and villages), but from an energetics (i.e., fuel)
and food web perspective, the pristine characterization may not be accurate. The commercial
catch of approximately 27.5 million fish each year (up to 70% of the total number of sockeye
produced) is a lot of calories that are not flowing through the ecological foodwebs of the
watersheds. Granted, some of the commercial catch (if not caught) might not enter the
watersheds, but some and perhaps a lot would, especially in good run years. While the harvest
level might be sustainable, the loss of energy to commercial fishing causes pause to characterize
the watersheds as pristine or nearly pristine. The potential impact of fisheries on energy flow has
been addressed by Pauly et al. (2000) and Libralato et al. (2008). I wonder if it is technically
possible that a reduction in the commercial fishery is a compensatory mitigation measure.
RESPOXSIC: Compensatory mitigation requirements address the needfor project proponents
to replace aquatic resources and ecosystem functions that their project has impacted Reduced
fishing harvests would not replace lost spawning and rearing habitat. Further, it would
remove the burden of compensation from the party that caused the damage. Determining
compensatory mitigation is outside the scope of this assessment; however, we have included a
discussion of compensatory mitigation in Appendix J in the revised assessment
PHWResvonse: Pointing out that reducing commercial harvest would not replace lost spawning
and rearing habitat is very good and I appreciate the insight. None-the-less EPA seems to have a
hard time acknowledging much less discussing the negative impact commercial fishing has on
ecosystem energetics and nutrient cycling. Acknowledging these past current and future impacts
would greatly improve the Cumulative Impact part of the EPA assessment. —( Formatted: Font: Not Bold, Not italic
Effluent treatment. Water quality information in the assessment for benchmarks, background,
and leachate is extensive. A thorough review of the water quality and toxicity information is
beyond the scope of work of this review. After several reads of this information, it appears that
the work is good for copper. For example, work on salmonid olfaction and copper conducted by
McCarthy et al. (2007) is potentially important and is cited. The inhibiting effects of copper on
olfactory receptor neurons cited by McCarthy et al. (2007) at or above 2 (ig/L are lower than the
Alaska hardness-based standards and the biotic ligand model (BLM) standard in Table 5-14, but
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 80
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
are above the biotic ligand model standard in Tables 5-15 and 5-16.1 assume this is due to
differences in binding of copper by dissolved organics but I am not sure. Whether one decides to
use the 2 (ig/L benchmark, or the even lower BLM benchmarks that are in some cases below
background values in Table 5-19,1 think the key question is whether proposed leachate
processing can cost-effectively achieve benchmarks that hover around background
concentrations. The answer is beyond my level of expertise.
RESPONSE: BLM-derived copper criteria are derivedfor the different leachates. The values
depend on the co-occurring ions, which differ considerably among leachates and, in the case
of ambient waters, on dissolved organic matter. The low copper benchmarks would be
achievable with treatment by reverse osmosis, which lias been used at other mines. Whether it
is economically feasible to achieve benchmarks close to background concentrations is outside
the scope of the assessment No change suggested or required.
I do not agree with the assessment's critical question - whether or not effects are observed at
these low levels (page 5-57, Exposure-Response Data from Analogous Sites, second sentence). If
effects are observed at background concentrations, it seems unreasonable to ask for an even
lower benchmark than background concentrations. The uncertainties assessment at the bottom of
page 5-57 also seems unreasonable. The possibility that background concentrations are not
protective in particular cases seems highly unlikely for one of the most productive salmon
communities in the world.
RESPONSE: The passages cited by the reviewer refer to the possibility of effects at copper
concentrations below criteria, not below background. Similarly, the studies mentionedfound
effects below criteria levels but above background concentrations.
nTTII, n t,i • ¦ ^ t 1 t- 1 i i r» 1 11 /\ Formatted: Font: Not Bold. Not Italic
PHWResponse: This isn t the way I read Table 5-19 - please recheck ^
I can think of many questions that are more critical than looking for effects on salmonids at
background or near background concentrations of copper. For example, it might be more
important to ask what concentrations of copper will result in a significant impact on the salmonid
populations and to ask what impact a mixing zone would have on salmonid populations. Last but
not least, what are the potential impacts of all toxics on the many other non-salmonid species?
RESPONSE: See response to previous comment regarding background concentrations versus
water quality criteria. Because water quality is so high at the site, the threshold for copper
toxicity is low based on the biotic ligand model Effects on salmonid populations from
exceeding toxic thresholds cannot be estimated became the available monitoring data do not
characterize salmonid demographics or productivity in the streams draining the site. However,
the analysis has been expanded to include estimates of kilometers of stream habitat that would
be exposed to copper levels sufficient to cause aversion, sensory deprivation, decreased
reproduction, or kills. Mixing zones are not allowed by the State of Alaska for water quality
compliance in anadromous streams, and the available stream data are not sufficient for
mixing zone modeling. However, a discussion of mixing zones, including the amount of
mixing that would be required to reach nontoxic levels, has been added Non-salmonidfish
are not included as endpoint species. However, because copper and most other metals are most
toxic to arthropods, the assessment implicitly addresses non-salmonidfish (which depend on
arthropods in the food web) as welL
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 81
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Question 4. Are the potential risks to salmonid fish due to habitat loss and
modification and changes in hydrology and water quality appropriately characterized
and describedfor the no-failure mode of operation? Does the assessment appropriately
describe the scale and extent of risks to salmonidfish due to operation of a
transportation corridor under the no-failure mode of operation?
David A. Atkins, M.S.
For the no-failure mode of mine operation, the risks to salmonid fish due to habitat loss and
modification in the vicinity of the project are described in terms of loss of lengths of stream or
areas of wetlands. Project proponents state that the mine will only impact a very small fraction of
the watershed (under a no-failure scenario). It is important to establish whether the modeled
impact (e.g., the loss of 87.5 km of streams) is significant, both in terms of the absolute impact,
as well as the effect on ecosystem fragmentation.
RESPONSE: Footprint effects on habitat loss are now characterized in relation to the
distribution of habitat conditions throughout the larger watersheds. Fragmentation effects are
not anticipated at the mine she, apart from blockage of headwater streams as described, but
are anticipated in the case of the transportation corridor (Chapter 10) and TSE failure
(Chapter 9).
In addition, project proponents often state they will preserve and even improve the fishery. As
mentioned in the answer to the previous question, it would be helpful to know what kinds of
mitigation efforts could be employed - minimization, reclamation and compensation - and have
some assessment of the potential effectiveness.
RESPONSE: Mitigation measures, including wastewater treatment and closure and post-
closure monitoring and maintenance, were included in discussion of the mine scenarios in the
draft assessment, and this discussion has been expanded in the revised assessment. The
mitigation measures proposed within the mine scenarios are those that could reasonably be
expected to be proposed for a real mine (they are a subset of options presented in Appendix I),
all of which were presented as appropriate for the Pebble deposit in Ghaffari et aL (2011).
Reclamation is not mitigation, but the revised assessment includes also some suggested
measures to be used in closure/post-closure to reclaim the disturbed areas. Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation would be addressed through a regulatory process that is beyond the
scope of this assessment. Nevertheless, in response to public and peer comments we have
included a discussion of compensatory mitigation in Appendix J of the revised assessment
The Assessment determines that construction of the transportation corridor could alter the
habitat, chemistry, and the migration path across the corridor for the over 30 streams that the
corridor will cross or come near. The report further states that the corridor could affect 270 km of
streams below the corridor and 240 km of streams above, but that there is no way to assess the
magnitude. Therefore, the impacts of the corridor on fish populations are unknown, and this
impact is not described in a way that can allow a reviewer to draw any conclusion.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 82
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The revised assessment states that "the exact magnitudes of changes in fish
productivity, abundance and diversity cannot be estimated at this time," but summarizes the
species, abundances, and distributions that would potentially be affected Also, the assessment
concludes that, assuming typical maintenance practices after mine operations, approximately
IS of 32 culverted streams with restricted upstream habitat would be entirely or in part blocked
at any time. "As a result, salmonidpassage—and ultimately production—would be reduced in
these streams, and they would likely not be able to support long-term populations of resident
species such as rainbow trout or Dolly Varden."
Further, the references for road design and construction practices seem to be more representative
of forest and rangeland roads than the type of road that would likely be constructed for this type
of project. It would be helpful to cite experience from other transportation corridors constructed
for mining and oil and gas projects and developed recently in Alaska.
RESPONSE: Because the proposed mining would take place in an undeveloped area, the
literature is necessarily from areas outside of Bristol. Hay. Further, we found no data
concerning the performance of culverts for mining or oil and gas projects in the region.
However, to the extent possible we used examples from representative environments. The
failure frequencies cited in the assessment are not restricted to forest roads. One of the papers
usedfor general information (l'urniss et aL 1991) focuses on forest and rangeland roads, but
it is a seminal publication on the potential effects oj roads, particularly as they relate to
salmon. The general conclusions of that paper should be applicable to the transportation
corridor considered in the assessment. Information on current design standards is now
included within throughout Chapter II). and relies on recent literature.
Steve Buckley, M.S., CPti
Risks to fish due to habitat loss and modification and changes in hydrology and water quality are
overly simplified given the broad parameters used to model these potential risks. More specific
details on the water balance would help define potential risks to fish from dewatering and habitat
loss, l-'or example, there is no attempt to identify groundwater flow paths or the specific response
of vanous landforms to seasonal changes in precipitation and runoff, yet 34 pages are dedicated
to an attempt to quantify these impacts. More detailed information is needed to accurately
quantify the changes in anticipated runoff and infiltration in the proposed area to determine
potential impacts to hydrology and water quality.
RESPONSE: A revised and much more detailed water balance and streamflow analysis is now
incorporated in the assessment (Chapters 6 and 7). The revised draft also includes estimates of
the specific changes in flow at individual gages. The commenter is correct that the assessment
does not specifically address the potential changes in landforms or, for that matter, to
vegetation in the land areas such as the drawdown zone which would be among the areas most
affected by the mine development We consider that any such changes would have only a
secondary impact on salmon relative to the impacts that the assessment does address.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 83
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Additional ecological information on the contributing watershed area for each fish bearing
stream crossing would help identify the potential impacts to fish due to the construction and
operation of a transportation corridor.
RESPONSE: Additional information on watershed attributes (discharge, channel gradient,
floodplain potential) of streams crossed by the transportation corridor, and their importance to
salmonids, is now included in the analysis presented in Chapter 10.
Courtney Carothers, Ph.D.
Six key direct and indirect mechanisms are identified to pose potential risk to salmonid fish
species: eliminated or blocked streams (87.5-141.4 km), reduced stream flow, removal of
wetlands (10.2-17.3 km), indirect effects of stream and wetlands removal (downstream effects
likely diminishing fish production), diminished habitat quality downstream of road crossings,
and blocked movement of salmonids at road crossings. These mechanisms are described clearly.
The report appears to appropriately describe the scale and extent of risks under a no-failure mode
of operation, although I have no particular expertise with which to evaluate this assessment.
RESPONSE: No change suggested or required.
Dennis /). Dauble, Ph.D.
The assessment describes the number of stream miles impacted under each mode of operation,
including miles blocked and eliminated. Less specific were descriptions of impacts due to
sedimentation and leachates. What is lacking is quantitative estimates of spawning and rearing
habitat that would be lost relative to the total habitat available. Having this information would
help provide perspective of overall risk to individual watersheds and the Bristol Bay watershed
as a whole. Risks to salmonid fish due to changes in water quality (i.e., toxic materials) need to
consider differences in sensitivity and behavioral response according to salmonid life stage.
RESPONSE: Stream habitat losses are now characterized in relation to the distribution of
habitat conditions throughout the larger watersheds (Chapters 3 and 7). The assessment of
risks from aqueous toxicity distinguishes overt toxic effects on early life stages from
behavioral eff ects on adults.
Surface water characteristics of site watersheds within the area of probable impact are detailed in
Table 5-17, but not so for other streams and lakes in the broader watershed. More information
should be presented where available. It is not clear whether potentially affected streams and
lakes might be nutrient limited (seems that they might be given their dependence on MDN). For
example, include N or P concentrations and some discussion about primary and secondary
productivity.
RESPONSE: We recognize that nutrient status, and more importantly prey availability, is a
critical component of habitat capacity for fish in these systems, and may be strongly driven by
salmon derived nutrients. We concur that more information is needed regarding potential
limiting factors for salmon productivity and capacity, and that food availability may be one
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 84
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
such factor. However, because water chemistry data may not provide a complete picture of
trophic status, particularly where direct consumption of salmon flesh, eggs, andfry is of such
high importance, and because nutrient status is a water quality or habitat parameter not
directly influenced by mining operations as outlined in our conceptual models (e.g., Figure 7-
1), we determined that nutrient status of area streams is outside the scope of this assessment
I found risks to salmonid fish due to operation of the transportation corridor well-described with
respect to spatial distribution of fish and their habitats.
RESPONSE: No changes suggested or required
Gordon H. Reeves, Ph.D.
The potential risks to the freshwater habitat of anadromous salmonids are appropriately
characterized and described for the no-failure mode of operation. The report considered the
primary potential impacts of mine development and operation that could impact habitat and
quantified the impacts where possible. The analyses seemed sound and logical, given the
acknowledged limitations about the actual mine location and operation.
RESPONSE: No change suggested or required.
One possible factor that could influence the results was the use of the USGS 1.63,360 maps for
developing the stream network. These maps generally underrepresent the amount of small
streams, which can be ecologically important contributors to the overall productivity of the
freshwater habitat of anadromous salmonids. This is acknowledged in the limitations (p. 5-46).
Thus, the potential loss and modification of habitat that the report describes could be considered
minimal at this time. It would be prudent to confirm the accuracy of the stream layer developed
from the 1:63,360 maps in any future analysis.
RESPONSE: The EPA agrees with this comment No change to the assessment suggested or
required.
The potential impact of the mine development and operation on the productive capacity of the
various river systems could be developed more fully to gain better insights into potential impacts
of the mine. The authors considered the amount of habitat that could potentially be impacted by
mine development and operation by estimating the stream length that would be impacted and by
considering the percent of spaw-ners of the various species (from ADF&G surveys) observed in
potentially impacted areas. However, the productive capacity of given stream reaches for a
given fish species can vary widely. Any additional analysis could consider using Intrinsic
Potential (IP) (Burnett et al. 2007. Ecological Applications 17:66-80), which considers local
geomorphic features to estimate the potential of a given stream reach to provide high quality
habitat for a given species. The concept, developed for use in the Pacific Northwest (PNW), has
been applied successfully for Chinook salmon in the upper Copper River (A. Bidlack, EcoTrust,
Cordova, AK, unpublished). The IP model for Chinook salmon from the PNW that was used in
the Copper River was modified after discussion with local biologists. Similar modification may
be needed for the PNW IP model for coho salmon to be used in Bristol Bay.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 85
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We now include a characterization of stream channel gradient, watershed
terrain (% flatland), and mean annual flow for all streams in the two watersheds. We are
unable to build a complete IP model, as this would require validation and more elaborate
construction of metrics appropriate to this region, but our preliminary characterization
provides the building blocks for assessing the distribution of key habitat-forming and
constraining features across these watersheds.
Another factor that I believe merits further consideration is the potential impact of altered
thermal regimes of discharge water from treatment facilities (p. 5-28). Warmer water could have
potential ecological impacts, particular during the time when eggs are in the gravel. Eggs could
develop more quickly and fry could emerge earlier as a result of even minor changes in water
temperatures (see: McCullough, D.A. 1999. A review and synthesis of effects of alternations to
water temperature regime on freshwater life stages of salmonids, with special reference to
Chinook salmon. US Environmental Protection Agency, Seattle, EPA 910-R-99-010. 279 p.; and
McCullough, D.A., J.M. Bartholow, H.I. Jager, and I I co-authors. 2009. Research in thermal
biology: burning questions for coldwater stream fishes. Reviews in Fisheries Science 17: 90-
113.). These changes could be significant ecologically.
RESPONSE: This is now addressed in Chapter 8.
The report noted in several places that the potential impact on groundwater flows was not
understood at this time but that disruptions of flow paths could have critical impacts on aquatic
resources. One impact that was not mentioned is the loss of over-wintering habitat. K.M.
Burnett (U.S.D.A. Forest Service, PNW Research Station, Corvallis, OR., draft report) found
that the major overwintering areas for coho salmon in the Nome River, AK were at points of
groundwater inputs. The groundwater influx created areas that were less likely to freeze during
winter.
RESPONSE: Overwintering effects from thermal changes are now described in Chapter 7.
Charles I Ves/ev Slaughter, Ph.D.
Yes, the risks to salmonids are well characterized with regard to the hypothetical mine operation
itself. Elowever, I suggest that the concept of "no failure," if taken as applying to the entire
operation from inception through operation, is not realistic.
RESPONSE: The "no failure" scenario was not meant to represent a realistic scenario.
Rather, it was meant to illuminate the effects that would occur solely from a mine footprint,
even in the absence of accidents or failure. The revised assessment no longer uses the term
"no-failure", but simply presents effects from scenarios having higher probability and lesser
magnitude (e.g, failure to collect or treat leachate water) and those having lower probability
and higher magnitude (e.g., TSFfailure).
The Assessment makes a fair start toward considering the risks to salmonids from the potential
transportation corridor. Elowever, the many issues regarding stream and wetlands directly or
indirectly affected by roads and pipelines are not fully explored. The extent (length, area) of
streams and wetlands affected, as outlined in the text, should be considered a very optimistic
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 86
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
lower estimate. The specific issues mentioned, such as bridge or road maintenance, culvert
blockage or failure, erosion from cuts, fills, and the roadway itself, are all significant. I simply
suggest that the potential consequences of imposition of the (hypothetical) transportation
corridor, and future expansions consequent to ancillary infrastructure development and further
additional resource extraction projects, would be broader, more severe and of more consequence
(and thus should receive more emphasis) than the Assessment indicates. I suggest more fully
incorporating Frissell and Shaftel's Appendix G into the body of the Assessment.
RESPONSE: The revised assessment notes that the characterization of both stream length and
wetland area affected likely represents a conservative estimate of the potential effects of the
transportation corridor on hydrologic features of this area. The cumulative risk section
(Chapter 13) has been expanded to include the transportation corridor, ancillary mining
development and secondary development. Additional inf ormation from Appendix G is
incorporated into the main text, and the appendix is referenced in a number of places.
John D. Stednick, Ph.D.
To address this question, a water balance needs to be developed lor the study area watersheds.
Develop a water balance that includes all the principal components and how they may vary in
time and space. The site characterization needs significant improvement, particularly as related
to hydrologic inventories and processes. Little to no data are presented on temperature,
precipitation, evaporation, frozen soils, soil moisture storage, and groundwater storage and
movements. The data that are presented often have unreasonable significant figures. The linkage
between surface and groundwater needs to be better demonstrated. Hyporheic exchanges are
recognized as being important, but the assessment does not demonstrate this linkage.
RESPONSE: . I complete annual water balance with patterns of temporal variability has been
developed and incorporated into the revised assessment. Detailed temporal variability is
beyond the scope of the assessment. The significant figures of reported data and analysis
results have been reduced to a reasonable number. Detailed data presentations on
temperature, precipitation, evaporation, frozen soils, soil moisture storage, and groundwater
storage and movements are generally not reported We cite sources of information, and data
are cited and reported as needed for assumptions of analyses conductedfor the assessment.
Additional information on surf ace water- groundwater interactions has been added
Iliamna Lake hydrology needs to be characterized. What are the inflows, outflows, and turnover
rates? What is the existing water quality in the lake? Aquatic life should be characterized as well.
What is the risk of pollutants entering the lake from the road corridor or upstream mine
development operations?
RESPONSE: The overall hydrology of the lake was not included because none of the
scenarios would result in a change in the hydrology or water quality of the lake as a whole.
Rather, any effects in the lake would be limited to the vicinity of outflows from the affected
streams. Risks from contamination of tributaries to Iliamna Lake are discussed in Chapters 8,
10 and 11 of the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 87
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Climate variability is recognized as a game changer. What are the potential future scenarios for
temperature and precipitation changes in southwest Alaska, and how will these scenarios affect
the water balance? How will climate change affect the availability of water for mine operations,
including processing and potable uses?
RESPONSE: Climate change projections and potential impacts are now discussed in Chapter
3.
Similarly, a complete water quality characterization is lacking. What is the water quality in
surface waters, groundwaters, in time, and in space? What is the definition of background water
quality? Numerous exploratory activities have taken place in the watershed and have the
potential to affect water resources. How were these separated or addressed? Given the geologic
and geomorphologic settings for the study area, are we comfortable that the watershed ridges
delineate the watershed area? Groundwater movements may ignore the physical watershed area
boundary and follow groundwater gradients. Streamflow measurements from the gauged
watersheds could be useful in answering this question. Similarly, the linkage of groundwater and
hyporheic exchange needs to be better demonstrated. Do these exchanges occur in all stream
segments and gradients? What effect does the groundwater have on stream temperatures? Are
depth to groundwater readings available? Is a groundwater monitoring program in place?
RESPONSE: The water quality described as background by the PLP ill their Environmental
Baseline Document was accepted as such in the assessment. Water quality in the three streams
that drain the site was presented in table 5- / ~ (now Table 3-4). The assessment now includes
information on estimated groundwater interaction strength across the project area streams
(Figure 7-14).
The tables and hydrographs (pages 5-32 to 5-39) are unclear. What streamflow changes are
associated with what salmon species and life stage? A boundary condition for adults is different
than for fry.
RESPONSE: Maps of species distributions in relation to affected stream segments are now
included in Chapter 7. The environmental flow analyses are not species or life stage specific,
but assume an overall risk associated with proportional deviations from the baseline flow
regime.
The proposed mine will use large quantities of water in ore processing and transport. How much
is required and how will this affect water resources; both surface and groundwater?
RESPONSE: The water balance has been extensively revised (Chapters 6 and 7), and updated
estimates are incorporated into the streamflow computations provided in Chapter 7. Effects on
groundwater resources are explicitly incorporated in the analysis of the pit dewatering,
associated cone of depression, leachate leakage from the TSF and the waste rock piles, and
interbasin groundwater transfers.
The no-failure mode of operation is predicted to change the watershed contributing area and
hence streamflow, and uses the boundary condition of a 20% change in streamflow as significant
salmonid habitat loss. The assessment assumes a liner response between watershed area and
streamflow contribution, and a linear response between habitat productivity and watershed area.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 88
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: A more comprehensive water balance is now used to estimate stream/lows,
which incorporates losses and additions due to pit dewatering, and wastewater treatment plant
processing and distribution, such that the relationship of watershed area and streamflow is not
linear. We do not assume a linear response between habitat productivity and watershed area
Upland settings are probably more productive in terms of productivity and should be addressed
as such.
RESPONSE: Relative productivity of aquatic habitats has not been extensively documented
across the region.
Toxin assessment focused on copper, and other metals can be presented to show the range of
metal concentrations for chronic and acute toxicity, i.e., arsenic, molybdenum, silver, barium,
and lead. Given the very clean waters (low hardness and organic carbon), the chronic toxicity of
various metals should be evaluated. Water quality varies in time and space in the study area, and
a better characterization of water quality could be developed. Metal loads could be calculated
with streamflow records. What is the proportioning of dissolved versus total metals? Are metals
transported with sediments? Do organic carbon fluxes change in space or time?
RESPONSE: The influence of receiving water chemistry was incorporated to the extent that
current science allows. The toxicity of copper was corrected for water chemistry using the
biotic ligand model, and the toxicity of other metals was corrected for hardness, when models
were available. In each case, water chemistry of the individual receiving stream was usedfor
the correction, Also, instream concentrations are based on streamflows, including changes in
streamflows due to mine operations. Both sediment and organic matter concentrations are
quite low in all three streams at the Pebble site, so copper remains dissolved in the model and
other metals in leachates are likely to remain dissolved as well
Salmonid risk from travel corridor: The proposed road location has the potential to affect 270 km
of stream between stream crossings and Lake Iliamna. The expected road erosion and sediment
production has known effects on salmonid resources. The discussion of the travel corridor does
not include the potential for road failures, landslides, blocked culverts, or ditch failure. The
discussion does not talk about traffic volume or the potential of hazardous material transport on
the travel corridor. Need to address road maintenance, fugitive road dust, and road chemicals
either dust or ice control.
RESPONSE: The original draft assessment included discussion of the potential for road and
slope failures, blocked culverts, and soil erosion from road cuts, borrow areas, road surfaces,
shouhlers, cut-and-fill surfaces, and drainage ditches. The revised assessment factors traffic
estimates into assessments of chemical spills from transport truck accidents (Section 10.3.3)
and impacts from dust (Section 10.3.5). Salts usedfor to reduce dust and improve winter
traction are discussed in Section 10.3.3. Potential mitigation measures for stormwater runoff
erosion, and sedimentation are discussed in Box 10-3.
There is no discussion of water processing after delivery of the slurry to the sea port and return
of waters back to the mine site.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 89
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Section 6.1.2.5 of the revised assessment discusses that the water would be
returned to the process water ponds. The scenarios indicate that slurry water would be
returned to the site without treatment other than removal of solids.
Roy A. Stein, Ph.D.
No-Failure Mode of Operation. My comments regarding the no-failure mode of operations and
their impact on salmon can be found under Question 3.
RESPONSE: See response to Question 3.
Road Use I: Page 5-60. Beyond calcium chloride, how can we be confident that the typical
chemicals that derive from highway use will not occur on this mine road (as noted on page 5-
60)? Is it because the low volume of traffic? If so, would not we expect accumulation through
time.. .over the 78 years of the mine operation (see Appendix G for some detailed analysis:
should some of this material be added to the main report?)? What about the impact of road dust
on nearby aquatic systems (wetlands, streams, rivers, etc.)?
RESPONSE: The text relating to traffic and contaminated runoff has been modified in the
revised assessment to read: "It is unlikely that the potential transportation corridor would
have sufficient traffic to significantly contaminate runoff with metals or oil, but storrmvater
runofffrom roads at the mine site itself might contain sufficient metal concentrations to affect
stream water quality." Though traffic-associated contaminants may be expected to increase
over time, they would probably not be as significant as stormwater runoff-associated metals
from the mine site., l.v noted in the revised assessment, the main impact of dust from the
transportation corridor on salmonids would likely be a reduction in riparian vegetation and
subsequent increase in fine bed sediment. The main impact of dust at the mine site would be a
direct increase in fine bed sediment due to mine construction and operation (the effects of
increased sediment loading are discussed in Section 10.3.4).
Road Use II: Page 5-62 to 5-63 (plus Appendix G: again, as with other appendices, include
more of this information in the main report). Will there be frost heave of the road bed such that
specific structures will have to be installed to prevent this movement of the road bed? These
roads will be treated with chemicals, such as calcium chloride, to keep the dust down and
contribute to an ice-free condition, but no data are available for the impacts of these chemicals
on nearby streams. How then do we deal with this issue (page 5-62 and 5-63)? The suggestion
is that one needs to have roads built at least 8 meters from streams, but this cannot be the case in
this situation, simply because of the large number of streams, rivers, and wetlands along the road
corridor? More detail as to the impact of the transportation corridor should be added, including
issues, such as truck accidents, fuel spills, other chemical spills, etc.
RESPONSE: The assessment does not address potential frost heave of the road bed, although
this factor will need to be considered during design of a road Additional information has been
added to the revised assessment on the potential impact of calcium chloride on nearby
vegetation, surface water, groundwater and aquatic species. According to the USDA Forest
Service (1999), application of chloride satis should be avoided within 8 m of water bodies. We
agree that the 8 m buffer zone for satis would be difficult to maintain, but it could be achieved
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 90
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
at some cost in dust suppression and winter road salting. It would not require keeping roads
out of that zone.
Additional information from Appendix G is incorporated into the main text, and the appendix
is referenced in a number of places. The revised assessment contains greater detail on the
potential impact of the transportation corridor. For example, the assessment now factors
traffic estimates into assessments of chemical spills from transport truck accidents (Section
10.3.3) and potential impacts from dust (Section 10.3.5). Fuel spills are covered in Chapter 11.
Road Use III: page 5-71 (plus Appendix G). The road will intersect multiple streams and rivers
along the northern end of Iliamna Lake, where as many as one third of the sockeye salmon in this
lake spawn. And this is where the causeway across Iliamna Lake will be built as well. From my
perspective, it seems that impacts on spawning sockeye will be large in this area (without saying
anything about causeway: will there be culverts or bridges to allow water and fish to
communicate with the rest of the lake)? I would argue this is important, given salmon are
attracted to certain odors and water-flows and these odors and water-flows are coming from
inlets streams into Iliamna Lake. Preventing any sort of blockage of water flow or salmon
migration would be the goal. Are there other issues that should be considered when building this
causeway?
RESPONSE: The assessment makes no mention of a causeway across Iliamna Lake. We
believe the commenter is referring to a proposed causeway over the upper end of Iliamna Bay,
which is part of Cook Inlet. Culverts or bridges would be built to allow fish access between
streams crossed by the proposed road and Iliamna Lake. The issue of culvert blockage is
discussed in the assessment.
Road Use IV. Points made by public testimony reinforces the idea that as this area is opened to
the public, the opportunity lor new, invasive species to colonize this pristine ecosystem increases
dramatically, likely to 100%. Simply put, invasive species will now be carried by humans via
the road, inadvertently, into this previously inaccessible watershed.
RESPONSE: EPA assumes that the proposed road would be closed to the public during
mining operations but potentially could become a public road after mining operations cease.
Even when not open to the public, construction and operation of the proposed transportation
corridor increase the probability that new terrestrial and aquatic species will be transported to
and potentially establish themselves in the Bristol Bay region If the road were opened to the
public, the probability oj coloni-ation by invasive species may increase further, but rates of
introduction by industrial and public vehicles cannot be distinguished given available
information Invasive species are addressed in Section 10.3.6 of the revised assessment
William A. Stubblefield, Ph.D.
The document appears to adequately address potential questions associated with habitat loss due
to hydrologic changes, especially considering the hypothetical nature of the mine and the lack of
specific detailed information regarding an actual proposed facility and all of the associated
operational details of the facility. The assessment of potential impacts and ecosystem protection
parameters is predominately based upon the publication of Richter et al. (2011). Additional
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 91
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
support and evaluation of these recommendations for fisheries populations in the Bristol Bay
area should be closely evaluated.
RESPONSE: Prompted by this comment, we consulted with regional biologists and
hydrologists to evaluate the suitability of the sustainability boundary approach for flows. We
asked them if there was any reason that fish populations in these streams, or the specific
hydrology of the area, made it exceptional with regard to this approach (e.g., was there any
reason to think that the Richter approach was not applicable here). We received uniform
support for applying this approach to Bristol Bay streams. We strengthen our emphasis that
this is a precautionary approach, and that the detailed hydrologic and habitat modeling work
that PLP contractors have begun will help provide a useful basis for more sophisticatedflow-
habitat modeling.
Dirk van Zyl, Ph.D., P.E.
Chapter 5 of the EPA Assessment is entitled: "Risk Assessment: No Failure". Chapter 5 presents
an evaluation of habitat loss and modification resulting from the hypothetical mine. A summary
of the "risks" associate with the "no failure" case is provided in Chapter 8. There is specific
focus on evaluating the magnitudes of the losses and modifications to the environment.
RESPONSE: No change requested or required
A risk assessment addresses three questions (Kaplan and Garrick, 1981):
• What can happen? (i.e., What can go wrong?)
• How likely is it that that will happen?
• If it does happen, what are the consequences?
There are a large number of risk assessment methods and it is common to express the magnitude
of risk as a combination of likelihood of occurrence and consequences (IEC, 2009). This is the
typical outcome for engineering assessments of systems. For example, in the case of a Failure
Mode and Effects Analysis (FMEA), it would be typical to develop a risk matrix to combine
likelihood of occurrence and consequences to express the level or magnitude of risk in
qualitative terms (Robertson and Shaw, 2012).
The EPA Assessment describes the two components of risk but does not provide any information
on the magnitude of the risk. For example, for the no-failure condition it describes the length of
streams, areas of wetlands, etc. that will be impacted by developing the mine, i.e. the
consequences. One may argue that the likelihood of occurrence of these consequences is unity
(or certainty) if the mine is developed, as this is not specifically addressed by the report.
RESPONSE: The risk assessment does address engineering risks in the manner specified,
including magnitudes of spills, leakage, etc., and the consequences for water and habitat
quality. We now explicitly clarify that losses of stream length are unavoidable for a project of
this magnitude (Chapter 7). The magnitude of the risk or the likelihood that the stream
lengths will be lost if the mine is constructed is 100%.
One would next expect an expression of the magnitude of this risk based on some comparison of
the consequences to a set of outcomes that could result in acceptable or unacceptable risks. The
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 92
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
EPA suggests this as an approach in its 1998 Guidelines for Ecological Risk Assessment (EPA,
1998): "In some cases, professional judgment or other qualitative evaluation techniques may be
used to rank risks using categories, such as low, medium, and high, or yes and no". Quantitative
approaches such as fuzzy logic has also been used to develop expressions of magnitude of risk as
described by EPA (1998): "For example, Harris et al. (1994) evaluated risk reduction
opportunities in Green Bay (Lake Michigan), Wisconsin, employing an expert panel to compare
the relative risk of several stressors against their potential effects. Mathematical analysis based
on fuzzy set theory was used to rank the risk from each stressor from a number of perspectives,
including degree of immediate risk, duration of impacts, and prevention and remediation
management. The results served to rank potential environmental risks from stressors based on
best professional judgment".
RESPONSE: Although the 1998 ERA Guidelines describe professional judgments as an
acceptable methodfor ranking risks in appropriate circumstances, it is not appropriate for this
assessment. The purpose of this watershed assessment is not to rank risks. It is to estimate, as
far as existing data and knowledge allow, the risks associated with proposed and potential
mining activities in the Bristol Bay watershed. EPA decided during the problem formulation
that this assessment would be based on published science. An assessment based on elicitation
of expert judgment could be performed in the future, if desired.
It is unclear to the reader how significant a loss of 87.5 km of streams in the Nushagak River and
Kvichak River watersheds is to the ov erall ecosystem. Are there any criteria that can be used to
develop such an expression? Can a multi-stakeholder workshop (as is often done) be used to
develop such criteria and expressions of risk magnitude? Without having such expressions of
risk magnitude it is impossible for those without specific expertise in salmonids to evaluate
whether this is a significant risk. Price et al. (2010) states that: "Between 1999 and 2008, 3,500
fish passage barrier culverts were replaced with fish-passable structures, reportedly opening
nearly 5,955 km of fish habitat in Washington streams (Governor's Salmon Recovery Office
2008)". Comparing the loss of 85 km to this gain of 5.955 km seems to imply that 85 km loss
may represent a relatively small risk, which may not be the case at all. However, the EPA
Assessment does not provide any insight in the magnitude of risk except to provide a value for
the consequences.
RESPONSE: The purpose of the assessment is not to assign significance to the risks, but to
provide information for decision-makers on the consequences of mining.
The EPA did conduct a multi-stakeholder conference to determine the significant endpoints
and exposure pathways for the assessment. However, the EPA decided during the problem
formulation that this assessment would be based on published science. Therefore, a multi-
stakeholder workshop would not be an appropriate mechanism to estimate risks or their
significance.
Comparing the potential loss of salmon-supporting streams in the Nushagak and Kvichak
River watersheds to restoration of streams in Washington for salmon recovery may not be very
useful The comment does illustrate that seemingly inconsequential or insignificant losses
have frequently led to diminished or even lost salmon stocks. In the example cited, apparently
3,500 seemingly inconsequential actions had to be remedied at public expense because of their
cumulative impacts.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 93
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Similar comments can be made with respect to the relative risks associated with the other losses
of ecological functions for other failure modes.
RESPONSE: See response to previous comment
It is recognized that it is important to maintain separation between the risk assessment and risk
management functions. As expressed by the National Research Council Panel in their report on
Science and Decisions (NRC, 2009): "The committee is mindful of concerns about political
interference in the process, and the framework maintains the conceptual distinction between risk
assessment and risk management articulated in the Red Book. It is imperative that risk
assessments used to evaluate risk-management options not be inappropriately influenced by the
preferences of risk managers".
RESPONSE: The EPA agrees with this comment and its implication that assessors should not
judge significance. No changes suggested or rei/nired.
Providing an expression of risk magnitude should not interfere at all in the separation of risk
assessment and risk management, but should provide the risk manager with one extra level of
analysis and insight from the expert assessor of the problem at hand. Multi-stakeholder
interaction will only serve to enhance the value of the risk ranking.
RESPONSE: See response to previous comment on this topic.
On p. 4-33, it is stated that after closure: "No PAG waste rock would remain on the surface". It is
also stated in Chapter 4 that PAG and NAG waste will be segregated. On p. 5-48, it is stated that:
"However, the primary concern during routine operation would be waste rock leachate. That
leachate would become more voluminous as the waste rock piles and uses of waste rock for
construction increased during operation. After mine closure, it would be a major source of
routinely generated wastewater along with water pumped from the TSF and pit. Leachate
composition from tests of the three waste rock types (Tertiary, East Pre-Tertiary, West Pre-
Tertiary) is presented in Tables 5-14 through 5-16". There is no specific indication which of
these waste rock types could be described as PAG or NAG and Chapter 5 seems to assume that
these 3 samples are representative of the total amount of waste rock, about 4 billion tonnes for
one mine scenario. If all the PAG material will be removed from the surface, as stated in the
scenario in Chapter 4, and the NAG will not generate acid drainage, then it is difficult to
understand why the waste rock piles and waste rock used for construction (supposedly all NAG
at this stage) would be the major source of "routinely generated wastewater."
Note that it is further unclear why there would be water pumped from the tailings and the pit if
the TSF were closed, as discussed above, and if it will take the mine pit 100 to 300 years to fill.
Some clarification is in order.
RESPONSE: This issue has been clarified in the revised document PAG and NAG waste
rocks would be identified during the course of mining, but the available test results indicate
that pre-Tertiary rock is PAG. The PAG waste rock would be segregated and none would
remain at mine closure. However, some will be on the surface during operation. In addition,
the NAG waste rock produces potentially toxic leachates that must be collected and treated
Water could be pumpedfrom the TSFs and the pit for treatment before discharge. Treatment
of pit water would occur once it is a source rather than a sink
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 94
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
A further reference to the fate of waste rock after closure is found on p. 5-77 of the EPA
Assessment: "Under the mine scenario, the mine pit, waste rock piles, and TSF would remain on
the landscape in perpetuity and thus represents permanent habitat loss." It should be noted that
the scenario states that PAG will not remain on the surface, whatever volume and area of land
surface that represents.
RESPONSE: Section 6.3.3 of the revised assessment clarifies that no PAG waste would
remain on the surface.
The descriptions of exposure and exposure-response resulting from the transportation corridor in
Section 5-4 of the EPA Assessment focus on potential impacts and make use of references that
are clearly not representative of the expected road construction. A number of these references
date from 1975 and 1976 (p. 5-59) and are not necessarily representative of road design and
construction practices in 2012. On p. 5-62, the following statement and reference is given:
"Sediment loading from roads can severely affect streams below the right-of-way (Furniss et al.,
1991 and references therein)". This reference is specifically focused on forest and rangeland
roads, clearly not representative of a major transportation road between a mine and the port
facilities from where its products are shipped. This publication contains many recommendations
specifically for forest and rangeland roads and some of them are indicative that it is not
applicable to the transportation corridor for a major mine access road: "Design cut slopes to be as
steep as practical. Some sloughing and bank failure is usually an acceptable trade-off for the
reduced initial excavation required" (p. 306); and "stream crossings can be considered dams that
are designed to fail. The risk of failure is substantial for most crossings, so how they fail is of
critical importance" (p. 310). The reference also refers to the application of oil as a dust
abatement additive on p. 312, which is hardly acceptable practice. In my review, I did not find
that any of the references used in the EPA Assessment refer specifically to mine roads such as
those considered for the transportation corridor at the Pebble Mine scenario.
RESPONSE: The information cited from the two publications noted by the commenter is still
true today. The first, use of Darnell (1976) was incorrect, and has been changed to Furniss et
al (1991). I /though Furniss et al. (1991) f ocuses on f orest and rangeland roads, it is a
seminal publication on the potential effects of roads, particularly as they relate to salmon. The
general conclusions of that paper should be applicable to the transportation corridor described
in the assessment. Furniss et aL (1991) lists a number of guidelines for road design and
construction that will help minimize adverse eff ects on salmonid habitats. It does not
specifically advocate the application of oil as a dust abatement additive. It merely states that
whatever chemicals are used, they should be applied so as not to enter streams, and that
subsequent transport of these substances into water courses should be evaluated
The failure frequencies cited in the revised assessment are from modern roads and not
restricted to forest roads. Because the proposed mining would take place in an undeveloped
area, the literature used in the assessment is necessarily from areas outside of Bristol Bay.
However, we used recent literature from representative environments to the extent possible.
Lastly, information on current design standards that would be used along the proposed
transportation corridor is now included within text boxes throughout Chapter 10.
It is further interesting that it is stated on p. 5-60 that there will be 20 bridges and 14 culverts
along the road without referring to this as an assumption, and no reference is cited for this
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 95
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
information. Will there be a change in impact if the decision is made to build 30 bridges and 4
culverts or 34 bridges and no culverts?
RESPONSE: The estimate of 20 bridges came from Ghaffari et a I. (2011). In the revised
assessment, crossings that would be bridged (now 18) are based on mean annual stream flows,
as explained in the text. If a decision was made to build more bridges and fewer culverts, there
would be a change in impact, but scenarios with 30 or more bridges are probably not realistic.
The discussion on the potential impacts of the transportation corridor on salmonids serves the
purpose of highlighting some aspects that engineers and fish biologists must take into account
when designing and maintaining the final transportation corridor lor the Pebble Mine and other
mines in the Bristol Bay area. However, this assessment does not appropriately describe the scale
and extent of the risks to salmonid fish due to operation of a transportation corridor under the no-
failure mode of operation.
RESPONSE: The "no failure" mode of operation was meant to illuminate the effects that
would occur solely from a mine footprint, even in the absence oj accidents or failure. This
term has been eliminated in the revised assessment. The no-failure scenario from the draft
assessment has been changed to a section on the effects of the footprint of a mining operation,
without regard for operational problems (Chapter 7). The revised assessment places the
streams along the transportation corridor into the context of the entire Nushagak and Kvichak
River watersheds with respect to important watershed attributes such as discharge, channel
gradient, andfloodplain potential Potential risl,s to fish habitats and populations associated
with the proposed corridor are then evaluated in some detail
Phyllis K. Weber ScunncU. Ph. I).
The no-failure model makes a number of assumptions about how the mine will be developed -
some may be accurate, some may be considerably different. It is important to take under
consideration that Pebble is currently a prospect, not a mine. Should this project proceed to mine
development, it will be incumbent on the mining company to develop a rigorous mine plan that
includes detailed information on all aspects of a future project. This mine plan will be reviewed
by state and federal staff with experience in large project development.
RESPONSE: The III' I agrees with this comment No changes suggested or required
The no-failure model discusses the amount of riverine habitat that will be lost to mining by the
mine pit, tailing storage facility, and waste rock dumps. Anadromous fish habitat is protected
under Alaska Statute 16.05.840-870. The statute requires review of a project potentially
affecting fish habitat and, where necessary, avoidance, mitigation, or compensation. A project
must provide free passage of fish; the project cannot be placed in such a way that fish are
prohibited from moving into the upstream reaches. Estimates of habitat loss from the mine
footprint are not possible without a more detailed plan of operations for the mine.
RESPONSE: The scenarios presented are meant to represent those expected as typical for
mining ofporphyry copper deposits of this type, and are based on preliminary mine plans from
NDM (Ghaffari et aL 2011). AUhough layout of mining components at a site may differ
somewhat from what we present in the scenarios, the main components of mining will remain
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 96
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
the same for open-pit mining. Given stream density in the area, direct losses of stream and
wetland habitat of a similar magnitude would be inevitable with projects of the specified
magnitudes.
There are many aspects of the development of a large mine project that need thorough review to
ensure that habitats are protected. These include, but are not limited to: classification and
storage of waste rock, lower grade ore, overburden, and high grade ore; development and
maintenance of tailings storage facilities; development and concurrent reclamation of disturbed
areas, including stripped areas and mine pits; collection and treatment of point and non-point
source water; quantity and timing of discharges of treated water; monitoring of ground water,
seepage water and surface water; and biomonitoring. The transportation corridor will require
review and permitting of every stream crossing of fish-bearing waters. In addition, plans should
be developed for truck wheel-washing to minimize transport of contaminated materials.
RESPONSE: The EPA agrees that these aspects would need to be subject to a thorough review
during the development and approval of a detailed mining plan. \o changes suggested or
required
Paul Whitney, Ph.D.
Material Resource Areas. Material resource areas, mentioned on page 4-34, for the road and
pipelines should be discussed in more detail. Will aggregate be required? If so, where are the
aggregate resources in relation to floodplains? I spent a summer surveying material resource
areas for a proposed arctic and subarctic pipeline and access road. Suitable material resource
areas are sizeable and are often important (e.g., aggregate) for wildlife (such as bears that
hibernate or survive the winter in dens) and fishery resources. Sometimes dens can only be
excavated in non-permafrost (i.e., aggregate) soils. It appears the project area is in a zone of
discontinuous permafrost, but permafrost could be more continuous in the higher elevations
along the road through the Kenai Mountains. An accurate assessment should determine the
permafrost location(s), as well as the area and importance of material resources for fish and
wildlife. In addition, Reclamation Plans for the material resource areas should be briefly
discussed to ensure that areas mined for aggregate will not avulse and capture streams.
RESPONSE: EPA agrees with this commenter that the impacts of material resource areas to
wildlife, fishery and other subsistence resources could be significant and must be addressed in
an environmental impact statement and as part of the 404 permit review. Review of potential
material resource areas was not included in the scope of this assessment.
Water for Dust Control. Dust control for the 86-mile proposed haul road will likely require a lot
of water. Where will this water come from? Withdrawal from streams crossed by the haul road
could have impingement and flow reduction consequences. Adequate screening could solve the
impingement issue. Some back-of-the-envelope calculations could determine if water
withdrawals for dust control could alter the projected hydrographs when salmonids are present in
the streams.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 97
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We expect water for dust control to be a small amount from any one source.
Permits would be required from the State of Alaska that would address impingement issues.
We do not expect this to be a major issue.
Question 5. Do the failures outlined in the assessment reasonably represent potential
system failures that could occur at a mine of the type and size outlined in the mine
scenario? Is there a significant type of failure that is not described? Are the
probabilities and risks of failures estimated appropriately? Is appropriate information
from existing mines used to identify and estimate types and specific failure risks? If
not, which existing mines might be relevant for estimating potential mining activities in
the Bristol Bay watershed?
David A. Atkins, M.S.
The Assessment focuses on some low probability, high impact failures (e.g., TSF failure), and
presents summaries of failures at existing mines. The majority of the focus is on catastrophic
failures, such as TSF, pipeline, water collection and treatment, and road and culvert. Anecdotal
information regarding mine failures is numerous, but often not well documented, so it is difficult
to get information on the details of failures of other projects. It is also difficult to extrapolate the
probability of failure from one site to the next, and the report stresses the wide range of
uncertainty, depending on design and environment. Without a more detailed understanding of the
mine plan and associated engineering, as well as additional detailed analysis, it is difficult to
determine if the failure probability estimates presented in the Assessment are reasonable.
RESPONSE: The authors concur with the conunenter that it can be "difficult to get
information mi the details of failures of other projects The statistics for historic tailings dam
failures are derived from the largest available database and include many tens of thousands of
dam-yean. The pipeline failure data cover millions of kilometer-years ofpipeline experience.
The data on failures of water collection and treatment systems and of culverts are less
extensive. We also recognize that even with detailed engineering and design information, the
prediction of failure probabilities is extremely difficult. Finally, since all of these low-
probability f ailures are statistical phenomena, the actual experience at any one site could be
vastly different than another similar site, even when the failure probabilities have the same
distribution.
The focus on catastrophic failures also takes attention away from what is probably a more likely
scenario. Every project is subject to accidents and smaller, non-catastrophic failures that have
varying degrees of consequence. Sometimes these failures are easily identified and fixed and
other times they can go un-noticed for periods of time.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative wastewater treatment plant failure, truck
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 98
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
accidents and spills, and refined leachate seepage scenarios) and explains why these
particular failure scenarios were chosen.
It would be helpful to describe some smaller-scale failures that have occurred at mine sites. A
partial list includes: accidents and spills along the transportation corridor or within the mine site;
unanticipated seepage of contaminated water that may be difficult to detect, collect and treat;
movement of water along preferential flow pathways that are difficult to characterize; temporary
failure of water collection and treatment systems; mistakes in engineering analysis that
underestimate the volume of water that must be collected and treated or overestimate the volume
of water available for use; and designing based on incomplete data and understanding of climate
conditions.
RESPONSE: There is a wide variety of failures that could occur, including those provided by
the commenter. Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., di esel pipeline failure, quantitative wastewater treatment plant failure, truck
accidents and spills, and refined seepage scenarios) and explains why tliexe particular failure
scenarios were chosen.
Steve Buckley, M.S., CPG
The engineering failures reasonably represent potential system failures outlined in the mine
scenario based on historic porphyry copper deposits of this type. It is less clear if the failures
reasonably represent a mine scenario based on state of the art engineering and mitigation
practices. Appendix I provides some information related to potential system failures and possible
mitigation measures designed to minimize these risks but these are not treated in any detail in the
assessment. It would be difficult to pull together the most modern engineering and mitigation
practices from around the world but it could help bound the risks associated with modern mine
development.
RESPONSE: Our purpose in the assessment is to evaluate the risks from hazards resulting
from a mine operated with appropriate mitigation measures for design, operation, monitoring
and maintenance, and closure. I cci dents and failures happen regardless of mitigation
measures: thus, effects of several failures are evaluated. Mitigation measures related to our
mine scenarios are now clearly discussed in Chapter 6.
The Red Dog mine in northwest Alaska might be relevant for estimating potential mining
activities in the watershed. Although the characteristics of the deposit differ significantly, at
roughly 150 million tons it is half the size of a reasonable minimum mine scenario and would be
helpful to characterize some minimum mine development scenario.
RESPONSE: A third mine size scenario, representing the worldwide median size porphyry
copper mine (Singer et al. 2008), is included in the revised assessment
Courtney Carothers, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 99
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The potential failures outlined in this assessment include: tailings dam failures, pipeline failures,
water collection and treatment failures, and road and culvert failures. These failures appear to
represent the key potential failures for this mining scenario, their risks appear to be estimated
reasonably, and statistics from existing mines appear to be used appropriately, although I have no
particular expertise with which to evaluate this assessment. As we discussed in our peer review
panel, the focus here is on catastrophic failure. More detail should be provided on likely non-
catastrophic failures, ones that would be more difficult to detect.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
leachate seepage scenarios) and explains why these particular failure scenarios were chosen.
We did include leachate seepage scenarios (Chapter X). which would be more difficult to detect
than catastrophic failures.
Dennis D. Dauble, Ph.D.
My experience in system failure of mines of the si/e and type outlined in the scenario is limited.
However, what does seem to be missing is the long-term effects of leachates to receiving water
bodies in any type of risk scenario, including both non-failure and failure modes. That is,
assuming no catastrophic failure, how might tailings constituents interact with aquatic habitats
seasonally, such as during periods of snowmelt and severe rainfall events?
RESPONSE: The original draft assessment contained a scenario in which tailings leachate
was not fully contained and reached a stream (Section 6.3 in the May 2012 draft). The revised
assessment includes estimates of leachate escaping from the TSEs andfrom the waste rock
piles, bypassing the collection systems, and entering the streams (Chapter 8). The estimated
loadings of copper and oilier elements from these leachate flows are included in stream
concentration estimates. The assessment discusses the impacts of these concentrations on the
aquatic habitat and biota. The couimenter is correct that we did not include a scenario in
which the dam does not Jail, but snowmelt and severe rainfall would result in overtopping and
release of untreated water. That is ivrr plausible, but there are just too many possible failure
scenarios to include more than a few of them
Gordon H. Reeves, Ph. I).
No comments on this question
Charles Wesley Slaughter, Ph.D.
Potential failures seem reasonable, based on history of other mining operations. However, the
consequences of hydrologic extremes during winter (frozen soil) conditions are not adequately
addressed. The possibility of the mining operation and the transportation network encountering
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 100
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
discontinuous permafrost is not mentioned, although at least some soils maps indicate permafrost
presence.
RESPONSE: Warhaftig (1965) reports that permafrost is sporadic or absent in the Nushagak-
Bristol Bay Lowland (Table 3-1). If permafrost were detected during project development or
construction, the designs would need to address any potential impacts on the infrastructure
and potential impacts of the infrastructure on the permafrost. Frozen soil could improve
vehicular access to parts of the project site and minimize the disturbance from such access.
The occurrence of an extreme hydrologic event, such as heavy rainfall on frozen soil or heavy
rain on an existing snowpack, could produce unusually heavy runoff and higher than normal
stream flows. We analyzed the impact of a tailings dam failure during the Probable Maximum
Precipitation (PMP) event, thereby generating the Probable Maximum Flood (PMF). The
increasedflow due to precipitation was, in most of the scenarios, small compared to the flow
releasedfrom the TSF. Any increases in the peak runoff due to frozen soil or melting
snowpack would also be small relative to the TSF release, so the modeled scenario can be
considered a reasonable bounding estimate.
Designs for the components of the mine infrastructure would need to consider the natural cold
region conditions and incorporate appropriate design features and safety factors to achieve an
acceptable level of performance, (ihaffari et a I. (2011) says "The Pebble deposit is located
under rolling, permafrost-free terrain in the Iliamna region of southwest Alaska..." and "The
deposit is situated approximately 1,000ft amsl, in an area characterized by tundra, gently
rolling hills and the absence of permafrost. " (ihaffari et aL (2011) describes the transportation
corridor thusly: " The road route traverses terrain generally amenable to road development.
... There are no significant occurrences of permafrost or areas of extensive wetlands."
Nevertheless, if sporadic areas of permafrost are discovered, the designs will need to address
the interactions between the infrastructure and the permaf rost.
The probability approach outlined lor potential TSF dam failure is unpersuasive. It is difficult to
relate to a number like "0.00050 failures per dam year," or to the implication on p. 4-47 that one
can expect a tailings dam failure only once in 10,000 to one million "dam years." This could
suggest to the casual reader that failure of the hypothesized TSF1 dam (for which one "dam
year" is one year) should not be anticipated in either the time of human occupation of North
America, or the span of human evolution.
RESPONSE: The commenter is correct. The proposed dams, if designed, built, and
maintained to current engineering best practices, would be anticipated to have a low annual
probability offailure. However, the failure probability would not be zero. The writers concur
with the commenter that these low probability numbers may be difficult for the casual reader
to grasp, so we now also present estimates ofprobability in terms ofprobability failure over
different time periods. The discussion of this issue has been expanded to clarify that the
failure rate is a design goal and is not based on empirical evidence.
Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-magnitude event at the
Pebble locale has an estimated return period of 200 years. Such a return interval probability is
difficult to interpret, given the lack of historical seismic records for the region; in any event, such
a return period estimate is in no way predictive of future seismic activity, in year 2012 or year
2212. (The suggested 200-year return period should also be viewed in light of the 79-year
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 101
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
suggested operating life of the hypothetical Pebble operation, probable longer-time operations at
other mineral extraction sites which would be developed following implementation of Pebble
and building from the infrastructure associated with Pebble, and also the projected very long
persistence of the TSFs following cessation of active mining).
RESPONSE: Box 9-2 in the revised assessment (Box 4-6 in the May 2012 draft of the
assessment) provided the OBE and return period determined by NDM in the Preliminary
Assessment A detailed engineering design and safety evaluation is outside the scope of this
assessment. The discussion of seismicity (Section 3.6) addresses the uncertainty in interpreting
and predicting earthquake magnitude and recurrence in the Pebble area
Box 4-6 does note that "The return periods stated in Alaska dam safety guidance are inconsistent
with the expected conditions for a large porphyry copper mine developed in the Bristol Bay
watersheds, and represent a minimal margin of safety."
RESPONSE: The return periods used are consistent with the Alaska Dam Safety Guidance,
however the operator could include additional margin of safety in the design for critical
structures. The return period and seismic safety factors do not inform the failure analysis in
this assessment, but would be important considerations during the review process for any
future mine plan
John D. Stednick, Ph.D.
The assessment reasonably addresses potential large system failures, but should include a variety
of smaller and perhaps more frequent failures (see Question 4). A large tailings storage facility
failure compared to a blocked road drainage culvert. The level of detail in the assessment of the
potential system failures varies considerably and baits the question—why? Does this demonstrate
lack of understanding of failure prediction, lack of failure prediction, or writing team expertise?
RESPONSE The El'. I does not believe that a blocked culvert requires or deserves the same
level of analysis as a tailings dam failure and spill The latter is a much more complex
phenomenon with multiple consequences that require evidence and analysis such as the
potential toxicity of the spilled tailings. In contrast, blockage of a culvert is a relatively
straightforward phenomenon. I'urther, a TSFfailure poses a much greater concern for
stakeholders and local communities due to the large magnitude of potential effects.
Tailings storage facility: The liquefaction phenomenon, internal and external erosion, seepage,
and overtopping are some of the main failure modes of tailings storage facilities. A large quantity
of stored water is the primary factor contributing to most tailings storage failures. The risk of
physical instability for a conventional tailings facility can be reduced by having good drainage
and little (if any) ponded water. Some suggest that the tailing pond freeboard should be able to
accommodate the 100-year, 72-hour storm/streamflow event. What are the State of Alaska
standards? Discuss the probability of failure of a TSF from other than overtopping by a
precipitation/streamflow event. The potential of seismic activity and its effect on tailings storage
and other earthworks needs to be addressed.
RESPONSE: The historic failure probabilities for tailings dams presented in the assessment
include all modes of failure, including slope instability, liquefaction, overtopping, erosion, and
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 102
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
seismic activity. Historical failures were discussed as supporting background information and
present a defensible upper bound on the failure probabilities. The failure probabilities used in
the assessment are based on Alaska's dam classification and required safety factors applied to
the method of Silva et al (2008). The data presented by Silva et aL (2008) consider only the
annual probability offailure from slope instability, but the methodology is equally applicable
to other failure modes. The discussion offailure probabilities in the revision (Chapter 9) is
expanded to clarify this issue.
Chemical transport spill: Mine development and ore processing will require significant loads of
petroleum and chemical products. Although the exact processing formulations are not given,
most copper porphyry mines use similar formulation in ore notation and processing. How will
chemicals be stored, transported, and recycled? What are the opportunities for accidents to
occur?
RESPONSE: Storage of chemicals is addressed in the assessment. Petroleum transport is by
pipeline and an assessment of the risk of spills lias been added in Chapter 11 of the revised
assessment. Process chemicals would be transported by truck. Discussion of truck transport
and potential for accidents has been included in Chapter 10, including a quantitative analysis
of the risk of wrecks and of wrecks that cause spills into streams or wetlands.
Roy A. Stein, Ph.D.
Failures Appropriate tor Mines nl" litis Si/.e and l \ pe. Given my background, I can't answer
with any authority, though the comparisons seemed appropriate, though clearly no extant mines
are as large as the one proposed herein. Some of the public testimony spoke directly to
comparisons with existing mines in dry areas would be completely inappropriate because it is the
hydrology of the Bristol I Jay watershed that would make it so very vulnerable to mining impacts.
RESPO\SE: No change suggested or required.
Failu res Not Described? I speak to failures associated with routine operations previously in
this review, as well as chemical spills along the transportation corridor. Also included herein
should be impacts of the Cook Inlet Port and potential spills, accidents, etc., on the marine
ecosystem.
RESPONSE: llecause the number ofpotential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The revised assessment includes more failure
scenarios (e.g., diesel pipeline failure, quantitative water treatment failure, and refined
seepage scenarios) and explains why these particular failure scenarios were chosen.
Discussion of transport and potential for accidents has been included in Chapter 10, but risks
are not evaluated Potential impacts in Cook Inlet are important and should be considered in
the regulatory process should a company submit a permit. However, became this is an
assessment of the Bristol Bay watershed, evaluation of impacts from the port (and
transportation corridor outside the Nushagak and Kvichak River watersheds) is outside the
scope of this assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 103
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Probabilities and Risks of Failures. These seemed reasonably well documented, though again
this falls outside of my expertise. Even though out of my realm, I still would have liked a more
quantitative assessment of these risks, developed in a rigorous, defensible way. I am
discouraged when I understand that history (in the eyes of the mining company) is not a good
predictor of the future because technology has taken us so much farther along, reducing risks of
whatever failure significantly. In my view, this is a specious argument and one that should be
roundly put to bed by the authors of this report. History is indeed the absolute best predictor of
the future and technological changes that have occurred since past mines must be absolutely and
critically evaluated to determine if indeed risks do go down. This is a serious issue and one that
should be addressed with some rigor by the authors.
RESPONSE: Historicalfailures were discussed as supporting background information. The
failure probabilities used in the assessment are based on Alaska's dam classification and
required safety factors applied to the method ofSilva et a I. (2008). The EPA has strengthened
the discussion of failure probabilities (Chapter 9) to the extent that available evidence allows.
Existing Mines as Comparisons. Given my background. I can't answer with any authority,
though the comparisons seemed appropriate, though clearly no extant mines are as large as the
one proposed herein (nor does it incorporate all we have learned in the mining business since the
last mine for which we have data). F.ven so, I again caution the authors that these existing,
historical mines provide the best data we have to estimate risks of failure and that, under no
circumstances, should one accept the indefensible argument that progress with mine safety
(speaking broadly, not just human health) has progressed to the point that these risks, previously
quantified in these older mines, are now small. As suggested above, we have no other
quantitative values for risk, except for existing mines and we cannot simply erroneously lower
the risk based on new, untested technologies.
RESPONSE: We agree that it is appropriate to include in our scenarios technologies that are
in current use and not those that are untested, and thus have done so. No change suggested or
required.
Proportional Losses i>l' Salmon. Is it possible to estimate the proportion of the salmon runs
compromised in the face of major failures in tailing storage facilities or other failures? In other
words, I would recommend adding a chapter that uses best estimates of salmon produced within
the Bristol Bay watershed and then assess the maximum impact of, let's say, a Tailing Storage
Facility failure—with this failure, might we lose 10%, 40%, or 75% of our salmon productivity?
In addition to this estimate, one might estimate the number of stocks or unique genetic units lost
with a major failure? I know these numbers are difficult to get, but if one begins with
escapement from these systems as well as insights from harvest, we may be able to bound these
impacts. Only in this fashion can we put these data into context. This exercise also will serve to
counter the argument by the mining company that they are only destroying some small
percentage of salmon habitat and hence (assuming a linear relationship between habitat lost and
salmon eliminated) only some very small percentage of salmon. Because losing 2% of critical
headwaters habitat may translate to huge losses of salmon (say 20%), one simply cannot assume
a linear relationship between habitat and salmon. Explicitly making this argument improves the
rigor of the main report.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 104
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The fact that salmon productivity cannot be assumed to be linearly related to
habitat has been more explicitly elaborated in the revised assessment. Due to lack of
comprehensive estimates of limiting factors across the impacted watersheds, population level
effects could not be quantitatively estimated, except in the most severe cases where total losses
of runs could be reasonably assumed Our ability to estimate population level effects was
limited to situations that were assumed to completely eliminate habitat productivity and
capacity in an entire watershedfor which estimates of escapement could be inferred For this
assessment, these conditions are only met in the TSFfailure scenario that completely
eliminates or blocks access to suitable habitat in the North Fork Koktuli River. In that case,
we estimate that the entire Koktuli portion of the run (- 2
-------
Dr. Paul Whitney
considered in the mine scenario include four major items: tailings management facility;
pipelines; water collection and treatment; and, road and culvert. These failure modes are
included in the risk assessments for a mine of this type and size, either qualitatively as is
typically done in FMEA, or quantitatively. However, the range of failure modes will also
consider many other types of failures that can also occur during regular mine operations. These
other types of failures may result in the full spectrum of risks (from insignificant to very high)
depending on site and mine life-cycle conditions. Performing extensive analyses for the four
major items implies that they could occur and that they are the only failure modes that will be
significant.
RESPONSE: It is true that any of the failures analyzed could occur, as based on historic and
current knowledge of mining; however, a focus on these failures does not imply that they are
the only failures that will be significant Because the number of potential failures is extremely
large, it is necessary to choose a representative set of failure scenarios. The revised assessment
includes more failure scenarios (e.g., diesel pipeline failure, quantitative water treatment
failure, and refined seepage scenarios) and explains why these particular failure scenarios
were chosen.
Responses to the other parts of the current question (Are the probabilities and risks of failures
estimated appropriately? Is appropriate information from existing mines used to identify and
estimate types and specific failure risks'? If not, which existing mines might be relevant for
estimating potential mining activities in the Bristol I Jay watershed?) are further discussed in my
responses to other questions below.
RESPONSE: No change suggested or required.
Phyllis K. Weber Scannell, Ph.l).
This section focuses on catastrophic failures; however, there are a number of non-catastrophic
failures that can occur at a mine site. Non-catastrophic failures include leakage of contaminated
water to ground or surface waters from PAG waste rock, the tailing storage facility, and exposed
ore surfaces, and from emergency discharge of untreated water from the TSF and ore spills from
trucking accidents. Such failures can be minimized or prevented with good site planning and
monitoring. An additional "failure" has been experienced at a mine in Alaska when the water
elevation of the tailing pond was sufficiently high to cause groundwater flow across a natural
divide into an opposite drainage.
RESPONSE: Because the number of potential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The review draft contained a scenario in
which tailings leachate was not fully contained and reached a stream (Section 6.3 in the May
2012 draft). The revised assessment document includes more failure scenarios (e.g., diesel
pipeline failure, truck accidents, quantitative water treatment failure, and refined seepage
scenarios) in Chapters 8,10 and 11 and explains why these particular failure scenarios were
chosen.
Paul Whitney, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 106
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Sediment Transport. The failure analysis indicates a sediment transport study was beyond the
scope of the assessment. Not only is such a study important for fish resources, it is important for
all ecological resources, especially plant community succession along the stream and a delta into
Bristol Bay.
RESPONSE: EPA agrees that a sediment transport study is important to fully characterize the
watershed and potential impacts from large-scale mining. The scenarios indicate that runoff
eroding sediment from the site would be directed to sediment retention basins. Sediment
runoff from the road is considered in Chapter 10. The development and implementation of a
model of sediment (i e., tailings) transport after the initial tailings outflow would require a
research and development effort that is beyond the scope oj this assessment
This might seem extreme, but the failure analysis indicates the Koktuli, Mulchatna, and
Nushagak Rivers would stabilize into a new channel after a failure and not continue to work their
way across the floodplain and eventually transport materials hundreds of miles down river. The
Mount St. Helens eruption, given as an analogy in the assessment (page 6-3). certainly moved
sediment into the Columbia River Channel, the Columbia River Estuary, and Pacific Ocean over
a hundred miles away. Copper concentrations in the Columbia River estuary as a result of the
eruption ranged from 1 to 43 (ig/L (Lee 1996). The upper limit of the range is approximately 20
times greater than the no-effect benchmark listed in the assessment. The down stream
consequences of changes in sediment transport and water/sediment chemistry for fish and
wildlife are sometimes very large, not anticipated (see Peace Athabasca Delta response below),
and costly to remediate. Solutions for the Peace Athabasca Delta involving check dam
construction may not be directly applicable to a tailings dam failure analysis in the assessment
but may have some value short of dredging. I'm not sure how it would work, but a mitigation
effort (page 4-32, para 3, last sentence) using bulk tailings would apparently be placed down
gradient to catch tailings in the event of a failure. Is this a safety check dam? It is possible that
the assessment could be improved if this and other redundant efforts to minimize risk could be
discussed in more detail and considered in the failure analysis.
RESPONSE: We have clarified this in Chapter 9 of the revised assessment. Our intent was to
describe a channel becoming reestablished in the deposited tailings and slowly reworking the
entire valley bottom, moving the fine material downstream. We agree that tailings would be
carried downstream to Bristol Hay over the long term with adverse impact to waters. The
mitigation addressed in the comment, where a levee was created downstream to capture the
flow of tailings, was not our intent. We described a reclamation scenario where the tailings are
dewatered and sloped to eliminate the chances of mass failure of the impoundment Mitigating
impoundment failure by creating a second dam downstream would greatly increase the mine
footprint on salmon habitat
I agree with the assessment's statement on page 6-11 (6 lines down) that impacts of a tailings
dam failure to fish would extend down the mainstem Koktuli River and possibly further. If the
Mount St. Helens analogy is a good one, the impacts could reach Bristol Bay. Even if the Mount
St Helens analogy is not a good one, I suspect sediments would continue to move down river as
the river(s) moved across their floodplains through time. I also agree with the assessment's
statement that the time to reach dilution approaching background would be very long, as the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 107
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
sediments in tributary rivers to Bristol Bay will be continually reworked (page 6-25) and
resuspended.
RESPONSE: EPA agrees with the commenter. No change suggested or required.
As a terrestrial ecologist, I have always been impressed by the impact of Bennett Dam on
sediment transport in the Peace Athabasca Delta ecosystem over 600 miles away (Cordes 1975).
The assessment does mention the potential impact of toxics on the likelihood of plant community
succession on deposited tailings (page 6-10), but this causal pathway was not assessed. Given the
increases in metal concentrations in the Columbia River Estuary cited by Lee (1996) for the
Mount St. Helens eruption, an assessment of down stream sedimentation and changes in
sediment chemistry should be addressed. The likelihood that far reaching impacts of a failure
could influence plant succession and wildlife habitat quality is probably not anticipated but given
the Mount St. Helens analogy and the lessons learned from Bennett Dam, the likelihood of such
impacts deserve more attention.
RESPONSE: This continent recommends expanding the scope of the assessment to include
direct effects on terrestrial plants and wildlife habitat. The scope was set during the planning
and problem formulation processes to encompass the expressed concerns of the Alaska Native
organizations that asked the Agency to address potential mining in the Bristol Bay watershed,
as well as the needs of the Agency's decision makers. This scoping process is in keeping with
the Guidelines for Ecological Risk Assessment (USEPA 1998). The commenter may be more
familiar with Environmental Impact Statements, which are full disclosure documents and
therefore are more broadly focused
PHWResponse: The reasons for not including direct impacts on wildlife given here do not align
with statements provided by natives that impacts on wildlife are more important than impacts on
salmon (page 5-77 in first draft). The reason(s) for not including wildlife should be consistent. I
can't help but wonder if the Alaska Native organizations are not interested in direct impacts to
wildlife. While living in Alaska I recall that a variety of birds and mammals and their parts were
and still are important for nutrition, warmth, adornment and trade. For example ermine and short
tailed weasels are used in a variety of ways on ceremonial dress.
Sediment Benchmarks. Once again, as a terrestrial ecologist with ecological risk assessment
experience, I know that sediment chemistry and determining toxic benchmarks for sediments is
very complex and subject to varied opinions. I admire the MacDonald et al. (2000) effort to
reach consensus on sediment benchmarks but I have three concerns. First, the consensus values
listed in MacDonald et al. (2000) are geometric means of values from several sources. The mean
consensus values likely do not equate to No Observed Effect Concentrations, which would
probably be lower than the mean values. Second, the lack of observed effects was sometimes for
a "majority" of sediment dwelling organisms, but not all (MacDonald et al. 2000, Table 1).
Third, some of the sources used for the mean values included interstitial water, but apparently
not all (Table 1). I will always remember Dr. John Stein (currently the acting director of
NOAA's Northwest Fisheries Science Center in Seattle) standing up at a workshop for the
Columbia River Channel Deepening Project. He had a small bottle of sediment and water in his
hand and, while shaking it, he said something to this effect: It's the pore water we are interested
in. Considering that a proposed mine, at some point, will be reviewed by NOAA, it seems
appropriate to consult with NOAA regarding benchmarks for all the species of sediment/pore
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 108
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
water-dwelling organisms likely to occur in the potentially effected watersheds addressed in the
assessment.
RESPONSE: The EPA agrees that pore water concentrations are generally useful predictors
of toxic effects of sediment, and sediment pore water toxicity is addressed in the assessment.
However, the MacDonald et al (2000) benchmarks are also a useful line of evidence. NOAA
has been consulted during the development of this assessment No change required
Question 6. Does the assessment appropriately characterize risks to salmonidfish due
to a potential failure of water and leachate collection and treatment from the mine site?
If not, what suggestions do you have for improving this part of the assessment? Are
significant literature, reports, or data not referenced that, would be useful to
characterize these risks, and if so what are tiny?
David A. Atkins, M.S.
Water treatment failures of varying scale occur at virtually every site that treats water, and mine
sites are no exception. The risk of failure of water treatment described in the assessment is useful
as background, but as the report states, the risk is highly uncertain. A non-catastrophic water
treatment system failure is fairly likely to occur at some point during the mine life, and, hence,
requires a detailed assessment. The treatment in the Assessment is cursory (less than one page).
This type of failure is much more likely than a TSF failure (which receives more than 20 pages
of analysis), and therefore requires a much more thorough treatment given the probability of
occurrence and likelihood of impact to salmon species.
RESPONSE: The wastewater treatment failure scenario has been expanded and is now
detailed and quantified in Chapter 8 of the revised assessment. Additionally, the non-
catastrophic failures of seepage collection from the TSE and waste rock piles also have been
included as scenarios with new and refined analyses.
The background water chemistry indicates mineral concentrations are very low. Therefore, water
treatment will be challenging if background conditions are to be met. Treatment will be
especially challenging given the sensitivity of the species of concern to concentrations of copper,
for instance, as well as the sensitivity to temperature that may be difficult to match in a water
treatment system.
RESPONSE: A mine would not necessarily have to match the low background concentrations,
but would at minimum be held to the water quality standards that protect the existing and
designated uses of the waterbodies where a discharge might occur. In the Bristol Bay area, all
waterbodies are protectedfor all uses and permit limitations would be derived to be protective
of the most stringent applicable Alaska Water Quality Standards (WQS). In addition, we
believe that compliance with the more recent Federal Water Quality Criterion for copper
would be necessary to protect aquatic life.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 109
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
During mine operation, a lapse in treatment would likely be identified and addressed quickly.
This type of treatment failure is ephemeral and would likely have a short-term impact on the
fishery, depending on the time of year of occurrence. It is likely that any impacts to the fishery
could recover in subsequent years after the problem is fixed. The site will require water treatment
long after closure, possibly in perpetuity. This period is more problematic, as a water treatment
failure could go unnoticed for some time or the resources may not be available to correct it
quickly, depending on how long after closure the failure occurs and the stewardship of the
treatment system.
RESPONSE: A failure during operation would be corrected quickly only if it did not require
extensive repairs or the manufacture and import of replacement equipment. The EPA agrees
that water collection and treatment failures after mine closure would be less likely to be
corrected in a timely manner. In addition, events after closure, such as filling of the pit, would
affect water quantities and qualities in ways that would aff ect treatment success. A discussion
of this issue has been added to Chapter 8 of the revised assessment. . I discussion offinancial
assurance has been added in Box 4-3.
Steve Buckley, M.S., CPG
Less than a page (4-39) is devoted to the failure of water and leachate collection and treatment.
This seems inadequate given it would be one of the main systems that could impact fish at the
potential mine site. In contrast, 20 pages are devoted to tailings dam failure (p.4-39 to 4-60).
RESPO.XSIC: The wastewater treatment failure scenario has been expanded and is now
detailed and quantified in Chapter 8 of the revised assessment.
Courtney Carothers, Ph.D.
The report concludes that wastewater and leachate treatment and collection failures could expose
local streams to mildly to highly toxic water harmful to invertebrates and fish species.
Depending on the type of failures, these exposures could last from a period of hours to years.
The report notes that in the case of Red Dog Mine, Alaska, the water treatment system was
inadequately designed, but does not discuss why such a design was approved and allowed to be
implemented, nor does it discuss the likelihood of replicating such a design flaw in future mining
scenarios.
RESPONSE: The Red Dog Mine's treatment system was inadequately designed because the
amount of water to be treated was underestimated Mine site hydrology is often difficult to
predict and treatment systems may fail despite the intent of mining companies and regulators.
Dennis D. Dauble, Ph.D.
More information on local hydrology, including seasonal runoff patterns (e.g., peak flows) and
groundwater movement would be useful. I found no description of existing water quality
characteristics of potential receiving waters, except what is included in Table 5-17 of the main
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 110
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
report. Are these values (such as hardness, which moderates metal toxicity) consistent
throughout the watersheds, including downstream lakes? Other questions include: What
volumes of leachates might be collected and treated versus volumes not captured and
subsequently released downstream? Is copper the only constituent of concern to aquatic animals?
Are there processing chemicals that would also be toxic?
RESPONSE: Monthly flow patterns for area streams are now presented in Chapter 7. The
three streams described in Table 7-17 are the three potential receiving waters for any site
effluents. The water balance, including leachate volumes, is now described in the assessment
(Chapter 6). Copper is the primary contaminant of concern. Others are described in the new
Section 6.4.2.3 and discussed in Chapters 8 and 11.
The assessment should also consider and discuss relative risk to aquatic ecosystems from
downstream transport of sediment-bound metals to Iliamna I ,ake. if deemed probable.
RESPONSE: Although metals in aqueous emissions would partition to sediment, and the
sediment would mobilize during high flows and eventually reach the lake, this route is not
judged to be significant. Toxicity is caused by dissolved metals, and concentrations from
release of metals from transported sediment to lake water are likely to be minor.
Gordon H. Reeves, Ph.D.
The report focused primarily on the lethal effects of the contamination from leachate and water
treatment and collection failures. However, could there be ecological consequences to fish and
invertebrates that are not directly lethal but that could have ecological consequences over the
long term? I suggest that this needs to be considered more fully in this assessment.
RESPOXSE: The report addresses both acute lethal effects and chronic lethal and nonlethal
toxic effects to both fish and the invertebrates on which they feed. It is highlighted in the text
when estimated concentrations are sufficient to cause acute lethality. Such effects are
important not only because tliey are severe but also because they could occur during episodic
exposures.
Sockeye salmon are most abundant salmon in Bristol Bay and a primary species of focus in this
analysis. The direct impacts of mine and mine-related activities have been considered but there
appears to be a lack of consideration of the impact on zooplankton, the food source for sockeye.
If this were a deliberate omission, then a statement about why it was omitted is required. The
revision should include this if it was an oversight.
RESPONSE: Any effluents would be released to the streams draining the site. In those
ecosystems, zooplankton are rare and the primary food organisms for fish are benthic
invertebrates. However, the toxicological data used to assess effects on invertebrates are
dominated by planktonic crustaceans. Therefore, if toxic concentrations of metals reach
Iliamna or other lakes, the toxicity assessment would actually be more relevant to the plankton
that occur there than to the insects in the receiving streams. Hence, the numerous references
to effects on aquatic invertebrates are relevant to zooplankton and no new analyses are
needed
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 111
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Charles Wesley Slaughter, Ph.D.
No. Text suggests that a monitoring well field downslope from the TSF (and presumably from
all hypothetical TSFs) would detect seepage; such seepage would then be intercepted and either
returned to the TSF or "treated and released to the stream channel." Either action presupposes
adequacy of monitoring seepage and subsurface flow (both spatially and temporally); returning
such water to the stream further presupposes fully adequate treatment to meet both regulatory
and aquatic biota requirements for water quality and flow regime.
RESPONSE: The water treatment and leachate capture discussions have been expanded and
are now detailed and quantified in Chapter 8 of the revised assessment.
Assumptions are very generalized and optimistic: "assuming no water collection and treatment
failures" and "excess captured water would be treated... and discharged to nearby streams..." -
this assumes both "no failures" over the life of the operation, and that such treated "excess
captured water" could be successfully treated before release to fully meet both regulatory water
quality criteria and the possibly more sensitive biological requirements of individual
invertebrates and fish stocks (Appendices A & B).
RESPONSE: Water management (mitigation) measures are more clearly described and
discussed in Section 6.1.2.5 of the revised assessment, and in sub-sections for the mine
components in the scenarios. Our intention for the "no-failure" scenario was to identify and
evaluate the unavoidable environmental effects if all systems and mitigation measures
operated perfectly, and to separate those effects from a scenario where systems periodically
failed However, the "no failure" scenario is no longer included The purpose of the
assessment is to describe the potential adverse environmental effects that could exist even with
appropriate and e//ective site mitigation measures, t he assessment is not intended to duplicate
or replace a regulatory process, which is where required permit discharge limits for water
quality would he determined.
John I). Sh'diiiili, Ph.D.
The TSF is designed to hold the tailings under water to minimize the oxidation of pyritic
materials and limit ARD or AMD production. The TSF will be underlain by hypalon to capture
leakage waters. There is the possibility of failure to collect waters from the TSF—either surface
runoff or leakage with or without storm (precipitation) events. There is also the possibility of
failure of the treatment plant to treat the wastewater. Such treatment systems in Colorado usually
have a bypass pond to temporarily hold waters for later pump back and treatment as a result of
power failure, plant going off-line, storm events, or plant maintenance.
RESPONSE: The original draft assessment contained a scenario in which tailings leachate
was not fully contained and reached a stream (Section 6.3 in the May 2012 draft); however,
this has been refined with new data in the revision for seepage from the TSF and waste rock
piles that escapes capture from the mitigation measures (Chapter 8). Additionally, the scenario
presents a suggested treatment option for mining influenced water and settling ponds for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 112
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
stormwater runoff. A wastewater treatment plant failure scenario in the revised assessment
assumes emergency storage capacity has been exceeded or the bypass system fails.
The waters in the study area have very low buffering capacity; metal toxicity would occur at low
concentrations and dilution of metals would require time and space. The maximum index counts
on page 6-39 are confusing and not well related to the risk characterization. Copper was used as
an example metal, but other metals are also toxic and further characterization of the waste rock
can be presented. Further analysis of a water and leachate collection failure can be made over
time: the effects of dilution flows over the various months with low flows, or when adult salmon
are present in the stream as opposed to juveniles - or, when juveniles are emerging. The toxicity
quantification is difficult and appears more of an academic exercise here, rather than site
specific.
RESPONSE: Undiluted leachate concentrations are used to calculate the hazard quotients (we
assume that is what the reviewer means by "maximum index counts "), to screen the
constituents for contaminants of concern, and because the State of Alaska does not allow
mixing zones in anadromous streams. However, dilution was considered in the failure
scenarios for waste rock leachates and tailings leacluites. Copper was emphasized because it is
by far the most toxic metal relative to concentrations in rock, tailings and concentrate
leachates. Site specific water chemistry was used to estimate the toxicological benchmarks for
copper andfor metals with hardness-dependent toxicity. More detailed analyses of dilution are
included in Chapter 8 of the revised assessment.
Leachate collection from the tailings area is only briefly described. What are the State of Alaska
standards for collection and treatment? What are the potential effects of not collecting or treating
the tailings leachate waters? Compare the detail and length of leachate discussion to the TSF
failure discussion (see earlier comments).
RESPONSE: There are no specific A K State standards for collection and treatment of tailings
leachate, but any discharge to waters of the United States requires a Clean Water Act Section
402 (CM. I$ 402) permit In Alaska, the Department of Environmental Conservation issues
these permits under the Alaska Pollutant Discharge Elimination System program Such
permits would contain effluent limitations that are protective of the State WQS. The original
draft assessment contained a scenario in which tailings leachate was not fully contained and
reached a stream (Section 6.3 of the May 2012 draft): however, this has been refined with new
data in the revision (Chapter 8) for seepage from the TSF and waste rock piles that escapes
capture from the mitigation measures.
Given the hydrologic connection between surface and groundwaters, what effect will
interception of all waters on the TSF do to the surrounding wetlands and groundwater levels?
Again the lack of a water balance does not let the reader determine if this water interception is
significant or will have significant resource effects.
RESPONSE: The water balance presented in the revised assessment quantifies the amount of
water captured at each TSF under each scenario. The assessment also presents the percentage
change in streamflow at existing gages in the North Fork Koktuli, the South Fork Koktuli,
and the Upper Talarik.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 113
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The water balance presented in the assessment looks at the overall amount of water dedicated
to consumptive uses, which therefore would no longer be available to contribute to
groundwater recharge or stream flow. Furthermore, the assessment attempts to quantify the
annual amount of and percentage increase or decrease in streamflow for individual stream
reaches. These stream flow results are presented in Chapter 7 of the revised assessment
Our analysis does not attempt to explicitly calculate groundwater levels or drawdown except
within the cone of depression around the mine pit. Our analysis assumes that all precipitation
that falls outside the actual mine footprint (i. e., the physically disturbed areas) would continue
to contribute to groundwater recharge and stream flow, although the exact flow patterns
would change due to collection of surface runoff and stream diversions. Areas within the
footprint would most likely see some decrease in the groundwater levels due to the collection,
management, and possible treatment of surface runoff with some potential increases
downstream of the points of water release from the wastewater treatment plant. Overall, the
cumulative reduction in streamflow would be approximately equal to the amount of water
retained on the site as tailings pore water.
Most, if not all of these failures are the result of human error. What safeguards will be in place?
What are the best mining practices to minimize human error?
RESPONSE: The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine (they are a subset of options
presented in Appendix I), all of which were presented as appropriate for the Pebble deposit in
Ghaffari et al. (2011). He assume that these types of measures would be applied throughout a
mine as it is constructed, operated, closed, and maintained through post-closure. Because the
possibility for human error is inherent in any activity in which a human is involved, accidents
andfailures happen. Human error-caused accidents and failures are minimized when the
humans involved follow careful and responsible management of the mining site. No change
suggested or required.
Roy A. Stein, Ph.D.
Groundwater Connectivity and TSF Construction I. Kxlensive connectivity between surface
water and groundwater means thai any failure will allow contaminated waters to flow quickly to
areas of importance for salmon. And, indeed how does one build a tailings pond (coarse textured
glacial drift in the Pebble Mine area) with this much permeability? Why would one only line the
tailings dam; shouldn't the entire Tailings Storage Facility be lined? Would not this be "Best
Practice"?
RESPONSE: Our estimates of the expected leakage from the full TSFs range from about 2 to
6 m3/min. If a mine at the Pebble deposit goes forward, the design of the TSFs should include
a more thorough flow analysis that would calculate the expected rate of flow and associated
flow paths from the TSFs. If the calculated leakage rates were unsatisfactory from an
environmental, operational, or economic perspective, the designer could incorporate other
design elements (e.g., a liner) to reduce the expected leakage rate. Full liners beneath TSFs
are not always used; however, there is a growing requirement to use liners to minimize risks of
groundwater contamination, with new mines in Australia being required to justify why one
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 114
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
wouldn't be required (Commonwealth of Australia 2007). Liners are not required in the US.
Whether something is "best" depends on the specifics of the site. The mitigation measures
proposed within the mine scenarios are those that could reasonably be expected to be proposed
for a real mine (they are a subset of options presented in Appendix I), all of which were
presented as appropriate for the Pebble deposit in Ghaffari et aL (2011). Evaluation of
different mitigation measures to determine if lining the facility would be "best" for this site
would be part of the regulatory process and thus is outside the scope of this assessment
Groundwater Connectivity and TSF Construction II: A water-impermeable barrier will be
installed only on the interior dam face and nowhere else. To prevent communication between
these facilities and the groundwater, is it feasible to map groundwater inputs before the facility is
filled, place barriers over these areas, and thereby reduce influx of groundwater into the facility
and perhaps prevent movement of toxic water into the groundwater? I make this point with some
hesitation, given the point made on page 5-29:
"Projecting specific mining-associated changes to groundwater and surface water
interactions in the mine area is not feasible at this time."
RESPONSE: See response to previous comment.
Failure of Leachate and Water Collections. See my comments under Question 12
below. In addition, I see this as a huge undertaking for which monitoring and response
(mitigation) are clearly as important as the actual plan to capture these wastes.
RESPONSE: EPA agrees with this comment. \ o change suggested or required
William A. Stubbhiifld, Ph.D.
The risk assessment attempts to consider the ell eels of metal discharges for water and leachate
from ihe mine site. This assessment is based on melals concentrations measured in potentially
"similar"' mine waters from other sites; concentrations of metals are likely to differ based on
source material and operational differences. The effects concentrations used in the evaluation
are based on US EPA ambient water quality criteria (AWQC) for metals, and this approach is
appropriate for "screening level" evaluations. It should be noted that exceedence of an AWQC
does not portend the occurrence of adverse effects. Ambient water quality criteria are derived in
such a way that they are intended to represent "safe concentrations." In other words, if
environmental concentrations remain below the AWQC, it is assumed that unacceptable adverse
effects will not occur; exceedence of an AWQC suggests that adverse effects may occur to some
species, but that this must be evaluated more closely. Salmonid species are not the most
sensitive organisms in the copper AWQC species sensitivity distribution (SSD); therefore, direct
effects on salmon are even less likely at concentrations in the range of the AWQC.
RESPONSE: Effluents or ambient waters from mines at other sites were not used The
leachate concentrations used (except for the product concentrate leachate) are from available
results of material leaching tests from the Pebble deposit. Otherwise, the Agency agrees with
these comments. The assessment used criteria for screening, but then examines the toxicity
data more closely, including field data to determine potential effects (e.g., aversion, sensory
inhibition, mortality and reduced reproductive success of salmonidfish). The greater
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 115
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
sensitivity of aquatic insects was described in the May 2012 draft of the assessment and is
further highlighted in the revised assessment. The protectiveness of the copper criterion is
considered in both the original and revised assessments.
It is interesting to note that the risk assessment document states that copper is one of the "best-
supported criteria. However, it is always possible that it would not be protective in particular
cases due to unstudied conditions or responses." Further, the document goes on to suggest that
organisms such as mayflies etc. are important to the aquatic ecosystem but are not considered in
the copper AWQC and therefore may not be sufficiently protective. It also suggests that because
an acute-chronic ratio approach is employed to correct the final acute value to obtain a final
chronic value, there may be increased uncertainty associated with the protectiveness of the
chronic criterion. This appears to be an area where F.PA might benefit by conducting research
(either alone or in concert with industry) to reduce uncertainty in the criteria to an acceptable
level. In addition, additional chronic toxicity data may be available from research conducted in
response to the European REACH regulations and consideration of this research may reduce the
level of uncertainty in the criteria. Bioavailability correction via the BLM approach is only
considered for copper in the risk assessment; biotic ligand models have been developed for a
number of metals (e.g., zinc, nickel) and these should be considered in the assessment as well.
Finally, the assessment approach seems to use a sum TIJ -based approach for assessing "metals
mixture" impacts. This is based on an assumption of additive interactions among the metals.
Although this is probably the best assumption in going forward, limited data are available to
support this approach.
RESPONSE: The EPA has examined the El 's 2008 I oluntary Risk Assessment of Copper
(the relevant REACH document). Although the authors could derive a chronic species
sensitivity distribution, that is because the way that they include and aggregate data differs
from the EPA 's method. They apparently did not generate new test data for the assessment. In
particular, they have no data for sensitive aquatic insects, so the EU assessment does not
resolve that problem The HIM was used for copper because it is the contaminant of greatest
concern and because the copper HLM has been approved by the EPA Office of Water. Other
metals with HIMs, such as zinc and nickel, occur at lower levels in leachates relative to their
toxicities, so HIM modeling was not justified
Areas where additional research would be beneficial include:
• Mixtures: Information regarding the potential interactive effects of multiple metal
exposures would be useful and would reduce assessment uncertainty.
• Species sensitivity concerns: there is extremely limited data (esp. chronic datat) on all of
the salmon species of concern in Bristol Bay
• Additional data, especially chronic toxicity data and data for additional metals for which
no water quality criteria exists, would be extremely helpful.
RESPONSE: The EPA agrees that these are good research topics. No change is suggested or
required
Dirk van Zyl, Ph.D., P.E.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 116
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to a potential failure of water and leachate collection and treatment from the mine site. It
only estimates the likelihoods of occurrence and the consequences. See discussion under
Question 4 above regarding suggestions for improving estimation and expression of the
magnitude of risks to salmonid fish due to potential failure of water and leachate collection and
treatment from the mine site.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows. We have added to the discussion of leachate
contamination, including estimation of the magnitude oj leachate escaping the collection
system and entering surface waters (Chapter 8).
Water collection and treatment failure likelihood. An estimate is presented for the amount of
seepage that may flow from the TSF. Similar estimates are not presented for the waste rock piles.
The effects of the effective exclusion of oxygen from the saturated or partially saturated tailings
should be considered in developing an estimate of the water quality of the resulting seepage. It
may be an important factor in reducing the oxidation of sulfide minerals remaining in the interior
of the TSF. This same effect could also mitigate the release of poor water quality in the long-
term following closure. The precipitation on the site may be sufficient to effectively retain a
suction saturated profile in parts of the TSF.
RESPO.XSIC: The revised assessment presents an estimate of the amount of seepage from the
waste rock piles (Table 6-3). The water quality of the leachate from the TSF was modeled with
the concentrations reported from the tailings humidity cell tests. The reported copper
concentration of0.00533 mg/l is lower than some of the reported groundwater concentrations,
and is the same order of magnitude as the highest reported stream concentration of 0.0013
mg/l The TSl at closure was modeled with free standing water at the surface and with
downward flow with an estinuited gradient between 0.07 and 0.12.
Water collection and treatment failure consequences. Water collection and treatment is being
done at a number of mines in North America. Past experience at the Red Dog mine is quoted;
however, there are many other examples that could have been examined. An important example
is that of the Equity Silver Mine in British Columbia (Aziz and Meints, 2012): "Acid Rock
Drainage (ARD) was discovered at the Equity Silver Mine in the interior of British Columbia in
1981. The latest water treatment plant was installed in 2003, 9 years after the mine closed in
1994, and is the fourth successive treatment plant for the site that has treated ARD for a period of
over 30 years. Discharge water quality was maintained since 1991 except during two high flows
associated with freshet conditions in 1997 and 2002. ARD collection and treatment system
upgrades were installed after 2002 and these have performed well through three large freshet
conditions in 2007, 2011 and 2012. The timeframe for treatment is perpetuity and financial
assurance is in place for a total amount of $56,291 million through a long-term security bond
(letter of credit) with the BC Provincial Regulatory Authority. The security bond is reviewed by
stakeholders every 5 year". Collection and treatment at Equity Silver indicates that companies
are committing to long-term water treatment of ARD and that regulatory frameworks are in place
to protect water quality in downstream streams and rivers. It is recommended that EPA perform a
more thorough review of other sites where water treatment occurs to better characterize this
failure mode.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 117
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We have been unable to obtain a copy of the cited report. The 2010
Sustainability Report, which is available, does not contain these specifics. This mine
apparently has begun perpetual water treatment. A comprehensive review of mine sites with
water treatment systems is beyond the scope of this assessment
Phyllis K. Weber Scannell, Ph.D.
This section of the report provides an in-depth discussion of possible sources and fates of
contaminated water. Chapter 6.3 discusses possible adverse effects from early mine closure or
prematurely shutting down a water treatment system. These issues highlight the need for a mine
plan that includes concurrent reclamation, sufficient bonding to conduct reclamation in the event
of an early shut down, and plans and specifications for collection and bypass of clean water and
collection and diversion to a water treatment system of contaminated water.
RESPONSE: The EPA agrees with this comment No change suggested or required
The Risk Characterization (Section 6.3.3) discusses possible contaminant loads to downstream
waters. As stated in this section, it "serves to indicate the large potential risk from improperly
managed waste rock leachate." This statement highlights the need for an in-depth mine plan
with sufficient monitoring and fail-safe provisions. An emergency discharge of untreated waters
from a tailings storage facility could be made to a collection pond for later treatment or the
tailings pond could be engineered to accommodate a higher flood event so the likelihood of
overtopping is minimized.
RESPONSE: The waste rock leachate scenario referenced by the commenter has been
eliminated and replaced with a more detailed and realistic scenario for waste rock leachate
collection and treatment (Chapter S).
Section 6.3.3 (Risk Characterization) states "Alternatively, water collection and treatment failure
could be a result of an inadequately designed water treatment system which could result in the
release of inadequately treated water as at the Red Dog Mine, Alaska (Ott and Scannell 1994,
USEPA 1998, 2008). In that case, the failure could continue for years until a new or upgraded
treatment system is designed and constructed." This statement is misleading and overly
simplistic; the water treatment system at the Red Dog Mine was designed to treat the predicted
flows. However, the stream bypass and collection systems were constructed in 1991 to intercept
seepage waters. The additional water that was collected and treated dictated construction of a
second water treatment system in 1992. Sand filters were added in 1993 to remove fine
particulate Zn. The issue was not that the water treatment was inadequate, but that the pre-
mining hydrologic data was insufficient and that state, federal, and mine officials lacked
experience in mine construction on permafrost soils.
RESPONSE: The water treatment failure at Red Dog was, as the commenter describes, the
result of an unintendedfailure to design a plant that was adequate for the mine and site. The
passage has been expanded to clarify the nature of the failure.
Overall, the discussions of risks to salmonid fish due to a potential failure of water and leachate
collection and treatment from the mine site highlight the need for more comprehensive
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 118
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
information on groundwater, including delineating flow pathways, depth to surface, and water
volumes. Additional information is needed on water collection, storage, and treatment at future
mine facilities.
RESPONSE: Discussions of wastewater collection and treatment have been considerably
expanded in Chapter 8.
Paul Whitney, Ph.D.
No comment on this question.
Question 7. Does the assessment appropriately characteri:,e risks to salmonidfish due
to culvert failures along the transportation corridor? If not, what suggestions do you
have for improving this part of the assessment? Are significant literature, reports, or
data not referenced that would be useful to characterize these risks, and if so what are
they?
David A. Atkins, M.S.
The description of culvert failure is necessarily general because there are currently no designs.
The general data on culvert failures presented lor the types of culverts described in the references
cited (principally for forest and range land) indicate a high probability of failure (30-66% failure
rate). It is probable, however, that the transportation corridor for the project would be
constructed to a higher standard than most of the roads included in these papers. It would be
helpful to know if similar data are available for highly engineered roads of the type likely to be
built lor the project.
RESJ'< > \ SIC: The assessment assumes modern mining technology and operations. We did not
find information explicitly from highly engineered roads, but to the extent possible we used
recent literature from representative environments. The failure frequencies cited in the revised
assessment are from modern roads and are not restricted to forest or rangeland roads.
Information on current design standards is now included in text boxes throughout Chapter 10.
Steve Buckley, M.S., CPG
The references provided in this section emphasize culvert failures in the Pacific Northwest and
Tongass National Forest. The streams and culverts in these regions are heavily influenced by
large woody debris loading. It would be more appropriate to classify the various potential stream
crossings by watershed and the amount of large woody debris available to be recruited to the
stream and influence culvert blockage.
RESPONSE: Many of the references in the assessment relate to the Pacific Northwest,
because much work on culverts and potential impacts on salmon have been performed there.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 119
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Flanders et til. (2000) has been deleted in the revised assessment. The failure frequencies cited
in the revised assessment are from the best available literature concerning modern roads, and
are not restricted to forest roads.
Courtney Carothers, Ph.D.
Culvert failures due to blockage and erosion are noted to be common and are likely to occur in
this scenario. Culvert failures would prevent the movement of fish, which could eliminate a year
class from blocked stream systems and fragment upstream and downstream populations,
increasing likelihood of localized population depletions and extinctions. Monitoring and
maintenance of culverts can be expected to decrease after mine operation, increasing the risks of
these failures. The report appears to appropriately characterize risks to salmonid fish due to
culvert failures along the transportation corridor, although I have no particular expertise with
which to evaluate this assessment.
RESPONSE: No change suggested or required.
Dennis D. Dauble, Ph.D.
Mitigation practices, such as new culvert design, was well described, as was bridging of
roadways and porous fills to mitigate risks due to culvert failure along the transportation
corridor. This assessment also included appropriate risk characterization for both the no-failure
and failure scenarios. There should be literature available from the Washington State Department
of Transportation on fish passage relative to culvert placement and design. Otherwise, I have no
suggestions for improvement.
RESPONSE: We have examined literature on fish passage relative to culverts from the
Washington State Department of Transportation. Culvert failure frequencies from their 2012
paper, in which approximately 62%of culverts were identified as total or partial barriers, were
not used in the assessment because we could not determine the age of examined roads.
Gordon H. Reeves, Ph.D.
The literature review of culverts and their potential impacts on fish and fish habitat is very
thorough and the presentation of results is accurate. However, most of the cited material is from
studies done in areas outside of Bristol Bay and the direct applicability of results is problematic.
This should be done in the revision.
RESPONSE: Because the proposed mining would take place in an undeveloped area, much of
the literature is necessarily from areas outside of Bristol Bay. However, to the extent possible
we used examples from representative environments, and applied the results to the proposed
mine.
I think that there are potential mitigation measures that were not presented. The primary one,
besides the use of bridges, as suggested in the report, is making all culverts be arch culverts.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 120
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
These culverts make use of the stream bottom, which reduces the potential for the culverts to
become perched and impede upstream movement, and are less likely to change the gradient than
other culvert types. All culverts could, as recommended in the report, be at least one bank width,
which is larger than required by the state of Alaska. This would minimize the possibility of
plugging with debris.
RESPONSE: A number of culvert types (e.g., arch culverts) may be used on the proposed
transportation corridor; it is not in EPA's purview to suggest which ones should be used
However, culvert design approaches specified in the Memorandum of Agreement between the
Alaska Department of Fish and Game and the Alaska Department of Transportation and
Public Facilities (ADF&G andADOT 2001) are described in litix 10-2 of the revised
assessment.
The review of potential road impacts lacked two possible consequences. One is that roads could
be corridors for the introduction of invasive species, plants and animals. The consequences of
the successful establishment of non-native species could have critical ecological impacts on
native species and the ecosystem. The second consequence is that a road will allow greater
access to streams where access was previously limited. Fish populations could be more easily
and intensively harvested in sport and subsistence fisheries, which adds additional stresses to the
populations. Lee et al. (1997. Assessment of the condition of aquatic ecosystems in the Interior
Columbia River basin. Chapter 4. Eastside Ecosystem Management Project. PNW-GTR-405.
Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research
Station.) found a direct relation between road access and the status of salmonid populations in
the Columbia Basin.
RESPONSE: We agree with the commenter that roads could be corridors for the introduction
of invasive species. The potential impacts of invasive species as a result of construction and
operation of the proposed transportation corridor are discussed in Section 10.3.6 in the revised
assessment. With respect to stream access, EPA assumes that the proposed road would be
closed to the public during mining operations but potentially could become a public road after
mining operations cease. If that were the case, there would be greater access to streams and
fish populations. However, the potentially important impact to fish would likely occur from
secondary (induced) development, or development resulting from the introduction of industry,
roads and infrastructure associated with mining. This is briefly discussed in Section 13.3 in
the revised assessment. Improved accessibility would increase hunting andfishing pressure, as
well as competition with existing subsistence users.
Charles Wesley Slaughter, Ph.D.
No. The Assessment does not adequately address the road/stream crossing/culvert issue. Given
the projected transportation corridor, Pebble locale to Cook Inlet, and the inevitability of a
further network of "minor" roads in the mine and TSF locale, plus additional infrastructure
linkages, road/culvert/stream crossings are a major concern for aquatic habitat and fisheries.
Readers of the Assessment should be directed to Frissell and Shaftel's Appendix G for a more
comprehensive discussion of this important topic.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 121
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The revised assessment addresses the road/stream crossing/culvert issue in
detail. Secondary development is now described qualitatively in Chapter 13 of the revised
assessment. The reader is repeatedly directed to Appendix G for more details.
The specific consequences of a failure on salmonid habitat and biology are portrayed well.
RESPONSE: No changes suggested or required
John D. Stednick, Ph.D.
No. It is unclear how the estimate that 50% of the culverts would fail was obtained, given that
the literature shows a range of 30 to 60% (Section 4.4.4). What literature was used? Are road
BMPs satisfactory in this environment? Have the Alaska BJVLPs been audited? Culvert repair
taking a week to several repairs in a month seems high. If the road crosses a critical salmon
rearing stream, conservative pipe sizing or bridgework could be considered. The direct loss or
inaccessibility of upstream salmon habitat does not necessarily translate to salmon loss. Timing
of culvert blocking event with salmon migration and duration of blockage should be considered.
Need to include references to Alaska Department of Natural Resources and Alaska Department
of Transportation.
RESPONSE. The literature showed a range of 30 to 66%, with an average of roughly 50%.
The literature used was noted in Section 10.3.2.1. However, Flanders et aL 2000 was deleted in
the revised assessment, bringing the average culvert failure frequency to 47%>.
Best management practices (HMPs) or mitigation measures would be used to minimize
potential impacts to salmon ecosystems from construction and operation of the proposed
transportation corridor. These HMPs, and their likely eff ectiveness, are now discussed in text
boxes throughout Chapter 10 (e.g.. Box 10-2). Environmental characteristics along the
transportation corridor would likely render the effectiveness of standard or even "state of the
art" mitigation measures highly uncertain. Further discussion on this is contained in Box 10-
5 of the revised Assessment
We are not aware of any audits oj Ilaska HMPs.
The text rehiring to culvert repair considers that the proposed mining would occur in an often
harsh, remote environment.
Design considerations (including sizing) for culverts are now discussed in Box 10-2. Culverts
would be designed in accordance with guidance in Alaska Highway Drainage Manual (ADOT
1995) and the Memorandum of Agreement between the Alaska Department of Fish and Game
and the Alaska Department of Transportation and Public Facilities (ADF&G and ADOT
2001). Both of these documents are cited within Box 10-2.
Blockage of a culvert by debris or downstream erosion would prevent the in-and-out migration
of salmon and the movement of other fish among seasonal habitats. The direct loss or
inaccessibility of upstream salmon habitat does not necessarily translate to loss of a
population of salmon, but production would ultimately be reduced If blockage of a culvert by
debris or downstream erosion occurred during in-migration of salmon and persistedfor
several days, it would result in the loss of spawning and rearing habitat If it occurred during
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 122
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
out-migration and persistedfor several days, it could cause the loss of a year class of salmon
from a stream
What are the design considerations for the culverts? What precipitation/streamflow relationships
will be used for sizing purposes? What are the usual casual mechanisms for culvert failure? How
much woody material do these streams carry? Do culverts fail from debris plugging, road
slumps, or overtopping by storm events? What road BMPs will be implemented?
RESPONSE: Design considerations (including sizing) for culverts are now discussed in Box
10-2. Culverts would be designed in accordance with guidance in Alaska Highway Drainage
Manual (ADOT1995) and the Memorandum of Agreement between the Alaska Department of
Fish and Game and the Alaska Department of Transportation and Public Facilities (ADF&G
and ADOT 2001).
As noted in Section 10.3.2, culverts are deemed to have Jailed if culverts (and thus fish
passage) are blocked (e.g., by debris, ice, or beaver activity) or if stream flow exceeds culvert
capacity, resulting in overtopping and potential road washout. The causal mechanisms for
culvert failure are briefly discussed in Section 10.3.2.
We do not have an estimate of the amount of woody material carried by streams in the
assessment area
BMPs (e.g., stormwater runoff and fine sediment mitigation) that would be implemented are
now discussed in text boxes throughout < liapler 10 of the revised assessment.
Roy A. Stein, Ph.D.
Sizing Culverts: Page 5-61. Here the suggestion is that "Culverts must be 0.9 times the ordinary
high-water width..and where the channel slope is less than 5%, the "the culvert is allowed to
be 0.75 times" this same metric. Does this take into account global climate changes, which
would mean higher flow rates than historically has been the case? Shouldn't culverts be sized
larger than what historical flow rates would suggest, given that Climate Change will likely result
in more intense storms and therefore greater stream flows than has historically been the case?
RESPONSE: The commenter makes a good point We do not believe that the sizing
suggestions noted by the commenter (found in the Memorandum of Agreement between the
Alaska Department of Fish and Game and the Alaska Department of Transportation and
Public Facilities (ADF&G and ADOT 2001) take into account global climate change. Climate
change projections and potential impacts are now included in Chapter 3. We note in Chapter
10 of the revised assessment that climate-related changes, such as increasedfloodfrequency
and shorter return interval for major flood events, would likely undermine the structure of the
proposed transportation corridor and stream crossings. The variability and magnitude of
stream flows could also enhance other impacts described in the chapter. Interactions between
climate change and mining risks are discussed in Box 14-2.
Culvert Failures: Page 6-42. Culvert failure rates of 30-66% suggest we are doing something
wrong with establishing these culverts to maintain stream flow under a road. How might we
reduce this rate of failure (larger culverts? placement issues? solutions of any sort?)? In fact, if
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 123
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
indeed 50% of the culverts will be blocked (see bottom of page 6-43), are we not dealing with an
unacceptable solution of running streams under roads? Might there be some replacement of
these culverts with bridges; certainly, bridges are more expensive to build, but they simply do
not have the failure (i.e., blockage) rate that culverts do. Might there be a trade-off here between
initial investment costs (high for bridges) and salmon protection (fewer blockage events)? What
would Best Management Practices tell us in this context?
RESPONSE: In the revised assessment, culvert failure frequencies are reported as 30 to 58%,
with an average failure estimate of 47%. The likelihood of extended blockages would be low
during mine operations because the roadway would be monitored daily to ensure that failures
could be rapidly identified and repaired However, the likelihood would increase after mine
operations cease, if inspection and maintenance frequencies declined to those of typical roads.
Best management practices (BMPs) or mitigation measures would be used to minimize
potential impacts to salmon ecosystems from construction and operation of the proposed
transportation corridor. These BMPs, and their likely effectiveness, are now discussed in text
boxes throughout Chapter 10.
Bridges would generally have less impact on salmon than culverts, but can result in the loss of
long riparian side channels if they do not span the entire floodplaiii. The actual decision as to
what type of structure (bridge versus culvert) would be constructed at each crossing would be
made by industry engineers in consultation with state permitting staff
William A. Stuhhlcfichl. Ph.D.
Potential effects on salmonid populations were evaluated due to culvert blockage and failures.
Culvert blockages will prevent salmon passage leading to possible effects on reproductive
success. Literature data for the incidence of culvert failures were used in assessing failure
probability. This seems to be an appropriate approach given the hypothetical nature of the mine
used in the assessment; however, this is not my area of expertise and I am not aware of
additional data that should be considered.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., I1.1:.
Road and culvert failure likelihood. The likelihood of road and culvert failures is discussed in
Section 4.4.4 (p. 4-62). This section relies on the paper by Furniss et al. (1991) for a number of
aspects. As was pointed out above, this paper is focused on forest and rangeland roads and is not
applicable to the access road for the Pebble Mine. It is recommended that further evaluations be
done of similar roads at mines constructed between mines and port facilities to update this
section.
RESPONSE: Furniss et aL (1991) does focus on forest and rangeland roads, but it is a
seminal publication on the potential effects of roads, particularly as they relate to salmon. The
general conclusions of that paper should be applicable to the transportation corridor proposed
in the assessment. The failure frequencies cited in the revised assessment are from modern
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 124
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
roads and not restricted to forest roads. Because the proposed mining would take place in an
undeveloped area, much of the literature used in the assessment is necessarily from areas
outside of Bristol Bay. However, to the extent possible we used recent literature from
representative environments. We found no literature concerning the operational success of
culverts on roads between mines or connecting mines and ports.
Road and culvert failure consequences. The failure consequences discussed in Section 6.4
seem to be based on almost total regulatory failure during and after operations. The information
also serves to highlight the aspects that should be considered when designing, operating and
maintaining the access road during operations and subsequently during closure.
RESPONSE: In the revised assessment, consequences of failures and routine operations are
covered in Chapter 10. The failure consequences are not based on almost total regulatory
failure. Rather, they take into account the use of best management practices (BMPs) or
mitigation measures that are discussed in text boxes throughout Chapter 10. Nonetheless,
environmental characteristics along the transportation corridor would likely render the
effectiveness of standard or even "state of the art" mitigation measures highly uncertain.
Further discussion on this is contained in Box I OS of the revised assessment.
Phyllis K. Weber ScanneU, Ph.D.
The risks to salmonid fish due to culvert failures would be minimized by implementation of
permits by Alaska Department of Fish and Game (ADF&G), Habitat Division. Under A.S.
16.05.840-870, Alaska has some of the most protective laws for fish and fish habitat in the
United States. Further, given the lack of specific information on road alignments, construction
methods and stream crossings, it is not possible to calculate lengths of affected streams, quantify
loss of fish habitats, or predict failures of culverts, side slopes, etc. The document would be
strengthened if it included specific information on locations of spawning and rearing habitats and
estimated the contribution of fish habitats in the Nushagak River and Kvichak River Watersheds
to the Bristol Bay fishery.
RESPONSE: Best management practices (HMl's) or mitigation measures that would be used
to minimize potential impacts to salmon ecosystems from construction and operation of the
proposed transportation corridor are now discussed in text boxes throughout Chapter 10. Box
10-2 specifically refers to fish habitat regulations under Title 16.
As noted in the revised assessment, uncertainty exists in the characterization of streams and
wetlands aff ected by the proposed transportation corridor. Based on the chosen road
alignment scenario (which agrees with that proposed in Ghaffari et a I. 2011) we feel that we
are justified in estimating the potential footprint of the proposed corridor and its potential
impact on fish habitats and populations. We note in the revised assessment that "Although this
route (the one proposed in the EPA scenario) is not necessarily the only option for corridor
placement, the assessment of potential environmental risks would not be expected to change
substantially with minor shifts in road alignment Along any feasible route, the proposed
transportation corridor would cross many streams, rivers, wetlands, and extensive areas with
shallow groundwater, including numerous mapped (and likely more unmapped) tributary
streams to Iliamna Lake (Figures 10-1 and 10-2)."
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 125
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Specific information on locations of spawning and rearing habitats along the proposed
transportation corridor is difficult to obtain; as noted in the revised assessment, the Alaska
Anadromous Waters Catalog and Alaska Freshwater Fish Inventory do not necessarily
characterize aU. potential fish-bearing streams became of limited sampling along the corridor.
Nonetheless, the revised assessment summarizes the species, abundances, and distributions
that would potentially be affected, and places the streams along the transportation corridor
into the context of the entire Nushagak and Kvichak River watersheds with respect to
important watershed attributes such as discharge, channel gradient, andfloodplain potential
As far as placing potential mining impacts in the context of the entire Bristol Bay watershed,
we are unable to build a complete IP model, as this would require validation and more
elaborate construction of metrics appropriate to this region. However, our preliminary
characterization provides the building blocks for assessing the distribution of key habitat-
forming and constraining features across these watersheds.
Paul Whitney, Ph.D.
Criteria for bridge versus culvert installations along the proposed haul road. The dynamic
process of beaver dams causing streams to move across the floodplain should also be a criterion
for determining if and where culverts are installed for a potential road (pages 4-36 and 4-63).
Even if salmonids are not present at a stream crossing, the mosaic of active and decayed beaver
ponds in the floodplain can be important rearing areas for forage fish and benthic drift that are
utilized by salmonids (Snodgrass and Meffe, 1997; Schlosser and Kallemeyn, 2000). If beaver
dams (but not salmonids) are present above proposed stream crossings, bridges or causeways that
allow the streams to move across the floodplain should be recommended versus a culvert.
RESPOXSF: Our revised assessment is based on the assumption that crossings over streams
with mean annual flows greater than 0. lim'/s would be bridged. However, the actual decision
as to what type of structure (bridge versus culvert) would be constructed at each crossing
would be made by industry engineers in consultation with state permitting staff. We agree with
the reviewer that beavers can have an important influence of channel location and
morphology. However, beavers move frequently, so over the life of the road the locations of
beaver dams would change and therefore seem unlikely to provide a good criterion for
crossing designs.
Beaver are known to block culverts at the upstream ends. Beaver-proof culverts are an option,
but all the designs I am aware of would certainly hinder, if not block, movement of forage fish
and benthic drift. Causeways or bridges are the best way to encourage beaver activity (i.e.,
functions) and all the benefits that accrue.
RESPONSE: The actual decision as to what type of structure (bridge versus culvert) would be
constructed at each crossing would be made by industry engineers in consultation with state
permitting staff.
Toxic plume. Spills of transported chemicals are not quantified in the culvert failure section of
the assessment. I have participated in a mine risk assessment and a landfill risk assessment where
spills of cyanide and landfill leachate have been modeled. While I did not conduct the plume
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 126
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
movement analyses, stream hydrologists readily calculated how far spilled materials would move
down stream until the concentrations of chemicals in streams reached acceptable benchmarks.
The longevity of a spill of chemicals for copper processing should be calculated. It appears that
the Water Treatment failure assessment on page 6-39 conducted some sort of plume analysis to
determine the potential for an impact on Iliamna Lake. Perhaps it is possible to use this analysis,
or at least the model, to address the consequence of a spill of transported chemicals.
RESPONSE: We have modeled the transport of spills from pipeline failures and emissions
from the wastewater treatment plant and leakage of leachates. In Section 10.3.3.1 of the
revised assessment we estimate the number of reagent spills that would occur over the roughly
25-year life of Pebble 2.0 scenario. We did not conduct a plume analysis for a spill of
transported chemicals, but given the toxicity of sodium ethyl xantluite (Section 6.4.2.3) we
expect that a spill of this compound into a stream along the transportation corridor would
cause a fish kill Given the uncertainty concerning the nature and magnitude of a truck
accident and spill, we decided that a quantitative analysis of transport andfate would not
materially contribute to the value of the assessment.
Question 8. Does the assessment, appropriately characterize risks to salmonidfish due
to pipeline failures? If not, what suggestions do you have for improving this part of the
assessment? Are significant literature, reports, or data not referenced that would be
useful to characterize these risks, and if so till a/ are they?
David A. Atkins, M.S.
The discussion of pipeline failures is based on published failure rates, principally for oil and gas
pipelines. This analysis results in a pipeline failure rate of one per 1,000 km per annum. This is a
pretty generic number that does not consider actual pipeline design. Rather it indicates that
pipelines designed using standard practices do fail with a fairly high frequency. The Assessment
does not apply this failure rate to the gas and diesel pipelines because "they are not particularly
associated with mining." Without the mine, there would be no pipeline. So given that this rate of
failure is quantifiable based on good data and that the pipeline would be built to serve the
project, this risk should be considered.
RESPONSE: The assessment assumes that pipeline design follows standard ASME practices.
A diesel pipeline failure and resultant spill into two creeks has been added in Chapter 11 in
the revised assessment. A gas leak is considered but is not analyzed because of the lack of
significant causal linkage to fish production.
A concentrate pipeline spill would have differing impacts depending on when and where the spill
occurred, with deposition in Lake Iliamna likely being the worst outcome. As noted in the report,
it is likely that a pipeline spill would be detected rapidly and that the volume of the spill would
be limited and amenable to remediation. A better description of how concentrate pipeline failures
have occurred would be helpful to better understand the risk for this project (e.g., the July 2012
Antamina concentrate pipeline failure, although this pipeline would operate under a much
different pressure regime due to extreme altitude change).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 127
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Consolidated, statistically representative data on concentrate pipeline failures
are not readily available, although anecdotal evidence from some case studies can be found
The pipeline failure statistics reviewedfor the assessment come primarily from oil and gas
pipelines, but also include some water and hazardous liquid pipelines. The performance of
mining concentrate pipelines is not expected to be better than the performance of oil pipelines,
because concentrate pipelines would be expected to be more susceptible to internal corrosion
and abrasion. The 2012 Antamina concentrate pipeline failure in Peru was reportedly caused
by the rupture of a pipe elbow in a valve station. The regulatory, geographic, and operating
conditions of the Antamina pipeline may differ greatly from those of the concentrate pipeline
in the assessment scenarios. A discussion of causes and probabilities ofpipeline failures is
included in Section 11.1 of the revised assessment. The revised assessment also includes
discussions of concentrate spills at the Bingham Canyon. I tali, and Alumbrera, Argentina,
copper mines (Section 11.3.4.2).
Steve Buckley, M.S., CPG
The assessment does generally describe the potential risks to fish from hypothetical pipeline
failures.
RESPONSE: No changes suggested or required.
Courtney Car others, Ph.D.
A pipeline failure would be expected to release toxic leachale into stream systems in the
transportation corridor, none of which would dilute the leachate enough to prevent severe toxic
effects (both immediate and long-term). The report discusses three pipeline failures in the Bajo
de la Alumbrera mine in Argentina. The largest pipeline failure lasted two hours (compared to
only two minutes of exposure hypothesized in the current mine scenario). The report could more
clearly describe this case and its likely effects. The report appears to appropriately characterize
risks to salmonid fish due to pipeline failures, although I have no particular expertise with which
to evaluate this assessment.
RESPONSE: The III' I agrees that it would be desirable to have more information on the
effects of the largest llajo de la I lumhrera spill, but we have included all information on that
failure that is available.
Dennis D. Dauble, Ph.D.
The risks to salmonid fish due to release of pipeline concentrate/slurry and leachates (as return
water) are well described. However, risks of a diesel fuel spill are not. More detail could be
provided on reclaimed water. For example, what toxic constituents (and at what volumes) would
be released to the environment if these pipelines failed?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 128
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: A diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11 in the revised assessment. New data on concentrate leachate in the slurry have
been added, and it is assumed to also describe the return water.
Gordon H. Reeves, Ph.D.
Assuming that characterizations of the pipeline failure are accurate, the potential impacts on fish
and fish habitat are appropriate and reasonable. It was clear that the effects of a pipeline failure
could be major, depending on the duration and timing of the spill, because of the concentration
of metals in the slurry, the particular life-stage present, flow conditions, and the reduced
potential to fully remove the material from a stream or wetland afterwards.
RESPONSE: No change suggested or required.
The one question that I had about this section was the potential impact on phytoplankton and
zooplankton in Lake Iliamna, particularly at the local scale. I assume that any spill from pipeline
failure would have potential impacts on the lake and on phytoplankton and zooplankton, the
major food for juvenile sockeye salmon. The ecological consequences would depend on the
extent and intensity of any spill and on how juvenile sockeye use areas near tributary streams. I
would expect that a spill could be particularly detrimental if juvenile sockeye use the area near or
adjacent to natal streams when they enter the lake. I think this should be considered in more
detail in any additional analysis.
RESPO.XSIC: The toxicity data for species sensitive to copper and most other potentially toxic
metals are derived from planktonic crustaceans, so it is already addressing planktonic lake
species more directly than stream benthic species. However, the more detailed analyses of
aqueous releases in the revised assessment (Chapter 11) provide a better basis for addressing
the risk of exposure in Iliamna Lake. Because the aqueous phase of the product concentrate
slurry would enter the stream briefly and would be rapidly diluted in the lake, it is not judged
to be a major risk to plankton relative to the potentially sustained direct effects on salmon eggs
and larvae of the deposited solid phase or relative to stream invertebrates, which would have
much less benefit of dilution.
Charles Wesley Slaughter, Ph.D.
No. Concerns with pipelines crossing streams, watercourses and wetlands are similar to those
earlier expressed for the road corridor. On-site investigation may well reveal many more
"watercourses," including intermittent and ephemeral streams, than the 70 crossings cited;
possible pipeline failures thus may have much wider potential for impacting salmonids than is
indicated in the Assessment.
RESPONSE: The document has been edited to indicate that 70 is a minimum value.
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and recently
in Montana (?), suggesting the probability (with what confidence limits?) that there would be
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 129
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
only 1.5 stream-contaminating spills or two wetland-contaminating spills over 78 years of
operation seems wildly optimistic (and what is half a spill?).
RESPONSE: If you think of the mine as a repeated experiment (i. e., many mines with
concentrate pipelines of that length and duration), we would expect a mean frequency of 1.5
spills. A more straightforward explanation is that we expect 1 or 2 spills, given the scenario.
These frequencies are based on a large data set, not just the TAPS experience (which is an
atypical pipeline design and much larger than the diesel line) or the spill in the Yellowstone
River (which is a single event). An explanation of the frequency has been added in Section
11.1 of the revised assessment.
Assuming that any spill (over the 78-year project span) would last only two minutes (p. 6-32, p.
6-34), with a consequent minimal volume of spilled material, also seems highly optimistic. Even
highly-automated systems, with redundant sensors and automatic responses, are susceptible to
error or failure, and the Bristol Bay watershed environment is not benign with regard to
mechanical apparatus. The authors appear to recognize this with their discussion of the
Alumbrera incident.
RESPONSE: The scenario has been modified to a 5-minute response. Also, more information
has been added about the possibility of system failure or human error in response to this and
similar comments. The uncertainty discussion in Chapter 11 indicates that the exposure
scenario is predicated on the successf ul operation oj a remote shuto ff There may be extreme
weather or geological events that render the remote shutoff system inoperable. We did not
evaluate those events, so the assessment may underestimate these risks.
The specific consequences of a failure on salmonid habitat and biology are portrayed well.
RESI'<> \ ST.: \ o changes suggested or required.
John I). Sh'diiicli. Ph.D.
The pipeline corridor consists of four pipelines over a distance of 86 miles. No information was
provided on pipeline structure or placement, other than mentioning of stream crossings. The
pipeline failure of concentrate slurry was modeled using chemistry from the Aitik (Sweden)
mine. Is this best approximation? That mine is about 80 years old and is processing ore from the
edge of the pit, with much lower sulfur content than Pebble.
RESPONSE: More iujormatiou oil the pipeline structure and placement was provided in
Section 6.1.3.2 of the revised assessment. We have replaced the USGS leachate data from the
Aitik product with analyses from actual concentrate slurry provided by Rio Tinto to describe
the aqueous phase of the slurry and the return water, which would be alkaline (Section
11.3.2.1). However, the Aitik data are used to estimate the risks from deposition of the product
in a stream or wetland The environmental leaching of the concentrate would resemble the
USGS's leaching test of the concentrate, not the alkaline solution in the pipeline slurry. To the
best of our knowledge, no other aqueous leach test of a porphyry copper product concentrate
is available.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 130
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Pipeline failures can be significant in any environment and spill or pipe break prevention
requires significant monitoring. Will automatic shutoff controls be included? Are workers
stationed 24 hours/day every day? Some of the past Alaska failures were in winter conditions,
when things were not easily visible—under ice or snow cover. How will this be addressed?
RESPONSE: Our scenarios include remotely-operated valves tied to a supervisory control and
data acquisition (SCADA) continuous monitoring system tied to a continuously staffed control
room The remotely-operated valves could be triggered either manually or through software,
but we expect that except in the event of a clear rupture of the pipeline, the SCADA signals
would trigger an alarm that would initiate manual intervention by the operator. Consistent
with best management practices, we have assumed that, the pipeline would be visually
inspected along its full length at least daily, and that the inspection protocols would consider
the difficulties of detecting a spill under snow or ice. Potential remedial actions for cleanup up
of diesel, product concentrate and return water spills are discussed in Chapter 11.
The toxicity approach seems reasonable. What is the anticipated chemistry of the return waters?
Diesel spill monitoring? The geometric mean of three values (which references) indicates that
there is a 14% probability of failure in each pipeline in each year. This is not acceptable at any
level.
RESPONSE: The return water is assumed to be the same as the aqueous phase of the slurry
water. . I diesel spill scenario has been added in Chapter 11. The pipeline failure rate reported
in the assessment is similar to the rates calculated in studies by others and represents several
large datasets. Some of the reported failures are due to corrosion, which tends to result in
small releases; some of the failures are due to mechanical impacts, which tend to produce
larger releases; some are due to other causes. The buried pipelines in our scenarios would be
expected to have a lower incidence of rupture from third party activities, but may have a
higher incidence of rupture due to landslides or earthquakes than the overall dataset
Corrosion or corrosion leaks could go undetected for longer periods of time compared to an
aboveground pipeline.
Roy A. Stein, Ph.D.
Pipeline Failures I. What dictates 14 km between automatic shut-off valves; shouldn't this
distance be shorter as the pipe becomes larger, i.e., related to the amount of liquid/slurry that
would be spilled upon pipe failure? Shouldn't all of these pipelines be double-walled? What
would Best Management Practices tell us in this context?
RESPONSE: The distance between valves appears to be consistent with current practices. The
pipelines are described as double-walled in above-ground reaches. A double-walled pipeline
along the entire length of a pipeline might be desired from a purely environmental protection
standpoint; however, it may not be feasible or cost effective to do this. Therefore, we have
proposed the most commonly used (and accepted method) of double-walled construction over
any water bodies. There are a large number offactors that go into standard practices for
design, construction, and testing of pipelines. In the draft assessment, we included a selected
number of these factors as examples of our design mitigation measures. In the revised
assessment (Section 6.1.3.2), we have included a statement that the design would follow the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 131
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
standards of the American Society of Mechanical Engineers, to indicate that modern
mitigation measures would be included
Pipeline Failures II. Like all other failures, it seems to me that "Standard Operating Procedures
(SOP)" for mitigation should be in place in anticipation of any future spill or contamination of
the environment. I do not think that these procedures need to be in this report, but an
acknowledgement of their presence and that mining companies will follow these SOPs in
response to any spills that occur, be it pipeline, TSF, truck, leachate bed, etc.
RESPONSE: The EPA agrees that a mining company should have an SOP for remediating
spills, but we have not been able to find such an SOP for remediating a product concentrate
pipeline spill
William A. Stubblefield, Ph.D.
Potential effects on salmonid populations were evaluated due to potential pipeline failures as part
of the risk assessment. This evaluation focused on potential failures associated with the pipelines
for the product concentrate slurry and return water. No consideration of the natural gas or diesel
pipelines was presented, stating that such pipelines "are common and the risks are well-known."
Although I would acknowledge the failures in natural gas and petroleum pipelines are common, I
would not discount the potential effects to salmon populations associated with such spills.
RESPONSE: I diesel pipeline failure and resultant spill into two creeks has been added in
Chapter 11. Natural gas is not a contaminant of concern because it would vaporize and, at
worst, burn, which would not pose a significant risk to salmonid fish.
Evaluation of potential impacts due to a spill of product concentrate slurry or return water was
based on extant data from an existing copper mine in Sweden; to the extent that this slurry and
return water is representative of similar materials coming from the Pebble mine, this approach is
appropriate. The assumptions used in the amount of material that might possibly be spilled seems
appropriate and based on past experience and realistic assumptions; however, these assumptions
need to be reconsidered if and when a real mine plan is prepared.
RESPONSE: The Aitik le achate is no longer used to estimate the aqueous phase of the slurry,
in response to other comments. Ilecause the slurry would be alkaline, appropriate analytical
data were obtained from Rio linto (Section 11.3.2.1). However, the Aitik data are still usedfor
the leaching of the concentrate in a stream or wetland where neutral water would be the
leaching agent. No other relevant data are available.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to pipeline failures. It only estimates the likelihoods of occurrence and the consequences.
See discussion under Question 4 above regarding suggestions for improving estimation and
expression of the magnitude of risks to salmonid fish due to pipeline failures.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 132
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows.
Pipelines failure likelihood. The EPA Assessment focuses on the failure of the concentrate
pipeline because "We do not assess failures of the natural gas or diesel pipelines here because
such pipelines are common, their risks are well known and they are not particularly associated
with mining". I find this statement puzzling because all pipeline failures should be of concern. It
is further puzzling because the likelihood of pipeline failures for the concentrate pipeline is
derived from the failure statistics for pipelines in the oil and gas industry (p. 4-60). Failure of the
Baja de la Alumbrera concentrate pipeline in Argentina is suggested as an analog, indicating that
such failures can occur; however, I disagree that "it suggests that concentrate pipeline failures
are common at a modern copper mine." This last statement is not supported by any further
analysis of concentrate pipeline failures at other modern copper mines. It is recommended that
such analyses be performed or that the text be edited to indicate this shortcoming.
RESPONSE: The revised assessment contains an analysis of the risks of a diesel pipeline spill
in Section U.S. Natural gas is not a contaminant of concern because it would vaporize and, at
worst, burn off which would not pose a significant risk to salmonidfish (Chapter 11). The
quoted statement is literally correct, liaja de la Alumbrera is a modern copper mine and
pipeline failures are common there. This suggests that failures are common. Further,
concentrate pipeline failures at the Antamina and Bingham Canyon mines are now discussed
However, the statement in the revised assessment has been weakened by changing "common"
to "not uncommon. " To support that statement, a review of pipeline failures at U.S. porphyry
copper mines has been added (Section II. I). However, the EPA has not found a sufficient
record of product concentrate pipeline operations to develop a probability offailure that is
specific to that pipeline type.
Pipelines failure consequences. Failure consequences are focused on the release of concentrate
into water. As indicated in Appendix H of the report, the analog concentrate from the Aitik mine
is dominated by chalcopyrite, a sulfide mineral which contains the copper. If the concentrate is
submerged under water in relatively slow flowing streams then very little long-term release of
the copper will occur, as the water does not contain sufficient oxygen to allow for sulfide
oxidation. It is only when the concentrate is transported to locations above the water level that
oxidation and release of metals will occur.
RESPONSE: Subaqueous oxidation of sulfides does occur in well oxygenated waters as this
comment recognizes by stipulating "relatively slow moving streams". However, as the flow
data in the spill scenario indicate, these are not slow moving streams. Subaqueous leachate
column tests are conducted to assess the potential for subaqueous oxidation. Tests conducted
by the Pebble partnership indicate the potential for subaqueous oxidation. Salmonids require
high dissolved oxygen levels, so salmonid streams are necessarily well oxygenated For
example, the EBD states that mean dissolved oxygen levels for the North Fork Koktuli River
was 10.2 ppiiL This oxygen would oxidize the sulfides.
Phyllis K. Weber Scannell, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 133
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
This section of the document focuses on effects of pipeline failures; however, without a viable
mine plan, descriptions of pipelines and estimates of possible effects are speculative. The
resource developer may opt to build a pipeline to transport fuel from the coast to the mine site or
slurry concentrate to the port. Construction of any pipelines would require review and approval
by state and federal agencies, such approvals would likely contain monitoring plans to ensure
pipeline integrity. However, the risks of pipeline failures should not be minimized; the Fort
Knox Mine near Fairbanks recently experienced a 45,000 gallon spill of cyanide solution after a
bulldozer struck a supply line (Fairbanks Daily News Miner, August 24, 2012).
RESPONSE: The EPA agrees that the specific locations of pipelines and requirements for
monitoring may differ from our scenario, which is haseil on preliminary mine plans (Ghaffari
et aL 2011). However, we believe our scenario is plausible and allows an evaluation of
potential impacts from pipeline failures. The Fort Knox spill has been added to the assessment
as an example of spills due to human error in Section 11.1.
The risks of a pipeline failure to salmonid fish depend on the duration of the spill, the type of
material spilled (return water or concentrate), the location of the spill (in the uplands or in a
waterway), and the timing. The effects of a pipeline failure in a waterway when juvenile salmon
are present would be far more severe than a pipeline failure in an upland area.
RESPONSE: EPA agrees with this comment. Discussion of spill location and life stage
exposed has been expanded in Section 11.3, anil the other issues have been carried over from
the May 2012 draft
Given that there currently is no information on road alignments or locations of future pipelines, it
is not possible to estimate the number of stream crossings (70, page 6-30) or an exact length (269
km, page 6-30) of potentially affected waterways. The risks from pipeline failures outlined in
the draft document should be revised when more specific information on the mine plan of
operations becomes available.
RESPONSE: The road alignment ami length in Northern Dynasty Mineral's preliminary
mine plan (Ghaffari et aL 2011) nvrc used in this assessment. The number of stream crossings
was also detailed in that plan, and nvn checked by the EPA using USGS data. We would
expect that risks would be re-evaluated as part of a future specific transportation corridor
plan No change required.
Paul Whitney, Ph.D.
Refer to comments/responses to Questions 2 and 7.
RESPONSE: See responses to those questions.
Question 9. Does the assessment appropriately characterize risks to salmonidfish due
to a potential tailings dam failure? If not, what suggestions do you have for improving
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 134
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
this part of the assessment? Are significant literature, reports, or data not referenced
that would be useful to characterize these risks, and if so what are they?
David A. Atkins, M.S.
The Assessment generically describes tailings dam failures and the potential impact in detail. It
also uses some site-specific information on tailings supernatant and humidity cell leachate. There
is no question that a tailings dam failure would be catastrophic for the fishery and the project,
and although low probability, is the single largest risk to the fishery. A tailings dam failure could
harm a very large area of the watershed for a very long period of lime and could require a
massive and expensive remediation effort.
RESPONSE: Agreed. No change suggested or required.
The tailings deposition and storage methods outlined in the Wardrop NI 43-101 report and
presented in the Assessment are conventional for the industry and comply with Alaska State
regulations. Because of the dire consequences of a failure in this highly sensitive and unique
environment, it would be necessary to employ state of the art methods for tailings management
and go 'beyond compliance' when designing and constructing this facility. This may include
employing methods that are novel, incur significant additional cost for construction, and lead to a
more stable and lower maintenance facility in the long term, such as dry stack or paste rock
tailings (blending waste rock in with tailings in the impoundment to provide extra geotechnical
stability). These methods, however, are not common practice and in some instances are still
under development.
RESPONSE: The assessment addresses state of practice methods to identify potential risks
that could result from such practices. I'lie III' I agrees that there is a possibility that a mining
proponent could design and propose practices that go beyond the state of practice in order to
reduce potential risks.
Steve Buckley, M.S., ( I'd
The assessment does generally describe the potential risks to fish from tailings dam failures.
RESPONSE: No change suggested or required.
Courtney Carothers, Ph.D.
In the event of a tailings spill, invertebrates and fish would be exposed to toxic tailings and
leachate. Actual tailings failure examples suggest the range of exposure would spread to an area
more than 100 km. Copper would be especially toxic to invertebrates, fish eggs and larvae.
Toxicity would last for decades. The report appears to appropriately characterize risks to
salmonid fish due to a potential tailings dam failure, although I have no particular expertise with
which to evaluate this assessment.
RESPONSE: No change suggested or required.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 135
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Dennis D. Dauble, Ph.D.
Tailings deposition is described in Chapter 4 of the main report, but I could not find anything
that described potential risks to fishes, including effects to aquatic food webs and loss of fish
spawning and rearing habitat.
RESPONSE: The TSFfailure description and the assessment of risks were presented in
separate chapters in the original draft assessment, but have now been moved into one location
(Chapter 9) for clarity
As noted in the text, the sediment transport model used could only simulate sediment transport
and deposition ~30 km downstream of the mine site. Thus, potential effects to fish habitats were
not well quantified for the mainstem Koktuli River (and beyond), in addition to the Mulchatna
and Nushagak rivers. Is there a likelihood that any tailings material might reach Lake Illiamna?
If not, say so in the document. It is equally useful to say where impacts will not occur (as it
relates to sensitive habitat) as it is to describe where impacts are likely and reasonable.
RESPONSE: None of the 3 TSFs are in the watershed of lliamna Lake. The hydrology of the
site has been clarified and better maps added to clarify this issue. Risks to lliamna Lake from
water treatment failures, even if they occur in the Nushagak drainage, are now noted (e.g.,
transport of toxic leachate from the South Fork Koktuli to Upper Talarik Creek via
groundwater exchange between theses basins: Chapter 8). However, sediment (i.e., tailings)
would not follow that route.
The assessment deemed that it was "not possible" to determine how far the initial slurry
deposition would extend, how far re-suspended sediments would travel, and how long erosion
processes would continue. It seems that information from other mine closure sites could be used
by assessment authors to infer effect by analogy. The statement alluding to potential sediment
run out distance at the bottom of page 4-56 of the main report should be included in the summary
of effects. This is an important point.
RESPONSE: The revised assessment now includes a clearer description of the magnitude and
duration of effects. He apply the runout distance equations of Rico et aL (2008) to conclude
that under Pebble 2.0 scenario dam failure conditions, runout distance exceeds 307 km (190
miles), reaching the marine waters of Bristol Hay (Section 9.3.2).
Gordon H. Reeves, Ph.D.
Assuming that characterizations of the dam failure are accurate, the potential impacts on fish and
fish habitat are appropriate and reasonable. Impacts, like those of a pipeline failure, are likely to
be widespread in the watershed and to be long lasting, resulting from inundation of areas by
sediments and contaminants in the water. I think that potential impacts across the broader scale
could be developed and highlighted more fully. Also, consideration of Intrinsic Potential (see
response to Question 4) could provide additional insights into potential impacts of a tailings dam
failure.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 136
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: While we were unable to conduct a full Intrinsic Potential modeling exercise, we
have quantified the distribution of classes of gradient, mean annual flow, and % flatlandfor
reaches of the site watersheds where TSFs would be placed As highlighted in the report,
streams - especially those downstream of the TSFs in the mainstem North Fork Koktuli and
South Fork Koktuli - are low-gradient, floodplain prone channels that currently support
spawning populations of several salmon species, rainbow trout, and Dolly Varden.
I think that potential consequences of climate change on hydrographs should have been
considered in this section. More precipitation is projected to occur as rain in the winter rather
than snow for many parts of AK. How would this potentially impact the tailings dam facilities?
This seems to be a key piece of information that is needed to better understand the risk of dam
failure and the potential for impacts on aquatic resources.
RESPONSE: Climate change projections and potential impacts are described in Chapter 3,
and referenced here as a key uncertainty for in perpetuity management of the TSFs and other
mine infrastructure. The probabilities for dam failure discussed in this assessment involve
dams constructed in a variety of climates. I ny potential increase in precipitation due to
climate change is but one of the conditions for which the dams would need to be designed A
design developed to handle a higher level of precipitation would be expected to have a similar
failure probabilUy to one designed for an area of lower precipitation as long as the same
design standards, e.g. the safety factor against overtopping, were used in both designs. Of
course, if the design did not consider the possibility of increased precipitation as a design
factor, and climate change did cause such an increase, then the probability of failure would be
higher.
I thought that results to date of the impacts of the volcanic eruption at Mt. St. Helens, while not
exactly the same as a mine operation, were not useful in considering long-term impacts and the
response of aquatic ecosystems to such major disturbances. The impacts on streams are still
more prevalent and extensive than what is described in the report. Most stream systems are
transporting large amounts of fine sediment and areas of exposed gravels are rare.
RESPONSE: Extensive discussion of the Mount St. Helens analogy has been removed at the
prompting of several reviewer comments including this one. We have retained key references
from the region that are illustrative for considerations of likelihood offine sediment transport
and recovery (or lack thereof).
Charles Wesley Slaughter, Ph.D.
Yes. Physical consequences of TSF dam failure are fairly portrayed. I would only suggest that
effects of initial sediment deposition and long-term remobilization and redeposition would
extend beyond the spatial and temporal limits of the modeling used in the Assessment.
RESPONSE: Agreed We now more clearly state that remobilization and deposition could be
extensive; potentially reaching Bristol Bay.
Employing advanced eco-hydraulic modeling tools such as MIKE-11, MIKE-SHE (DHI,
Copenhagen), and consultation with state-of-art practitioners (IAHR-International Association
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 137
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
for Hydraulics Research, UI Center for Ecohydraulics Research, and others), along with
improved high-resolution input data such as LIDAR survey of the complete Kvichak and
Koktuli/Nushagak systems, would allow a more complete estimate of potential hydrologic and
sedimentation (and consequently biotic) consequences of TSF dam failure for the entire river
system, headwaters to Bristol Bay.
RESPONSE: We agree that LIDAR survey data would greatly improve the understanding of
the project area topography. Alternative modeling platforms coupled with LIDAR have the
potential to improve the estimates provided in this assessment, but these data collection and
modeling efforts were not within the scope of this project.
John D. Stednick, Ph.D.
The tailings dam failure was modeled and the distance of sediment transport was estimated. The
modeled tailings dam failure used an estimate of 20% mobilization from the tailings ponds. How
was this value determined? The model was run to a stream length of 30 km (the rivers
confluence), yet the report acknowledges thai a sediment pulse could run for hundreds of
kilometers. The moisture content of the tailings is estimated to be 45% by volume (page 4-50);
the 20% volume of sediment may be underestimated. This initial risk to salmonid fish is clear,
but the persistence of the sediment affect could be discussed.
RESPONSE: 20% was selected as a reasonable estimate, falling within the range of historic
failure release volumes (e.g., Azam and Li 2010 state 1/5 (i.e., 20%) andDalpatram 2011
states 20-40",,). We agree that the total volume released during a failure would vary depending
on water content, consolidation of tailings, and meteorological conditions in the valley during
the time of the failure. We also agree that the sediment initially deposited on the valley
floodplains and the sediment that remains downstream of the failed dam would become a
continuous source of tailings, with the potential to re-suspend during each subsequent
rainfall-runoff event that occurs before any sort of mitigation/clean-up efforts could be
implemented to control this process. The 20",, used was based on recent literature for the
amount that is generally released from these types of dam failures. The tailings dam failure
was intended to be a conservative analysis to shed light on whether a failure is a significant
concern in the Nushagak River watershed.
The Mount St. Helens analogy is inappropriate for a variety of reasons and such comparisons
should be removed from the assessment.
RESPONSE: Use of the Mount St. Helens analogy has been removed
The probable maximum precipitation (PMP) value was extrapolated from Miller (1963) and the
assessment commented how this value might be reduced upon further analysis. Conversely,
additional data could increase this value. There was no discussion of the recurrence interval of
this 24-hour storm.
RESPONSE: The assessment used the published PMP from Technical Paper No. 47 (the
current guidance available from NOAA, at mnv.nws.iioaa.gov/oh/hdsc/studies/pmp.htitd).
Within TP 47 the 24- hour PMP values presented exceed those provided for the 100-year 24-
hour event. While actual planning, design, and construction of a TSF dam will require
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 138
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
additional study, for the purposes of this assessment we relied on readily available information
and current published guidance for the determination of a design storm event with the
magnitude that would likely be considered during design of the dam The published PMP was
applied to a Type-1 storm within HEC-HMS to determine a flood hydrograph that would
result in the watershed located above the TSF as describedfor the purposes of this assessment.
This hydrograph is expected to peak at 291 cms and was applied to the hydraulic model to
cause an overtopping and subsequent failure of the TSF dam. The flood wave generated by the
dam failure had a peak flow of149,263 cms and 11,637 cms (large and small dam failures
respectively). The hydrology discussed and applied related to the dam failure represents a
precipitation and runoff event of the magnitude that the dam would be expected to
accommodate, but in this assessment was used as a mode oj failure.
No hydrologic data were provided. The streamflow gauging stations operated by the US
Geological Survey near the study area suggest peak streamflow rates from snow melt and from
rain events. The hydrograph shape and magnitude help determine if rain or snowmelt
dominated. In the assessment, the peak flow estimate from the Natural Resources Conservation
Service runoff method used a Type la storm distribution, the least intense precipitation
distribution, but the literature would suggest that a Type 1 distribution would be more
appropriate for Alaska. How does this storm event compare to the measured flood at Ekwok
(page 4-50)? The curve number (CN) was not identified, nor the methods used to calculate that
value. Similarly, the watershed slope, time to peak, hydraulic length, channel routing functions,
and channel resistance methods or results were not presented. What precipitation data are
available? The design of culverts, bridges, and storm water ponds all require good precipitation
records and the confidence in that estimate is based on record length.
RESPONSE: We agree that the Type-1 storm is appropriate for Alaska This has been
corrected and updated. Additional, hydrologic parameters were also included in the update to
better describe the development of the 11EC-IIMS model It is important to note that this
hydrologic event was used to provide a mode of failure for the dam. The flood wave generated
by the dam dwarfs the PMP hydrograph and the runoff from the storm event is not relevant.
The comparison is unclear for a 3,313 m3/s flow in a 2,551 km2 watershed area to the TSF flow
of 1,862 m3/s and an area of 1.4 km2. What was the precipitation and recurrence interval for the
Ekwok storm? The relation of groundwater flows to streamflow during storm events needs to be
evaluated. The flood producing precipitation events in this area no doubt add to groundwater
flows.
RESPONSE: This discussion was provided to help the reader understand the magnitude of the
dam failure flood wave relative to an event that was experienced by a local community. We
also wanted to draw attention to the fact that such a failure would occur in the upper end of
the watershed where typical storm events do not generate floods of such magnitude. We
acknowledge that the two watersheds are not similar and the intent was not to draw
comparisons of runoffpotential from one watershed to another, just to help the reader gain
context.
With a new estimate of precipitation depth of known recurrence interval, the design storm could
result in a higher flood event with greater velocities and greater sediment transport ability, along
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 139
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
with a greater sediment volume released from the TSF, resulting in a greater risk to salmonid fish
and habitats.
RESPONSE: We agree. The tailings dam failure was intended to be a conservative analysis to
shed light on whether a failure is a significant concern in the Nushagak River watershed No
change suggested or required
Roy A. Stein, Ph.D.
TSF Failure and Remediation. The text on pages 6-1 to 6-2 states:
"Remediation may occur following a tailings spill, but it is uncertain. A spill
would flow into a roadless area and into streams and rivers that are too small to
float a dredge, so the proper course of remediation is not obvious."
At this juncture in time, this statement points to the fact that we do not have the technology, or
the appropriate operating procedures, in place to remediate a TSF spill. Does this essentially let
the mine operator "off the hook"? Should we be promulgating mining activities in locations
where we cannot remediate spills, given our current state of knowledge or ability to apply current
techniques? What guidance would "Best Management Practices" provide for this situation?
RESPONSE: The intent of the statement is to point out the challenges of remediation of a
large-scale tailings spill in a remote area. Large-scale tailings cleanup is challenging even in
urban and more developed rural areas where there is road/rail, access for equipment and
transport of contaminated sediment. In the assessment watersheds, the only current access to
downstream areas is provided by the rivers themselves, which are generally too small for
large-scale dredging equipment.
William A. Stubblefield, Ph.D.
Potential effects on salmonid populations were evaluated due to tailings dam failures. Tailings
dam failure would potentially result in the release of large volumes of mine tailings and
associated contaminated waters, leading to possible acute and long-term effects on salmon
populations. It is also important to note that direct effects on salmon may be very species
dependent, due to life-cycle differences, and the time at which the dam failure occurs. Potential
effects due to sediment inundation/impaction can adversely affect habitat, leading to decreased
spawning. Evaluation of the potential for tailings dam failure effects considered acute and
chronic risks due to aqueous exposures, chronic risks due to sediment exposures, and risks due to
dietary exposures. All of these seem to be appropriate exposure pathways and all were
adequately considered, although site-specific information will improve risk predictions.
RESPONSE: Agreed No change suggested or required
Dirk van Zvl, Ph.D., P.E.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 140
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The EPA Assessment does not identify or appropriately characterize the risks to salmonid fish
due to a potential tailings dam failure. It only estimates the likelihoods of occurrence and the
consequences. See discussion under Question 4 above regarding suggestions for improving
estimation and expression of the magnitude of risks to salmonid fish due to potential tailings dam
failure.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. In this case we have
estimated those for a tailings dam failure to the extent that existing information allows.
Although fish abundance data are limited for this region, we did identify the potential
magnitude of a tailings dam failure to the Chinook run oj the Xusliagak River.
TSF failure likelihood. The failure statistics given on p. 4-45 are based on tailings failure
statistics over the last 50 years or so. Was there also a review of the operational histories, and
therefore failures, of tailings impoundments designed and constructed in the last 10 to 15 years?
It is recognized that one of the failures identified in Box 4-4 (Aurul S.A. Mine, Baie Mare,
Romania) falls in this category. However, many of the failures included in the analyses are
associated with older tailings facilities, especially those associated with large releases of tailings
solids. A significant improvement in tailings management is the implementation of an
Independent Tailings Dam Review Board (ITRB) for large mining projects (Morgenstern, 2010).
An example of the activities of an ITRB is given in Minera Panama (2012). Morgenstern (2010)
provides a listing of tailings failures from 2001 and 2010 and comments that "in no case, to the
knowledge of the Writer, was there systematic third party review" of the failed facilities as
would be the case when an ITRB is active. I expect that a tailings review board will also be used
for the Pebble Mine and the behavior of a tailings management facility designed and operated
under these conditions will be more representative of the potential failure likelihoods expected
for such a facility. It is expected that this likelihood will be much lower than those used in the
evaluations of the scenario in the EPA Assessment.
RESPONSE: The probabilities for dam failure used in the assessment were not derived solely
from the historical record. Historical failures were discussed as supporting background
information and present a defensible upper bound on the failure probabilities. The failure
probabilities used in the assessment are based on Alaska's dam classification and required
safety factors applied to the method ofSilva et aL (2008) which compares dams designed,
constructed, and/or operated under different standards. The discussion of failure probabilities
in the revision (Chapter 0) is expanded to try to clarify this issue.
TSF failure consequfiui-s. 11 is difficult to estimate the volume of tailings that will be released
when a tailings impoundment fails. The release of 20 percent of tailings from a slurry deposited
TSF may be realistic when it contains a large pool and is subjected to a large flood, but it is
unrealistically high for a TSF containing a small or no pool (such as in the case of a filtered dry
stack). I would consider the assumption that a release of 20% of the tailings material for the
Pebble mine scenario is on the high side, even during operations. When the mine is closed and
the tailings reclaimed I would consider the 20% release assumption as unrealistic, especially if
the closure implementation included a diversion system designed for the PMF. It is further
unrealistic to assume that the released tailings will remain in the downstream channels and flood
plains following the failure. In the case of the Aznalcollar Tailings Dam failure in Spain, all the
released tailings downstream of the mine were removed. While such a removal action will
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 141
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
impact parts of the watershed, it will help to recover the area faster than leaving all the tailings in
place and will also reduce the longer-term impacts on downstream water quality. I therefore
disagree with the assumption on p. 6-2 that "the assessment assumes that significant amounts of
tailings would remain in the receiving watershed for some time and remediation may not occur at
all." Box 6-1 provides "background on relevant analogous tailings spill sites" and three historic
sites are used as analogs. These are not realistic analogs, as they all relate to historic mining
under completely different scenarios. While the material historically released in these streams
were from base metal mines, the circumstances of their release, especially in the case of the
Clark Fork and the Coeur D'Alene Rivers, were very different. Long-term uncontrolled releases
occurred in these river systems due to regulatory circumstances or historically acceptable
practices that differ significantly from those in the 21s Century.
RESPONSE: We concur with the commenter that the estimation of the potential release
volume is difficult and inexact The release of 20% of impounded tailings used in our analysis
is well below the reports of 30% to 66% in the historic record. Dry stacking was not proposed
in the Ghaffari et al. (2011) report and has not been proposed in our scenarios. Ghaffari et aL
(2011) proposed maintaining a pond on the top of the TSF to keep the pyritic tailings
submerged, implying that the bulk of the tailings would remain saturated. H e acknowledge
that there are other tailing management strategies that could reduce the potential risk of
failure.
The remediation of the 1998 Aznalcollar Tailings Dam failure in Spain was completed within
the 6 months between the April failure and the onset of heavy rains in October. The
Aznalcollar area, near Seville, has a much drier and warmer climate than the Bristol Bay
area. It is also a heavily farmed area with flatter topography and far better access. The success
at removing the tailings from the Aznalcollar failure would be difficult to replicate in the
Bristol Hay area, and significant amounts of tailings would remain in the receiving watershed
for some time.
Box 6-1 in the original draft assessment (now Box 9-1) clearly states that the analogs
presented "provide evidence concerning the nature of exposures to aquatic biota". They are
not used to address the probability of a tailings dam failure or other aspects of the release of
tailings.
Phyllis K. Weber Scannell. Ph.D.
The assessment considers two possible failures of the tailings dam: a partial-volume failure
occurring during mine operations and a catastrophic failure occurring during or after mine
operations. The partial-volume failure (as modeled in the assessment) would result in a greater
than 1,000-fold increase in discharge and the catastrophic failure in a greater than 6,500-fold
discharge.
RESPONSE: No change suggested or required
The discussion of tailings dam failures describes possible changes in channel and floodplain
morphology and briefly mentions that the tailings deposition would be a source of easily
transportable, potentially toxic material.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 142
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: No change suggested or required
The potential for increased metals loadings to river and lake systems is understated. Although
there are no current predictions of tailings water quality, the water quality of tailings water from
similar mines could be used to model increases in metals loading from dam failures.
RESPONSE: The original draft and the revised assessment use tailings leachate data from the
PLP EBD (see Chapter 8). We believe that is more defensible than use of tailings leachates
from other mines.
In addition to the partial-volume failure and the catastrophic failure, there are other possible
sources of metals loadings from the tailings pond. Examples are emergency releases of untreated
tailings water, seepage of tailings water into the groundwater, and flow from the tailings pond to
groundwater in an adjacent drainage as the head (i.e. hydrostatic pressure) is increased as the
tailings pond is filled. The last example was experienced at the Red Dog Mine when the
increased elevation of the tailings pond caused water to flow underground into the Bonns Creek
drainage instead of the Red Dog Creek drainage. Interception ditches were installed after the
increases in metals loading to Bonns Creek were detected.
RESPONSE: Because the number of potential failures is extremely large, it is necessary to
choose a representative set of failure scenarios. The original and revised drafts of the
assessment include "seepage of tailings water into the groundwater, andflow from the tailings
pond to groundwater in an adjacent drainage" but not emergency releases of tailings water.
Paul Whitney, Ph.D.
Duration. I agree with the assessment that it would take a "very long time" (page 6-25, first full
para, last line) to reach concentrations that would not exceed threshold exposure levels. A "very
long time" could mean hundreds of years to one person or geological time (i.e., millions of
years) to another person. The assessment could be improved if some sidebars are put on the time
likely required for no risk dilution or "more normal channel and floodplain." One suggestion
would be to estimate the amount of time it would take the river/stream to move across the
floodplain in the "relatively undisturbed" Bristol Bay watershed. I would also like to know
whether reclamation or rip rap or rock weirs in areas with spilled tailings would reduce or extend
the time to reach "more normal" conditions.
RESPONSE: We agree that geomorphic analyses of channel/floodplain recovery following a
TSFfailure would help improve the estimation of recovery time, but such an analysis would
be a research and development effort beyond the scope of this assessment. Analogous
examples of recovery following massive contributions of sediment to stream systems were
explored (e.g., Mount St. Helens) but were not uniformly valued by reviewers, and were
droppedfrom the assessment.
Question 10. Does the assessment appropriately characterize risks to wildlife and
human cultures due to risks to fish ? If not, what suggestions do you have for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 143
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
improving this part of the assessment? Are significant literature, reports, or data not
referenced that would be useful to characterize these risks, and if so what are they?
David A. Atkins, M.S
The assessment does a good job analyzing the importance of fish resources to other wildlife and
to Alaska Native communities. The lack of site-specific information in the report results in only
very general conclusions that there 'would be some effect.' Of course, wildlife in the project area
and any traditional use of these lands would be affected by project construction under the no-
failure scenario. However, due to the lack of information, it is unclear if this is an area rich in
other wildlife or if there are traditional native land users that rely on the area. Although the
conclusion of this section is necessarily general, it would be helpful to have more detailed
characterization of wildlife and native use in the project area.
RESPONSE: More information about subsistence harvest use ill the area of the mine
scenarios and other areas of the watershed that may be affected by mining activities has been
added to the report Because the scope of the assessment includes potential effects to Alaska
Native communities related to salmonids, this has been the focus of the evaluation ofpotential
effects on indigenous cultures. However, EPA recognizes that there are potential effects due to
loss of subsistence harvest areas for other species, as well as potential ejfects on non-Alaska
Natives who practice a subsistence way of life. This is clarified in the revised assessment
(particularly in Chapters 12 and 13).
As the commenter notes, there is limited information about the use of specific mining claim
areas by wildlife. The Pebble Limited Partnership Environmental Baseline Data Report
includes data on the presence/absence of wildlife species around the Pebble claim, but we are
not aware of any data on abundance of wildlif e, so a more detailed characterization of a
specific mine claim area may not be possible. There is some information about subsistence use
of the mining claim areas which has been published by Alaska Department of Fish and Game.
General areas of wildlife harvests have been added as a figure to the revised assessment
Under the failure scenario, a tailings dam failure, in particular, would be catastrophic for wildlife
and Alaska Native communities that use the area.
RESPONSE: A new Figure 5-2 illustrates that subsistence use offish is extensive in areas
downstream of the TSE in the Miilcliatna and Nushagak Rivers, and this fact is now
referenced. Effects of a TSE failure on wildlife are now discussed in Section 12.1.
Steve Buckley, M.S., CPG
No comments on this section.
Courtney Carothers, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 144
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Wildlife: The sections discussing risk to wildlife resulting from effects on salmonids are fairly
short. Those animals that directly feed on these fish are likely to be impacted, as well as those
that depend on other resources enhanced by the marine-derived nutrients supplied by salmon
carcasses. The report concludes that the primary aquatic contaminant is copper (5-75), but notes
that the ore processing chemicals are unknown, as are their toxicities (5-59). These unknowns
could be noted as potential contaminants.
RESPONSE: Section 8.2.2.5 now includes a discussion of ore processing chemicals and their
toxicity.
Human cultures'. Overall, the main report (and Appendix I)) describes the central role that
salmon play in both Yup'ik and Dena'ina culture, both traditionally and in contemporary
communities. As noted above, the scope of the assessment focusing on these two cultural groups
should be made more clearly. Appendix E, for example, focuses on other human groups local to
this region, and those who migrate to the region for commercial fishing and recreation, who may
also be affected by risk to fish in this region. The vulnerabilities listed in Appendix D (p. 4-5)
could be listed in the main report more clearly as risks.
RESPONSE: The assessment text has been expanded to identify these vulnerabilities.
Literature on the effects of contaminated or declining resources on subsistence communities
could be utilized to describe likely impacts in more detail. For example, the report notes: "the
actual responses of Alaska Native cultures to any impacts of the mine scenario is uncertain" (ES-
26). While the specific responses are uncertain, likely responses can be predicted (and many are
articulated in Appendix D). There are data on the psychological, social, cultural, and economic
disruptions caused by the Exxon Valdez oil spill (e.g.. Braund and Kruse 2009; Palinkas et al.
1993), the cumulative effects of oil and gas development in the North Slope region (e.g., Braund
and Associates 2009; NRC 2003), and social impacts related to mining development in Alaska
(e.g., TetraTech 2009; Storey and Hamilton 2004). Drawing on some of this literature could help
provide likely scenarios for impacts to Alaska Native subsistence-based communities from
decreased quality, quantity, or diversity of salmonids. Current and recent responses to salmon
shortages in the Yukon-Kuskowkim region may also be helpful to include.
RESPONSE: The references provided have been reviewed and information from case studies
used where appropriate to provide data on responses of Alaska Native communities to
disruptions in subsistence resources (see Chapter 12).
Clearly the impacts to subsistence are not just lost food sources, but loss of healthy subsistence
lifeways, loss of practices, loss of cultural connections to the past, loss of connection to specific
places, loss of teaching and learning, loss of sharing networks, loss of individual, community,
and cultural identity, among others as detailed in Appendix D. This point could be made more
forcefully. As noted above and detailed in the specific comments below, subsistence is framed at
times in the report as primarily important for physical health and economic necessity. The
cultural, social, psychological, and spiritual aspects of subsistence livelihoods should also be
consistently highlighted.
RESPONSE: The assessment has been expanded to bring forward more information on the
psychological, cultural, social, and spiritual connections between Alaska Native cultures and
fish.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 145
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
As discussed Appendix D, Alaska Native cultures in this region and other regions in the state are
also dependent upon the cash economy, both for subsistence production and for other needs. The
role of commercial salmon fishing or other wage engagements related to salmon in the study
communities, while discussed in Appendix E, is not given much discussion in the main report.
How dependent is the subsistence economy upon commercial and recreational fisheries and in
this region?
RESPONSE: Additional information about the connections between the subsistence way of
life and the commercial and recreational fisheries has been added to the text of the revised
assessment.
There is a brief mention of non-fish related impacts to A laska Native communities in the main
report (5-77). Unless a full treatment of these impacts (positive and negative) is included, these
paragraphs should be removed. While in general, I am supportive of an increased scope (i.e., it is
incredibly difficult to isolate only salmon-mediated impacts to Alaska Native communities),
these other economic, social, and cultural impacts are not presented fully in the analysis, nor was
the ethnographic research designed to investigate these impacts, so passing mention of them here
does not seem appropriate.
RESPONSE: The text of the assessment has been clarified to indicate the scope of the
assessment does not include an evaluation oj direct effects on indigenous culture or human
welfare, because it is limited to aquatic pal/mays and impacts. However, the fact that there
would be direct effects is acknowledged in < hapler 12, consistent with comments from peer
reviewers and the public.
Dennis /). Double. Ph.D.
There is considerable detailed information in Appendices D and E relating to impacts of the
project to the economy. This information includes how salmon affect all segments of the
population, such as cultural resources of Native Peoples. However, not addressed in detail were
long-term impacts to Native Peoples that might occur after losing a way of life that includes
salmon. The description of potential impacts to their health and welfare should be expanded.
There are numerous examples of how Columbia River tribes have been negatively impacted due
to loss of fish resources (and lishing as a lifestyle) as a result of dam construction. These
impacts go beyond simple economics.
RESPONSE: EPA recognizes that there is a great deal of information about cultural and
health effects on indigenous populations from loss offish resources. We have referenced some
case studies in the assessment (with a primary focus on Alaska), and have expanded the
discussion ofpotential effects from a loss of salmon resources in Chapter 12.
The report should include a discussion of effects specific to unique user groups. That is, some
communities rely almost solely on sockeye; kings are more important to others. These impacts
could be segregated by watershed, for example. Also, some groups have more option for
subsistence gathering if sockeye and Chinook salmon resources are impacted. Potential impacts
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 146
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
of a declining salmon population due to mining operations would be less for them than groups
"on the edge" who currently rely mainly on salmon.
RESPONSE: The text of the revised assessment (Section 5.4.2.2) has been expanded to
acknowledge the differences between communities with regard to use and reliance on salmon
and other fish resources.
Disturbance of wildlife from noise and roadways should be included with respect to migration
corridors and critical habitat. Highlight species most likely at risk from human disturbance,
habitat loss/displacement (from the project footprint), and loss of salmon. For instance, do some
piscivorous species have the ability to shift their diet to include another source of protein? If so,
how would this shift affect the human culture with reference to a subsistence lifestyle?
RESPONSE: EPA recognizes that there are potential direct effects on wildlife from large-
scale mining operations, including disturbances from noise and roadways; however direct
effects on wildlife are outside of the scope of the assessment, as explained in Chapter 2. We
would expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
direct effects on wildlife.
Gordon H. Reeves, Ph.D.
These topics are outside my area of expertise so I cannot comment 011 if the report adequately
characterizes the risk to wildlife and human cultures. I am familiar w ith some of the literature on
the importance of salmon to wildlife and the report represents the finding fairly and accurately. I
cannot provide any additional literature, reports, or data.
RESPONSE: No change suggested or required.
Charles Wesley Slaughter, Ph.D.
No. The Assessment clearly qualified that its objective was to consider risks to salmonids, and
only inferentially consider "salmon-mediated" effects.
RESPONSE: No change suggested or required
Appendix C provides a comprehensive discussion of non-fish wildlife and the relation of those
populations to salmon. However, the Assessment itself (Volume 1) provides only a brief
summary in Chapter 2.2.3, which could allow a cursory reader to perhaps conclude that wildlife
populations have little risk of impact from the hypothetical Pebble project. Is this the intent of
the Assessment authors? A more in-depth reading of Appendix C allows inferring potential
consequences to wildlife and birds of "salmon-mediated" impacts of mining development.
RESPONSE: We have clarified assessment endpoints and provided additional background
information from Appendices A and B in Chapter 5. Direct effects on wildlife are outside the
scope of the assessment (as stated in Chapter 2), but their potential importance is
acknowledged in Chapter 12.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 147
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The assessment has been expanded to include more information from Appendix C of the draft
report (now an independent U. S. Fish and Wildlife Report).
The scope of the assessment is focused on potential risks to salmon from large-scale mining
and salmon-mediated effects to indigenous culture and wildlife. EPA recognizes the
complexity of potential direct, secondary, and cumulative effects on wildlife. We would expect
that a full evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
complexities and utilize the information in U.S. Fish and Wildlife report (Appendix C of the
draft report)
USFWS RESPONSE: We acknowledge the comment about the comprehensive discussion of
wildlife and the relation of selected species to salmon in Appendix C. Appendix C could be
used as the basis of an assessment ofpotential impacts of mining on wildlife, but decisions
related to assessment scope with regard to potential impacts of mining on wildlife is the
responsibility of EPA.
Appendix D provides a comprehensive and useful discussion of the indigenous people of the
Bristol Bay region and of their traditional ecological knowledge and cultures. Appendix D
clearly lays out the vulnerabilities and risks (summarized on p. 4-5 and pp. 150-153) associated
with the (hypothetical) major resource extraction projects. However, the Assessment (Volume 1)
provides only cursory consideration of these human aspects of the potential project - on p. ES-
23, and in Section 2.2.5. Presumably, this is because the EPA mandate is to conduct an
ecological risk assessment, rather than assessment of consequences for human populations,
whether indigenous, native, resident, non-native, non-resident, or the larger cash economy world
as represented by the State of Alaska, Northern Dynasty Minerals, or Pebble Limited
Partnership.
RESPONSE: The scope of the assessment related to Alaska Native cultures is limited to
salmon-mediated effects—that is, potential effects on indigenous culture if there are negative
effects on salmon. The assessment draws on the more comprehensive information in Appendix
D to evaluate these risks, Ilie text oj the revised assessment has been expanded to include and
discuss the vulnerabilities and risks to I laska Native culture from potential large-scale
mining.
John D. Stednicli, Ph. I).
The effect on wildlife section largely focuses on the return of nutrients to the land in various
shapes and forms and adds little to the risk discussion. Other than marine-derived nutrients,
other stressors exist. What are the consequences of the mine operation on other wildlife habitats?
Habitat fragmentation? Noise and light disruptions, etc.?
RESPONSE: The scope of the assessment is focused on potential risks to salmon from large-
scale mining and salmon-mediated effects to indigenous culture and wildlife. The EPA
recognizes the complexity of potential direct, secondary, and cumulative effects on wildlife and
would expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider these
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 148
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
complexities. The final assessment acknowledges the potential for direct effects on wildlife
(Chapter 12).
Mine development will require the use of explosives. What are the effects of the nitrate and
ammonia in the air following each detonation? The National Trends Network data suggest that
the area receives about 1 kg/ha/yr of nitrogen in precipitation. Thus, the increase in atmospheric
inputs from the explosions may exceed the marine-derived nutrients. What are the
consequences?
RESPONSE: See response to previous comment Also, stressors evaluated were based on how
they would significantly affect our primary endpoint of interest (the region's salmon
resources) and their relevance to EPA's regulatory authority and decision-making context
Although atmospheric inputs from mining operations are a potential source of contamination
to streams, they are not considered as significant as other stressors and are not regulated
under the Clean Water Act. Finally, because of terrestrial nitrogen fixation, nitrogen is not
believed to be the most important marine derived nutrient.
The potential loss or change in lifestyles of indigenous peoples is important, but this information
seems relegated to Appendix D. Include more of this information in the main body of the text.
Actually, there is a significant amount of information in the appendices that should be brought
forward.
It is unclear why there is such variability in the detail or depth of assessment of each of the
stressors. Why does the TSF failure section have 34 pages while the section on potential effects
on native peoples only has a few pages with a reference to an appendix? The unevenness of the
coverage needs to be addressed,
RESPONSE: The text of the revised assessment has been expanded to incorporate more of the
background information from Appendix 1) and a more comprehensive evaluation ofpotential
salmon-mediated risks to indigenous culture, the assessment's primary endpoint is the
abundance and production of sahtumidfish. so effects on those receptors are treated in more
depth tliau the secondary eudpoiuts. fisli-inedialed effects on wildlife and Alaska Native
cultures.
Roy A. Stein, Ph.D.
Risks to Wildlife. The importance of salmon in bringing Marine Derived Nutrients (MDN) into
these freshwater ecosystems and watersheds and their role in influencing wildlife and associated
interactions between wildlife and human cultures was well described. Loss of these nutrients
would severely compromise wildlife, and thereby human (through reductions of subsistence
harvest), populations. The appendix dealing with wildlife was quite detailed and well done and
it would serve the main report well for the authors to include critical information from this
appendix.
RESPONSE: No change suggested or required
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 149
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
USFWS RESPONSE: We acknowledge the comment regarding the quality of Appendix C.
Comments related to the scope of the assessment are not specific to Appendix C and are the
responsibility of EPA.
Risks to Native Alaskan Culture I. I thought that the assessment could be approved if the same
approach used for the mine, i.e., a case history approach, was used for human cultures. Surely,
there exist situations where salmon have declined or have been reduced by
development/exploitation (the Fraser River, perhaps?) where subsistence by Native Alaskans
was historically paramount. Once the salmon were reduced, what was the impact on the Native
Alaskans subsistence culture? How did the Native Peoples respond? From whence did they get
sustenance, cash, etc.? What sort of displacement occurred?
RESPONSE: The EPA recognizes that there is a great deal of information about cultural and
health effects on indigenous populations from loss offish resources. We have referenced some
case studies in the assessment (with a primary focus on Alaska), and have expanded the
discussion ofpotential effects from a loss of salmon resources.
Risks to Native Alaskan Culture II. The only risks reviewed here came via salmon-mediated
impacts on the culture of Native Alaskans. Direct impacts of the mine through jobs (potentially
positive), wildlife (likely negative), etc. all could be discussed briefly, serving to broaden the
overall impact of the mine and its associated activities on Native Alaskans.
RESPONSE: The EPA acknowledges that the scope of the assessment is limited to salmon-
mediated risks to indigenous culture and that there would be numy other potential effects
(positive and negative) from development of large-scale mining in the region. Direct impacts to
Alaska Native culture have been acknowledged in the revised report, but are outside of the
scope of the assessment.
Risks to Other Cultures. I was a little surprised that little text was spent on recreational
anglers, commercial fishers, subsistence users (other than Native Alaskans), etc. Appendices
provide some guidance here and this text need not be voluminous, but mentioning these impacts
of the mine on these groups would improve the main report.
RESPONSE: The discussion in the revised assessment has been expanded to acknowledge
potential eff ects on non-Alaska Native subsistence users. The EPA acknowledges that there
are potential impacts on recreational anglers and commercial fishers. However, because the
assessment focused on fish-related effects to Alaska Native culture, an evaluation ofpotential
impacts to the recreational and commercial sectors is outside of the scope of this assessment.
William A. Stubblefield, Ph.D.
Potential effects to wildlife and human cultures are briefly addressed in the risk assessment. No
"quantitative" assessment of potential effects is provided. For the most part, it appears that
potential affects to both wildlife and human cultural endpoints are directly proportional to the
injury suffered by salmon populations as a result of any spills or failures. Given the level of
detail available at this point in time regarding mine operations and closure, that is probably about
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 150
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
as far as any assessment could go. I'm not aware of any literature reports or data that would
assist in further characterization of these potential injuries.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment does not identify or appropriately characterize the risks to wildlife and
human cultures due to risks to fish. It only estimates the likelihoods of occurrence and the
consequences. See discussion under Question 4 above regarding suggestions for improving
estimation and expression of the magnitude of risks to wildlife and human cultures due to risks to
fish.
RESPONSE: The EPA believes that the likelihood oj an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. II c have estimated those to
the extent that existing information allows.
Phyllis K. Weber Scannell, Ph. D.
The document focuses on effects to wildlife that would occur from failures - tailings dam failure,
pipeline failure, etc. There are other sources of disturbance to wildlife that should be addressed
in a future mine plan and agency review and permitting. Other mines in Alaska limit truck
traffic on the haul road during caribou migrations, incinerate all kitchen waste, educate workers
on bear safety, and prohibit inappropriate disposals of food containers or other wildlife
attractants. Other factors that might need to be addressed to protect wildlife are limiting air
traffic and noise during certain times of the year. Unless addressed, these issues are more likely
to cause detrimental effects to wildlife than dam or pipeline failures.
RESl'i > \ SIC: No change suggested or required. The EPA agrees with the reviewer that any
future mine plan would require an evaluation of and mitigation for, direct effects on wildlife.
Paul Whitney, Ph.D.
Problem Formulation for Wildlife. If one addresses the problem formulation for wildlife at face
value, the answer is pretty straight forward. The assessment tells us that the consequences of loss
and degradation of habitat on fish populations could not be quantified because of the lack of
quantitative information concerning salmon, char, and trout populations (page ES-26, third
bullet). Furthermore, we learn that indirect effects, such as risks to wildlife, cannot be quantified
(page 5-75, para 1, last line). Stating that reduced salmon production would reduce the
abundance and production of wildlife (page 5-75, para 1, last sentence) is accurate but not
appropriate for a document that is intended to provide a scientific and technical foundation for
future decision making (page ES-1, para 1, last sentence). It is certainly possible to provide a
more complete scientific and technical response to a question regarding impact to wildlife. I
respectfully suggest that the first step in developing a more complete scientific and technical
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 151
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
response is to modify Question 10 by deleting the words "due to risks to fish" and to separate the
risk analysis for wildlife from the risk analysis for human culture. A revised wildlife question
should address both indirect and direct risks. A revised human culture question should address
(among many other risks) both direct and indirect risks to fish and wildlife.
RESPONSE: As the reviewer notes, the scope of the assessment is focused on potential risks to
salmon from large-scale mining and salmon-mediated effects to indigenous culture and
wildlife. The EPA agrees with the reviewer that direct effects on wildlife are likely to be
important. We would expect that a full evaluation of any future mining permit applications
and subsequent National Environmental Policy Act Environmental Impact Statements would
consider these direct effects. The revised assessment iichnowledges the potential for direct
effects on wildlife as well as risks due to fish.
Mixed Messages. The assessment and questions given to the review committee give mixed
messages regarding the scope of work. A variety of statements and conclusions in the assessment
are inter-related, as one would expect for a document that addresses ecosystem issues. While
acknowledging the overlapping issues and responses, I offer the following discrete points that
give me mixed messages regarding the scope of work and an appropriate characterization of risks
to wildlife. Since these points are related, there is some repetition of information but each of the
following items address a discrete issue.
1. The PREFACE (para 2) clearly states "Our goals in conducting this assessment are to
complete an objective assessment of the potential impacts of large scale mining on
aquatic resources in the Bristol Bay watershed." This statement is subject to wide
interpretation by this reviewer and by the document itself. For example, aquatic resources
include many species of wildlife and the Cederholm papers indicate that there are 137
species of wildlife that are associated or closely associated with fish. Wildlife species and
their associations are aquatic resources but have been glossed over in the assessment. The
lack of follow-through in implementing the above goal statement (i.e., not addressing
aquatic wildlife and wildlife associated with fish) is problematic. One can readily
understand that timing and budget constraints result in some statements indicating that
data are just not available or that certain types of modeling, while possible, were not
conducted. Elowever, it is difficult to understand why potential impacts to wildlife are
limited to indirect effects due to loss of fish if the assessment was truly conducted as an
assessment of aquatic resources. The message is mixed because the goal is to assess
aquatic resources, but many aquatic resources are not assessed.
RESPONSE: Hie assessment has been revised and reorganized to better explain the
goals and objectives of the assessment. As stated in previous responses to comments,
the focus of the wildlife assessment is limited to salmon-mediated effects.
2. The assessment was supposedly conducted as an ecological risk assessment (Abstract,
line 5). One of the first steps of an ecological risk assessment is to state well defined
endpoints. Suter (1993, page 22) presents 5 criteria that any endpoint should satisfy. My
notes are in brackets.
a. Societal relevance (Assessment appears to do a good job of identifying
concerns of people living near the area of the proposed mine. Yet, many of the
potential endpoints listed in site models (e.g., Figure 3-2E) are not addressed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 152
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The reason for not assessing endpoints in the site models (such as wildlife
quality and quantity) should be provided. Better yet, clearly state both direct
and indirect endpoints for wildlife.)
b. Biological relevance (The biological - especially ecological - relevance is
incomplete. Wildlife and the functions they provide are relevant in that
wildlife comprise many of the secondary and tertiary consumer species in the
ecosystem. The upland and aquatic habitats utilized by wildlife are the indirect
and direct sources of many nutrient and energy inputs to fish in the Bristol
Bay ecosystem. The upland components of the Bristol Bay ecosystem should
be assessed in more detail to provide a biologically relevant assessment. If
not, the assessment should directly indicate why not).
c. Unambiguous operational definition (Overall a good job for a hypothetical
mine. See response to Question 2 lor suggestions).
d. Accessibility to prediction and measurement (The inability to estimate the
impact on salmon numbers is problematic. Is the inability due to a lack of
population data, the unwillingness to utilize peer-reviewed methods other than
demographic population modeling, or some other reason? Furthermore, the
potential reduction in marine-derived nutrients and the potential impact on
wildlife can be assessed and predicted - see below).
e. Susceptibility to the hazardous agent (The assessment does a good job on the
hazards for aquatic resources but not susceptibility to wildlife biotic (e.g., loss
of habitat) and abiotic stressors (e.g., ice dams and scouring, as well as
hydrographs and sedimentation influence on ecological succession).
Furthermore, an adequate ecological risk assessment should address ALL
ecological stressors (e.g., loss of habitat and disturbance) not just toxics).
Suter's five criteria should be addressed when selecting endpoints. All the endpoints
selected and illustrated in figures should be assessed. The incomplete assessment of
quantitative estimates of all the impacts on salmon, non-salmonid fish, wildlife, as
well as community and ecosystem parameters should be remedied.
RESPONSE: The criteria above apply to the specific assessment endpoints considered
in the assessment, as described in Chapter S. Each of these endpoints meets the above
criteria, and additional information describing endpoints within the ecological risk
assessment process has been added to the beginning of Chapter 5.
3. A complete assessment of aquatic resources as stated in the PREFACE should address
both direct and indirect risks to the many species of wildlife and wildlife habitat that are
aquatic resources and closely associated to aquatic resources. The list of all the aquatic
species and habitat is lengthy. A partial list of categories of wildlife and wildlife habitats
that are aquatic resources includes grebes, ducks, shorebirds, beaver, muskrat, otter,
mink, and riparian, emergent, and aquatic vegetation. The impact of mining on all of the
aquatic resources is given scant coverage in the assessment, with the possible exception
of impact to wetlands. The PREFACE provides wide ranging goals but the assessment
has many gaps, and I wonder why.
RESPONSE: The latest draft of the assessment has been revised and reorganized to
better explain the goals and objectives of the assessment. Those goals and objectives
focus the assessment on salmonidfish and a limited qualitative discussion of the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 153
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
sahtwn-mediated effects on wildlife and Alaska Native cultures. The EPA
acknowledges and agrees with the reviewer that there will be direct and indirect effects
of mining on wildlife, but the detailed evaluation of those effects will be left to the
NEPA and permitting processes should a mine be proposed
4. The Executive Summary (para 1) indicates . .USEPA launched this assessment to
determine the significance of Bristol Bay's ecological resources and evaluate the
potential impacts of large-scale mining on these resources." The emphasis on salmonids
does not fairly represent the entirety of ecological resources. To me, ecological resources
should include ecological parameters such as plant community succession, species
diversity, energy flow, and structure and function, in addition to the information on
salmonids.
RESPONSE: The scope of the assessment has been clarified in Chapter 2, and the
specific endpoints considered in the assessment are described in Chapter 5. We
recognize throughout the text that other ecosystem components are important and
potentially will be affected by large-scale mining, but fall outside this assessment's
scope. Ecological risk assessments focus on endpoints that are important to decision
makers and stakeholders, are susceptible to the stressors, and meet other criteria; they
do not attempt to cover all ecosystem components (Suter 1993, EPA 1998).
5. The Scope of the assessment (ES-,1 last para) indicates that "wildlife ... as affected by
changes in the fisheries are additional endpoints of the assessment." This informs the
reader that only wildlife species that are affected by changes in the fisheries will be an
endpoint. Yet many species of wildlife that are affected by changes in the fisheries are
not addressed. In addition, such an endpoint glosses over the importance of assessing fish
that are affected by changes in wildlife. For example, beaver modify aquatic habitat and
terns prey on fish. The scope of the assessment seems to brush over many of the
important fish and wildlife interactions discussed by Cederholm et al. (2000) and
Cederholm et al. (2001) and to focus mainly on marine-derived nutrients. I am pleased
with the discussion of the marine-derived nutrients but am not clear how the emphasis on
this one interaction, which is not quantified in the assessment, furthers the understanding
of potential mining impact on the Bristol Bay ecosystem.
RESPONSE: The EPA agrees with the reviewer that there are many complex fish and
wildlife interactions in the study area and the revised assessment acknowledges this
fact. The assessment focused on loss ofMD.X to the system and on potential loss offish
as a food source for wildlife as important potential salmon-mediated effects on wildlife.
We would expect that a full evaluation of any future mining permit applications and
subsequent National Environmental Policy Act Environmental Impact Statements
would include analysis of other potential causes of effects on wildlife.
6. Gende et al. (2004) use a variety of methods to quantify both the marine-derived nutrients
and energy that transfer from salmon to terrestrial wildlife and habitats. For example,
they determined that bears moved nearly 50% of the salmon-derived nutrients and energy
from streams by capturing salmon and dragging the carcasses from the stream. Gende has
also worked with University of Alaska Associate Professor Mark Wipfli (Gende et al.
2002) on the relationship of salmon to terrestrial habitats. Alaskan and national experts
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 154
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
are likely available to assist EPA in quantifying the movement of marine-derived
nutrients to the terrestrial ecosystem.
RESPONSE: The references are appreciated. The EPA acknowledges that a more
quantitative evaluation of salmon-derived nutrients to the terrestrial ecosystem may be
possible. However, this type of evaluation would require further data collection and
significant analysis, which is beyond the scope of this assessment. A detailed
evaluation of the movement of marine-derived nutrients to the terrestrial ecosystem is
not needed to assess the impacts of mining on salmonids, and is more detailed than is
neededfor a qualitative discussion of the salmon-mediated effects on wildlife and
Alaska Native culture. We would expect that a full evaluation of any future mining
permit applications and subsequent National Environmental Policy Act Environmental
Impact Statements would address potential effects to salmon-derived nutrients.
7. As mentioned above, the assessment does a fine job of emphasizing the importance of
marine-derived nutrients transported to the terrestrial environment but has relatively little
information on the importance of terrestrial-derived allochthonous nutrients transported
to the aquatic environment. Doucett el al. (2007) discuss methods for measuring
terrestrial subsidies to aquatic food webs using stable isotopes of hydrogen. The potential
loss of terrestrial subsidies due to mining might be as great as a potential reduction of
marine-derived nutrients. It would be informative to discuss the relative importance of
marine-derived, autochthonous, and allochthonous nutrients. Such information might
influence "best mining practices" and reclamation that can partially compensate for lost
allochthonous inputs.
RESPONSE: See response to previous comment.
PHWResponse: The revised EPA text does address this issue but not in a relative way. ^--{formatted. Font. Not Bold, Not italic )
8. Figure 3-2E indicates wildlife quality, quantity or genetic diversity, as well as wildlife
predation are important to Alaska Native Cultures. Tribal elders are said to have concern
for "potential direct effects on other subsistent resources" (which includes wildlife and
vegetation). In addition, Appendix D comments on concerns regarding wildlife (e.g.,
caribou) and vegetation (e.g., berry gathering) that are not directly linked to the fisheries.
An assessment of the direct impact of potential mining and the ongoing exploration on
wildlife and vegetation would address the concerns mentioned above. Once again, I
wonder why the site model figures outline so much detail that is not addressed in the
assessment text.
RESPONSE: The conceptual models were drafted to gain an understanding of all
potential sources, stressors, and effects from large-scale mining and have been revised
to clarify the scope of the assessment The EPA focused the assessment on potential
risks to salmonids and subsequent salmon-mediated effects on indigenous culture and
wildlife. There are many other potential risk pathways which are outside the scope of
this assessment, including potential loss of plant and wildlife subsistence and other
cultural resources (see Figure 2-1).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 155
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The conceptual models have been modified to more accurately reflect issues within the
assessment's scope, and to more clearly identify those pathways considered outside of
scope when they are included
PHWResponse: The revised models more clearly identify those pathways considered
outside the scope. EPA might want to revise para 2 on page 5-1 that states "(2) the
abundance, productivity, or diversity of the region's wildlife populations" is one of three
endpoints in the revised assessment.
9. Comments in the main assessment Volume 1 (page 5-76, first lines) indicate that
information on wildlife and potential direct impacts are being collected but are not
included. Including measures of direct wildlife impact in addition to indirect impact as a
result of changes in fisheries would greatly improve the assessment. There are many
ways to estimate the impact of habitat loss on wildlife (Morrison et al. 1998). A better
understanding of impacts on wildlife would presumably provide a better technical basis
for designing a reclamation plan.
RESPONSE: The scope of the assessment is focused on potential risks to sahnon from
large-scale mining and salmon-mediated effects to indigenous culture and wildlife.
EPA agrees with the reviewer that direct effects on midlife are important and that
there are ways to estimate effects related to habitat loss. We would expect that a full
evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
direct effects. The revised assessment acknowledges the potential for direct effects on
midlife. The assessment does not include a reclamation plan and is not intended to
support the development of a reclamation plan.
10. The assessment (page 5-77, last sentence) states: "Although this assessment is focused on
sahnon, the non-salmon-related impacts on native cultures from routine mine operation
are likely to be more significant..." If this is in fact the case, what better reason could
there be for increasing the scope to include wildlife, vegetation, community, and
ecosystem structure and function (i.e., non-salmon)? Methods and mine-related examples
for assessing such an increased scope are available from the Northwest Habitat Institute
(nwlii.org) in Corvallis. Discussions with Tom O'Neil, Executive Director, indicate that
available data from the Alaska GAP Analysis Project and the Alaska Natural Heritage
Program could likely be used to make such assessments.
RESPONSE: See response to previous comment. This type of evaluation would be
valuable but would require further data collection and significant analysis, which is
beyond the scope of this assessment.—._
PHW Response: If the statement in item 10 above is true and the non-salmon-related
impacts are more important to native cultures it seems inconsistent to say that Alaska
Native organizations only asked for the assessment to address salmon-mediated impacts
on wildlife.
11. The risk assessment's focus on "indirect effects on wildlife" (page 5-1, para 2) is
consistent with some of the several goal statements that lean toward fish influence on
wildlife. I have difficulty reconciling this emphasis because further on in the text (page 5-
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 156
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
75, para 1, last line) it is stated that the "indirect effects cannot be quantified." If the
indirect effects cannot be quantified, one should think it is even more important to get a
handle on the direct effects, which can be assessed by any one of several wildlife
methodologies (Morrison et al. 1998).
RESPONSE: Wildlife is not a main focus of this assessment. We agree that a more
detailed assessment of direct and indirect impacts of mining to wildlife will have to be
done as part of the NEPA and permitting processes. This assessment is focused on
sahnon because of the world class, outstanding salmon fisheries in Bristol Bay.
Further, as a keystone species, salmon serve as an excellent indicator of overall
impacts to the ecosystem.
PHWResponse: If you are a hammer, everything looks like a nail. When the hammer
hits the nail all is well. If you are a fisheries biologist everything looks like
harvest/escapement.When harvest and escapement hit goals all is well. If vou are a
fishery ecologist. the welfare of the salmon ecosystem looks like biodiversity, structure
and function. If it is possible that biodiversity, structure and function have changed, vou
want to know why and what the consequences might be. The above EPA responses are
definitely made by a fishery biologist.
Does the above EPA response assume that all species of salmon are keystone species and
that no wildlife species are keystone? The keystone species concept may be a valid
reason for a fisheries biologist but various wildlife species such as sea otter, wolf, and
beaver have been considered keystone species. While the keystone species concept
allows ecologists to communicate with other ecologists and the public, the keystone
concept is a hypothesis and oversimplifies the complexity of ecosystem biodiversity.
structure and function.
{ Formatted: Font: Not Bold, Not Italic
A
The Clean Water Act, as far as I know, doesn't focus on fish populations but rather
ecosystem function. It is hoped decision makers give due consideration to the importance
wildlife has to ecosystem functions as thev relate to salmon.
12. The assessment indicates the exceptional quality of the fish populations and their
importance to the region's wildlife is due to five key characteristics. The fourth
characteristic is "the increased ecosystem productivity associated with anadromous
salmon runs" (page 2-20, Section 2.3). Similar statements are made by Woolington
(2009). How is this increase measured, and what is the baseline for the increase? I am
sure the biologists that wrote about the increase are trying to convey the concept that the
energy and marine-derived nutrients provided by salmonids is incorporated into the
primary and secondary production in the terrestrial ecosystems in the Bristol Bay
watersheds. I am not sure that such a statement considers the loss of energy and marine-
derived nutrients as a result of commercial fishing. Is it possible that commercial harvest
may decrease ecosystem productivity compared to productivity prior to pre-European
(e.g., 200 hundred years ago) harvests? Perhaps it would be more accurate to delete the
word "increase." On the surface, this might seem like a small edit but this edit is one of
many edits needed to address, not only the existing conditions in the watersheds, but also
the existing conditions relative to "pre-European conditions." My understanding of a
Cumulative Impact Analysis is that it includes past assessments of ecosystem resources
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 157
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
and functions as well as current and future conditions (see discussion of Cumulative
Impacts in response to Question 11).
RESPONSE: The purpose of the assessment is to evaluate potential effects of large-
scale mining on the fishery, not the effects of commercial fishing. Therefore, the
potential effects on wildlife associated with loss of marine-derived nutrients considers
the present condition of the watershed andfishery as a baseline for an evaluation of
potential future effects of large-scale mining.
13. Pauly et al. (2000) and Libralato et al. (2008) address the energy impacts of fisheries. I
would be interested to know if the impact of fisheries on marine-derived nutrients and
energy available to the wildlife and terrestrial ecosvstems in the Bristol Bay watersheds
could be compared to the potential impact of the example mine on marine-derived
nutrients and energy available to the watersheds. I s the potential impact of the mine small
in relation to the impact of commercial fishing or very large compared to commercial
fishing? Such information would be informative for determining acceptability of risks. It
is possible that such a comparison could stimulate discussion of the possibility of
compensatory mitigation for losses due to mining.
RESPONSE: Compensatory mitigation for losses of aquatic resources Jrom mining is
governed by a regulatory process outside the scope of litis project. Nevertheless, in
response to public and peer comments we have included a discussion of compensatory
mitigation in Appendix J of the revised assessment.
PHWResponse: This EPA response is consistent with previous responses but does not
address mv question.
14. The assessment could be greatly improved if more of the linkages and pathways
illustrated in the various Conceptual Models were addressed and if impacts on ecological
parameters, such as community succession (down gradient, in the lake and on tailings and
waste rock) and aquatic and upland structure and function, were addressed. For example,
Site Reclamation is illustrated and highlighted as orange boxes in three places on Figure
3-2C (I'm not sure what the orange highlighting signifies, for there is no legend on this
figure). Information on possible wildlife limiting factors (e.g., calving and nesting
habitat) and plant communities in the watershed could improve the quality of Site
Reclamation. A Reclamation Plan needs to address the likelihood that reclamation can be
implemented and if so w hat benefits it might accrue through time. This type of
information would likely result in a better risk assessment.
RESPONSE: EPA agrees that more detailed information on wildlife limiting factors
would need to be collected and evaluated as a part of a site reclamation plan. We would
expect that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would include
this information as a basis for a site reclamation plan.
15. The discussion of wildlife species in Appendix C is very good but little of the insight
provided in these descriptions of hunted or trapped species is reflected in the assessment
of impact of potential mining practices on wildlife. I cannot help but wonder (i.e., mixed
message) why Appendix C is so thick and so little of the good information in Appendix C
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 158
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
is reflected in the assessment document. For example, the impact of noise and human
presence related to mining and roads is addressed on page 54 (Appendix C). Such an
impact on certain sensitive species could be equal to or greater than the loss of wildlife
habitat in the mine footprint. Woolington (2009) indicates that the Red Dog Mine in
Alaska has implemented certain measures that have reduced the impact of the haul road
on wildlife. It would be good to know what these measures are and if such measures
would be equally effective for a road relatively close to Anchorage.
RESPONSE: The purpose of the appendices vs. the main assessment document has
been clarified in Chapter 1. The former Appendix C has now been published as a
separate US Fish and Wildlife Service document because it serves purposes for the
USFWS beyond the scope of the EPA's assessment.
USFWS RESPONSE: We acknowledge the comment about the quality of Appendix C.
Comments related to the scope of the watershed assessment are not specific to
Appendix C and are the responsibility oj T.I'. I.
Will the proposed haul road to the mine be closed to the public and hunting? If so, the
impact of the road may not be high. I can remember Val Geist saying that wildlife behave
the way one teaches them to behave. Examples are deer in the streets of Banff Park
during the daytime and big horn sheep and elk foraging along the Trans Canada Highway
in the Park. The big horn sheep often stop traffic as people stop to observe them. Outside
the Park, where deer and elk are hunted, one is less likely to see deer and elk along the
highway and deer take on a more nocturnal existence.
RESPONSE: Tor purposes of the assessment, EPA assumes that the road would be
closed to the public during mining operations but potentially could become a public
road after mining operations cease. About 80",, of the transportation corridor is on
private land owned by various Alaska Xative I illage Corporations, with which the
Pebble Limited Partnership has existing commercial partnerships. Alaska Native
Regional Corporations are charged with managing land and resources within their
control in the best interests of their shareholders. Chapter 13 includes a discussion on
the potential impacts of increased access due to additional roads.
The scope of the assessment is focused on potential risks to salmon from large-scale
mining and salmon-mediated eff ects to indigenous culture and wildlife. EPA has not
evaluated direct eff ects from a transportation corridor on wildlife. We would expect
that a full evaluation of any future mining permit applications and subsequent
National Environmental Policy Act Environmental Impact Statements would consider
the direct effects of the transportation corridor on wildlife.
16. As an arctic and sub arctic small-mammal wonk, I anticipate that any discussion of
Alaskan ecosystems will include some discussion of the less charismatic fauna (small
mammals and songbirds) and the functions they provide. Such discussion is given little
attention in the assessment.
RESPONSE: These issues are outside the scope of the assessment, as specified in
Chapter 2. Appendix C (now an independent US Fish and Wildlife report) selected key
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 159
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
wildlife species for characterization, based on their direct dependence on salmon or on
their roles in distributing marine derived nutrients through the ecosystem. It was not
possible to fully characterize every species. The EPA recognizes that there are multiple
species that were not fully characterized and that have important functions in the
terrestrial and aquatic ecosystem We would expect that a full evaluation of any future
mining permit applications and subsequent National Environmental Policy Act
Environmental Impact Statements would consider all species present in these
watersheds.
USFWS RESPONSE: The USFWS selected key species to include in Appendix C and
the selection method is described in the report. II e certainly concur that small
mammals are important from an ecosystem perspective, but they are less linked to fish
or to public concerns than the selected species. Songbirds were included in the wildlife
report under landbirds.
Question 11. Does the assessment appropriately describe the potential for cumulative
risks from multiple mines? If not, what suggestions do you have for improving this part
of the assessment?
David A. Atkins, M.S
According to the Assessment, cumulative risks result from the potential development of at least
five additional prospects: Humble, Big Chunk, Groundhog, Sill, and 38 Zone. Exploiting these
prospects would amount to development of a mining district (see discussion for Question 2 in
regards to appropriateness of the mining scenario).
RESPONSE: No change suggested or required.
The Assessment quantifies the loss of stream lengths and wetland areas that potentially support
salmon and resident fish from the development of these projects under a 'no-failure' scenario.
The assessment is highly speculative given that mine development plans are not available for
these prospects.
RESPONSE: ( limn/alive impacts assessments evaluate past, present and reasonably
foreseeable future actions that are temporally and spatially linked to the project under
consideration, the potential mines in the revised assessment are reasonably foreseeable based
on State of Alaska planning documents and industry exploration activities. We have expanded
the discussion of cumulative impacts in Chapter 13.
As with the Pebble scenario, it would be helpful to put this loss of resource in perspective in
terms of the fish resources as a whole. It would also be helpful to describe any mitigation
measures that are feasible to offset the impact of loss of streams and wetlands. Furthermore, it
would be helpful to better understand the role these developments could have in further
fragmenting salmon populations.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 160
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Due to lack of comprehensive estimates of limiting factors across the impacted
watersheds, population level effects could not be quantitatively estimated except for the most
severe cases where total losses of runs could be reasonably assumed Our ability to estimate
population level effects was limited to situations that were assumed to completely eliminate
habitat productivity and capacity in an entire watershedfor which estimates of escapement
could be inferred For this assessment, these conditions are only met in the TSFfailure
scenario that completely eliminates and blocks access to suitable habitat in the North Fork
Koktuli River. In that case, we estimate that the entire Koktuli portion of the run (-28% of
Nushagak escapement) could be lost. Higher proportional losses would occur if significant
downstream effects occurred due to transport of toxic tailings fines beyond the Koktuli as
modeled under the full TSFfailure.
Compensatory mitigation is governed by a regulatory process outside the scope of this
assessment but will be an important part of any permitting process. Nevertheless, in response
to public and peer comments we have included a discussion of compensatory mitigation in
Appendix J of the revised assessment
We have expanded the cumulative effects discussion (Chapter 13) to give a broader description
of streams, wetlands andfish at risk and have incorporated a discussion of the possible effects
on biological complexity andfragmenting salmon populations.
The following potential subsidiary impacts from development of a mining district of this scale
should also be described in more detail or at least mentioned:
• The extensive road network required to support mines in the area and the attendant
development associated with this network.
• The camps associated with the project, in migration of workers to the project areas, and
the demand for resources to be imported from outside the area.
• Invasive species that may follow this scale of development.
RES1 SIC: The discussion about the cumulative effects of roads (including invasive
species), secondary development, and mining camps has been expanded in Chapter 13.
Steve Buckley, M.S., ( I'd
The assessment does describe the potential for cumulative risks from the development of
multiple mines in the area; however, the section is misleading in that it describes specific mine
footprints and tailings disposal sites at these prospects where there is no information on the size
or character of the potential future mine sites. The assessment does not describe the cumulative
effects of mine development.
RESPONSE: We have modified the descriptions ofpotential future mines to make them less
specific (e.g., recognizing that mining infrastructure could go anywhere in the subject claim
blocks). We have expanded the discussion of potential cumulative effects to include the
cumulative effects of multiple mines and development induced by the advent of large scale
mining in the watershed (Chapter 13).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 161
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Courtney Carothers, Ph.D.
In general, the report suggests that effects from multiple mines would increase the prevalence
and cumulative impacts of the risks described for the one-mine scenario. Again, for the cultural
assessment, the conclusion is made that effects on humans would be primarily "direct and
indirect loss of food sources" (7-15). As the number of large-scale mines increases in this region,
the entire subsistence way of life could come under threat. This would be a much larger impact
than lost food sources.
RESPONSE: The text of the revised report (Chapter 12) lias been expanded to clarify that a
loss of the subsistence way of life goes well beyond the low oj jood resources.
Dennis D. Dauble, Ph.D.
Individual risk is described in varying levels o I" detail with overall risk or effects considered to be
largely additive. However, the relative magnitude of the effects of mining each ore deposit is
difficult to discern. It is possible that one of the smaller ore sites could be developed within an
acceptable risk scenario, but it is difficult to determine given that the assessment is largely built
on potential impacts of the Pebble Mine. To put things in perspective (individually and
cumulatively), there should be a discussion of habitat lost given each individual mine footprint,
during normal operation (includes water treatment and withdrawal) and as a result of pollutant
exposure. Also, Section 7.4.1 of the main report provides estimates of stream miles affected due
to blockage and elimination, but provides nothing quantitative for other direct and indirect
impacts of mine operation. The cumulative risk discussion in Chapter 7 could be expanded to
link up with the conceptual model described in Chapter 3.
RESPONSE: We cannot reliably predict the habitat loss due to additional mines. We have
tried to describe plausible examples of where additional mines could be developed on the basis
of active exploration on existing claim blocks. In the revised assessment (Chapter 13) we
predicted aquatic resource impacts based on a typical mine footprint being constructed
anywhere in the block and an average stream and wetlands density. We also expanded
discussion of the impacts of ancillary mine infrastructure and induced development This
provides a conservative estimate of the cumulative effects of multiple mines and sheds light on
whether cumulative effects are a significant concern. We have developed a specific conceptual
model for cumulative impacts and used it to enhance the discussion.
Gordon H. Reeves, Ph.D.
I found this chapter well done, even though the analysis was less extensive than what was done
for the Pebble Mine. It is clear that multiple populations would be put at varying degrees of risk
simultaneously if mine development occurs as is portrayed in this report. This certainly could
compromise the "portfolio effect" (Schindler et al. 2010 Nature 46513 June
2010|doi: 10.1038/nature09060), which has maintained the long-term productivity of sockeye
salmon in Bristol Bay.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 162
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We have added the portfolio effect to the discussion of the potential effect of
multiple mines.
Charles Wesley Slaushter, Ph.D.
Yes - but a qualified "yes." The Assessment appropriately outlines the probability of additional
resource extraction projects beyond Pebble itself, and recognizes that additional resource
opportunities (beyond the claims depicted in Figure 4-6), currently unknown or unverified, could
become viable or desirable to some interests in the future. Section 7.4 summarizes many of the
risks. However, the brief coverage (16 pages) accorded the entire subject of "cumulative risks"
is not consonant with the very long-term, spatially dispersed (and presumably linked by
transportation and communication corridors) impacts and risks of multiple mines (and associated
infrastructure) in many different sectors of the Bristol Bay watershed.
RESPONSE: We have added to the cumulative effects section (no»• Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative eff ects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources and /isli. It is intended to shed light on whether
cumulative effects are a significant, concern.
John IX Steel nick, Ph.D.
Cumulative risks can result from multiple risks (effects) from a single mine or individual risks
from multiple mines. This chapter identified potential risks from proposed mine activities;
without consideration of design standards or performance criteria, which is difficult without
specific mine designs/plans. Environmental risks were weighted equally - a TSF failure as
compared to a blocked culvert. The simple addition of stream length, as affected by various mine
footprints, does not represent a cumulative risk.
RESPOXSE. We have added to the cumulative eff ects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative eff ects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
Roy A. Stein, Ph.D.
Cumulative Risks from Multiple Mines. Clearly, as amply demonstrated in the Environmental
Risk Assessment, cumulative risks would be greater than those from just the Pebble Mine, even
though these risks are difficult to quantify. Important points made in this section deal with the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 163
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
economies of scale that would benefit additional mines coming to the Bristol Bay watershed.
After the first mine, new mines become more profitable simply because some of the
infrastructure (roads, power, fuel pipelines, etc.) has already been provided, thus reducing cost
outlays for the establishment of new mines. To me, this seems as quite an insidious process, for
once the door is swung open for the first mine, then many more will follow owing to
infrastructure considerations; with these additional mines come far greater cumulative
environmental risks. Quantifying these risks would help the reader and the public understand
what the ramifications of allowing one mine to begin operations might be. More text attempting
to quantify these cumulative impacts would be useful and instructive.
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative effects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andJisli. It is intended to shed light on whether
cumulative effects are a significant concern.
William A. Stubblefield, Ph.D.
The potential for cumulative risks associated with the development of multiple mines in the
Bristol Bay watershed is not treated with a great degree of detail. Although each of the potential
stressors (e.g., water withdrawal, habitat illumination, road and stream crossings) are
acknowledged and addressed, little quantitative consideration is given to the potential effects
associated with development of multiple mines. This, however, is probably appropriate given the
hypothetical nature of the single mine scenario and the potential for greater impacts associated
with the development ot multiple lines. Short of concluding that "failures at one mine could be
bad and failures at multiple mines could be worse," little else could be concluded. It is noted that
the multiple mines scenario leads to multiple tailings impoundments, more roads and culverts,
increased discharge potential of contaminated waters and increased habitat loss and reduction of
water resources and all of these lead to potentially greater environmental injury as a result of
failures.
RESPONSE: We have added to the cumulative effects section (now Chapter 13) by discussing
the effects of induced development, including transportation corridors, and the effects of
increased overall access in the watersheds (Box 6-1 discusses cumulative effects in terms of a
single mine). The cumulative eff ects assessment is not a definitive, quantitative evaluation. It
is, instead, a plausible example of how a mining district could develop and a conservative
estimate of the impacts to aquatic resources andfish. It is intended to shed light on whether
cumulative effects are a significant concern.
Dirk van Zyl, Ph.D., P.E.
The EPA Assessment does not appropriately describe the potential for cumulative risks from
multiple mines. In fact, the assessment does not identify the risks, only the likelihood of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 164
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
occurrence and the consequences. See discussion under Question 4 above about estimating and
expressing the magnitude of risks and what is required to appropriately describe the potential for
cumulative risks from multiple mines.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk. We have estimated those to
the extent that existing information allows.
The cumulative assessment is very conceptual at best, as there are no specific proposals from any
of the other potential resource areas. Cumulative impacts can only be evaluated once further
details about other potential mines and their plans are available. At this time, this section can at
best be seen as speculation.
RESPONSE: The cumulative effects discussion is not meant to be a definitive, quantitative
evaluation. We have presented a plausible example of lion- a mining district could develop and
a simple estimate of the impacts to aquatic resources andfish. It is intended to shed light on
whether cumulative effects are a significant concern.
It is impossible to improve this part of the assessment with the information on mine development
currently available; it can only be done when further information is published by the various
mining companies.
RESPONSE: Definitive quantitative risk assessments of future mines will need to wait for
more information on those mines and on the resources they could potentially impact This
assessment is meant to identify potential issues of concern with development of large scale
mining in the two watersheds. Based on the information available now, is the possibility of
cumulative impacts from multiple mines and induced development a significant concern? We
have utilized the existing information on the number of mining claims, the extent of recent
exploration activities, local and state land use plans, and the ubiquitous nature of fish habitat
in the watersheds to shed light on that question. Future en vironmental impact statements will
provide more definitive analyses. II e have tried to improve the cumulative analysis by
discussing the potential impacts from entire mine footprints, transportation corridors, induced
development and increased access.
Phyllis K. Weber Samuel/, Ph. I).
There are two issues that should be considered: cumulative effects from a single mine and
cumulative effects from multiple mines. Cumulative effects from a single mine might include
aquatic habitat degradation from non-point sources, including run-off from exposed mineralized
rock, seepage from the tailings impoundment, contaminated dust, noise, and other forms of
disturbance.
RESPONSE: We added a discussion on the cumulative effects from a single mine (Box 6-1),
as well as discussion of multiple transportation corridors, induced development and increased
access (Chapter 13).
Cumulative effects from multiple mines are difficult to predict because there are too many
unknowns. It is frequently to the advantage of a mining company to take advantage of existing
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 165
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
infrastructure, without building new camps, new mills, etc. It is also possible to use an old mine
pit for tailings or waste rock disposal from a new site; however, none of these features can be
determined until there is sufficient exploration to determine if mining is feasible, to characterize
the deposit, and to develop a detailed mine plan. To date, there is not sufficient information to
predict cumulative effects from multiple mines.
RESPONSE: We have expanded the discussion of cumulative impacts from multiple mines in
Chapter 13. The analysis of cumulative effects is largely qualitative rather than predictive. It is
a plausible example of future cumulative effects that can shed light on whether cumulative
effects is an important topic for consideration as plans for mining in the Nushagak and
Kvichak River watersheds move forward
Paul Whitney, Ph.D.
Need to Address Past Impacts. EPA's guidance for reviewing cumulative impact analyses (EPA
1999 - most recent guidance on EPA's website) asks that past, present and reasonably
foreseeable actions be considered. The assessment's coverage of cumulative risks from multiple
mines certainly addresses "reasonably foreseeable" mines. Not addressed in the analysis is past
impact(s) and how such impact(s) might be additive to current and foreseeable impacts. As
commented above, commercial fisheries remove a lot of the salmon (up to 70% - over 10 million
fish) annually from the drainages associated with the mine. The continual annual reduction, or
loss of energy and nutrients that might otherwise return to the ecosystems, should be considered
a past impact as part of the cumulative impact analysis.
RESPONSE: The purpose of the assessment is to evaluate potential effects of large-scale
mining on the fishery, not the effects of commercial fishing. Therefore, the assessment
considers the present condition of the watershed andfishery as a baseline for an evaluation of
potential future effects of large-scale mining. We have expanded the discussion on past,
present and reasonably foreseeable actions in Chapter 13. Since the fishery has sustained
itselffor over 100 years despite commercial fish harvesting, it is unclear to what degree
reduced escapement to the watersheds has had negative impacts. As Hilborn (2006) states,
"Recent paleoecological analysis of returns to Bristol Bay show no indication of decreased
production since commercial exploitation began. " Furthermore, the fishery is tightly managed
by the Alaska Department of Fish and Game, which reduces the harvest to increase
escapement when necessary.
PHWResponse: I'm just citing EPA guidance for Cumulative Impact Analyses . I'm taken back
that it is "unclear" to what degree reduced escapement to the watersheds has had negative
impacts. Page ES-24 of the revised assessment clearly states that "The loss of these nutrients
from a reduction in salmon would likely reduce the production of riparian or upland species." So
it must be clear to EPA that the loss of MDNs due to a mine would likely reduce ecosystem
productivity in the watersheds. If this is clear for mining. I can't understand why it is not clear to
EPA that reduced escapement to the watersheds from past, current and future commercial fishing
has had negative effects.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 166
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Several citations (including Hilbom 2005. Paulv et al. (2000. and Libralato et al. (2008) were
provided in my original comments that address the negative impact of fisheries on biodiversity,
energy flow and nutrient cycling in the ecosystem. If these citations are "unclear" and mv
previous comments about the impact of reduced MDN are unclear, there is not much else I can
do to explain that reduced escapement will result in less energy and MDN in the watersheds. To
me this is an obvious impact or as stated in Table 9 11 (9-50) a "Logical Implication". Perhaps
the fishery has sustained itself but a sustainable fishery is not evidence that there is no impact of
the fishery on the aquatic, riparian and terrestrial ecosystems of the Bristol Bay watershed., Formatted: Font: Not Bold, Not italic
I have prepared/managed cumulative impact analyses of fish and wildlife for a relatively small
(e.g., a 200-acre aggregate mine) and a relatively large 7,100-acre coal mine in Washington. A
cumulative impact analysis of past, present, and future fish and wildlife losses for the coal mine
expansion in a large watershed indicated cumulative impacts of the expansion were less than one
percent of past mining, agriculture, urban and forest activities in the watershed. In addition, the
cumulative loss could be fully mitigated by compensatory restoration. I would be interested to
know what the estimated fish and wildlife loss is due to the example copper mine in comparison
to the loss related to the commercial fishery. In addition, I would be interested to know if
potential fish and wildlife losses due to mining could be fully mitigated. If the watershed is
pristine or nearly pristine, the opportunity for compensatory mitigation may be low. If the past
impact of the commercial fishery is large and the watersheds are not pristine, there may be
opportunities. There is not much degraded habitat that could be improved by a mitigation plan. If
such a cumulative impact analysis were conducted, it may stimulate conversation about reducing
commercial fishing to compensate for impact losses due to mining.
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment. Appendix J includes a discussion of the challenges of creating equivalent
habitat for anadromous and resident fish and why this would be particularly challenging in
the context of the Bristol Bay watershed Compensatory mitigation requirements address the
needfor project proponents to replace aquatic resources and ecosystem functions that their
project has impacted Reducedfishing harvests would not replace lost spanning and rearing
habitat. Further, it would remove the burden of compensation from the party that caused the
damage.
PHWResponse: Mv intent for mentioning a reduction in commercial harvest to compensate for
potential impat(s) due to a large mine is more rhetorical than substantive. Mv intent was to
"stimulate conversation about reducing commercial fishing" or "wondering if it was technically
feasible" not to recommend it. The above EPA response is a good start at a discussion. Potential
impacts of the proposed mine could definitely reduce harvest and escapement and what is the
recourse? Some options for "out of kind" and "off site" compensations are discussed in the
recently added mitigation discussion and this is good.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 167
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Question 12. Are there reasonable mitigation measures that would reduce or minimize
the mining risks and impacts beyond those already described in the assessment? What
are those measures and how should they be integrated into the assessment? Realizing
that there are practical issues associated with implementation, what is the likelihood of
success of those measures?
David A. Atkins, M.S.
The Assessment describes what is considered to be conventional 'good' mining practice, but
does not adequately describe and assess mitigation measures that could be required by the
permitting and regulatory process. A thorough analysis of possible mitigation measures as
employed for other mining projects and the likelihood that they could be successful in this
environment would be necessary.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed modern
mining technology and operations. The terms are qualitative when generally interpreted, or
have a regulatory meaning. The term "best management practices " is a term generally applied
to specific measures for managing non-point source runoff from stormwater (40 CFR Part
130.2(m)). Measures for minimizing and controlling sources of pollution in other situations
are referred to as best practices, state of the practice, good practice, conventional, or simply
mitigation measures. H e have added a text box in the revision Chapter 4 to discuss terms.
Mitigation measures considered feasible, appropriate, and 'permittable' (as per Ghaffari et aL
2011), were considered in the assessment, and these are measures common to other copper
porphyry mines. Evaluation of alternative strategies (e.g., other options presented in Appendix
I for the mitigation of the same issue) should be part of the permitting process for a specific
mining plan. The State of I last,a does have statutory and/or regulatory requirements for an
approved Plan of Operations (II I I (
-------
Dr. Paul Whitney
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment. Appendix J includes a discussion of the challenges of creating equivalent
habitat for anadromous and resident fish and why this would be particularly challenging in
the context of the Bristol Bay watershed
It is also becoming common practice to offset impacts from project development with
preservation of equivalent habitat areas that are also at risk from development (http://bbop.forest-
trends.org/). It is unclear if this is a feasible consideration lor this project as this could involve
allowing one development (e.g., Pebble), while potentially taking away the development rights
of others (presumably for proper compensation).
RESPONSE: See response to above comment. The potential efficacy of using habitat
preservation as a form of compensatory mitigation for impacts in the llristol Bay watershed is
discussed in Appendix J.
Steve Buckley, M.S., CPG
There are many reasonable mitigation measures that could reduce the risks and impacts beyond
those described in the assessment. Some of these are contained in the Appendices and referenced
therein but not discussed in detail or described in the assessment. It is beyond the time
constraints provided in this review to develop an exhaustive research list of these potential
mitigation measures; however, EPA could include measures designed to: reduce the mine
footprint and limit the number of potentially affected watersheds; reduce, isolate, or eliminate the
amount of potentially acid generating waste; provide secondary containment measures for all
pipeline corridors; and use natural streambed arch culverts and bridges at fish bearing stream
crossings.
RESPONSE: The mitigation measures proposed within the mine scenarios are a subset of
options presented in Appendix /, all of which were presented as appropriate for the Pebble
deposit in (ihaffari et al. (2011). Appendix I includes multiple options for each source of
contamination, with some having specific applicability (e.g., site and material characteristics)
and others being more broadly applicable. Some additional mitigation measures have been
added to the scenarios in Chapter 6 of the revised assessment (e.g., processing PAG waste rock
over the life of the mine and selective flotation to minimize pyrite in tailings). Mitigation to
compensate for effects on aquatic resources that cannot be avoided or minimized by mine
design and operation (e.g., by reducing the footprint or limiting the number of affected
wetlands) would be addressed through a regulatory process that is beyond the scope of this
assessment. Nevertheless, in response to public and peer comments we have included a
discussion of compensatory mitigation in Appendix J of the revised assessment
Courtney Carothers, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 169
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
While I do not have knowledge of mitigation measures, a more thorough discussion of mitigation
measures could be included. Even if mitigation measures are largely deemed to be ineffective in
this case, they should be presented and evaluated as such.
RESPONSE: The EPA agrees with this comment and Appendix I is intended to provide that
discussion. The mitigation measures proposed within the mine scenarios are those that could
reasonably be expected to be proposedfor a real mine (they are a subset of options presented
in Appendix I), all of which were presented as appropriate for the Pebble deposit in Ghaffari
et aL (2011). Evaluation of alternative strategies (e.g., other options presented in Appendix I
for the mitigation of the same issue) should be done during a permitting process for a specific
mining plan. The assessment assumes that measures chosen Jor the scenarios would be as
effective as possible.
Dennis D. Dauble, Ph.D.
Potential mitigation measures are well described in Appendix I. I have no suggestions for
additional measures. Implementation of mitigation measures is entirely dependent on the
regulatory framework for operations and the oversight and monitoring practices that would be
mandated as a condition of the mining activity. Thus, some discussion of how/which mitigation
practices would be most applicable in the Bristol Bay watershed (and limitations thereof), given
constraints and characteristics of local hydrology and geology, is warranted.
RESPONSE: The mitigation measures proposed within the mine scenarios are those that
could reasonably be expected to be proposed for a real mine, are a subset of options presented
in Appendix I, and were presented as appropriate for the Pebble deposit in Ghaffari et aL
(2011). Evaluation of alternative strategies (e.g., other options presented in Appendix I for the
mitigation of the same issue) should be done during a permitting process for a specific mining
plan Permit applicant mitigation measures that reduce the risks identified in the assessment
would be welcome during the application process.
Gordon H. Reeves, Ph.D.
I identified one potential mitigation for culverts - the use of open arch types that are at least one
bankfull width in si/e. As described in my response to Question 7, this could reduce many of the
potential impacts raised by the authors.
RESPONSE: Suggestion noted. In our scenarios (Chapter 6), we assumed state-of-the-art
practices for design, construction, and operation of the road infrastructure, including design
of bridges and culverts for fish passage.
Charles Wesley Slaughter, Ph.D.
If it is assumed that the PLP project, or some similar development, were to go forward, I cannot
suggest mitigation measures beyond those discussed above. Since a major concern for salmonids
- perhaps THE major concern - is with consequences of the transportation corridor, simply
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 170
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
having the mine without the roads/pipelines would alleviate much potential risk. However, there
is presumably no practical, economically feasible way to not have the transportation corridor; air
transport of all materials to and from the site might technically be possible, but would not be
economically feasible.
RESPONSE: We agree that it would be infeasible to have a mine in this location without also
having road access. No change suggested or required
John D. Stednick, Ph.D.
The purpose of this assessment is not to identify mitigation measures. This suggests that things
can be fixed by mitigation. Risks were identified for a variety of situations, and the preventative
measures would better address the mining impacts. Mitigation measures are also a mining cost
that needs to be determined by the mining company and compliance with state and Federal
regulatory authorities.
RESPONSE: Comment noted We agree that mitigation measures required for an actual mine
would be determined through the regulatory proces s. No change suggested or required
Roy A. Stein, Ph.D.
Mitigation: Complete Tailings Storage Facility (TSF) Failure. In some ways, some of the
failures reviewed herein are not really subject to mitigation. For example, if a Tailings Storage
Facility (TSF) completely fails, options for mitigation become limited very quickly. With a
complete failure, it is "game over", with toxic sediments flowing into the Nushagak River all the
way to Bristol Bay, thereby destroying the entirety of salmonid spawning habitat in this river by
redirecting the channel and inundating the gravel/cobble steam bed with sediment (meters in
depth). Mitigation under these circumstances is impossible, in my view. Given this scenario, I
was surprised that the impacts were only assessed 30 km downstream of the TSF failure; is this
realistic? I think not, given what we know of other mines.
RESPONSE: We acknowledge the limitations of the hydrologic modeling. We now emphasize
more clearly the potential for the effects of TSFfailure to extend to Bristol Bay. The potential
for remediation (and associated risks) are discussed in Section 9.7.
The amount of text dedicated to a TSF failure is large (36 pages) as compared to other failures. I
suggest this section be shortened to bring it more in line with other sections. Briefly
summarizing impacts would focus the text and help the reader appreciate what it would mean to
have a TSF failure without having to wade through so much text.
RESPONSE: We have revised this section of the assessment. However, given its importance to
stakeholders, local communities, and decision makers, it was not significantly shortened
Other accidents andfailures are now addressed in more detail
Dry Stacking Mine Tailings: Appendix I, page 9. Given the horrific impact of a TSF dam
failure, should mine operators consider a relatively new technique incorporating "paste tailings
technology"? Here, tailings are thickened by water removal (down to 20% water) and filtering;
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 171
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
tailings are then dry stacked onto a lined disposal site. These stacks "have a lower potential for
structural failure and environmental impacts" (Martin et al. 2002). I would encourage the
authors to argue for this substitute for the more traditional TSF for this solves at least two
problems: 1) eliminates the possibility of a TSF failure (a huge gain!) and 2) reduces the amount
of monitoring and maintenance of the waste tailings "In Perpetuity" (another major gain).
Finally, what does "Best Management Practices" have to say about these two approaches to the
storage of tailings waste?
RESPONSE: We agree that dry stack storage and paste storage are options for mine
proponents to consider. Although the option of using paste tailings is one that might be
explored, the type of tailings disposal that operators propose is generally based on two things:
the type of mine and the quantity of tailings. A balance would need to be considered between
the low risk/high impact of a tailings dam failure and the certainty of increased area of
disposal and the increased cost of tailings handling for paste tailings. A "rule of thumb" for
design of dry stack tailings is to allot 25 acres for every thousand dry tons of tailings per day
over the life of a 20 year operation (SME Mining Engineering Handbook 1973). This would
amount to 4,900 acres (25 *200000 * 99/1000). Since paste tailings are more voluminous than
dry and the scenario contains a 25 year mine life rather than 20 years, the area of a paste
tailings facility would exceed the calculated amount resulting in over a 1/3 increase in the
area occupied by tailings than contemplated in the current mine scenario. With a paste
tailings f acility, the monitoring and maintenance of a large dam is removed but that is
replaced by the monitoring and perhaps treatment of the leachate draining through the
tailings pile (that contains 20",, water) as well as the long-term maintenance of the cap that is
placed over the tailings at the end of the mine life. The choice of a tailings disposal
methodology may be affected by the need to avoid or minimize effects to wetlands or the
technical feasibility of constructing the structural components of the project.
Mitigation: Partial TSF Failure. If a TSF partially fails or is discovered beginning to fail, then
I believe mine operators have a chance to save the dam and thereby protect the river, but only if:
1) the appropriate Standard Operating Procedures (SOP) for an emergency response are in place,
2) the necessary equipment (my presumption here is that heavy, earth-moving equipment would
be required), materials, and supplies are onsite near the facility, and 3) trained personnel
(meaning that they have practiced these repair SOPs in the preparation for such an event) are
available for immediate action. One might argue that these procedures are more proactive than
mitigating and I would agree, reflecting the near impossibility of invoking any mitigation
measures associated with TSF failure.
RESPONSE: Our scenarios presuppose that there are mitigation measures in place, but
examines the effects of a failure of those measures. EPA agrees that SOPs are part of a
mining operation and plan; however, inclusion of SOPs and details for emergency responses
to a failing dam were outside the scope of this assessment. No change suggested or required
Dredging: Post TSF Failure. In the text (pages 6-1 to 6-2), a reference is made to dredging
materials out of the river post spill. I can't imagine this would mitigate any losses of spawning
substrate for salmon. Indeed, because only 5% fines in gravel substrates compromise salmon
reproductive success (and perhaps even selection of these areas for spawning in the first place),
removal of meters of sediment with a dredge doesn't seem to be a solution. Whereas dredging
might, in a best-case scenario, reduce the time to recovery of the substrate, I don't believe it will
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 172
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
hasten recovery significantly. Dredging also serves to bring toxic sediments up into the water
column perhaps compromising all organisms in the system. By not dredging, we allow the
natural system to recover, which, in my view, would be preferable to any sort of "dredging
mitigation". Some reflection by the authors on this issue would be valuable.
RESPONSE: Agreed. Adverse effects of dredging are discussed in Section 9.7.
Mitigation: Pipeline Failure. With automatic shut-off valves stationed along all four pipelines,
we would expect to know precisely just how much effluent will be spilled during any single
event. With this information in hand, mine operators can easily anticipate spill size, toxicant
characteristics and thereby judge what equipment, materials, and supplies would be necessary for
mitigating any spill. As with the TSF failure, mine operators should have in place: 1) the
appropriate Standard Operating Procedures (SOP) for an emergency response, 2) the necessary
equipment, materials, and supplies onsite near the pipeline (given the length of the road, these
items should be cached at several locations along the road, such that response time is
minimized), and 3) trained personnel (meaning that they have practiced these mitigation SOPs in
the preparation for such an event) are available for immediate action. Shouldn't pipes be double-
walled? Again, what would "Best Mining Practices" say in this context'?
RESPONSE: The draft assessment included pipeline failure scenarios that quantified the
magnitude of spill events. The draft included the most commonly used and accepted practice
of using double walled pipes where above ground and water. There is, to the best of our
knowledge, no standard SOP for a product concentrate slurry spill. As is discussed in the
assessment, terrestrial spills would be excavated. Spills to streams might also be excavated,
although the physical damage to habitat would be considerable and the efficacy of excavation
would depend on how far stream flow had spread the material An assessment of diesel spills
has been added to the revised assessment in Chapter //, and it refers to a NOAA and API
report for standard remedial practices.
Mitigation: Failure of \\ aler and Leachate Collection. This failure differs from TSF and
pipeline failure, where failures are more akin to catastrophic, for here failure is somewhat more
gradual in coming (my guess is). Proactive vigilance is the watch phrase here where continual,
careful monitoring will indicate when failure begins. Because the "spill potential" is relatively
small (certainly compared to a TSF failure), less urgency is required on the part of mine
operators. However, as pointed out previously, just because the potential is small, over time the
impacts could be great. Hence, the mitigation undertaken with water and leachate collections
would require (one would hope) j list the tweaking of the collection system in place to eliminate
leakage through time. Again, personnel trained in how to respond to these gradual increases in
water and leachate leaks are required to stay ahead of this issue, thus preventing any toxic
materials from flowing downstream into the Nushagak and Kvichak rivers.
RESPONSE: The revised assessment acknowledges that the leachate capture system is not
likely to capture all leachate during routine operations and shows that water quality standards
would be exceeded in a considerable stream length as a result. Mitigation measures that
ensure or maximize the detection of leachate in groundwater before it reaches surface waters
would be vital
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 173
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
William A. Stubblefield, Ph.D.
I'm sure there are number of technological/engineering measures that could be implemented to
reduce the potential for environmental injury associated with development of mining in the
Bristol Bay watershed. The development of this a priori risk assessment provides useful
information in identifying were potential risks may exist and should provide mine development
professionals with the degree of guidance about the types of risks and potential consequences of
mine activity failures. Perhaps by recognizing the magnitude of adverse consequences associated
with potential failures, steps can be taken to implement safety measures early in the planning
process that would render mine development more acceptable. In addition, using the assessment
to define areas of uncertainty my provide direction for future research that would be beneficial
for the project. Again, because of the lack of detail associated with the hypothetical mine
scenario, it is impossible to estimate the likelihood of success of any mine control activities.
RESPONSE: One purpose of this assessment is to inform future decisions concerning mine
design and required mitigation. The purposes of tlic assessment are clari/ied in the revised
introduction (Chapter 1).
Dirk van Zvl, Ph.D., P.E.
Yes, there are reasonable mitigation measures that would reduce or minimize the mining risks
and impacts beyond those already described and incorporated by the EPA in the assessment.
There are a host of measures thai are not addressed in the assessment and lists of these are
identified below under the headings of regulatory and engineering. This list is by no means
exhaustive.
RESPONSE See response under the coinmenter's headings of "Regulatory" and
"Engineering " below.
While the EPA Assessment presents a series of potential mitigation measures in the main report,
the majority were rejected. Appendix J [sic] to the report also includes a generic discussion of
mitigation measures. The Main Report does not address the application or implications of these
in any project specific details, e.g. compensatory mitigation for wetlands, streams and other
aquatic resources.
RESPONSE: The El'. I did not reject the majority of potential mitigation options mentioned in
the report The mitigation measures proposed within the mine scenarios are those that could
reasonably be expected to be proposedfor a real mine. They are a subset of options presented
in Appendix I, were presented as appropriate for the Pebble deposit in Ghaffari et al (2011),
and are common with mining ofporphyry copper deposits. Mine design for the scenarios is
closely based on that presented in Ghaffari et aL (2011) and incorporates the mitigation
measures included in that mine design. Mitigation to compensate for effects on aquatic
resources that cannot be avoided or minimized by mine design and operation would be
addressed through a regulatory process that is beyond the scope of this assessment.
Nevertheless, in response to public and peer comments we have included a discussion of
compensatory mitigation in Appendix J of the revised assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 174
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Multi-stakeholder engagement processes, such as Failure Mode and Effects Analysis, can be
used to further expand on these mitigation measures. It is recommended that EPA recognize
these and potentially other measures that may be proposed in the public comments and make a
serious effort in including the potential effects of these on failure likelihoods, consequences and
risk magnitudes. It is an important aspect of improving on the range of potential outcomes.
RESPONSE: The EPA recognizes that other mitigation measures and mine designs have the
potential to address impacts and risks identified in the assessment. Evaluating the risks of all
potential mitigation measures and mine designs is not a goal of this assessment, which
identifies potential risks of concern from plausible mine scenarios based on current practices
and a recent plan (Ghaffari et aL 2011). A permit applicant \ mitigation measures that further
reduce the risks identified in the assessment are appropriate /or the application process.
Regulatory. The EPA Assessment neglects the typical outcomes resulting from the permitting
and regulatory processes for new mines, where permit stipulations may require specific actions
resulting from discussions, public comments and regulatory frameworks. The following is a
partial list of these:
• Section 404 of the CWA (See discussion under Question 3 above).
• Permitting stipulations and requirements for monitoring. Using permit stipulations for
monitoring can reduce a large number of the consequences identified in the report. For
example, the consequences associated with blocked culverts, etc. can be significantly
reduced by permit monitoring requirements and subsequent enforcement.
• Financial assurance will be required for mine closure. Financial assurance can be a very
beneficial tool during operations, premature closures, mine closure, as well as post-
closure monitoring and maintenance. Experience gained during operations can help
develop the closure and post-closure financial assurance requirements.
RESPONSE: The assessment is intended to identify impacts and risks of concern which may
inform future NEI'A and permitting processes. The revised assessment presents basic
discussion of financial assurance and listings of requirements for permitting a mine in
Chapter 4. The assessment is not a mining plan and is not meant to replace the regulatory
process for development of mining plans for these prospects, but can inform that process
about important risks that should he addressed
Engineering. A number of engineering options are mentioned in the report but discounted in
many cases. Many of these engineering mitigations are currently used in the industry. The
following may repeat a number of those mentioned in the report:
• Redundancy, e.g. additional embankments may be considered downstream of the TSF to
contain tailings and supernatant that may be transported as a result of TMF failure. While
this may result in a larger local surface impact, it will protect downstream waters in the
case of a failure resulting in tailings discharge.
• Further processing of tailings to remove the remaining sulfides.
• Tailings management options other than slurry deposition, such as production and
management of filter cake.
• Waste rock management options to reduce releases during operations, e.g. addition of
lime to the PAG rock.
• High standards implemented for road design, construction, monitoring and maintenance.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 175
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
• Double containment of all pipes containing concentrate and other materials. This is
already required under the International Cyanide Code for all pipelines containing
cyanide solutions.
RESPONSE: The revised assessment presents background information on mining, including
a sub-set of mitigation measures from Appendix I, in Chapter 4 and the scenario-specific
information in Chapter 6. The option of selective flotation to reduce the amount of sulfide in
the bulk tailings, as well as processing of PAG waste rock over the course of operations, have
been included in Chapter 6. In the draft assessment, selective flotation was discussed in the
background information, but was missed in the scenarios. A double-walled pipeline along the
entire length of a pipeline might be desiredfrom a purely environmental protection
standpoint; however, it may not be feasible or cost effective to do this given the length of the
pipeline. Therefore, the assessment proposed the most commonly used (and accepted method)
of double-walled construction over any water bodies. While the EPA opted not to include
redundancy in tailings embankments in the scenarios, this option, as well as full pipeline
double-wall construction, could be evaluated J or appropriateness during the regulatory
process for any future permit application.
The likelihood of success of these proposed and other potential mitigation measures can be
evaluated by considering their impacts on the overall project. A range of alternative project
technical alternatives and facility-siting locations will have to be developed instead of using only
the hypothetical scenario.
RESPONSE: The purpose of the assessment is to describe the potential adverse environmental
effects that could exist even with appropriate and eff ective site mitigation measures.
Considering alternative options is part of a permit application process and outside the scope of
this assessment No change required.
Phyllis K. Weber Samnell, Ph.D.
There are many avoidance or mitigation measures that would be implemented to reduce or
minimize mining risks. I have described some possible approaches when answering the previous
questions. To summarize:
The two most important questions for reducing or minimizing mining risks are:
• Can a mine in this area be designed for closure?
• Is it acceptable to develop and operate a mine that will require essentially perpetual
treatment?
RESPONSE: EPA agrees that these are key questions that must be addressed in the regulatory
process. Our purpose in the assessment is to evaluate the risks resulting from a mine operated
with modern conventional mitigation measures for design, operation, monitoring and
maintenance, and closure. The regulatory process addresses significant and unacceptable
risks. Chapter 4 of the revised assessment discusses regulatory andfinancial assurance
requirements for mining in Alaska
Specific Measures that can be taken to minimize risk include:
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 176
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Limiting metals contamination and acid drainage:
• Design the mine pit to limit oxidation on pit walls. Where feasible, conduct concurrent
reclamation.
• Develop plans for classification and storage of waste rock, lower grade ore, overburden,
and high grade ore.
• Develop and maintain tailings storage facilities with fail-safe provisions. An emergency
discharge of untreated waters from a tailings storage facility could be made to a
collection pond for later treatment or the tailings pond could be engineered to
accommodate a higher flood event so the likelihood of overtopping is minimized.
Consider alternate methods for tailings disposal (dry slack following sulfide removal,
etc.).
• Implement concurrent reclamation of disturbed areas, including stripped areas and mine
pits.
• Collect and treat point and non-point source water.
• Design and implement plans for the quantity and timing of discharges of treated water;
especially if the treated water is high in total dissolved solids. Monitor ground water,
seepage water, and surface water.
• Design system for collection and bypass of clean water and collection and diversion of
contaminated water to a water treatment system.
• Require stations for truck wheel washing.
RESPONSE: Many of the measures presented here were included in the scenarios and in
Appendix I and are mitigation measures commonly included for mining of this type. Other
bullets noted here are good suggestions for things to address during the regulatory process,
shoulil a permit application be submitted.
Protection of Eish Habitat:
• Review all in-stream activities in waters important to the spawning, rearing, or migration
of anadromous and resident fish.
• Design and implement a biomonitoring program.
• Review every road crossing of lish bearing waters to ensure free passage of fish.
RESPONSL: Suggestions noted for consideration during any future regulatory permitting
process.
Possible Measures to Limit Lffects to Wildlife:
• At the planning stages, design aspects of the project to create or enhance wetland and
aquatic habitats for fish, bird, and wildlife species.
• Limit truck traffic on the haul road during migrations.
• Incinerate all kitchen waste.
• Educate workers on bear (or other wildlife) safety.
• Limit air traffic and noise during critical times of the year.
RESPONSE: Suggestions notedfor consideration during any future regulatory permitting
process.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 177
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Paul Whitney, Ph.D.
Comments on Mitigation. The key word here is "reasonable." What is reasonable to a person not
involved in mining on a day to day basis will likely not be reasonable to a mining company
executive or mining engineer. The likelihood that "reasonable" means different things to
different people is exacerbated by the mixed messages regarding "best" mining practices (e.g.,
page ES-10, five lines from the bottom) versus "not necessarily best" mining practices (e.g., page
4-17, four lines from the top). As mentioned above, both of these statements can't be accurate.
For purposes of this discussion, I am assuming that there are many more mining practices that
could be proposed to address many of the uncertainties mentioned in the assessment. First of all,
most of the mitigation measures mentioned in the assessment provide one line of protection and I
suspect there are many types of redundant mitigation thai could be implemented. Redundant
protection such as: double-walled pipes in all sections lhal cross floodplains (above and below
grade); poly liner/vegetated caps to soak up and capture run oil' water before it contacts waste
rock; redundant clay, glacial till (waste360.com/mag/waste_landfills_glacial_till); poly liners;
and secondary liquid collection systems.
RESPONSE: The EPA has revised the assessment to better explain the qualifying terms. What
was intended is that we have assumed modern conventional mining technology and
operations. The terms are qualitative when generally interpreted, or have a regulatory
meaning ("best management practices "). and thus we have eliminated their use in the revised
assessment. Measures for minimizing and controlling sources of pollution in other situations
often are referred to as best practices, state of the practice, good practice, conventional, or
simply mitigation measures. He assume that these types of measures would be applied
throughout a mine as it is constructed, operated, closed, and post-closure, regardless of the
qualifier that one wishes to place with it. To remove any ambiguity and subjectiveness of terms
"good" or "best", we have removed them in the revision and have added Box 4-1, which
includes definitions for several terms used. Double-walled pipelines were included in the
scenarios for sections above or beneath water bodies. Redundant mitigation measures are
appropriate for consideration during the permitting process.
1 appreciate the assessment's discussion of the potential difficulties of using poly liners, but the
discussion might benefit from a review of Koerner et al. (2005) that provides another perspective
on the life of HDPE liners. The relationship of liner life to temperature is presented in their Table
2 and a summary in the text indicates that covered liners have a "half life of 446 years at 20
degrees Centigrade." This is a lot different from the 20 to 30 year estimate of service life cited in
the assessment on page 4-1 l(last para). Appendix I does cite a 2011 version of the Koerner et al.
(2005) paper and appears to misquote it (page 9, last para, last full sentence). The Koerner et al.
(2005) paper estimates a "halflife of 446 years" not a "lifetime" of 446 years, as cited in
Appendix I. Perhaps Koerner updated his 2005 estimate in the 2011 version.
RESPONSE: In the 2011 Koerner reference, there is conflicting text - the 446years is listed
as the "lifetime", and later also as "half-life"; "449"years is also stated as the half-life and
one is referred to the Table in the report with the title of "Lifetime", where there is no such
number. It appears the author is using the terms half-life and lifetime to mean the same thing.
Later in the white paper, there is text that states "... its predicted lifetime (as measured by its
half-life)..." The following text has been used in Appendix I and is consistent with the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 178
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
reference: ..Koerner et ah (2011) presents that a nonexposed HDPE liner could have a
predicted lifetime ("as measured by its half-life") of 69 years at 40 °C to 446years at 20 °C."
The discussion of liners has been moved to Chapter 4 in the revised assessment and includes
69 to 600 years as half-lives, based on Rowe 2005 and Koerner et aL 2011. The personal
communication reference has been removed
PHWResponse: Thanks for looking into this issue and updating the text. EPA might want to
consider under what circumstances a non-exposed liner would experience 40 degree C
temperatures in the Bristol Bay watershed. 20 degree C might even be wanner than a non-
exposed liner would experience. A Formatted: Font: Not Bold, Not Italic
I assume a mining engineer (if asked) could design a series of smaller impoundments or
innovative lined impoundments that could avoid a lot of the problems cited in the assessment.
The trade-off might be increased loss of natural resources due to a larger footprint and increased
construction cost, but the risk cited in the assessment and the risk of a catastrophic failure might
be greatly lowered.
RESPONSE: An engineer could design a series of smaller impoundments that would likely
decrease the risk of catastrophic failure, but would impact a larger footprint and increase
construction costs. However, cumulative risk of failure would increase. More than one dam,
even if they are smaller, means a higher risk offailures, even if they contain less volume, as
well as more points where water management systems couldfail, even if the dam itself does
not.
I would be interested to know what THE BEST mining practices are. If the mitigation measures
mentioned in Appendix I are, in fact, the best, it should be so stated and taken into consideration
in the assessment. For example, page 4-21 of the assessment indicates the TSF would be unlined
and not have an impermeable barrier between the tailings and groundwater. Appendix I, page 9
indicates TSFs can be lined if problems are expected. There is a lot of good information in the
Koerner et al. (2005) paper and Appendix I. It seems the types of mitigation measures in
Appendix I could be better captured in the main report.
RESPONSE: The terms "best", "good" or other are qualitative when generally interpreted, or
have a regulatory meaning (best management practices when referring to storm water
measures). Measures for minimizing and controlling sources ofpollution in other situations
often are referred to as best practices, state of the practice, good practice, conventional, or
simply mitigation measures. The multiple measures presented in Appendix I are options that
might be "best", depending on the site-specific conditions and any other constraints. More
discussion on mitigation measures has been added to the scenarios in Chapter 6 of the revised
assessment.
Compensatory mitigation and reclamation are briefly mentioned in the assessment. A more
detailed discussion of the opportunities and feasibility of reclamation and compensatory
mitigation might reduce the likelihood of potential impacts of the example mine plan (see
response to Question 3).
RESPONSE: Mitigation to compensate for effects on aquatic resources that cannot be avoided
or minimized by mine design and operation would be addressed through a regulatory process
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 179
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
that is beyond the scope of this assessment. Nevertheless, in response to public and peer
comments we have included a discussion of compensatory mitigation in Appendix J of the
revised assessment.
Question 13. Does the assessment identify and evaluate the uncertainties associated
with the identified risks?
David A. Atkins, M.S
The Assessment states: the 'range of failures is wide, and the probability of occurrence of any of
them cannot be estimated from available data.' Uncertainly is addressed throughout the report,
typically with a qualitative discussion. There is a high degree of uncertainty with respect to how
the mine would be developed, operated and closed, as well as how any impacts would be
mitigated. This large uncertainty makes assessing risk difficult.
RESPONSE: No change suggested or required.
Steve Buckley, M.S., CPG
The assessment identifies some of the uncertainties associated with the identified risks but does
not evaluate these in great detail.
RESPONSE: No change suggested or required.
Courtney (druthers, Ph.D.
The report includes specific sub-sections to discuss uncertainties for the risks associated with
habitat modification (Section 5.2.4), pollutants (5.3.4), and water collection and treatment failure
(6.3.4). Uncertainties related to abundance and distribution of fish in the watershed draining the
mine site, road and stream crossings, salmon-mediated effects on wildlife, salmon-mediated
effects on human welfare and Alaska Native cultures, tailings dam failure, pipeline failure, and
road and culvert failures are not discussed in separate sections; however, several uncertainties
related to these risks are noted throughout the report, and in summary sections (Sections 8.5 and
8.6).
RESPONSE: We have added separate uncertainty sections to each of the major topics in the
risk assessment chapters.
The "sensitivity relative to overall results" of the key assumptions and uncertainties presented in
Table 4.8 in Appendix E (pp. 193-195) would be a helpful model to employ in the main report.
For non-experts in the technical dimensions of mine construction and operation, uncertainty
rankings would be useful. For example, "We are 'highly uncertain' about the accuracy of these
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 180
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
predictions given this unknown factor," or "We expect this uncertainty has a negligible effect on
the model we employ to calculate this risk."
RESPONSE: The assessment has been restructured to consider certain impacts resulting from
the mine footprint (Chapter 7) separately from less certain impacts resulting from potential
failures (e.g., Chapters 8 and 9). For uncertainties considered in each chapter, we also have,
when possible, indicated a general categorization of (1) the level of uncertainty, and (2) the
sensitivity of our conclusions to this uncertainty.
Dennis D. Dauble, Ph.D.
The most likely scenarios and probabilities of failure are described based on assumptions of
project size and magnitude. For the most part, estimated risks are conservative (i.e., effects are
stated as "likely" if no further information is available). A weakness of Integrated Risk
Characterization (Chapter 8 of the main report) is having a long list of identified uncertainties,
which leads one to speculate, "so what do we know?" Not being familiar with the formal risk
assessment process, it appears this "assessment" (which is loosely based on a risk assessment
framework), falls short of providing something with any degree of certainty.
RESPONSE: The assessment does not state that effects are likely if no further information is
available. The assessment, as far as the available information allows, identifies potential
events and their eff ects, their probabilities o f occurrence, and possible ranges. Uncertainties,
inherent in any risk assessment or mine plan, are dearly identified in the assessment
Gordon H. Reeves, Ph.D.
Uncertainties and limitations are explicitly identified and acknowledged for topics that I am
familiar with (fish and aquatic ecology and fish habitat) throughout the report. These are
summarized succinctly and clearly and the consequences to the findings are articulated.
RESPONSE: Yo change suggested or required.
Charles Wesley Slaughter, Ph.D.
Yes. The authors fairly attempt (pp. ES-24-26, and in each chapter) to note the various
uncertainties and assumptions incorporated into the Assessment. Sections 8.5 and 8.6 briefly
summarize those uncertainties. A question remains concerning the "uncertainties" associated
with assigning probabilities to various failure scenarios; I remain unconvinced that those
probabilities have real meaning or significance for decision-making (see response to Question 5,
above).
RESPONSE: As explained in the assessment (particularly in Chapter 14), the probabilities
have various sources and different interpretations, which we have tried to make clear. Some of
them are more useful in decision making than others, but all are the best values that could be
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 181
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
derivedfrom available information. Most are based on empirical frequencies, but some are
engineering goals or regulatory targets.
John D. Stednick, Ph.D.
The uncertainties are presented adequately.
RESPONSE: No change suggested or required
Roy A. Stein, Ph.D.
Uncertainties. I think that the Environmental Risk Assessment did a nice job of identifying
uncertainties surrounding this presentation, but a relatively poor job oI" quantifying them (at least
partially due to a lack of information). As a consequence, I found this section more than
disconcerting. Certainly, the authors have worked hard to present an accurate portrayal of the
impact of a large-scale open pit mine in the watershed of Bristol Bay. Even so, upon review of
the list of uncertainties with regard to this effort (pages 8-10 to 8-13), I conclude that we know
little of what the impact of this mine will be in any quantitative sense. Clearly, from the
Environmental Risk Assessment, we do know qualitatively what is likely to occur when a mine
of this size and type will be put into operation in this environment.
RESPONSE: The goal of the assessment is to evaluate risks from large-scale mining to
salmon populations and secondary effects of salmon losses. Where applicable information was
available, we quantified our uncertainties to the extent possible. Where information was
lacking, we provided a qualitative assessment. I 'liis approach also reveals missing information
that will be needed for any future environmental assessment.
However, from the list of uncertainties, we are operating at the outside edge (and beyond in
man) cases ) of the semi-predictive models used in anticipating the impacts of the mine footprint,
the routine operations of the mine, and the impacts of failures of TSF, pipelines, and
water/leachate collections on extant salmon populations. And our knowledge of the baseline
populations of the seven species of salmonids is no better, for we do not know the size, diversity,
distribution, or vital rates (i.e., recruitment, growth, and survival across life stage) of these fishes.
RESPONSE: The EPA agrees that we do not know everything we would like to know about
fish populations in the project area, but we do know the approximate extent of various species,
and we do know that spawning and rearing by salmon, Dolly Varden, and rainbow trout occur
in the project area. We summarize the types of risks to fish from plausible mine scenarios.
Couple these two sets of uncertainty and the prognosis outlined in the report is suspect, at the
very least, and somewhat anticipatory at best (I cannot bring myself to use the word
"predictive"). I fully realize that these are the cards the authors were dealt (I do applaud the
authors for making the best of an information-poor environment), but it seems to me that we are
on tenuous ground when we attempt to predict the impact of the Pebble Mine on salmon,
associated wildlife, and Native Alaskan cultures in the Bristol Bay Watershed.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 182
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: EPA agrees that the assessment is not a definitive quantified prediction of all
impacts to salmon, wildlife, and Alaska Native cultures of large scale mining. However, the
revised draft does develop more quantitative estimates of the implications of the mine
scenarios than the first review draft The point of a risk assessment is to make the best
prognostication possible given the available information. The complexity of a large mine and
of the receiving ecosystems, as well as the unpredictability of the natural processes and human
errors that would challenge the integrity of a mine, make prediction tenuous even with the
most detailed mine plan. However, the assessment evaluates the likely impacts of a set of
plausible mine scenarios, thereby highlighting potential risks and impacts of concern. This
will inform decisions going forward and any NEPA or permitting processes in the future.
William A. Stubblefield, Ph.D.
The risk assessment attempts to identify and evaluate the uncertainties associated with each of
the recognized potential risks. The authors have, for the most part, successfully identified a
number of uncertainties that may affect the accuracy and conclusions of the risk assessment.
Clearly, this information should provide a basis for prospective mine planners and regulatory
authorities to focus their efforts to minimize potential environmental risks. In some cases the
uncertainties identified are probably best addressed through the development of additional data
and this should guide future research efforts undertaken prior to mine development and
operation.
RESPONSE: No change suggested or required.
Dirk van Zvl, Ph.D., P.I:.
The EPA Assessment does not identify the risks, only the likelihood of occurrence and the
consequences. See discussion under Question 4 above about risk. Uncertainties are identified and
evaluated for the likelihoods of occurrence and in some cases for the consequences. However,
because the magnitudes of the risks are not expressed, their uncertainties are also not explicitly
expressed.
RESPONSE: The EPA believes that the likelihood of an occurrence and its consequences
constitute an appropriate and generally accepted definition of risk We have estimated those to
the extent that existing inf ormation allows.
The report identifies uncertainties in a number of sections, including in Chapter 8. In many cases,
these uncertainties are expressed in qualitative terms and are not quantified. The biggest
uncertainty/variability in the evaluation of a hypothetical project is associated with the potential
range of design features, waste management options and operational details that could be
included. This was completely overlooked in the analysis by assuming a specific design for the
hypothetical mine. The failure likelihoods and consequences on salmonid fish are very
dependent on the assumptions for the hypothetical mine. These uncertainties are neither clearly
identified nor included in the evaluations. This is a major shortcoming of the present analysis.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 183
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The comment is correct in stating that the scenarios are a major source of
uncertainty with respect to what might be implemented if an actual mine were developed.
However, the scenarios are intended to represent a mine using modern conventional practices
and are based on the preliminary mining plan put forward by Northern Dynasty Minerals as
"permittable" (Ghaffari et ah 2011). Any future plans put forward would undoubtedly differ,
and a permitted mine plan would differ from that put forward by the mining companies. The
actual operation of a mine, in turn, inevitably differs from the approved plans. Rather than
attempt to estimate the uncertainties in predicting all of those changes in plan, the EPA chose
to put forward reasonable and typical scenarios, estimate some associated risks, and describe
the uncertainties associated with the risk estimation. Identification of the risks associated with
a typical mine can inform initial mine design and the A//' I and permitting processes. The
revised draft makes this clear.
Phyllis K. Weber Scannell, Ph.D.
The important features of the Environmental Assessment are to describe the fish, wildlife, and
human use of the subject area and to define possible risks from development of a large porphyry
copper mine. There are many uncertainties associated with the identified risks and most were
identified in the document. The document could be strengthened by putting a greater emphasis
on sources of contamination (such as mine seepage, poorly designed collection systems, exposed
pit walls, etc.) in relation to the permeability of the soils.
RESPONSE: The revised assessment includes more emphasis on sources of contamination
(e.g., diesel pipeline failure, quantitative wastewater treatment plant failure, and a refined
seepage scenario) and their potential hydrologic transport including through permeable soil
and rock.
The 5th bullet on page 8-11 outlines important uncertainties for protecting fish species. These
uncertainties include life-stage-specific sensitivities to temperature, habitat structure, prey
availability, and sublethal toxicities. These factors must be considered should a mining project
go forward.
RESPONSE: The EPA agrees with this comment. No change suggested or required
The 6th bullet on this page discusses the preliminary nature of leaching test data. These tests
must be sufficiently comprehensive to predict both short term and long term water quality from
all sources, including PAG and NAG waste rock, pit walls, and pyritic tailings.
RESPONSE: The EPA used the available leaching test data and agrees with the comments
about future tests.
Paul Whitney. Ph.D.
Uncertainty summary. The discussion of uncertainties in the assessment is, in most cases,
appropriate. It seems that one could use these discussions as a scope for additional work needed
prior to an assessment that would properly assess risk of the example mine. As the uncertainty
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 184
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
discussion appears in the assessment, this reader wonders what to make of it. There is a lot of
uncertainty in this world that we find acceptable; the ultimate goal seems to determine if the
cumulative uncertainty is acceptable or not. Such an evaluation remains to be made and I'm not
sure how it could be made based on the level of information presented in the assessment and the
current state of the uncertainty discussions.
RESPONSE: The goal of the risk assessor is to describe the uncertainty, not determine
whether it is acceptable. The acceptability of existing uncertainty is a judgment made by the
risk manager, who must decide whether to make a decision or defer the decision until more
information is available.
The summaries of uncertainty included in Sections 8.5 and 8.6 could be improved if some sort of
realistic and useful conclusion(s) could be presented. The Section 8.6 summary seems to "pile
on" uncertainties, rather than summarize the uncertainties in the assessment. While piling on is
informative, it is not the sort of summary I was looking for. Conclusions in the Section 8.5
summary of conclusions remind us that the effects of mining on fish populations could not be
quantified and, as a substitute, the effects on habitat were used as a surrogate. So we are left with
an estimate of 87.5 to 141.4 km of streams that would be removed and this would cause an
adverse effect. Based on this very general risk conclusion, we learn that "In summary, it is
unlikely that there would be significant loss of salmon subsistent resources related to the mine
footprint" (page 5-77, last paragraph). First, it's not clear how this conclusion was reached.
Second, if such a conclusion was possible for subsistent resources based on the data available,
why couldn't such a conclusion be reached for sport and commercial fisheries? Third, does it
follow that no significant loss to salmon subsistence resources would result in no significant loss
to wildlife that utilize this resource? If so, such an indirect analysis is not informative regarding
an environmental assessment for the example mine. Alternatively, a direct assessment of the loss
of habitat using habitat-based population models for both fish and wildlife would be much more
informative.
RESPONSE: Uncertainties are now discussed in each chapter of the assessment so they can
be more closely associated with a particular topic.
Additional information on the geographic scope of subsistence use in the area has been added
to the assessment and the conclusions clarified accordingly. Although there is some
(incomplete) information on subsistence use areas, similar data are lacking for the geographic
scope of sport and commercial fishing in the watershed
PHWResponse: The above response doesn't answer my question. Does EPA stand by its
conclusion of unlikely significant loss of salmon subsistent resources related to the mine
footprint? | ( Formatted: Font: Not Bold, Not Italic
There is not sufficient data to quantify the loss of salmon as a subsistence resource under
various scenarios or to quantify the fish-mediated effects to wildlife. Discussions of these
topics have been expanded to illustrate the complexity of the interactions and provide further
qualitative assessment. Direct effects on wildlife from habitat loss related to large-scale mining
are outside of the scope of this assessment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 185
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Adaptive Management. Holling (1978) in his Adaptive Environmental Assessment and
Management book discusses political uncertainty and how adaptive management might be able
to address the issue. Considering that the mine being proposed is a multi-century system, it's
poignant to realize that Alaska was owned by Russia about 150 years ago and Oregon was being
claimed by the Spanish about 200 years ago.
RESPONSE: The commenter is correct in pointing out the inapplicability of adaptive
management to decisions like whether to permit large scale mine development in the Bristol
Bay watershed Not only is the time horizon inappropriate, but mine permitting decisions
cannot be remade iteratively like the resource management decisions for which Hollings and
Walters developed adaptive management (e.g., setting annual fishing limits). However, careful
monitoring and revision of permit conditions, although not literally adaptive management,
could have some of its benefits. Monitoring requirements and provisions for modifying permits
are beyond the scope of this assessment
Adaptive management is a tool designed to deal with uncertainties in risk evaluations (Ruhl and
Fischmann 2010). If implemented properly, w ilh testable hypotheses of risk, adaptive
management may be something to consider for the example mine. My experience with adaptive
management is that adequate funds are seldom allocated to learn by doing, to test hypotheses and
to implement new management if hypotheses are not met. Formalizing financial instruments to
ensure that funds are available is equally difficult to negotiate. Nonetheless, I agree with Ruhl
and Fischmann that the theory of adaptive management is sound and may be the only way to deal
with uncertainties such as climate change. Considering the number of uncertainties identified in
the assessment, for the scope of work to clearly state hypotheses, to address the uncertainties, to
fund studies to test the hypotheses, and to fund alternative management if hypotheses are not met
is cumulatively daunting. For example, there are about 50 state variables in each of the Site
Model Figures (3-2A. B. C and D). All totaled, that's about 200 (50 x 4 figures) state variables
for salmon alone. Considering there are 100s more species, that's about 20,000 (200 x 100
species) state variables. Then there are fluxes/linkages between the state variables and that is
another 20,000 fluxes/linkages that should be monitored. For adaptive management to work,
clear goals and hypotheses to assess whether observed data meet the goals should be stated for
approximately 40,000 variables and linkages. Then there is the task of defining the monitoring
methods and statistics to determine whether or not goals for the variables and linkages are being
met. I acknowledge that there are probably ways to trim down the monitoring effort, but one can
start to imagine the enormity of implementing a monitoring plan for adaptive management. Then
there is also what to do if goals are not realized. I am not aware of any alternative to adaptive
management other than contingency planning which often lacks the "learn by doing" feature of
adaptive management.
RESPONSE: See response to previous comment
Important wordsmithing. So many of the uncertainty evaluations make statements about certain
parameters that "could not be predicted" (page ES-20, para 4); "could not be quantified" (e.g.,
page 6-11, second full para); or are "unpredictable" (page 5-44, para, line 8). These are just a few
of many examples. It would be more acceptable, at least to me, to state that estimates were not
included in the assessment. This type of wording occurs in some parts of the assessment and
might be more accurate.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 186
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Changes have been made in Chapter 7 (formerly Chapter 5) and Chapter 9
(formerly Chapter 6).
Vague wording. The assessment includes a lot language that seems vague, at least to me. The list
is long but includes: "highly pure water"; "other ecological responses"; "key wildlife"; "essential
wildlife"; "overall ecosystem functioning"; "serious population-level consequences"; "different
thermal characteristics"; "could be locally significant"; and "very long time." Lackey (2001)
acknowledges the need for scientists to communicate with the public using normative science but
expresses concerns that normative descriptors such as ecosystem health are subject to wide
interpretation. He suggests that the most direct alternative to using normative science is to simply
and clearly describe what is being discussed. The assessment would benefit if the normative type
words used above (any many more) were quantified w ilh estimates, a range or some type of
measureable or testable parameter.
RESPONSE: The authors have tried to be precise with the language of the assessment and
avoid normative terms. Changes to the text of the revised assessment have been made with this
comment in mind
Question 14. Are there any other comments concerning the assessment, which have
not yet been addressed by the charge i/uestions. which /uinel members would like to
provide?
David A. Atkins, M.S.
Long-term risks from development of an open pit have not been characterized. It is difficult to
predict the chemistry of the lake that will form in the open pit, but there is some potential that
water quality will be poor, which may be exacerbated by pit backfilling with waste rock. The pit
lake could impact waterfowl and may have some impact on groundwater if there is outflow when
the lake reaches an equilibrium level.
RESPONSE: The potential pit was not assessed because it is not anticipated that salmon or
other fish (the primary assessment endpoint) would reach it. Therefore, it is out of scope.
However, a pit lake as a potential source of water to streams is briefly considered in the revised
Chapter 8. The scenarios no longer include the placement of acid generating waste rock in the
pit at closure.
Steve Buckley, M.S., CPG
None.
Courtney Carothers, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 187
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
None.
Dennis D. Dauble, Ph.D.
Based on public comments and discussions that took place by panel members in Anchorage
August 7-9 of this year, this report confuses in both intent and approach. Is the intent of EPA's
assessment to characterize potential impacts to the Bristol Bay watershed (title) or does it address
a more defined portion of the Nushagak River and Kvichak River watersheds (objective
statement)? Was the approach an "assessment" (a fairly broad term) or an "ecological risk
assessment" (suggests a specific scientific framework was applied to the risk/effects analyses)?
These shortcomings should be addressed in the final assessment document.
RESPONSE: We have revised the discussion of purpose, scope and endpoints in response to
this and other similar comments (see Chapters I through 5 of the revised assessment). The
assessment addresses multiple spatial scales, as detailed in Chapter 2. It is an ecological risk
assessment, but that term is usually shortened to assessment to make the document more
readable.
Gordon H. Reeves, Ph.D.
The major issue that was not considered in the assessment was the potential impact of climate
change, particularly regarding the form and liming of precipitation. Admittedly, there is
uncertainty about the magnitude of changes that will result from climate change, which makes it
difficult to consider. However, the potential consequences of climate changes on such topics as
tailing site facilities, water availability, and culvert failure seem appropriate. It will also be
important to consider potential impacts of climate change so their signal can be distinguished
from potential mine impacts during any monitoring that occurs.
RESPONSE: Climate change projections, including increases in precipitation, are addressed
in Chapter 3 of the revised assessment. Projected precipitation increases are used as the basis
for a qualitative discussion of the link among changes in precipitation, hydrology, and mining
including site facilities, water availability, and culverts (Box 14-2). This box also discusses the
importance of monitoring impacts of climate change to distinguish between potential impacts
of climate change and potential impacts from the mine.
Charles Wesley Slaushter, Ph.D.
I would simply re-emphasize that a truly comprehensive assessment of the potential
consequences of a large-scale mineral extraction project, be it the "hypothetical" Pebble-like
project or a different endeavor, should fully consider both the immediate project-specific
impacts, and the long-term watershed-wide consequences of "ancillary" developments - such as
other mines, which might become economic once the primary project's infrastructure is in place.
There should be full recognition of the irreversible nature of such developments, and of the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 188
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
potential and limitations of possible reclamation or mitigation measures for the full suite of
resources and ecosystem "services" involved, both short-term and long-term.
RESPONSE: Ecological risk assessments generally have more limited scopes than
comprehensive environmental impact statements. This document is not meant to be a
comprehensive evaluation of all impacts potentially stemming from a large-scale mine, and its
scope has been more clearly defined in Chapter 2. We have added discussion to the cumulative
assessment chapter (Chapter 13) to emphasize the long-term watershed-wide consequences of
ancillary or induced development We have added a discussion of compensatory mitigation in
Appendix J.
John D. Stednick, Ph.D.
There are several references to streamflow measurements that would be especially helpful to
better characterize the site. The US Geological Survey has some streamflow gauging stations and
precipitation records that would complement the analysis. Annual precipitation values were
derived apparently from a computer model used to analyze global climate change at University
of Alaska Fairbanks. How do these data compare to field measurements? The prediction of a 10,
50, 100, or larger event using a short-term precipitation record, results in a larger error term on
the predicted streamflow. How common is the occurrence of rain on snow (ROS) streamflow
events?
RESPONSE: U.S. Geological Survey and PLP data are presented in Chapter 7 of the
assessment. We believe the SXiP data used in the assessment to be the best available for
summarizing precipitation inputs to large areas; these data have been updatedfrom the May
2012 draft of the assessment. We used accepted, published data to estimate the size of the
probable maximum flood. II e do not make estimates of 10, SO, or 100 year events.
Dust production and transport: A variety of mining processes will generate dust. What are the
wind patterns, chemical composition, and opportunity to land in surface waters or wetland areas?
What potential is there for metal or toxin transport? Overburden removal will require explosives
that leave nitrate, ammonia, and often sulfur in the air. What about this transport? Or rain out?
RESPONSE: Stressors evaluated in the assessment were based on how they would
significantly affect our primary endpoint of interest (the region's salmon resources) and their
relevance to EPA's regulatory authority and decision-making context Although fugitive dusts
from mining operations are a potential source of contamination to streams, and should be
considered in a regulatory permitting process, they are not considered as significant as other
stressors chosen for the assessment and are not regulated under the Clean Water Act.
Discussions of transport dust and associated potential impacts are now included in Chapter
10.
The literature cited is often dated or lacking. The technical review panel has proposed numerous
references that can be used to strengthen the document.
RESPONSE: We have reviewed all references suggested by the peer reviewers to the extent
possible, and have incorporated information where applicable.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 189
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Roy A. Stein, Ph.D.
• IN PERPETUITY
Sustainable Salmon vs. One-Time Mine. Some irony exists as one considers the trade-off
between salmon and this mining operation (and make no mistake, we cannot have both
mining and productive salmon stocks in the Bristol Bay watershed). We are trading
sustainable salmon stocks that, with science-driven management, rigorous regulatory
oversight, and limited exploitation, should provide salmon literally 1000s of years into the
future against the development of a mine that will provide minerals in the relative short term
(within 25 to 78 years). As a result of the mining operation, the government (and likely it
will be the state or Federal government) will be saddled with a 1000 years (at a minimum,
based on the assessment) of monitoring and maintenance of this closed-mine site.
RESPONSE: No change suggested or required.
• MINERAL NEEDS
• Strategic Needs. I was surprised that no section of the Environmental Risk Assessment
included any justification for why copper (does mining copper fulfill a strategic need for the
United States; most all sites I researched do not list copper as a strategic mineral for our
country), gold, molybdenum, and some additional rare earth elements were needed within the
context of our economy. Are other sources available? Are these specific elements in short
supply? Are they required for the United States to compete in worldwide markets regarding
cell phones, other electronic devices, or solar panels? For example, rhenium is used in the
aviation industry and some web sites suggest it is critical to our defense industry. Some
justification would have helped me understand this huge undertaking.
RESPONSE: The EPA agrees that the strategic need for these resources should be
discussed in the broader XICI'A environmental assessment process, but this topic is beyond
the scope of this assessment.
• ORGANIZATIONAL ISSUES
Page 5-59. I struggled throughout the document with organizational issues. As I read more
and more text, I had the sense that I had read these facts or these perspectives previously. I
mention this above but it is here on page 5-59 that the issue is nicely summarized. Note the
text in the last paragraph of the page, where it discusses all of the important issues associated
with roads and stream crossings. My suggestion would be that these sections (for every
topic) be combined such that one section would exist for Roads, one for the Pipelines, etc. In
so doing, the reader can capture all of the relevant information about a specific aspect of the
mine in a single section of the report. I believe this would improve impact, readability, and
shorten the report substantially (and also serve to reduce what seems to be a fair amount of
redundancy).
RESPONSE: We have reorganized the document in response to this and similar comments
to improve readability and clarity.
Pages 6-10 to 6-11. These pages reflect another example of redundancy. Text to this point
discussed and reviewed just how long we might expect the fine sediments to persist in rivers
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 190
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
and streams post tailings dam failure. Yet, here again, on pages 6-10 to 6-11, these numbers
are reiterated. Combining these sections would help the reader and reduce redundancy.
RESPONSE: The TSFfailure analysis was formerly split between two chapters and has
now been combined into one (Chapter 9).
Section 6.3, Pages 6-36 to 6-42. Not to beat a dead horse, but in this section, nearly all of
the citations to tables and figures are to those tables and figures that are found in sections of
the report other than Section 6. This organizational scheme is what makes the report
cumbersome to read and follow the logic and the argument.
RESPONSE: The assessment has been completely reorganized, as suggested by reviewers.
William A. Stubblefield, Ph.D.
None
Dirk van Zvl, Ph.D., P.E.
The EPA Assessment mentions twice that "interactions with regional stakeholders" and
interactions with members of the Intergovernmental Technical Team were used to refine the
analysis, etc. (p. ES-2 and p. 3-6). A robust stakeholder process includes careful documentation
of the stakeholders identified during the project (which may include stakeholder mapping),
records of meetings (attendee lists, meeting notes, etc.), resolution of differences, etc. The EPA
Assessment does not contain any references to any such materials, which implies to me that the
stakeholder process was informal and not robust.
RESP(> \ SE: Details on stakeholder involvement in the assessment process have been added
in Box I-I.
Phyllis K. II chcr Scanncll, Ph.D.
At present, Pebble remains a prospect and there is no plan of operations for the mine. Should the
project move forward to development of a mine, it will be necessary to develop an in-depth
mining plan of operations. The mining plan should include the following:
• Transportation - of equipment and personnel and for shipping ore. Transportation of
ore, including loading facilities, wheel washing, and other measures to prevent ore
spillage and contamination.
• Siting of mine facilities, including tailings ponds, waste rock storage areas,
concentrate storage area, bypass systems for clean water, and collection systems for
contaminated water.
• Mill operations, including a description of the process for concentrating ore.
• Chemical and fuel storage and Spill Prevention and Contingency Plans.
• Personnel housing, including handling of domestic waste (sewage, garbage).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 191
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
• Water treatment plant. Processes that will be used, anticipated concentrations of
metals and TDS, anticipated discharge volumes, and predicted mass loadings.
• Monitoring plans for seepage from tailings ponds, waste rock storage areas, etc.
Monitoring likely will include a series of wells and possibly, a pump-back system.
• Predictions for acid rock generation and measures that will be put in place during
mining to minimize future seepage from the mine site.
• Plans for concurrent reclamation and future closure of the mine.
• Specifications for sufficient bonding to provide site stabilization and water treatment
in the event of a premature or temporary shut-down and reclamation at closure.
RESPONSE: The commenter is correct that these are important points that should be
considered in evaluation of a mining plan once submitted. \ o change suggested or required
After the Mine Plan of Operations is developed, an environmental assessment plan should be
developed that identifies potential effects to fish and wildlife and their habitats from specific
components of the mine (as listed above). In addition, the assessment should include cumulative
effects of nearby mines (if appropriate) on fish and wildlife habitats and water quality.
RESPONSE: The commenter is correct and we would expect these things to be considered
during the regulatory permitting process. So change suggested or required.
Among the most important issues that must be addressed are transportation, potential for acid
mine drainage and metals leachate, control of point and non-point pollution, and developing the
mine for future closure.
RESPONSE: The commenter is correct and we would expect these things to be considered in
more detail during the regulatory permitting process. No change suggested or required
Paul II liitnev, Ph.D.
Here are a number of other comments that do not fit neatly under the first 13 questions:
1. Unquantifiable wetland and riparian loss. There are many parameters that are supposedly
unquantifiable. An example "unquantifiable" parameter is the area of riparian floodplain
(page ES-14). Another example is the "unquantifiable area of riparian floodplain and
wetland habitat that would be lost..." (page 8-2, first sentence). There are many methods
to characterize or delineate riparian floodplain and wetland. For example, information in
Table 5-23 and associated text on page 5-69 characterize wetland loss within 100 and 200
meters of the road. This is a rather crude method but it is at least an estimate. Perhaps it
would be accurate to say that the area of riparian floodplain and wetland habitat lost was
not characterized. Such a statement would be more internally consistent with statements
that the natural system is "incompletely characterized" (page ES-25). Better yet,
characterize the wetland and riparian habitat losses using one of the many existing
methods. Section 8.2 is a good start. An explicit list of wetland and riparian loss estimates
seems important for an EPA review. Once a list of potential wetland and riparian losses is
tallied, the next question is whether or not it would be possible to compensate for such an
impact. If it is not possible, it might not be possible to achieve no net loss of wetland
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 192
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
resources.
RESPONSE: The EPA agrees that greater quantification of wetland and riparian
losses is essential for the NEPA and permitting processes that oversee compensatory
mitigation and "no net loss". This assessment is based on existing information
(including wetland inventories) and, as the commenter has pointed out, reveals the
areas where existing information is lacking.
2. Combinations and Permutations of analyses. The review of the assessment is somewhat
complicated by the relationship of certain analyses to following analyses and so on. Take,
for example, the use of the Mount St. Helens eruption as a surrogate for a tailings dam
failure. One could assume that such an analogy is inappropriate and recommend that it be
deleted from the assessment. If this recommendation was followed for an initial baseline,
all subsequent information on distance moved, toxicity, remediation, and duration of
impact would no longer apply and would presumably be deleted. Alternatively, EPA
might further develop the Mount St. Helens eruption analogy and demonstrate why it is a
useful part of the assessment. If this is the case, the related following (i.e.. secondary)
information presented on distance moved, toxicity, remediation, and duration would be
retained. Then, it is appropriate for a reviewer to comment on the adequacy of the
following secondary information. Then, it is a possibility that the secondary or following
information is not accurate and the reviewer has to consider that it could be dropped or
retained. Soon, a reviewer of the assessment document is wondering how to address an
unwieldy and hard to follow combination of initial, secondary, and tertiary events. More
likely than not, most reviewers do not consider all the combinations and
recommendations. As a result, the reviewers' comments could fall short of adequately
addressing the restructuring of the main assessment. In short, EPA should not assume that
following the reviewers' comments will result in an assessment that addresses the next
iteration of initial, secondary, and tertiary comments.
RESPONSE: The EPA believes that this comment raises an important point. We did
delete the analogy to Mount Saint Helens ash, as recommended by other peer
reviewers. However, as the comment indicates, that leaves the assessment with no
analysis of recovery oj salmonid streams from the deposition of large volumes offine
particulate material.
3. Address all levels of ecology. Ecological resources can be characterized at many levels of
organization. Populations are often characterized by birth and death rates. Communities
are often characterized by species diversity, succession, and associations of species.
Ecosystems are often characterized by structure, function, nutrient cycling, and energy
flow. From a wildlife perspective, the assessment does a fine job of discussing marine-
derived nutrients but concludes that. .the fish mediated risk to wildlife - cannot be
quantified given available data..." The assessment could be improved if information
regarding community and ecosystem parameters were quantified (or at least addressed at
the level that marine-derived nutrients were addressed). Woolington (2009) comments on
the importance of certain serai stages for wildlife. I am confident that an interview with
him, reclamation specialists, and others at ADF&G could provide a lot of information on
community succession, plant diversity, and wildlife habitat relationships. This
information, in addition to the insight provided in Appendix C, would provide a much
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 193
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
better understanding of possible mine impacts and opportunities for compensatory
mitigation. It is also possible that traditional knowledge of villagers might provide insight
to understanding plant community and ecosystem parameters.
RESPONSE: Ecological risk assessments are not intended to address all levels of
biological organization. Rather, as described in the Guidelines for Ecological Risk
Assessment (USEPA 1998), they focus on a limited number of assessment endpoints
that are most likely to influence a decision because of their importance and
susceptibility. The endpoints were selected in consultation with stakeholders and
decision makers. Further, even if more consideration of the effects of marine-derived
nutrients on terrestrial plant communities were iiichided, the salmon-mediated effects
of the mine on wildlife still could not be quantified. II e have clarified the assessment
scope, and recognize that numerous other important ecosystem components also could
be affected
4. Ecosystem evaluation. The Ecosystem Integrity Section (2.3.5) seems to miss the mark. It
mentions the "nearly pristine conditions," with the caveat that approximately 70% of
salmon returning to spawn are commercially harvested. This is then described as a
managed and sustainable landscape. Maintaining a sustainable resource is not an accurate
characterization of a nearly pristine ecosystem. First of all, sustainability means a lot of
different things to different people. For example, forests in the Pacific Northwest are
managed in a sustainable way but the ecosystem is hardly pristine. Hilborn (2005)
discusses the multiple definitions of sustainability. Instead of deciding which definition
he likes the best, he indicates that sustainability is like good art - it's hard to describe but
we know it when we see it. Second, Hilbom (2005) states: "The record shows clearly that
almost all forms of human activity - agriculture, forestry, urbanization, industrialization,
and migration - reduce biodiversity of natural flora and fauna. This is almost certainly
the case with fishing as well." Hilborn acknowledges the sustainability of the Bristol Bay
fishery but I doubt if he would claim that there is no impact of the fishery on the
ecosystem of Bristol Bay.
RESPONSE: Additional information on commercial fisheries management has been
added to Chapter S of the revised assessment. However, the purpose of this assessment
is not to assess the relative effects of potential mining and commercial fishing—it is to
evaluate potential effects on endpoints if a mine were to be developed, given existing
conditions and activities in the region.
PHW Response: Hilborn clearly states that fishing most certainly has a negative impact
on biodiversity which is a negative impact. The negative past, current and future impact
of the Bristol Bay Fishery should be addressed in the Cumulative Impact Analysis.
Rather than emphasizing the nearly pristine conditions and sustainability of the fishery, I
suggest that other measureable characterizations of the aquatic and terrestrial ecosystem
be measured and quantified. For example, a description of plant communities and
succession would provide the reader with a better understanding of how plant
communities are naturally maintained, and subsequently, how possible mining activities
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 194
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
might alter the successional processes. Failure to address and understand such
relationships led to unexpected consequences for down stream plant communities,
wildlife diversity, and village residents in the Peace Athabasca (Cordes 1975) Delta when
the Bennett Dam was built hundreds of miles up stream on the Peace River. The drastic
changes in the ecosystem function were related to changes in just a few inches of water in
the delta at certain times of the year. The Bristol Bay assessment would be improved if
the expected changes in the hydrographs and sediment transport were related to the
successional processes that maintain early successional plant associations such as the
alder-willow association, which is important for moose. The write up for moose
(Appendix C, page 33) indicates that these early successional associations are maintained
by bank scouring. It would be good to know the role of ice dams and anchor ice on
current levels of scouring and how changes due to milling might alter these important
sustainable functions. The assessment's emphasis on fish and fish mediated impacts runs
the risk of missing possible impacts on the terrestrial environment and down stream
sedimentation that would not only influence human culture but also the aquatic
environment. The impact on terrestrial resources and wildlife is not inconsequential to
aquatic resources, at least in my opinion. Throughout my evaluation, I cite references that
discuss the importance of specific associations between fish and wildlife.
RESPONSE: A comprehensive assessment of potential impacts of large-scale mining
would need to consider all levels of ecological organization. The scope of this
assessment, however, is much narrower, as described throughout the problem
formulation chapters (Chapters 1-0). II e clarified the assessment scope, and recognize
that numerous other im/wrtant ecosystem components also could be affected
5. Multi-directional fish and wildlife relationships. The assessment's emphasis on marine-
derived nutrients and the reduction of this salmon-derived resource only looks at a one-
way fish to wildlife interaction. Wildlife to fish functions, such as beaver dam building,
are very important wildlife to fish interactions. Both salmonid fish, as well as forage fish,
receive benefits from beaver functions such as tree felling, dam building, and food
storage (Snodgrass and Meffe, 1997; Schlosser and Kallemeyn, 2000). This issue is
briefly mentioned in Section 5.2.1.2 (para 1) and discussed on pages 5-19 and 5-22, but
needs to be expanded to address the benefits for fish. The dynamic process of dam
construction and dam decay is important, not only for moving streams across the flood
plain, but also for creating a mosaic of plant associations and wildlife (e.g., moose)
habitat in and near the floodplain. In addition, such activities create a mosaic of habitat
for forage fish. An accurate characterization of the impact of a potential mine and road
necessitates, not only an assessment of the loss of fish on wildlife, but also the loss of
wildlife and their functions on fish. The influence of wildlife and terrestrial processes and
functions on fish discussed in the Cederholm papers needs to be included in the problem
formulation step. The dynamic process of beaver dams causing creeks to move across the
floodplain should also be a criteria for determining if and where culverts , versus bridges
or causeways) are installed for a potential road (pages 4-36 and 4-63).
RESPONSE: Additional text has been added discussing the functional role of beaver
dams in Chapter 7. The discussion of wildlife in Chapter 5 has been expanded to
acknowledge the complexity of fish/wildlife interactions in the watersheds. Please note
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 195
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
that the assessment does not use beaver activity as a criterion for determining what type
of structure (bridge is. culvert) would be constructed at each crossing. The revised
assessment assumes that crossings over streams with mean annual flows greater than
0.1Sm3/s would be bridged However, the actual decision as to what type of structure
would be constructed at each crossing would be made by industry engineers in
consultation with state permitting staff
nTrn, n t-v ¦¦¦ "VT* • 1 sA. Formatted: Font: Not Bold. Not Italic
PHWResponse: .Go Beavers!!! Nice iob. ^ '
6. fi.e.Measuring fish wildlife interactions. Understanding and addressing the multi-
directional interaction of fish and wildlife are facilitated by an approach developed by the
Northwest Habitat Institute, introduced in Johnson and O'Neil (2000) and elaborated on
the Northwest Habitat Institute web site (nwhi.org). The Interactive Biodiversity
Information System (IBIS) database allows an assessment of fish and wildlife
interactions and functional characterizations (i.e., ecosystem functions). The IBIS
database is a logical extension of the Jack Ward Thomas Wildlife Habitat Relationships
and the GAP analysis for the Pacific Northwest. Such an analysis could build on
information in the Alaska GAP database and the Natural Heritage Program database in
Alaska.
RESPONSE: A comprehensive assessment of potential impact of large-scale mining
would need to consider all levels of ecological organization and interaction. The scope
of this assessment, however, is much narrower, as described throughout the problem
formulation chapters (Chapters 1-6). We clarified the assessment scope, and recognize
that numerous other important ecosystem components also could be affected
7. Ecological risk assessment for toxic chemicals. The amount of information in the
assessment pertaining to ecological toxics risk is impressively large. The amount of time
and team expertise needed to adequately review this information is well beyond the scope
of the proposed review. 11" a team was assembled to review this information, I would ask
them to consider:
a. The applicability of a probabilistic risk assessment to address some of the
uncertainty associated with the deterministic information presented in the
assessment;
b. Stressors in addition to toxics (e.g., habitat loss, noise, human disturbance, light,
and water warming in the winter); and
c. The relationship of stressors to populations in addition to no observed effect
levels.
I have worked on large scale impact assessments where the stress of habitat loss far
exceeds the potential stress of toxics. I would think EPA would want to know whether or
not this is the case for the example mine in their assessment.
RESPONSE: The revised assessment evaluates the potential impacts from habitat loss
and toxic releases from plausible mine scenarios based on existing information about
the watersheds. The review panel includes an expert on aquatic toxicology who did
provide an excellent review of that material
PHW Response: Stubblefield is an excellent resource but I don't recall him assessing the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 196
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
relative importance of toxic risk to loss of habitat risk. I remember asking Stubblefield
during one of the Anchorage Meetings with the peer group: Should risk to habitat be
addressed in an ecological risk assessment". He said ves. So it seems that a salmon
endpoint based on habitat should be considered jointly with the toxicological component
of the risk assessment.
The first draft mentioned gold as a resource but after talking to other peer reviewers I was
assured that gold ore would not be processed on site. The revised draft indicates that
pvrite tailings would be processed on site using a cyanide heap leach process. Given this
addition, it seems appropriate to mention the possibility of an accident while transporting
cyanide to the processing site. Similar analyses for the Crown Jewel Project in
Washington indicated a cyanide transportation spill into or near a creek would have acute
short-term impacts for many miles downstream. I would like to know if ethyl xanthate.
the primary chemical of concern in the revised assessment, is more toxic than cyanide? Formatted: Font: Not Bold, Not italic
8. My only regret in this review is that I could not visit the proposed mine site and
watersheds during the summer and the winter. I have flown over the Bristol Bay
watersheds while working for other mines in Alaska, but I have never been on the ground
in these watersheds. I am concerned that if I did visit the proposed sites my assessment
points above might change and even change dramatically. It is hoped that all the EPA
staff and biologists working on the assessment are able to visit the watersheds in the
summer and winter.
RESPONSE: No changes suggested or required.
3. SPECIFIC OBSERVATIONS
[NOTE: in the page notations below, S = Section, P = Paragraph, L = line]
David A. Atkins, M.S.
1. Global: All significant figures should be reviewed to make sure they are reflective of the
level of uncertainty (i.e., using an estimate of 141.4 km of streams eliminated when this
value is probably realistically +/- 50%).
RESPONSE: All significant figures in calculations and measurements made by EPA have
been reviewed and changed to reflect the uncertainty of the analysis. For example km of
stream is reported to the km. Significant figures from cited documents are as reported in
the document.
2. Global: Many references cited in the text are not included in the reference list.
RESPONSE: References have been updated
3. Global: The executive summary, main report, and appendices, in many instances, present
different information with sometimes different implications. These three levels of detail of
information should be more cohesive.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 197
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: We have re-written the executive summary and main report to make them
more consistent with each other and the appendices.
4. Page ES-24 (P2): Suggest changing 'Cumulative Risks' to 'Cumulative Effects of Multiple
Mine Development.'
RESPONSE: The title of the chapter (now Chapter 13 was changed
5. Page 3-2 (PI): The justification for excluding ancillary development from the assessment
should be better explained. In some instances, opening up an area for natural resource
development has had as much or more impact on the environment and ecosystems as the
development itself (for example, oil and gas development in some areas of the Amazon
Basin)
RESPONSE: Discussion of ancillary mining infrastructure (in addition to the pit, TSFs,
waste rock piles, roads and pipelines) has been added to liox 6-1, and discussion of
induced development (development which follows initial development in an unimproved
setting) has been added to Chapter 13.
6. Page 3-7 to 3-11: The conceptual models are quite helpful, but are not referenced or utilized
sufficiently when discussing impacts. It would be helpful to more fully incorporate them into
the assessment.
RESPONSE: Sub-components of the conceptual models have been included at the outset
of the risk analysis chapters, to better frame the pathways evaluated in each chapter.
7. Page 4-1 (PI, L9): Why does the assessment describe current 'good' and not 'best' practice?
The rationale for this decision needs to be described. In addition, it is likely that anything
other than 'best' practice would not be permitted in this context.
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. The terms are qualitative when generally
interpreted, or have a regulatory meaning. The term "best management practices" is a
term generally applied to specific measures for managing non-point source runofffrom
storm water (40 CER Part 130.2(m)). Measures for minimizing and controlling sources of
pollution in other situations are referred to as best practices, state of the practice, good
practice, conventional, or simply mitigation measures. We have added a text box in the
revision Chapter 4 to discuss terms. Mitigation measures consideredfeasible, appropriate,
and 'permittable' (asper (ihaffari et aL 2011) were considered in the assessment, and
these are measures common to other copper porphyry mines.
8. Page 4-11 (P3, L10): The liner lifetime is quite low, and given the importance of this
assumption, would warrant more than a personal communication that does not appear in the
references (North pers. comm.).
RESPONSE: The discussion of liners has been expanded and moved to Chapter 4 (Section
4.2.3.4) in the revision. The personal communication reference has been removed
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 198
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
9. Page 4-21 (PI, L4): Is the assumption about the TSF locations from the authors or from the
Wardrop 43-101 report?
RESPONSE: The location of the TSFs is a combination of alternative sites described in
Ghaffari et aL 2011 (the Wardrop report), in the NDM 2006 Water Rights Application (see
reference list for Chapter 6), and our knowledge of site characteristics suitable for tailings
impoundments.
10. Page 4-26 (S4.3.7-Water Managment): It would help to provide appropriate ranges for
numbers (e.g., the precipitation at the mine and TDF is 803 and 804 mm/yr, respectively,
which implies an unrealistic degree of certainty). It would also be helpful to include a
diagram. I am uncertain why 'cooling tower' water losses would be included in the mine
water balance since power generation would likely be at a remote location and other impacts
from power generation are not considered in this assessment.
RESPONSE: In general, we have presented our best estimates for the parameter values
discussed in the assessment and have used these estimates in our analyses and
calculations. A full uncertainty analysis is beyond the scope of this assessment, so we have
provided ranges only for critical parameters such as the mine size. We followed the
Ghaffari et aL (2011) mine plan in placing power generation on the site, which is the
reason for the natural gas pipeline. Therefore, water use for cooling is appropriate.
Net precipitation at the mine site has been recalculated. The monthly mean flows for each
gage were summed across the year, producing an area-weighted average of860 mm/yr.
Ghaffari et aL (2011) describe the construction of a combined-cycle natural gas-fired gas
turbine power plant at the mine site and estimate the cooling tower evaporative and drift
losses from the plant. The cooling tower losses were included because they would
constitute a substantial consumptive water use.
11. Page 4-31 (P3, L6): I low was the filling lime of 100 to 300 years for the pit estimated? What
constitutes full (e.g., within x% of the pre-mining water table)?
RESPONSE: The filling time was calculatedfrom the estimated inflow rate to the empty
pit. Most of the groundwater infiltration would come from the uppermost 100 m and the
direct precipitation rate would be constant, so the filling rate would not decrease
significantly until the water level was close to the surface compared to the pit depths. We
recognize that as the water level approached the surface, the cone of depression would
shrink and the filling rate would decrease. We have not attempted to refine our analysis to
account for this decrease in rate and the available data do not merit such a refinement
The filling time was simply calculated by dividing the pit volume by the inflow rate. In
practice, when the pit was full the pit level would be maintained below the overflow level by
pumping if treatment were required or be allowed to establish a natural outlet at its low
point if treatment was not necessary.
12. Page 4-52: The PMP is based on the Miller (1963) reference (not included in the reference
list). How was it estimated?
RESPONSE: The PMP was based on Technical Paper No. 47
(http://www.nws.noaagov/oh/hdsc/PMP_documents/TP47.pdf). This PMP was used as the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 199
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
precipitation input to a Type-1 (for Alaska) HEC-HMS simulation to calculate a runoff
hydrograph. 291 cms was determined to be the peak discharge from this 24-hour design
storm
13. Page 5-22: Should refer to 'recapture efficiency' rather than 'recovery rate'. How were the
values of 16% and 63% derived?
RESPONSE: The words "recovery rate" were meant to be descriptive and not as a defined
technical term There is no standard term in widespread usage for the parameter reported
(ie. the percentage of excess water available for reintroduction into the streams versus the
total amount captured by the mine processes). In the revised assessment, we have changed
the wording to "reintroduced" (Table 6-3, Section <>. /. 2. >). The reintroduction percentages
were calculated as part of the water balance. The water balance totaled all the sources of
water captured per year and then subtracted all oj the annual, consumptive water losses.
The remaining annual volume of water is excess water that the mine does not needfor
operations and which is available for reintroduction to the streams. The ratio of the
annual reintroduced volume to the annual captured volume yields the reintroduction
percentage.
14. Page 5-23: The mean annual unit runoff values are not reproducible from the values given for
drainage area and measured mean annual How.
RESPONSE: These values have been corrected.
15. Page 5-32 to 5-39: Tables showing How changes lor different mine sizes and figures for
minimum mine size are difficult to interpret. A different presentation method and/or narrative
description would help.
RESPONSE: Tables, figures, and text for this section have been revised
16. Page 5-45 (S5.2.3): The preceding section (Section 5.2.2) focuses on 'Effects of Downstream
Flow Changes' and Section 5.2.3 focuses on wastewater treatment. It is not clear why this is
the sole focus of Section 5.2.3. In addition, only a short paragraph is included in this section.
Certainly there are oilier possible risks beyond water treatment, and even this discussion is
too cursory given the importance of the issue.
RESPONSE: Discussion of water treatment has been expanded in Chapter 8 (a new
chapter in the revised assessment).
17. Page 5-45 (S5.2.4): This section states a number of assumptions that could result in under or
over estimating impacts on stream flow from the mine footprint. This section leaves the
impression there is a lot of uncertainty, both with assumptions behind the estimation and with
how successful any attempts at mitigation may be. Therefore, we are left with very large
error bars on estimates that should be reflected in the numbers presented for loss of length of
streams and areas of wetlands.
RESPONSE: We have improved the rigor of the analyses regarding the impact of mining
on streamflow. We have attempted to be very explicit about our assumptions and
approaches, and we believe the analyses are appropriate and defensible.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 200
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
18. Page 5-46 (P3): This paragraph discusses the possibility that estimates of stream length
blocked by mine construction may be overestimated if engineered diversion channels are
successful. This is an important form of mitigation that needs to be evaluated further. It
would be helpful to evaluate mitigation efforts in similar types of systems to determine if
reconstructing streams is feasible and could be successful.
RESPONSE: Appendix J includes a discussion of the efficacy of constructed spawning
channels. There is little evidence in the scientific literature to suggest that such channels
are effective at creating sutiable spawning habitat Furthermore, there is nothing in the
scientific literature that suggests salmon streams could be successfully reconstructed in
landscapes where significant alternations in area soils, hydrology and groundwater flow
paths have occurred, such as would be the case with the mine scenarios described in the
assessment.
19. Page 5-59 (Bullet 4): Whole effluent toxicity (WIT) testing and downstream biotic
community monitoring would likely be pari of any discharge permit. This requirement would
not preclude developing a better understanding of protective discharge chemistry and
temperature requirements before permitting, especially given the quality of the receiving
water.
RESPONSE: Appropriate WE T testing and downstream biotic monitoring would be
desirable if a mine were permitted. Additional studies of site specific chemistry and
temperature tolerances may also be desirable. \ o change required.
20. Page 5-60 (S5.4.2): The statement that mine traffic will not be a large enough volume to
affect runoff needs support. Do we know the road will only be used for mine traffic? Can we
estimate the volume of mine traffic for a mine of this size and then look at runoff from an
analog system?
RESPONSE: This statement has been deleted from the assessment However, we found no
analogous data that would allow us to quantify runoff.
21. Page 5-60 (S5.4.4.1): How was the number of stream crossings determined (e.g., what
criteria were used to define a stream vs. a channel)?
RESPONSE: All l:SC,'S designated streams were included No artificial channels were
identified and all natural channels are streams.
22. Page 5-65 to 5-68 (S5.4.8.2): Text states that 240 km of stream upstream of the transportation
corridor has a gradient greater than 10% and, therefore, is likely to support fish. Should this
be less than 10% (as stated in the header to Table 5-22 and Table 6-9)? If so, how was the
<10% value chosen?
RESPONSE: The comment is correct, though in the revised assessment we use <12% as
the criterion and cite a relevant source.
23. Page 5-74 (S5.5): Salmon-mediated effects on wildlife seems under-analyzed in the report,
particularly when compared to the information presented in Appendix C.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 201
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The discussion of potential salmon-mediated effects on wildlife has been
expanded (Chapter 12).
24. Page 5-75 (S5.6): As above, salmon-mediated effects on Alaska Native Cultures seems
under-analyzed in the report, particularly when compared to the information presented in
Appendix D.
RESPONSE: The discussion of potential salmon-mediated effects on Alaska Native
cultures has been expanded (Chapter 12).
25. Page 6-3 (S6.1.2.1): I don't find the Mt. St. Helens analogy useful.
RESPONSE: It has been removed
26. Page 6-9 (PI, LI): Why would 'present' resident and anadromous lish not suffer habitat loss
in the event of a TSF failure? Are they upstream of the area inundated?
RESPONSE: Yes, these populations are upstream of the TSE-affected area. However, they
may be impacted by barriers to seasonal movement or fragmented by loss of connectivity.
This has been clarified in the text.
27. Page 6-28 (S6.1.5): The 'Weighing Lines of Evidence' section is not well developed and not
particularly useful. It also does not inform the risk characterization and uncertainty
discussion.
RESPONSE: The EPA believes that, it is important to weigh all relevant lines of evidence.
An explicit weighing allows readers and reviewers to see what was weighed and how it was
weighted. Such transparency is desirable in general Therefore, the weighing of evidence
has been expanded and the method is better explained
28. Page 6-30 (PI, L2): The statement that risks of failure of the gas and diesel pipelines are not
considered because they are not particularly associated with mining makes no sense. Without
the mine, there would be no need for the pipeline. And if the types of failures and risks are
well known, then this is one of the areas that could actually be assessed with some degree of
certainty.
RESPONSE: Diesel pipeline failure has been added. A gas pipeline is still not assessed
because it is judged to not pose a potentially significant risk to fish (see Chapter 11).
29. Page 6-36 (S6.3, P2): Designating closure as 'premature closure,' 'planned closure,' and
'perpetuity' with water treatment ceasing immediately, continuing until permits are
exhausted or water is nontoxic, or until institutions fail does not seem reasonable. Any of
these closure scenarios would be planned, and would involve regulatory compliance reviews,
bonding, etc. Walking away without continuing to collect and treat water would be an
unlikely scenario. The issue of treatment in perpetuity is a larger issue that needs to be
treated in detail.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed It highlights the needfor adequate bonding andfor plans that
include closure by the bond-holding agency.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 202
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
30. Page 6-37 (P4, L4): The report states that (acid generating) waste rock could be left in place
in the event of premature closure. This scenario should be addressed in the mine closure plan
and during the closure bonding process.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed It highlights the needfor adequate bonding andfor plans that
include closure by the bond-holding agency.
31. Page 8-1 to 8-2: Given that this chapter discusses overall risk to salmon, it would be helpful
to put the estimates of km lost in terms of the total stream or watershed available habitat. We
need some context and metric for assessing significance.
RESPONSE: In Table 2-1, we have explicitly expressed the relative areas of each spatial
scale as % of the scale above it (e.g., the % of the entire Hristol Hay watershed made up of
the Nushagak and Kvichak River watersheds, the % oj the A ushagak and Kvichak River
watersheds made up of the mine scenario watersheds, etc.).
32. Page 8-2 (S8.1.2): The focus on a few types of catastrophic failures does not reflect the
current typical mining scenario. Based on experience at other mines, it is more likely that
smaller-impact failures and accidents would occur during the mine life. It would be helpful to
use some current case studies to illustrate this point.
RESPONSE: The number of possible types, magnitudes and combinations offailures and
accidents is virtually infinite. Therefore, we used a bounding strategy. The potential effects
are bounded by those of routine operations and a realistic severe failure or accident. We
also cite case studies from recent reviews of mining failures and individual failure cases to
indicate some of the possibilities.
33. Page 8-3 (18-1): Why would most concentrate pipeline failures occur between stream and
wetland crossings?
RESPONSE: Most oj the transportation corridor is not on or adjacent to a stream or
wetland. Therefore, most failures would occur in those areas between stream and wetland
crossings.
Steve Buckley, M.S., (7Hi
1. Page xii: ICF is referred to in the document page xvi, but not listed as acronym or
abbreviation.
RESPONSE: ICF is the name of one of the contracting companies that has worked on the
assessment, and is only referred to as ICF in the introductory material (thus is not
included as an acronym or abbreviation).
2. Page ES-2 (P2, L5-6): "altered by geologic processes by would not degrade..." is unclear
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 203
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Metals do not degrade but their chemical form may be altered (via
speciation, binding, formation of salts, etc.). This is explained in the revised Executive
Summary.
3. Page ES-24 (P4, L2): "geologic defects" in unclear.
RESPONSE: This was clarified in the revised assessment.
4. Page 1-1 (P3, LI): "17 existing mine claims..should read "existing claim blocks"
RESPONSE: Corrected
5. Page 4-11 (P3, L10): "North pers. comm." There is no reference, date, or information on
North.
RESPONSE: The discussion of liners has been expanded and moved to Chapter 4 (Section
4.2.3.4) of the revised assessment, and includes additional material and references on liner
lifetime. The personal communication ref erence has been removed.
6. Page 5-20 (P4, L3): The "northeastern United States" not comparable to western Alaska.
RESPONSE: We agree that these two regions diff er in many important ways, but the
general patterns of headwater contributions of nutrients and other resources is shared
between them, as reflected in the numerous citations of studies from Alaska that follow in
this paragraph.
7.
Page 7-1 (PI, L5): "claims blocks" shoii
ild read "c
R1CSPO.XSE: Corrected.
8.
Page 7-3 (P3, L4): As above
R1CSPOXSE: Corrected.
9.
Page 7-6 (P2, L5): As above
RESPONSE: Corrected.
10.
Page 7-7 (P2, P4): As above
RESPONSE: Corrected.
11. Page 8-1,2 (P2, P3): In the discussion of removal of stream kilometers and wetlands it would
be helpful to express these numbers in a percentage of the overall watershed stream
kilometers and wetlands to put perspective on these numbers.
RESPONSE: In Table 2-1, we have explicitly expressed the relative areas of each spatial
scale as % of the scale above it (e.g., the % of the entire Bristol Bay watershed made up of
the Nushagak and Kvichak River watersheds, the % of the Nushagak and Kvichak River
watersheds made up of the mine scenario watersheds, etc.).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 204
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Courtney Carothers, Ph.D.
1. Page ES-2 (P3): "wildlife and the Alaska Native cultures of this region."
RESPONSE: The Executive Summary has been rewritten to incorporate this change.
2. Page ES-5 (P2): "Chief among these resources are world-class commercial, sport, and
subsistence fisheries for Pacific salmon.
RESPONSE: Language added to the revised assessment
3. Page ES-8 (last P): 1. Should Alutiiq (Sugpiaq) cultural group also be included? Alutiiq
residents noted in Igiugig and Kokhanok (Appendix 1), p 15). 2. Change 2nd sentence to: "In
contrast, the salmon base upon which indigenous peoples in the Pacific Northwest depend is
severely threatened."
RESPONSE: 1. Although there are residents oj other cultural groups in the watersheds,
the predominant cultures are Yup 'ik and Dena "ma, so we have f ocused on these groups
but have added mentioned of the Aleut/Alutiiq. 2. This sentence has been edited
4. Page ES-9 (PI): "Salmon are integral to the entire way of life in these cultures as subsistence
food, fishing and subsistence-based livelihoods, and as the foundation for..."
RESPONSE: Language added to the revised assessment.
5. Page ES-9 (P2): "52% of the subsistence harvest, although for some communities this
proportion is substantially higher" (e.si.. noted to be as hiah as 82% on pg 93 of Appendix
D).
RESPONSE: Language added to the revised assessment
6. Page ES-10 (PI): Could also add replacement value for subsistence resources or for salmon,
and the range of estimates for economic valuation of subsistence presented in Appendix E,
noting of that economic valuations do not fully capture the value of these practices.
R1CSPOXSIC: Language added to the revised assessment.
7. Page ES-14 (#1): Are these all the fish spp. at risk, or only the one deemed to be
commercially, recrealionallv valuable? Subsistence spp also include others. Should make
clear what the focus is.
RESPONSE: This sentence refers to the sockeye being particularly at risk because they
spawn and rear below the road Commercial or recreational value is not considered
8. Page ES-23 (P3): As noted above, other mines in Alaska (e.g., Red Dog) and oil and gas
development studies on North Slope may be useful to include predictions about how
subsistence practices will change with mining development and perceived impacts. Including
citations with these statements would be helpful.
RESPONSE: Chapter 12 has been expanded to include discussions of and references to
studies of impacts to subsistence practices from resource extraction industries.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 205
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
9. Page ES-23,24 (last P): "if salmon quality or quantity is adversely affected (or perceived to
be affected)"
RESPONSE: Language added to the revised assessment
10. Page ES-26 (last bullet): There is much data on cultural disruptions caused by the Exxon
Valdez oil spill, and cumulative effects of oil and gas development in North Slope region,
current salmon shortages in Yukon-Kuskowkim. Clearly subsistence is not about lost food,
but about lost lifeways, loss of practices, loss of teaching/learning, and loss of identity. This
point could be made more forcefully. While the specific impacts may not be entirely
predictable, there are likely outcomes that could be included based on experiences in other
regions of the state and/or world.
RESPONSE: The point suggested was made more J orcejully in < liapter 12 of the revised
assessment.
11. Page 1-2 (P2): "this assessment does not provide an economic or social cost/benefit
analysis..."
RESPONSE: Language added to the revised assessment.
12. Page 2-15 (PI): Other important subsistence fish spp not listed in Table 2-5, e.g., whitefish
and winter freshwater fish are listed as integral subsistence species in Appendix D. Again
make focus here clear.
RESPONSE: Scope of the assessment and the specific assessment endpoints considered
have been clarified in Chapters 2 and 5. Table 5-1 has been incorporatedfrom Appendix B
to give a more complete view of the fish species found in the watershed
13. Page 2-18 (S2.2.4): The net economic valuation ranges presented in Table 73, Appendix E
would be helpful to include here.
RESPONSE: Because the assessment is not meant to be a cost-benefit analysis, we only
briefly consider the economic value of the assessment endpoints. In addition, Table 73
deals with the economic value of moose, caribou, and brown bear, which are only
considered in terms offish-mediated effects.
14. Page 2-19 (last sentence): "because no alternative food sources are economic viable." This is
a bit of a misrepresentation. The point is that people choose to live subsistence lifestyles.
Even if food at the stores was cheap, many would choose not to substitute for subsistence
hunting, fishing and gathering. This narrow economic framing misses the cultural and
lifestyle component of subsistence, and frames it merely as food procurement. This is not the
case throughout the document, but in this instance I would suggest changing this sentence to
reflect the irreplaceability of the subsistence lifestyle (dependent on access to high-quality
foods) rather than the economic viability of substituting alternative food sources.
RESPONSE: The revised assessment expands the discussion of the cultural and lifestyle
importance of subsistence (Chapters 5 and 12). The economic viability of alternative foods
is mentioned in Chapter 12, but in the context of a variety of benefits of subsistence.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 206
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
15. Page 2-20 (first sentence): Here and in Appendix D, the legal framework for federal and state
definitions of subsistence should be clarified. Several times in Appendix D an indigenous
subsistence priority is noted (e.g., pg 88: "No other state in the United States so broadly
grants a subsistence priority to wild foods to indigenous peoples as does Alaska."). The
authors should clarify what they mean by indigenous preference (i.e., as opposed to rural
preference?) in state and federal subsistence management. They should include particular
references and additional clarifying information.
RESPONSE: Citations were included in Appendix D in response to this comment
16. Page 3-2 (PI): "would be benign or have no effect on the environmental or social systems,"
RESPONSE: Discussion of scope has been rewritten (< liapter 2).
17. Page 3-4 (PI): ".. .provide subsistence for Alaska Natives and others." Particularly because
subsistence is defined as a rural right in Alaska, all subsistence users should be included as
potentially affected groups.
RESPONSE: References to non-Alaska Natives who practice a subsistence way of life
have been included in the assessment. However, the focus remains on Alaska Natives.
18. Page 3-11 (Fig 3-2E): This conceptual model appears less developed than the others. It
would interesting to work on expanding it out to include missing dimensions; e.g., add health
and healing activity (in addition to nutrition), cultural continuity (alongside social relations
and linked to language and traditional ways of teaching). With a decrease in economic
opportunities comes an increase in reliance on transfer payments. Overall it is a nice
illustration, but strikes me as less complete than the others.
RESPOSSE: The conceptual model has been revised and expanded in response to this
comment.
19. Page 4-15 l i able 4-3): Estimation of 200,000 metric tons of ore processed per day is much
higher rate than any of the other mining operations listed in Table 4-4. Is this due to the
low/moderate quality of the ore?
RESPONSE: Initially, EPA utilized the ore production rate that was available in Ghaffari
et aL (2011), but we have added a smaller sized scenario to the revised assessment to be
similar to the worldwide median-sized porphyry copper mine. Processing throughput is
generally a function of the quality of ore and the cost of mining and processing the ore.
Ghaffari et al (2011) state: "The process is based on conventional grind-crush-float
technology, using proven plant equipment of the largest sizes that have been industrially
installed", which suggests that mining rates are limited also by the size of proven mineral
processing equipment
20. Page 4-21 (last P): 208 m high dam is "much higher than most existing tailings dams." What
are average dam heights? Or how much higher than most existing tailings dams? Does this
high height affect probability of failures?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 207
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Table 4-1 presents dam heights at other Alaska mines, all of which are
significantly lower than 209 m Higher dams mean larger impoundments, storing more
tailings; thus, if failure occurs, more waste material can be released
21. Page 4-23 (P2): "a well field spanning the valley floor." This is unclear. Could it be added to
Fig 4-7? How often would groundwater be monitored?
RESPONSE: The revision clarifies this sentence, adding that installation would be at the
downstream base of all tailings dams to monitor groundwater flowing into the valley. It
was not added to the revisedfigures primarily to keep the figure from becoming too
"busy".
22. Page 5-48 (PI): "effluents would be required to meet criteria." I low different is treated
discharged water from unaffected water?
RESPONSE: Estimates of wastewater composition have been added to the assessment
(Table 8-9). Background surface water data are presented in Table 8-10.
23. Page 5-59 (2nd bullet): Is there any information available on ore processing chemicals, how
much are used, and likely toxicities?
RESPONSE: The toxicity of sodium ethyl xanthate, the primary processing contaminant
of concern, is now considered (Section 8.2.2.5). However, its concentration in tailings and
concentrate slurries cannot be estimated with any reasonable confidence due to the lack of
publicly available information.
24. Page 5-76 (bullet list): The list of cultural factors that may be negatively impacted could
include others: individual, community, and cultural identity; sense of place and place
attachments; community sustainability; cultural unity/conflict avoidance.
RESPONSE: The list of cultural factors that could be negatively impacted was expanded
in the revised assessment.
25. Page 6-46 (bullet list): In addition to the two listed, another should be added noted that
subsistence practices (harvesting, processing, sharing, consuming) are important for
psychological, social, emotional, and cultural health and well-being.
RESPONSE: This section was clarified to focus on nutritional benefits. The discussion of
health benefits from subsistence practices was expanded elsewhere.
26. Page 6-47 (PI): ".. .the physical, psychological, social, and cultural benefits of engaging in a
subsistence lifestyle..
RESPONSE: Revision made.
27. Page 6-47 (PI): References should be added (and were included earlier in the report) for the
statement: "would likely employ a small fraction of Alaska Natives."
RESPONSE: This section has been revised and references the experience of other Alaska
Native communities.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 208
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
28. COMMENTS SPECIFIC TO APPENDIX D
a Single-space for consistency with the rest of appendices
RESPONSE: Appendix D has been single spaced.
b. The title is a bit misleading. Only eight pages in the report discuss traditional
ecological knowledge, and here not in much depth.
RESPONSE: The TEK section was expanded
c. The research design, methods, and data analysis should be described in more detail.
Clarify sampling procedure (both for communities and individuals). For example, it is
unclear if younger generations, particularly active subsistence harvesters, were
targeted as well as elders and culture bearers. Interview protocol should be included
clearly as an Appendix.
RESPONSE: The methodology section has been expanded. Villages were asked to
identify Elders and culture bearers for interviews. Younger generations and/or
subsistence harvesters were not targeted but could have been included by villages
for interviews.
d This section may make a few overstatements (e.g., "only in Alaska are wild salmon
abundant").
RESPONSE: comment noted and taken into consideration during revisions.
e. P12 - "those outside of the state." Change to "outside the region," as many urban
Alaskans are not familiar with subsistence communities.
RESPONSE: The text has been revised.
f. P12 - "Since the questions dealt with a cultural standard, there were few alternative
points of views." Should cultural agreement be a matter of investigation rather than
assumed? This statement needs to be justified. Perhaps with the authors' 40+ years of
experience working with these communities they have come to expect cultural
agreement, especially among elders. If this is the case, that should be clarified. To
what extent did group interviews (2-6 people interviewed together, except for one
single interview) also contribute to cultural agreement? These details are important
given that the results are given on an agree/disagree format.
RESPONSE: The text has been revised to clarify this point
g. P17 - 2,738 is listed in Table 2 and 2,329 is listed here.
RESPONSE: The discrepancy has been corrected
h. PI 9 - Here is perhaps another example of overstatement - 100% of the population
has access to waters of the rivers and lakes. What is meant here? For subsistence, this
access depends upon having transportation and gear or social relations. Do 100% of
people have this in this region?
RESPONSE: The text has been revised and a table added to clarify that the topic is
access to drinking water.
i P20 - Reword "the archaeological work is largely due to five projects."
RESPONSE: The text has been revised to clarify.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 209
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
j. P26 - "located along a salmon stream indicates salmon were likely a primary
resource."
RESPONSE: Text has been revised.
k P31-32 - Several of these quotes focus on social changes (e.g., elimination of dog
teams, relationships to commercial fishing changing over time). People likely harvest
less fish now because they do not support dog teams, yet now they need more money
for fuel and equipment. These are important considerations for understanding
contemporary mixed economy. These points are mentioned in this cultural
characterization, but perhaps could be made a bit more clearly. At times, even the
contemporary characterization reads a bit like "timeless" traditional cultural
relationships to the land and resources, yet it is important to accurately characterize
the subsistence-based communities in their full contemporary realities and
complexities.
RESPONSE: The quotes in the section Jocus on history anil culture.
I P34 - "Large disruptions to the population have not been documented to occur until
epidemic..."
RESPONSE: Text has been revised to clari/y.
m. P34/5 - Both kashgee and qasgiq are used for men's house - it is also defined three
times over these first few pages of this section.
RESPONSE: The text was revised.
n. P35 - "earlier bow and arrow wards" should either be explained or omitted.
RESPONSE: Reference omitted from the revised text.
o. P38, first full paragraph, last sentence - What is meant by "observe the practice?"
This general statement is not adequately supported. Authors should provide specific
instances or more discussion if this point is to be included. As written, it risks
conveying a static view of TEK and practice and culture. Many indigenous
communities in Alaska, e.g., Kodiak villages, while exploited by a colonial economic
system, also strategically adapted to benefit from those systems in ways compatible
with their village lifestyles (e.g., cannery and village co-dependencies that elder
fishermen in this region remember fondly; Carothers 2010). It would be helpful to
have more information on this context in this region (e.g., Hebert 2008; Donkersloot
2005).
RESPO \ I'cxt omittedfrom the revised draft
p. P40 - More information would be useful on Alaska Native participation in
commercial fishing in this historic period up through the present.
RESPONSE: Additional reference to Alaska Native participation in commercial
fishing has been added to the assessment text
q. P47-48 - Ellam yua and tnughit are defined twice.
RESPONSE: Comment noted No change made.
r. P81-84, Table 9 - Second/third part of questions no explained. Since this is an
agree/disagree table, remove other questions for which no information is presented.
All questions would ideally be contained in an interview protocol attached as an
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 210
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
appendix.
RESPONSE: This change has not been made. The authors chose to both identify
the interview questions and the summary responses in this table. As identified in
the introduction the semi-structured interview format was not a "survey" in which
one would ask everyone the same set of questions (as required by structured
interview protocol), but the protocol called for respondents to talk about the subject
elicited by the question. This is a very standard interview technique in
anthropology.
s. P87 - 'non-monetize' - but important to note that the modern subsistence economy
now depends upon cash inputs (ATVs, boats, snow machines, gas, parts, repairs,
guns, nets, etc.).
RESPONSE: The relationship between the modern subsistence way of life and the
market economy has been acknowledged in the appendix and the assessment text.
t P88 - First full sentence, last sentence poorly worded.
RESPONSE: This section has been slightly rewordedfor clarity.
u. P89-90 - The subsistence discussion is confusing.
RESPONSE: No suggestions for revisions were made.
v. P92-93, Tables - Update with recent data if possible.
RESPONSE: Tables 12 and 13 contain the most recent data available to the public.
w. PI 00 - If percentage of working age population not in labor force is a better measure,
it should be included rather than official employment rates (or in addition to).
RESPONSE: These data were added to the appendix.
x. PI 10 - "Villagers in the study also eat store-bought foods, but do not prefer them" -
make clear again that most residents interviewed were elders or identified culture
bearers. A concern for many subsistence villages in other regions of Alaska is the
displacement of younger generations from fish camp and other subsistence practices,
and preferences for store foods, particularly candy and soda. If this region is unique in
that regard, make that clear here.
RESPONSE: The text was revised to make it clear that the interviewees expressed
preference for subsistence foods. The section on nutrition was also expanded.
y. Section C "Physical and Mental Well-being" - Subsistence for emotional/mental
health should be added as a sub-section here. Given the high rates of social problems
in Alaska Native villages (e.g., suicide, violence, addiction), many cultures talk about
subsistence practices as being healing activities or producing emotion, spiritual and/or
mental health. This important aspect isn't covered in the other sub-sections.
RESPONSE: The discussion of suicide was expanded and a discussion on
behavioral and mental health was added to the "Social Relations" section. The text
of the assessment was expanded to acknowledge the social and spiritual aspects of
the subsistence way of life.
Z. PI 13 - Makhoul et al. is listed as 2010 in references.
RESPONSE: This error was corrected.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 211
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
aa. PI 14 - Change Local Wild Fish and Local Practices, and "ecologically, socially,
culturally, spiritually, and possibly even evolutionarily." Point is that subsistence
salmon are not just vehicles for protein and nutrition, but form the basis of incredibly
important subsistence ways of life that are irreplaceable.
RESPONSE: The revision was made and the discussion of social and spiritual
values of the subsistence way of life was expanded in the appendix and the
assessment.
66. PI 15 - Add 'cultural and social disruption' to the list of risks.
RESPONSE: This revision was made.
cc. PI 52, 2nd and last bullet points - These are risks of mining development, not of
decreased quality/quantity of fish (defined as outside the scope of this assessment).
The last bullet point would apply to fish-effects if reworded - some community
members may decide it is not safe to eat fish, causing factions of those who express
concern and those who do not. Others to possibly include: cultural loss as younger
generations do not learn the practices of subsistence; stress on other areas and
communities of the region where people may target subsistence resources; and health
risks of eating contaminated fish.
RESPONSE: Comment noted. these topics imv discussed in the assessment text
(Chapter 12).
dd PI56 - Sing to sign.
RESPONSE: Correction made.
ee. Several grammatical errors throughout.
RESPO \ \A.': Every effort mis made to correct grammatical errors.
29. COMMENTS SPECIFIC TO APPENDIX E
~. P9 - Components of total value should include indigenous homeland for Alaska
Native cultural groups.
RESPOS SE: The value of indigenous homeland and the value of being able to
bequest that homeland and cultural traditions onto future generations is included
in the net economic value analysis presented in Section 5. Section 5.2 discusses
and attempts to quantify the cultural value of engaging in a subsistence livelihood,
referred to as the "activity value". Additionally, Section 5.6 discusses the
importance of the existence and bequest value of the resource, and the challenges
encountered in estimating these values.
~. P12 - Clarify usage of Aleut (Alutiiq/Sugpiaq?).
RESPONSE: The Bristol Bay region includes Alutiiq and Sugpiaq Native
Alaskans. This has been clarified in the text of Appendix E.
c. P22 and 26 - Change Boraas citations to Boraas and Knott.
RESPONSE: This change has been incorporated
d P32 - Much of recreational use is non-market and could be included in the list at end
of 2nd paragraph.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 212
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Because we do estimate the value of recreational fishing, sport
hunting, and non-consumptive wildlife viewing, these are not identified as non-
market values in this paragraph.
e. P96 - Citation for typical crew share of 10%?
RESPONSE: This estimate is based on the author's interviews with crew members.
Appendix E has been revised to make this clear.
f. PI22 - Reasons for differences in earnings between local residents and others is
important. The mixed subsistence-cash economy and cultural ideas about commercial
work in this region may offer an explanation. See: Koslow 1986, Langdon 1986,
Carothers 2010.
RESPONSE: These references have been added to Ippendix E and a brief
discussion has been incorporated in the document.
g. PI34 - Ugashik, Egegik, and South Naknek have over 30.
RESPONSE: Thank you for noting that these comimiiiities have over 30 permit
holders per 100 residents. We have emphasized this in the text.
h. PI 36, last paragraph - This paragraph seems abrupt/misplaced. A more thorough
discussion is needed here to include these points.
RESPONSE: This paragraph describes information in Table 39, which appears
immediately belmv the paragraph. The earnings data in Table 39 is from the US
Census as well as the Alaska Commercial Fisheries Entry Commission. As
earnings data is limited and the reasons for variation in earnings are not apparent,
an additional discussion was not incorporated,
i P178, Section 4.3 - No discussion of role of regional and village Native corporations
or the Community Development Quota program for federally-managed fisheries.
RESPONSE: Section 4.3 provides a broad overview of the economy of the Bristol
Hay Region. Although specific entities and programs like the Native corporations
and Community Development Quota program are not named, the relative
importance of government at a state, regional, and local level is considered There
are many other entities that participate in the economy and government programs
that support it. This model of the regional economy does not identify specific
government entities or programs,
j. P191 - While the majority of formal sector jobs are taken by nonresidents, may want
to note that local economy - subsistence - is all local and highly dependent on
resources of the region.
RESPONSE: We appreciate this comment, and this section of Appendix E
discusses the local and regional importance of subsistence in paragraph 3. Because
this information is covered in detail in paragraph 3, additional information was not
incorporated into the paragraph summarizing jobs captured by the market.
k PI 93 - 2009 is mentioned as an unrepresentative year and given a sensitivity ranking
of "high." More information should be included on the anomalous 2009 - in what
direction should we expect to interpret data from this year compared to more average
years, or those at other ends of the extremes?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 213
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The yearly variation in the commercial fishing industry is described
in detail in Chapter 3. The 2009fishing data is given a sensitivity rating of high as
the data presented in Chapter 4 estimating commercial fishing wages andjobs rely
on inputs from the 2009fishing season.
There is no reason to believe that 2009 is inherently an unrepresentative year for
the fishery, but it is impossible to know if it is a representative year for the future
performance of the fishery. In order for the wage andjob estimates presented in
Chapter 4 to be representative of future performance, the 2009 data used to
represent the base year would need to also be representative of future performance
of the commercial fishery.
To clarify that this is a potential bias as the flit lire of the commercial fishery cannot
be predicted, the text associated with this item has been rephrased
I PI 95 - Number of households engaged in subsistence - ADl'&G data should provide
estimates.
RESPONSE: The analysis relies on data from ADF&G: however, the item listed
here identifying the potential bias of the number of households engaged in
subsistence acknowledges that the data from Al)lr&(i is purely an estimate. Should
the data relied on from ADF&G be found to be inaccurate for some reason, this
analysis that relies on those data would be biased accordingly.
m PI 98 - ATV, snow machines, should be added to "boats and trucks"; work by Robert
Wolfe and others (Wolfe et al. 2009) suggests that about one third of households in
Alaska Native villages harvest the majority of subsistence foods (and share,
especially with the least active households). How does this finding affect these
estimates?
RESPONSE: We recognize that I It s and snow machines are frequently used in
the subsistence harvest The original data source used to develop these expenditure
estimates did not specify expenditures for these machines, and thus they are not
identified in the quantified expenditures. In this analysis the total expenditure
estimates drive the analysis, not the composition of what these expenditures
purchased. Differences in the composition of expenditures would have a negligible
bias, as identified in Section 4.8.
The finding from the Wolfe 2009 citation provided above would not affect these
findings. This analysis relied on the assumption that all native households are
participating in the subsistence harvest, and estimated expenditures based on that
assumption. If some native households are more active than others, the
expenditures by less active native households will be lower than those estimated
here, but this will be offset by higher expenditures by more active native
households.
it. P202 - Explain why % of adults with 4+ years of college is used in this model? The
model was not explained clearly enough for me to understand it.
RESPONSE: The percentage of adults with four or more years of college
education is used in the model along with other relevant regional variables to
predict adjusted gross income per capita for each community. This instrumental
variable is included in the analysis because it is a variable that is relevant for
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 214
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
predicting adjusted gross income per capita in each community. For additional
information on the analysis used, please see Duffield (1997).
o. Some fisheries, e.g., crab fisheries, are not included in the economic analysis, yet
depend in part of Bristol Bay ecosystems, as discussed in Appendix F.
RESPONSE: As noted in the assessment, there are other species present that
support a commercial fishery, such as the crab fishery the commenter notes.
Although Appendix E is predominately focused on the importance of the salmon
fishery, some considerations for the commercial and subsistence value of these
fisheries is included
Section 4.3 of Appendix E provides an overview of the regional economy, which
fish harvesting and processing. Table S3 estimates employment in the fish
harvesting and processing sector. As identified in the footnote to Table S3, these
employment estimates include fisheries in addition to the salmon fishery.
Appendix E also acknowledges and considers that subsistence harvests may include
species other than salmon. In Section 2.2 of Appendix E the general distribution of
subsistence harvest by Bristol Hay residents is presented.. I.v identified in this
section, Bristol Bay subsistence harvest includes salmon, non-salmon fish, birds
and eggs, vegetation, marine mammals, marine invertebrates, and land mammals.
Subsistence expenditures are estimated based on the number of native households
and expenditure estimates from Goldsmith (1998). Goldsmith's expenditures
consider subsistence expenditures more broadly, thus these other species are
included in the economic analysis.
p. References - Peterson et al. 1992 and Brown and I Jurch 1992 not included in
references.
RESPONSE: Change has been incorporated.
30. COMMENTS SPECIFIC TO APPENDIX G
a. Mitigation measures are largely concluded to be ineffective. Would be helpful to
compare mitigation measures and their success/failure in other mining examples.
RESPONSE: Appendix G contains available information about mitigation
measures generally related to roads and pipelines. Mitigation measures are not
specific to mining operations.
31. COMMENTS SP1 iCl 1TC TO APPENDIX H
«. P7 - Exposure of groundwater and waterfowl to chemical contaminants are listed as
main environmental concerns from tailings storage facilities. Impacts to human health
from ingesting contaminated water or birds. Clarify in report that direct risks to
human health are not accessed (only through reduction or elimination of subsistence
harvests?).
RESPONSE: The assessment language has been clarified to acknowledge that
there are potential direct risks to human health, but that the scope of the
assessment is limited to potential effects to indigenous cultures related to salmon
An evaluation of human health effects from ingestion of contaminated non-salmon
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 215
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
subsistence foods is not within the scope of the assessment. EPA expects that a full
evaluation of any future mining permit applications and subsequent National
Environmental Policy Act Environmental Impact Statements would consider these
direct effects.
Dennis D. Dauble, Ph.D.
1. Appendix B, Page 30: Table 1.1 suggest adding a column to indicate relative abundance, for
example, if individual fish species listed are abundant, common or rare. Also, are there
known differences in distribution and abundance for the Nushagak and Kvichak watersheds
relative to those watersheds unlikely to be affected by mining activities?
RESPONSE: This information, when known. has been added to Appendix B.
2. Appendix A, Page 42: The statement that diminished salmon runs present a "negative
feedback loop" where spawner abundance declines, appears to conflict with the last
paragraph on page 41.
RESPONSE: The text describes bottom-up effects ofMl)S on stream ecosystems and
points out that these linkages have not been empirically established The negative feedback
loop mentioned in the following paragraph could result from reductions in salmon-based
resources that promote either bottom-up (i. e., a reduction in salmon-derived N and P) or
direct (i.e., a reduction in salmon eggs and flesh that can be consumed by fish) benefits to
juvenile fishes. Further, the fact that bottom-up effects ofMDN have not been firmly
established does not negate considerable circumstantial evidence for bottom-up nutritional
deficits in Columbia Basin spawning streams. Edits have been made to help clarify.
3. Appendix F, Page 3: Is there significant sediment transport from the Bristol Bay watershed to
the Nushegak and Togiak Bays/estuaries'.''
RESPONSE: A new section was added to „ ippendix E that discusses the importance of
estuary habitat to salmon populations.
4. Appendix F, Page 7: What is the juvenile salmon resident time in Bristol Bay? How quickly
(and at what size/time of \ ear) do they move from shallow nearshore to offshore habitats?
RESPONSE: A new section was added to Appendix F that discusses the importance of
estuary habitat to salmon populations.
5. Page 8-15 (L2): Suggest deleting "likely." There will be impacts.
RESPONSE: We believe this language is consistent with the uncertainties explained in the
assessment. No change has been made.
Gordon H. Reeves, Ph.D.
1. Page 6-9 (PI): Why would resident fish not "suffer" immediate loss of habitat as a result of
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 216
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
dam failure like anadromous fish would?
RESPONSE: This statement refers to fish located upstream of TSFfailure and affected
downstream reaches. This has been clarified in the text.
2. Couldn't fish move to another stream if a culvert is blocked and prevents upstream
movement? This may eliminate the fish from a particular stream for a year but it might not
reduce the overall productivity.
RESPONSE: This comment is based on the assumption that spawning and rearing habitat
are not limiting. We believe that they area
Charles Wesley Slaughter, Ph.D.
1. Global: Provision of full color versions of all figures would have been helpful to this review.
The selected color versions supplied here were useful - we should have had them all.
RESPONSE: Comment noted
2. Page 2-4: Color codes are confusing - use of different colors for same "moisture state" in the
five regions doesn't make sense (to me).
RESPONSE: Figure 2-2 (Figure 3-1 in the revised assessment) depicts hydrologic
landscapes across the Xushagak and Kvichak watersheds separated into their
physiographic divisions as defined by Wahrhaftig (1965), combined with their climate
class as defined by Feddema (2005). Each color group (e.g., yellow, green, red, etc)
represents a separate physiographic division while color intensity (e.g., light green to dark
green) represents the climate class within a given physiographic division The use of
different colors is intended to show the variety of climate classes within a given
physiographic division, while also depicting similarities in climate class across different
physiographic divisions (e.g., darkest red and darkest green show that two areas having
diff erent physiographic characteristics are alike in their climate class).
3. Page 3-4 (S3.4, LI): .. .when mine is active
RESPONSE: Correct (now in Section 4.2.4).
4. Page 4-5 (last P): Refers to Fig. 3-1, but "existing road segments" listed are not shown on
Fig. 3-1, nor are several cited locales: Williams Port, Pile Bay, King Salmon, andNaknek.
RESPONSE: This reference has been deleted
5. Page 4-11 (first P): "The vast majority of tailings dams are less than 30 m in height.
DOES THIS REFER TO TAILINGS DAMS AT ALL KNOWN MINING OPERATIONS,
OR TAILINGS DAMS ENVISIONED OR PROBABLY TO BE USED IN BRISTOL BAY
WATERSHED?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 217
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The sentence refers to all known mining operations. The text was changed
from "The vast majority of tailings dams..." to "The majority of existing tailings dams..."
(Chapter 4).
6. Page 4-11 (first P): "Although upstream construction is considered unsuitable for
impoundments intended to be very high or to contain large volumes of water or solids... .this
method is still routinely employed." ARE THE TSFs SUGGESTED IN THIS
ASSESSMENT CONSIDERED "UPSTREAM," "DOWNSTREAM," OR
"CENTERLINE"? Para. 1, Section 4.3.5 (p. 4-21), states that "the most plausible sites" for
TSFs are "the higher mountain," which suggests that these TSFs would be "upstream"
facilities, therefore "considered unsuitable..."
RESPONSE: The TSFs suggested in the scenarios would be constructed using the
downstream method and then switched to a centerline method. It is true that in this
location, the upstream construction would be unsuitable. The discussion of the upstream
method was intended for background and has been moved to Chapter 4 in the revision,
with the scenario-specific suggestions discussed in Chapter 6.
7. Page 4-23 (P2): Text suggests that a monitoring well field downslope from the TSF (and
presumably from all hypothetical TSFs) would detect seepage; such seepage would then be
intercepted and either returned to the TSF or "treated and released to the stream channel."
Either action presupposes adequacy of monitoring seepage and subsurface flow (both
spatially and temporally); returning such water to the stream further presupposes fully
adequate treatment to meet both regulatory and aquatic biota requirements for water quality
and flow regime.
RESPONSE: Yes. the well fields would be downstream Jrom all embankments for all TSFs
and this has been clarified in the revision. It is assumedfor the scenarios that mitigation
measures are operated appropriately. Text has been clarified in the new Chapter 6 to
indicate that water would be treated to meet permit requirements. Scenarios in the revised
assessment evaluate effects from uncaptured seepage during routine operations from both
the TSFs and waste rock piles.
8. P4-26 to 4-28 (S4.3.7): This "Water Management" section seems cursory, highly
generalized, and optimistic. Statements such as "uncontrolled runoff would be eliminated";
"water from these upstream reaches would be diverted around and downstream of the mine
where practicable"; and "Precipitation...would be collected and stored...." do not indicate
actual (proposed) practices or techniques, nor inspire confidence that actual runoff events
during "normal" conditions, let alone during hydrologic extremes (such as a rain-on-snow
event with underlying soils still frozen) would be planned for or actually managed
adequately.
RESPONSE: Water management measures are more clearly described and discussed in
the revision (Section 6.1.2.5), and in sub-sections for the mine components in the
scenarios. Measures are standard and common to existing mine sites.
Collection and diversion structures would need to be designed and built to handle the
anticipatedflows over the life of the mine, including during extreme weather events. The
details of these collection and conveyance systems are beyond the scope of this assessment
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 218
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
and are more properly in the domain of the permitting process when a specific mine plan
is proposed
9. P4-29 (PI): Suggests that 20% more water than available would be required "during startup,"
and that difference would be satisfied "from water stored in the TSF"; if 20% more than is
available would be needed, where would it come from to be available from the TSF?
RESPONSE: The water balance has been revised in Chapter 6. The mine operation would
capture more water than needed during all phases of operation. Water could be stored in
the TSF as soon as the starter embankment is operational Ghaffari et al (2011) further
explain: "Once the TSF starter embankment construction is complete, site water will be
diverted into this facility to ensure adequate make-up water for process plant start-up. At
this time, advanced open pit dewatering will commence. I 'liis water will either be treated
and discharged or diverted to the TSF, depending on environmental requirements."
10. Page 4-29 (P3): Assumptions are very generalized and optimistic: "assuming no water
collection and treatment failures" and "excess captured water would be treated.. .and
discharged to nearby streams..." - this assumes both "no failures" over the life of the
operation, and that such treated "excess captured water" could be successfully treated before
release to fully meet both regulatory water quality criteria and the possibly more sensitive
biological requirements of individual invertebrates and fish stocks (Appendices A & B).
RESPONSE: Water management measures are more clearly described and discussed in
the revision in Section 6.1.2.5, and in sub-sections for the mine components in the
scenarios. Measures are standard and common to existing mine sites. The revised
assessment considers water treatment failure quantitatively and includes refined scenarios
with new data for seepage from TSFs and waste rock piles.
11. Section 4.3.8: This and the previous section mention (but in my view, do not adequately
stress) the extremely long time frame for post-mining active management and oversight.
Many hundreds of years of active management is a longer time than many industrial,
corporate or governmental entities are capable of really embracing - witness the current US
Congressional practice of "kicking the can down the road" - a human trait.
RESPONSE: This issue is addressed in Section 4.2.4 of the revised assessment.
12. Page 4-32 (S4.3.8.2. P3): Suggests that pyritic tailings could be "shipped off site" - i.e., to
where? Deep ocean dumping, or Yucca Mountain?
RESPONSE: For smaller amounts of tailings, the option of shipping them to an off-site
location for disposal might be an option; however, this is less likely with large mines with
high volumes of tailings, so has been removed in the revised assessment.
13. Page 4-34 to 4-37 (S4.3.9): This reviewer finds the short Transportation Corridor sub-chapter
to be succinct, but inadequate and superficial in view of the long-term consequences of
imposition of the transportation corridor as portrayed. These deficiencies are addressed, in
part, in other sectors of the Assessment, most comprehensively in Appendix G.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 219
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The transportation corridor analysis has been substantially expanded, and is
now found in its own chapter (Chapter 10).
14. Page 4-38 (Box 4-3, P2): Para. 2 states that the southwest extension of the Lake Clark Fault
is currently understood to extend to perhaps 16 +/- km from the Pebble ore deposit; however,
this is not reflected in Figure 4-11, which suggests that the Lake Clark Fault terminates
perhaps 100 km northeast of the Pebble locale. Elsewhere in Box 4.3, there is
acknowledgement that, while there is no evidence of recent tectonic activity in the immediate
Pebble vicinity, there is relatively little site-specific data or long-term historical seismic
record. I infer that any predictions concerning seismicitv or earthquake occurrence of any
magnitude would have very high uncertainty.
RESPONSE: Figure 4-11 (now Figure 3-15) lias been revised to more accurately illustrate
the current understanding of the terminus of the Lake < lark Fault. Section 3.6 describes
the uncertainties in predicting seismicity in the I'ebble region.
15. Page 4-41 (Box 4-4): Note that in each of the four tailing dam failure examples, the failed
structure was roughly an order of magnitude smaller (in height) than the hypothetical TSF-1
structure, yet those failures had major negative consequences.
RESPONSE: Text was added to Section 9.1.1 in the revised assessment to indicate that the
dams in these failure examples were significantly smaller than the dams proposed in our
mine scenarios."
16. Page 4-45 to 4-47 (S4.4.2.2): The probability approach to tailing dam failure is unpersuasive
as presented. It is difficult to relate to a number like "0.00050 failures per dam year," or to
the implication (p. 4-47) that one can expect a tailings dam failure only once in 10,000 to one
million "dam years." This could suggest to the casual reader that failure of the hypothesized
TSF1 dam (for which one "dam year" is one year) should not be anticipated in either the time
of human occupation of North America, or the span of human evolution.
RESPONSE: The proposed dams, if designed, built, and maintained to current
engineering best practices, would be anticipated to have a low annual probability of failure
but the failure probability would not be zero. We concur with the commenter that these low
probability numbers may be difficult for the casual reader to grasp, so we also present the
estimates ofprobability in terms ofprobability failure over different time periods. For
example, an annual probability of failure of0.0005 equates to a 5% probability of failure
over 100 years and a 39% probability of failure over 1000years. Text was also added to
clarify the basis for our failure rates.
17. Page 4-48 (Box 4-6): Box 4-6 suggests that the Operating Basis Earthquake (OBE) for a 7.5-
magnitude event at the Pebble locale has an estimated return period of 200 years. Such a
return interval probability is difficult interpret, given the lack of historical seismic record for
the region; in any event, such a return period estimate is in no way predictive of future
seismic activity, in year 2012 or year 2212.
Box 4-6 does note that "The return periods stated in Alaska dam safety guidance are
inconsistent with the expected conditions for a large porphyry copper mine developed in the
Bristol Bay watersheds, and represent a minimal margin of safety..
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 220
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The return periods used are consistent with the Alaska Dam Safety
Guidance, but the operator could include an additional margin of safety in the design for
critical structures. The return period and seismic safety factors do not inform the failure
analysis in this assessment.
18. Page 4-50 to 4-60 (S4.4.2.4): The modeled hydrologic consequences of overtopping/flooding
of the hypothetical TSF1 dam/reservoir seem reasonable, given the relatively limited
hydrologic data set available for model input. Probable Maximum Precipitation and Probable
Maximum Flood results should be approximately "correct" and within the same order of
magnitude of potential storm and flood events. The potential consequences outlined (peak
flow volume, sediment transport and deposition, length of stream corridor impacted) appear
realistic for the scenario. I suggest that this topic and hypothetical result should be given
more visibility and emphasis in the assessment.
RESPONSE: The dam failure was modeled a\ an overtopping event since the possibility
could exist. However, the magnitude of the I'M I' run-off as com/Hired to the resulting
failure flood wave is very small We agree that an actual dam design will consider
additional information, but this assessment did not investigate the hydrology of the PMF
in greater detail
19. Page 4-62 to 4-63 (S4.4.4): While accurate, this section does not adequately address the
road/stream crossing/culvert issue. Given the projected transportation corridor, Pebble locale
to Cook Inlet, and the inevitability of a further network of "minor" roads in the mine and TSF
locale, plus additional infrastructure linkages, road/culvert/stream crossings are a major
concern for aquatic habitat and fisheries. This issue receives more attention in Sections 5.4
and 6.4, and is mentioned elsewhere in Volume 1 (e.g., Table 8.1, Box 8-1, para. 8.1.2.4.).
Readers of the Assessment should be directed to Frissell and Shaftel's Appendix G for a
more comprehensive discussion of this important topic.
RESPONSE: This issue is now analyzed in more detail in its own chapter (Chapter 10).
The appendix is cited.
20. Chapter 5: Assumes scenario of "no failure" for entire project, over complete project life. Is
this a realistic scenario, given experience with industrial developments in real-world settings
subject to vagaries of equipment, landscape, geology, weather, local climate, and human
judgment and decision making/execution?
RESPONSE: The "no failure" scenario has been eliminated in response to several
comments, and is non> considered in terms of mine footprint impacts in Chapter 7.
21. Page 5-1 (S5.1.1): Question: is "sampling extensively for summer fish distribution over
several years" adequate for characterizing fish populations, given the wide fluctuations in
salmon escapement and return note elsewhere in the Assessment (e.g., Table 5-1, para.
5.1.2)?
RESPONSE: The answer to this question is of course dependent upon the objectives of the
characterization. For the purposes of this section of the assessment, identifying species
distributions within the study area, sampling was adequate for minimally characterizing
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 221
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
the distribution of salmon, Dolly Varden, and rainbow trout within the project area,
including known spawning areas.
22. Page 5-12 to 5-48 (S5.2): Estimates of habitat, wetland and stream blockage or loss seem
reasonable, but, as noted in the text, are probably conservative or "at the low end."
RESPONSE: Agreed No change suggested or required
Estimates of probable streamflow diminution (p. 5-25) seem reasonable, but make no
reference to seasonality.
RESPONSE: We assume a constant seasonal demand, thus streamflmv changes follow
baseline seasonal patterns under this analysis. 11 e clari/y that water managers may alter
holding and release of water to address environmental flow needs.
23. Page 5-29 to 5-30 (thermal regimes): This section makes no mention of aufeis or "nalyds,"
ice accumulations which can exert major control on spring and early summer habitat
availability and thermal conditions. For examples, see Slaughter (1990), among many other
references.
RESPONSE: Citation and text hare been added in Section 7.3.2.
24. Page 5-30 to 5-31: Concur thai maintenance of natural llow regime is the desirable target; the
"sustainability boundary" approach is a way to attempt managing within the "natural" bounds
of variability. Note: Figure 5-9 is map of streams and wetlands lost, not predicted flow
alteration hydrograph (see last sentence, p. 5-31).
RESPONSE: Figure references have been updated.
25. Page 5-37 (Figure 5-10): LJT100]) - predicted flow is ALWAYS below lower 20%
sustainability boundary. IJT.I.00C. IJT100CI. UTC100B -predicted flows are always within
the 20% +/- sustainability boundaries.
RESPONSE: Figure has been revised, and data now presented in Table 8-1.
26. Page 5-38 (Figure 5-11): Predicted flow for two upper gages (SK100G, SK100F) is always
below the 20% sustainability boundary. Predicted flow at other gages appears to be near or
within the 10% and 20% lower sustainability boundary.
RESPONSE: Figure lias been revised, and data now presented in Table 8-2.
27. Page 5-39 (Figure 5-12): Predicted high flow for NK119A is far below the lower 20%
sustainability boundary throughout the open water season. From onset of snowmelt, flow at
other gages is roughly at or within the lower 20%) sustainability boundary.
RESPONSE: Figure has been revised, and data now presented in Table 8-3.
28. Page 5-41: GageNK119A - is the estimated decrease of streamflow (minimum mine size)
63% (Table 5-13) or 73% (text)?
RESPONSE: Values have been updated
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 222
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
29. Page 5-41: In any case, text is clear: predicted flow reductions for North and South Forks
Koktuli River are materially below the lower 20% sustainability boundary. Text suggests
that while upper Talarik Creek would be essentially obliterated in this hypothetical scenario,
lower gaging stations on Talarik Creek might have partial augmentation of reduced flows,
from small tributary flows and groundwater; without supporting data, this suggestion seems
unsupportable.
RESPONSE: Revised text illustrates important trans-watershed groundwater contributions
to lower Upper Talarik Creek.
30. Page 5-42 to 5-45: These pages fairly summarize the potential for substantive alterations to
streamflow regime and surface water/groundwater relationships.
RESPONSE: No change suggested or required.
31. Page 5-44 (P3): "Once the mine is no longer a net consumer of water, we assume that flow
regulation through the water treatment facility could be designed to somewhat approximate
natural hydrologic regimes, which could provide appropriate timing and duration of
connectivity with off-channel habitats." I suggest that this is a highly optimistic
assumption, and does not address water quality questions (which are raised elsewhere in the
Assessment).
RESPONSE: We emphasize that the ability to manage flows will be dependent upon
sufficient infrastructure andflexibility in miter management.
32. Page 5-45 (S5.2.3): This entire paragraph should receive greater emphasis.
RESPONSE: This section (now Section 7.3.3) has been expanded
33. Page 5-46 (P4): Ignores variable-source-area concepts, which are widely accepted in
hydrologic and watershed analysis.
RESPONSE: We respectfully disagree. The variable source area concept presents the idea
about the expansion and contraction of saturated areas that are the immediate source of
streamflow during storms. In does not invalidate the notion that the total drainage area
contributes with long flowpaths to streams. These longflowpaths ultimately contribute to
saturated zones that support perennial, streamflows..
34. Page 5-46 (P5): Assumes requirement for more water than is available, but leaves hanging
the question of where that more water might be sourced. Given that the site is a watershed
headwaters, what might be tapped as additional water supply, and what might be the impacts
on that source(s)?
RESPONSE: The revised water balance indicates that the mine would have the capability
to capture more water than it needs during all operational phases, including start-up, from
within the mine footprint.
35. Page 5-59 to 5-63 (S5.4): See earlier cautions concerning stream crossings, culverts. Note
that road cuts and culverts are particularly susceptible to development of aufeis ("icings"),
often resulting from blockage or alteration of subsurface water movement during cold
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 223
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
conditions, as witnessed by long-standing AKODT maintenance issues - Richardson, Steese,
Dalton highways, for example.
RESPONSE: The potential for culvert blockage by aufeis is discussed in Chapter 10 of the
revised assessment.
36. Page 5-60 (S5.4.2, PI): The statement that".. .it is unlikely that a mine access road would
have sufficient traffic to significantly contaminate runoff with metals or oil" is unsupported;
it might be instructive to look at traffic loads for the access road from the Steese Highway to
the Ft. Knox mine, a much smaller operation than the proposed Pebble development.
RESPONSE: This statement no longer occurs in the assessment.
37. Page 5-60 (S5.4.2, P2): First sentence is correct. Second sentence is unsupported and
probably incorrect (see Appendix G). Yes, runoff from roads is location-specific; that does
not mean that runoff from roads would be insignificant to salmonids, given the very large
number of streams (perennial, intermittent, and ephemeral), and wetlands, which would be
intersected by the total road system of the Pebble project. This also seems to be contradicted
by Section 5.4.3.
RESPONSE: This issue is now given more attention in Chapter 10.
38. Page 5-60 (S5.4.4.1): There are many more water or seep crossings than 34 - see USGS
topog sheets, ACME Mapper, or Google I iarlh.
RESPONSE: The number oj crossings lias been updated in the revised assessment
39. Page 5-61 (S5.4.4.2 and 5.4.4.3): Development of aufeis ("icings") consequent to partial or
full culvert blockage, or induced by soil mantle compaction (i.e., by roads or off-road vehicle
traffic) can partially or wholly block stream channels. Such blockage, in association with ice
on the streambed. may last long past snowmelt and persist well into early summer, possibly
affecting fish movement.
Rl'.St'ONSE: The potential impact of aufeis on stream channels and potential existence
past breakup is discussed in < liapter 10 of the revised assessment.
40. Page 5-62 (S5.4.5): While I don't have the specific citations at hand, there are published
analyses of dust effects associated with the North Slope haul road and Prudhoe Bay road
network. Obvious effects include accelerated snowmelt along the road corridor, and nutrient
or pollutant contributions to road corridor environs.
RESPONSE: The revised assessment includes information from published literature on
dust effects in the Arctic (Chapter 10).
41. Page 5-63 (S5.4.6.3): Should itread "...impacting 270.3 km of stream..."? (Incidentally, it
is interesting that this coarse assessment finds it possible to state impact to within 100-meter
resolution.)
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 224
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The revised assessment notes that risks to salmonids from de-icing salts and
dust suppressants could be locally significant, but we cannot state that the entire stream
length between potential road crossings and Iliamna Lake would be affected
42. Page 5-65 (S5.4.7.3 and 5.4.8.3): Viewing the transportation corridor landscape, via maps,
Google Earth or ACME Mapper, gives me the impression that the estimate of 4.9 km2
wetlands directly impacted is a very low number. It is easy to play with the numbers given
on pp. 5-69 - 5-70, regardless of their accuracy; 66 km of road impacting wetlands, assuming
a 10-meter roadway footprint, yields 0.66 km2 of wetlands "under the road" (vs. 0.18 km2 in
the text). The 200-meter proximity to wetlands cited, over 66 km of road, yields some 13
km2 of wetland impact (vs. 7.3 km2 in the text). 80 km of road within 200 m of streams or
wetlands yields 16 km2 of road/wetland impact. Since these are all assumptions and
estimates, it is not possible to conclude that any of these figures would be the "true" area
impacted.
RESPONSE: The 0.18 km2 refers to the area oj wetlands which would be filled by the
roadbed "intersecting" wetlands (19.4 km length of roadbedx 9.1 m width of roadbed),
and does not include roadbed adjacent to wetlands (note that this area is ft 11 km2 in the
revised assessment due to a slight change in methodology for measuring road length). The
4.9 km2 (not 7.3 km2) area of wetland within 200 in of the length of road (on both sides)
within the study area is based on actual National IVetland Inventory (NWI) data (the
calculation methodology is described in liox 10-1 of the revised assessment).
Though NWI data were utilized in this analysis, the 200 m road buffer was derivedfrom a
literature estimate oj the road-ej/ecl -one for secondary roads.
43. Page 5-74 (S5.4.10): Should this say impact, rather than "risk"?
RESPONSE: "Impact" could be used here, but we chose to use "risk" to note the
probability rather than certainty that salmon would be affected by the corridor-associated
activities/events listed
Text implies that even this "no-failure" scenario will impact salmonids; however, it is
apparently not possible to estimate specific changes or the magnitude of such changes.
RESPONSE: The no-failure scenario has been eliminated in response to several
comments: effects resulting from the mine footprint are now discussed in Chapter 7. The
exact magnitudes of changes in fish productivity, abundance and diversity cannot be
estimated at this time, but the species, abundances, and distributions that would potentially
be affected are summarized in Chapter 10 of the revised Assessment.
44. Page 5-74 to 5-77 (S5.6): This section seems cursory and understated, particularly in view of
the extensive discussion of Appendix D.
RESPONSE: The discussion of fish-mediated effects on Alaska Native culture has been
expanded and is now in Chapter 12.
45. Section 6.1: Concur with general overview statements, and with conclusions regarding
immediate consequences of TSF dam failure, which would likely be as severe as or more
severe than stated in 6.1.2.1.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 225
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
P. 6-6, first sentence - note that the failure scenario predicts over 70% fines < 0,1 mm, vs.
the 6% "natural" fines concentrations.
P. 6-13, last para. - the assumption that overtopping would not occur in winter is not
warranted, as the authors admit when citing the Nixon Fork Mine incident. In the Bristol
Bay environment, a major rain-on-snow event in winter or spring is within the realm of
possibility, and of course human error is, if not inevitable, always possible.
RESPONSE: We clarify that the scenario we chose to illustrate overtopping did not occur
during low-flow conditions, but was in response to a flood We acknowledge that failure at
other times, and in response to other events, is possible (Section 9.3).
46. Page 6-15 (Box 6.2): Box text implies that human error, lack of timely oversight and
correction was responsible - but never directly savs "human error." The apparent
assumption that there is no hydrologic activity after free/.e-up (or perhaps, after an ice cover
forms on the pool) was naive and incorrect. At least in that case, it appears that both dam and
spillway design (not adequately considering winter and ice conditions) and
operation/inspection (human error) were responsible.
RESPONSE: The comment is probably correct. However, we have no evidence that the
commenter's conclusion that both engineering error and operator error were responsible
was correct.
47. Section 6.1.4.1: Appropriately recognizes the long lime period for exposure, over extended
stream lengths, through both initial deposition and multiple re-mobilization and redeposition
events.
RESPONSE: Agreed. No change suggested or required.
48. Page 6-28 (S6.1.5 and Table 6-6): Seems jargon-laden and does not add to strength of the
Assessment.
RESPONSE: Opinion has been divided concerning this explicit weighing of evidence.
Given that the analysis of risks to fish from exposure to spilled tailings involves six
separate lines of evidence, the EPA intended to show and discuss how the evidence was
weighed
49. Page 6-29 (S6.1.7): Concur that remediation ".. .would be particularly difficult and
damaging..."
RESPONSE: No changes suggested or required
50. Page 6-30 (S6.2): Even though "We do not assess failures of the natural gas or diesel
pipelines...," those pipelines would be equally susceptible to failure as the slurry line.
Concerns with pipelines crossing streams, watercourses and wetlands are similar to those
earlier expressed for the road corridor. Similarly, I suspect that careful inspection would
reveal many more "watercourses," including intermittent and ephemeral streams, than the 70
crossings cited.
RESPONSE: A diesel spill has been added to the assessment, but it was judged that a gas
leak would not pose a potential risk to fish (Chapter 11).
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 226
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
The "probability" argument on p. 6-32 is an understandable attempt at quantification, but is
unpersuasive. Given the spill history of TAPS, pipelines in the Prudhoe Bay field, and
recently in Montana (?), suggesting the probability (with what confidence limits?) that there
"should be" only 1.5 stream-contaminating spills in78 years of operation seems wildly
optimistic.
RESPONSE: These probabilities andfrequencies are based on a large data set, not just the
TAPS experience (which is an atypical pipeline design and much larger than the diesel
line) or the spill in the Yellowstone River (which is a single event).
Assuming that any spill (over the 78-year project span) would last only two minutes (pp. 6-
32, 6-34), with a consequent minimal volume of spilled material, also seems highly
optimistic. Even highly-automated systems, with redundant sensors and automatic responses,
are susceptible to error or failure, and the Bristol Bay watershed environment is not benign
with regard to mechanical apparatus.
RESPONSE: The EPA agrees with this comment and has increased the response time to 5
minutes.
The authors appear to recognize this with their discussion of the Alumbrera incident.
RESPONSE: No change suggested or required.
51. Section 6.4: Potential road/culvert failures are recognized (again, not that ice issues are not
discussed). Extended periods for repair/rebuilding might be anticipated - witness the
repeated problems with the highway to Eagle, AK over the past several years - and that is a
State of Alaska responsibility, not that of a private company.
RESPONSE: We agree with the commenter that extended periods for repair/rebuilding
might be anticipated. Chapter 10 of the revised assessment notes that long-term fixes to a
road that was damaged by erosionalfailure of a culvert may not be possible until
conditions are suitable to replace a culvert or bridge crossing. Eurther, multiple failures,
such as might occur during an extreme precipitation event, would likely require longer to
repair.
Potential for multiple simultaneous or concurrent failures is appropriate. Non-Alaska
examples would be the Pacific Northwest flood events of 1964 and 1996, both major
precipitation events with widespread flooding and road failures, in a region with much more
developed infrastructure and response capacity.
RESPONSE: We agree with the commenter with respect to the likelihoodfor multiple
simultaneous or concurrent f ailures during extreme precipitation events. This is alluded to
in the assessment.
52. Chapter 7: Recognition of probable additional mining activity, in the wake of a Pebble
project, is appropriate. Assessment is necessarily limited to currently-known potential
mining projects. The cumulative and irreversible consequences of multiple developments,
with associated road, power, housing, communications infrastructure ("secondary
development") should be more heavily emphasized, even though it is not possible to quantify
all those consequences.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 227
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: The cumulative impacts of multiple developments, road corridors, and
secondary or induced development have been emphasized through additional discussion in
Chapter 13.
53. Chapter 8: Section 8.1.2 - note many potential failure modes not analyzed; the lack of
analysis in this Assessment should not be taken to mean that such failure could not or will
not occur.
RESPONSE: Because the number ofpotential failures is extremely large, it is necessary to
choose a representative set offailure scenarios. The focus of the assessment was on
potential effects to salmon. The revised assessment includes more failure scenarios (e.g.,
diesel pipeline failure, quantitative water treatment failure, and refined seepage scenarios)
and explains why these particular failure scenarios were chosen. Section 14.1.2 in the
revised assessment explains that the assessment only presents a few failure scenarios. The
EPA agrees with the commenter that lack oj analysis should not imply a failure could or
will not occur.
54. Table 8-1 (Row 2): The reasoning behind the statement that "Most [product concentrate
pipeline] failures would occur between stream or wetland crossing [sic] and might have little
effect on fish" is hard to understand; stream crossings, whether via elevated utilidors or via
sub-channel borings or utilidors, are locales of angular change, piping connections and
joints, and subject to stresses of hydrologic extreme events - so why would such sites be less
subject to potential pipeline failure?
RESPONSE: Most of the corridor is not at or adjacent to a stream or wetland Therefore,
failures are most likely to occur between streams or wetlands rather than at streams or
wetlands. The statement is not about probabilities per unit length.
55. Page 8-4 (Box 8-1): As noted elsewhere, the probability arguments for TSF dam failure are
not persuasive, and seem designed to imply that a TSF dam failure would not occur within
the next 10,000 years (or 3,000 to 300,000 years with three TSFs operational). This
implication is difficult to square with information on actual past failures presented in Box 4-4
and Table 4-8.
RESPONSE: The discussion of this issue has been expanded to clarify that the failure rate
is a design goal and is not based on empirical evidence.
56. Chapter 8: The potential risks and impacts are fairly and succinctly stated. Given the
extremely long-term nature of the projected Pebble project, and the irreversible changes
which would be imposed to the region, the risks seem, if anything, understated. I attribute
this to the decision to focus this Assessment on salmon and anadromous fisheries, with some
attention on salmon-mediated impacts - i.e., effects on indigenous culture, on wildlife other
than salmon, etc.
RESPONSE: No change suggested or required
John D. Stednick, Ph.D.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 228
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
1. Page ES-9: Economics of Ecological Resources - section seems weak.
RESPONSE: Economic data were included as background only. Economic effects of
mining are not assessed
2. Page ES-14: Overall risk to salmon and other fish. Never really separates fish species out in
other discussions. Dolly varden more sensitive to metals?
RESPONSE: The distributions of the salmonid species can be and are distinguished by
species, but not by their sensitivities to metals. For copper and some other metals, data are
available for rainbow trout that can be applied to other salmonids. Dolly Varden have not
been used as a test species.
3. Page 2-3 (P3): Four climate classes. Why this classification system'.' Perhaps easier to
identify by watershed maps?
RESPONSE: Climate classes are used as dejined from Feddema (2005) to give the average
annual moisture index. The Feddema Index was used as it broadly characterizes a given
landscape location on a spectrum of wetter to drier conditions due to the effects of
precipitation and evapotranspiration. Climate classes are used independent of watershed
boundaries, as this classification system allows one to compare regions that are either
alike or dissimilar across watershed boundaries and highlight that, depending on scale,
watersheds can be quite heterogeneous in their degree of water availability from
headwaters to mouth. The revised assessment now has five classes ranging from semiarid
to very wet as shown in Figure 3-1.
4. Page 2-5: Precipitation values - significant figures'?
RESPONSE: Precipitation values as shown in Figure 2-3 (Figure 3-2 in the revised
assessment) are calculated from S V I /' data. S \ IP data report monthly values of
precipitation to the nearest mm. During calculations, all precipitation values are rounded
to the nearest whole mm
5. Page 2-23: Monthly values of streamflow. Would be nice to see average or daily streamflows
somewhere.
RESPONSE: The purpose of this figure is to show the general pattern of annual
streamflows in this region. More detailed information (e.g., daily flows) are readily
available for the i'SGS gages, hut those data are not necessary to illustrate general annual
trends.
6. Pages 3-7 to 3-11: Would like to see more discussion of conceptual models. If a picture is
worth a 1000 words
RESPONSE: Conceptual models have been revised and new models developed, and these
models now appear in each of the risk analysis chapters to clarify the pathways considered
in each chapter.
7. Page 4-1 (PI): ...represent current good, but not necessarily best, mining practices. Why not
use the best methods or state of the art methods?
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 229
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: With regard to the terminology of "best", "good", or other terms for the
practices used, what was intended to be conveyed in the assessment is that we assumed
modern mining technology and operations. The terms are qualitative when generally
interpreted, or have a regulatory meaning. The term "best management practices" is a
term generally applied to specific measures for managing non-point source runofffrom
storm water (40 CFR Part 130.2(m)). Measures for minimizing and controlling sources of
pollution in other situations are referred to as best practices, state of the practice, good
practice, conventional, or simply mitigation measures. We have added a text box in the
revised Chapter 4 to discuss terms.
8. Page 4-18: Shaded relief. Perhaps contour lines in another figure?
RESPONSE: This figure is showing shaded relic/ of the landscape in Figure 4-6 and is
consistently used across multiple figures as a common base map for figures in this
assessment. It was decided to use shaded relief as opposed to contour lines to keep the
figure(s) cleaner in appearance, while also showing general topographic features. The
focus of this figure is on significant mineral deposits within the assessment watersheds and
not topography, therefore we have decided to keep the background as currently shown
9. Page 4-21 (PI): ...mostplausible sites given geotechnical, hydrologic, and environmental
considerations. Can this be elaborated?
RESPONSE: This sentence was intended to convey that these locations have similar
geotechnical, hydrologic, and environmental conditions to the site ofTSF 1. A more
comprehensive overview of the geologic and geomorpliic setting of the Nushagak and
Kvichak River watersheds is now provided in Chapter J.
10. Page 4-21 (P2): The TSF would be unlined other than on the upstream dam face and there
would he no impermeable barrier constructed between tailings and underlying groundwater.
Is this correct? 1 thought I read the whole TSF would be underlain by liner?
RESPONSE: The assessment states that TSF embankments would have an engineered
liner, but the impoundment is not lined in our scenarios.. I full liner is possible, but with
very large TSFs, stability with a liner may become an issue and benefits must be weighed
against other potential adverse outcomes, as well as costs. This is something that could be
considered in a regulatory permitting process. No change required
11. Page 4-24: Leachate Recovered. This refers to only the leachate collected from the dam face?
RESPONSE: No, this box refers to leachate from tailings and waste rock The text
"leachate recovered" in the formula included in the box refers to the leachate recovered in
testing conducted by the PLP. No change required
12. Page 4-26: Water management. This is confusing. Collect precipitation for processing, yet
divert upstream waters around the mine and not use? Where are the leachate recovery wells,
and are they just a safeguard?
RESPONSE: Water management measures are now clearly described and discussed in the
revision (Section 6.1.2.5), and in sub-sections for the mine components in the scenarios.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 230
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
Upstream water sources would be diverted around a mining site to minimize losses of that
water source to the environment, as it could remain uncontaminated The leachate
monitoring wells are to monitor groundwater downstream from the TSFs and waste rock
piles that would experience infiltration of their surfaces by precipitation falling on them
Those wells could be converted to recovery wells, or new recovery wells placed, if it were
found that seepage from these structures had contaminated the groundwater. Measures
are standard and common to existing mine sites.
13. Page 4-28: Significant figures on precipitation estimates? What is the ET and how is it
calculated?
RESPONSE: The water balance is now discussed in detail in Section 6.2.2 and the water
inputs discussed in Section 6.2.2.1. In the revised assessment, we used the three known
USGS gages draining the Pebble deposit site /NK100A (USGS 15302250) on the North
ForkKoktuli, SK100B (USGS 15302200) on the South Fork Koktuli. and UT100B (USGS
15300250) on the Upper Talarik] to calculate monthly mean flows at each gage over their
period of record. Monthly mean flows for each gage were summed across the year and
then averaged across the three gauges weighted by their watershed area to produce an
area-weighted average of net (Precipitation - ET) yearly runoff of860 mm/year for the
general deposit area. All data were presented to the nearest whole mm
14. Page 4-31 (P3): ...the mine pit would take approximately 100 to 300 years to fill. From
groundwater inflow only? Why such a large range of the estimate?
RESPONSE: The time for the pit to fill was based on groundwater infiltration and direct
precipitation into the pit and was presented as an approximate range. The lower end of the
range of estimates pertained to the Pebble 2.0 scenario and the higher end pertained to the
Pebble 6.5 scenario. Current estimates are approximately 20years, 80years, and 300years
for the Pebble 0.25, Pebble 2.0, and Pebble 6.5 scenarios, respectively.
15. Page 4-50 (P3): This peak flow calculation and discussion is confusing and needs
clarification.
RESPONSE: Presentation of the peak flow calculations and Itydrologic modeling of the
TSF failure scenarios has been revised and clarified. This is now presented in Section 9.3.
16. Page 4-52: What is the recurrence interval of the 356 mm?
RESPONSE: The recurrence interval is not reported in the reference cited. The PMP 24-
hr is greater than the reported 100-yr 24-hr (see Technical Paper No.47).
http://www. nws. noaa.gov/oh/lidsc/studies/piiip. html
http://www. nws. noaa.gov/oh/hdsc/PMP_documents/TP4 7. pdf
17. Page 4-60 (P4): Why a geometric mean using three values?
RESPONSE: The three values were chosen because they represented subsets of the
available data that better represented the conditions of the proposed pipelines (i. e.,
pipelines less than 20 cm in diameter, pipelines in northern climates, and pipelines run by
small operators). Focusing on these three subsets reduced the range of the estimated
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 231
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
probability of failure. The geometric mean was used because the comparison involved
ratios with potentially significant differences in magnitude in the denominators. Using the
geometric mean typically reduces the bias that a small dataset can introduce into an
arithmetic mean. In this case, the difference between the two means (0.0010 vs. 0.0011) is
immaterial and small with respect to the range of failure probabilities.
18. Page 5-10: Define highest reported index spawner.
RESPONSE: Section 7.1.2 defines this as the highest count of salmon reported in the EBD
for selected index reaches.
19. Page 5-20 (PI): ...salmon abundance related to pool size... .and heaver ponds provide
particularly large pools. Are data available to characterize the stream type? Are beaver
present?
RESPONSE: We now provide a characleri-alion of stream gradient, mean annual flow,
and % flatland to characterize general stream and valley characteristics. Beaver are
present in the area (Chapter 7).
20. Page 5-22 (P2): Assuming that no natural flow or uncontrolled runoff would be generated
from the mine footprint. Is all precipitation is intercepted or does this refer to the subsurface
streamtlow generation mechanisms?
RESPONSE: In our scenarios, nv assume that, all precipitation that falls within the mine
footprint (except the fraction oj the IS I' and waste rock leachate that bypasses the
groundwater interception and collection systems) is managed by the operator by collection
and routing leading to either storage in a TSE or testing, potential treatment, and release
to the environment. Any water consumed or stored on site would reduce the total
streamflow below baseline conditions, although some stream reaches would receive
increased stream flows due to discharges from the wastewater treatment plant.
Groundwater and streamflows are inextricably interdependent throughout the site, so one
could anticipate that areas with lower streamflows would also have lower groundwater
levels.
21. Page 5-23: It appears that the mean annual unit runoff is calculated incorrectly.
RESPONSE: This lias been corrected
22. Page 5-24: Table show s flow relumed from footprint. Does not fit with page 5-22?
RESPONSE: Water balance descriptions, tables, andfigures have been extensively
revised
23. Page 5-29 (P2): Groundwater-surface water connectivity. Are data available to show this
connection throughout the watersheds or does the groundwater only return to the hyporheic
in the low gradient areas? Similarly, where are the temperature data that suggest the lake and
groundwater connection or this reference by incorporation?
RESPONSE: Figure 7-14 is now included to illustrate areas of modeled upwelling
strength. We provide several examples of groundwater-moderated temperatures from the
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 232
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
EBD data. But water temperature data are not generally summarized here and we refer the
reader to the extensive data provided in the EBD.
24. Pages 5-32 to 5-39: The tables and hydrographs illustrating the potential flow changes are
difficult to appreciate or interpret. Another means of presentation?
RESPONSE: We now include figures (7-10 through 7-17) illustrating the locations, extent
and magnitude of flow changes.
25. Page 5-41 (Table 5-13): The value of 0.15 km affected by the maximum mine size is
questioned.
RESPONSE: Stream lengths are now rounded to I km
26. Page 5-52 (Table 5-17): Can we see summary statistics on water quality, not just means? Plot
of concentrations vs. streamflow?
RESPONSE: Coefficients of variation have been added to the table (now Table 8-10). Plots
of concentrations versus other variables are available in PLP's Environmental Baseline
Document, Chapter 9.
27. Page 5-55: These effluent specific values are higher than those for background surface water
because of the higher content of mineral ions. This sentence needs clarification.
RESPONSE: The sentence has been rewritten to clari/y that it refers to differences in the
BLM-deri ved criteria.
28. Page 6-6 (Table 6-1): Last line. What is the +/- value after the mean?
RESPONSE: + or - one standard deviation are reportedfor the basin-wide samples; this
has been clarified in the table.
29. Page 6-13 (P3): ... an intense local storm. Why use the Type la distribution for precipitation
distribution?
RESPONSE: This has been corrected to Type I per SCS guidance for Alaska
Roy A. Stein, Ph. I).
1. Global (see Table 5-3 as example): Some thought should be given to significant figures for
the numerical values given in the report, especially so for Chapter 5, where much uncertainty
exists regarding stream lengths blocked (for example) by the mine footprint: "...34.9 km of
first- through third-order streams..will be eliminated. Rounding in this context makes
sense for we really do not know this impact to the nearest 0.1 km. I would encourage a
thorough review of these values throughout the report.
RESPONSE: Significant figures for calculations and measurements by EPA have been
changed to reflect uncertainty (for example, stream length is reported to the km).
Significant figures from literature are not revised
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 233
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
2. Page 2-17: Salmon populations are closely managed by Alaska Department of Fish and
Game; how closely and what do we know of the populations that are managed regarding
numbers, resilience, variability, etc.? A box summarizing the fishery management practiced
by Alaska Department of Fish and Game would help put biological resources in perspective.
RESPONSE: Box 5-2 has been added in Chapter 5, providing an overview of commercial
fisheries management in Bristol Bay.
3. Page 2-17: The importance of salmonids to marine predators is likely not an issue, given the
very large numbers of salmon stocked in the Pacific.
RESPONSE: No changes suggested or required.
4. Page 2-20: Nushagak and Kvichak rivers contain >58,000 km of streams; 13% is
anadromous fish habitat, but this is likely underestimated. How do we know this is an
underestimate? What proportion of this >58,000 km has been surveyed?
RESPONSE: ADE& <7 estimates that the current listing of Waters Important for the
Spawning, Rearing and Migration of Anadromous Fishes represents "less than 50% of the
streams, rivers and lakes actually used by anadromous species" (ADE&G A WC database,
available at: www. adfg. alaska.gOv/sf/SARR/A WC/)). This clarification has been added to
the assessment in Box 7-1.
5. Page 2-25 (bottom of page): I wonder if a bit more couldn't be written about the idea of a
salmon sanctuary, fleshing out the ideas of Rahr and Pinsky here.
RESPONSE: The creation of salmon sanctuaries in the Bristol Bay region is a
management decision, and thus outside the scope of this assessment; therefore, no change
is necessary.
6. Page 3-2: With all the items in the first full paragraph eliminated from consideration (e.g.,
power generations, worker housing, Cook Inlet Port), might the analysis herein be considered
minimal impact on the I Jrislol I Jay watershed?
RESPONSE: The scope of the assessment has been clarified throughout the document,
particularly in Chapters / and 2. It is true that this assessment only considers a subset of
potential impacts from large-scale mining.
7. Pages 3-7 to 3-11: These conceptual models might best be placed in the chapter to which
they refer; in so doing, it is easier for the reader to follow along. In turn, these models did
not seem to be discussed in text to the extent that they drove the impacts generated. Add
more explanatory text.
RESPONSE: In the revised assessment, conceptual models are presented and discussed in
the appropriate chapters.
8. Page 4-5 (bottom of page): Typically, when citing a figure in another chapter, i.e., one not
near the current text, the format should be "see Figure 3-1" to keep the reader on track.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 234
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: Comment noted. Figure references have been retained in parentheses.
9. Page 4-8: Is there some chance of "block caving" here? Some text clarifying this point here
would be appropriate.
RESPONSE: Additional information on block caving is provided in Chapter 4 (Section
4.2.3.1 and Box 4-4). Block-caving is a possibility for mining deposits that are deeper.
Sources of hazards assessedfor risks, however, would be similar, as block-caving and
surface mining both create tailings, waste rock (block-caving does have a smaller amount
of waste rock), and sources of contamination to the water.
10. Chapter 4: Could these data and insights be productivelv moved to Chapter 5, thus reducing
redundancy? The organization would be 1) a description of the mine features relevant to the
text, then 2) a discussion of salmon and the mining impacts on their habitat. Separating these
into chapters seems artificial.
RESPONSE: Revision of the assessment has included organization of material into
additional chapters and the EPA believes this new organization is easier to follow.
11. Page 5-20 (P3): What does the phrase, "free-water area" mean? Ice free, perhaps?
RESPONSE: Removed
12. Pages 5-20 to 5-21: Paragraphs on these two pages reflect some of the redundancy that I saw
throughout the report. In the 3"' full paragraph on page 5-20, text begins with
"...groundwater inputs may be critical..." and in turn on page 5-21 in the first full paragraph,
the topic sentence begins "...groundwater-influenced stream flow.. .likely benefit fish.
Both deal with the same topic. Hence, these two paragraphs could easily be combined,
serving to shorten the text, reduce redundancy, and improve readability.
RESPONSE: These paragraphs have been combined
13. Page 5-22: Similarly, the text on page 5-22 includes citations (at the end of the same
sentences) to both sections and tables and figures in Chapter 4. This suggests to me that the
text is not only overlapping, it is redundant and better overall report organization would serve
its presentation well.
RESPONSE: Chapters have been reorganized, but some cross-chapter referencing
remains necessary. In this case, it is important that the reader be able to refer back to the
mine scenarios and operations chapter if desired, to better understand the complex water
management system and budget that drives the analysis in Chapter 7.
14. Page 5-23 (Table 5-5): Some explanation in the table title would benefit the reader, as to
where these gages are placed in the stream. The labeling is arcane at best UT100D; letters
suggest Upper Talarik Cr., but what is 100D. Now, I figured out from text that the first ones
were high in the watershed and then they proceeded downstream (increasing drainage area
gave me a hint as well). Use better descriptors (such as SKI through SK5 from low to high
stream order) or explain the ones that are being used.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 235
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
RESPONSE: New figures illustrating gage locations are included. Gage names remain
consistent with EBD usage for consistency
15. Page 5-26 (Table 5-7): I cannot find any bold in this table that would reflect the pre-mining
condition. Is that the same column as "Pre-Mining" as it is in Table 5-8 on page 5-34 (and
the next few tables as well)?
RESPONSE: Corrected
16. Page 5-27: Cite Figure 5-8 in text any time the stream gages are mentioned. The reader then
has the ability to easily refer back to stream gage locations.
RESPONSE: Corrected
17. Pages 5-32 to 5-39: How do these tables and ligurcs differ? Might they be 10% reduction in
flow, 11-20% reduction inflow and >20% reduction in flow, as suggested in text? If so, then
these table and figure titles need to reflect this information and be better described in the
legends of the figures. Finally, do we need both tables and figures?
RESPONSE: Figures have been removed, with exception on one illu\tration to clarify
process.
18. Page 5-31: The Richter et al. (2011) reference, which underpins this section is incomplete in
the Literature Cited (Chapter 9), suggesting only March as the publication date. Update in
any revision.
RESPONSE: Citation has been updated.
19. Page 5-31: It should be made clear that the Richter et al. (2011) sets quite specific bounds for
all rivers regarding ecosystem function and does not provide any specific insight into salmon
production (made somewhat clear in the last sentence of the next to the last paragraph on
page 5-43). I think an additional caveat stating this explicitly on page 5-31 would improve
the text.
RESPONSE: Additional clarification has been added
20. Page 5-46: Stream flow losses due to the mine footprint are discussed first as underestimated,
then as overestimated, and then a conclusion that they are underestimated (i.e., because the
mine will need more water than is available from surface run-off). If this is the case, why go
through the other scenarios.. .to seem even-handed? I am not sure all of this text is required.
RESPONSE: Uncertainties and assumptions sections have been revised
21. Page 5-46: Mine start-up will require more water than is available from the footprint of the
mine itself. Hence, water will be captured from other streams than just those associated with
the mine, which will influence stream flow and groundwater supplies. Are there estimates of
the amount beyond the surface water that will be required?
RESPONSE: An extensively-revised water balance description and accounting are now
used Indirect effects of groundwater loss to pumping are incorporated
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 236
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
22. Pages 5-53 to 5-58:1 am a little confused by this section. The implication throughout is that
copper toxicity will be based on the response of aquatic invertebrates (which would then be
protective of direct effects on salmon). However, near the end of this section, there is a
discussion about zooplankton being most sensitive and a comment about the reliance of
juvenile sockeye rearing in lakes on these zooplankton. Yet, there is no resolution of what
criteria will be used for toxicity values.. .will it be aquatic invertebrates or the more sensitive
zooplankton? Clarification is required here.
RESPONSE: Zooplankton are aquatic invertebrates. The same criteria apply to both
invertebrates in the benthos and those that drift, in the lake.
23. Page 5-59: "Discharge permits for mine in the Bristol Bay watershed should include relevant
whole-effluent toxicity testing and monitoring of biolic communities in receiving streams".
Does this quote solve the problem mentioned in the previous comment? Is this a realistic
expectation for mine operators before permits are issued?
RESPONSE: The statement reflects the hopes of the assessment authors. They do not
constitute an obligation upon the State of Alaska. This is clarified in the revised
assessment.
24. Page 5-68 (Table 5-22): >10% rather than <10% in the table title.
RESPONSE: The reference was correct in the draft. Table 10-8 now shows streams with
gradients <12% likely to support salmon.
25. Chapter 6: Rivers is lower case when multiple rivers are listed.
RESPONSE: Comment noted.
26. Page 6-1: Why assume that only 20% of the tailings stored would be mobilized with any dam
failure? Is there a justification for this important assumption? Yes, see Appendix I, page 14
for c i tation to Dalpatram (2011); this should be cited in the main report in Chapter 6.
RESPONSE: Citations oj Dal/uitram (2011) and Azam and Li (2010) have been added,
drawn from Appendix I.
27. Page 6-4: Add map showing the impact of the failed TSF 1, i.e., distribution of sediments and
impact downstream.
RESPONSE: Because we are not predicting specific sediment depths in this revision, we
have determined that this information was best presented by a description in the text and
tables.
28. Page 6-42: The sentence in mid-paragraph (3rd complete paragraph on this page) ".. .multiple
failures such as might occur..." My guess is that an example should follow the phrase "such
as."
RESPONSE: Culvert failures are described in the preceding paragraphs (in greater detail
in the revised assessment), so an example is not needed here.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 237
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
29. Page 7-2: The sentence at the end of the last full paragraph on the page makes little sense:
"The overall consequences are diminished and extinct salmon populations."
RESPONSE: This sentence has been clarified.
30. Page 8-6 (last sentence of 2nd complete para): "further" should be "farther"; sorry, I just
couldn't help myself...
RESPONSE: Text revised.
William A. Stubblefield, Ph.D.
None
Dirk van Zvl, Ph.D., P.E.
I have indicated a number of specifics in the text above and do not have any others to list.
Phyllis K. Weber Scannell, Ph. I).
1. Page 4-11: The document states "geomembranes are generally estimated by manufactures to
last 20 to 30 years when covered by tailings (North pers. comm...)." Unless North is a P.E.
with experience in geomembranes. the statement needs a stronger reference. For example,
Erickson et al (2008)* discuss the quality issues with geomembranes related to manufacture,
installation and application of a soil-based cover (such as bentonite).
RESPONSE: The reference is appreciated, but others were chosen for use related to time
of life for liners. The discussion of liners has been expanded and moved to Chapter 4
(Section 4.2.3.4) in the revision, and includes additional material and references on
lif etime. The personal communication reference has been removed
2. Page 4-23 (LI): This first sentence is confusing and implies that oxygen has low solubility
because it is in the tailings pond. Suggested change: eliminate the first phrase "In a TSF".
RESPONSE: The change has been made in the revised assessment.
3. Page 6-36 (P3): Last sentence states: [water] treatment would continue until institutional
failures ultimately resulted in abandonment of the system, at which time untreated leachate
discharges would occur. This statement is not supported by any documentation and is not
clear what is being implied. Failure of governments? As stated in my response to questions,
any mine plan must include sufficient bonding and plans for reclamation, including necessary
water treatment.
RESPONSE: Bonding and a perpetual trust could result in treatment for some period, but
it possible that long-term funding will not last to the end of time.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 238
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
-------
Dr. Paul Whitney
4. Page 6-37 (P4): End of paragraph states "premature closure could leave waste rock piles in
place." Again, there is a need for plans for mine closure, concurrent reclamation and
sufficient bonding.
RESPONSE: The premature closure scenario is intended to address the possibility that the
closure plan is not followed It highlights the needfor adequate bonding andfor plans
that include closure by some agency.
5. App G, Page 5 (PI, last line): Document states: ".. .other short road segments connect
Dillingham to Aleknagik and Naknek to King (Figure I)."
Shouldn't King be King Salmon?
RESPONSE: Corrected
6. App G, Page 5 (P3): Dolly Varden should be capitalized throughout the document. The list
of fish species should contain scientific names or reference a table of common and scientific
names.
RESPONSE: Corrected
The document states "In the most comprehensive published field inventory, Woody and
O'Neal (2010) reported.'' Because the authors have not reviewed other documents on field
inventories, the phrase "most comprehensive" should be changed to "a comprehensive"
RESPONSE: Corrected.
Paul Whitney, Ph. I).
A lot of page and paragraph comments are included the above responses to charge questions. I
have no further comment.
EPA DRAFT RESPONSE TO PEER REVIEW COMMENTS on Page 239
An Assessment of Potential Mining Impacts on Salmon Ecosystems of Bristol Bay, Alaska (May 2012 draft)
------- |