&EPA
United States
Environmental Protection
Agency
Office of Water
Office of Wetlands, Oceans, and Watersheds
Handbook for
Developing Watershed TMDLs
DRAFT
-------
U.S. Environmental Protection Agency
Office of Wetlands, Oceans, and Watersheds
1200 Pennsylvania Avenue, N.W. (4503T)
Washington, D.C. 20460
-------
Handbook for Developing Watershed TMDLs
DRAFT
December 15, 2008
U.S. Environmental Protection Agency
Office of Wetlands, Oceans & Watersheds
-------
Handbook for Developing Watershed TMDLs December 2008
Disclaimer
This document provides technical information to TMDL practitioners who are familiar
with the relevant technical approaches and legal requirements pertaining to developing
TMDLs and refers to statutory and regulatory provisions that contain legally binding
requirements. This document does not substitute for those provisions or regulations, nor
is it a regulation itself. Thus, it does not impose legally binding requirements on EPA or
States, who retain the discretion to adopt approaches on a case-by-case basis that differ
from this information. Interested parties are free to raise questions about the
appropriateness of the application of this information to a particular situation, and EPA
will consider whether or not the technical approaches are appropriate in that situation.
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Contents
/. Introduction 1
1.1. Rationale 2
1.2. Purpose 3
1.3. Audience 4
1.4. Organization 5
1.5. Background 5
1.5.1. History of the TMDL Program 5
1.5.2. History of the Watershed Approach 6
1.5.3. Integration of Watershed Approach into Core Water Programs 7
2. Understanding Watershed TMDLs 11
2.1. Environmental Benefits 12
2.2. Financial and Programmatic Benefits 14
2.3. Implementation Benefits 20
3. Identifying Candidates for Watershed TMDL Development 23
3.1. Identifying Opportunities based on Impairment and Related Issues 25
3.1.1. Impairment and Source Type 25
3.1.2. Waterbody Type 27
3.1.3. Technical Approach Considerations 28
3.2. Accounting for Programmatic Commitments and Priorities 28
3.2.1. Consent Decrees and Settlement Agreements 29
3.2.2. Priority Rankings 29
3.2.3. Availability of Resources 30
3.2.4. Public Concerns 34
3.3. Leveraging Existing Watershed-based Programs or Efforts 34
4. Developing Watershed TMDLs 37
4.1. Stakeholder and Public Involvement 37
4.2. Watershed Characterization 40
4.2.1. Data Analysis for Problem Identification 41
4.2.2. TMDL Target Identification 48
4.2.3. Source Assessment 49
4.3. Linkage Analysis 51
4.3.1. Factors Affecting Selection of Technical Approach for Watershed TMDL
Development 51
4.3.2. Practical Applications of Various Approaches for Watershed TMDL
Development 56
4.4. Allocation Analysis 69
4.4.1. Scale or Resolution of Source Allocations 69
4.4.2. Priority and Feasibility of Source Allocations 71
4.4.3. Stakeholder Priorities and Implementation Goals 73
4.5. Development and Submittal of TMDL Report 73
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
5. Supporting Implementation of Watershed TMDLs 81
5.1. Multiple Program Involvement/Coordination 81
5.1.1. Watershed-based Permitting 81
5.1.2. Water Quality Trading 83
5.1.3. Watershed-based Planning 84
5.2. Structuring the TMDL to Support Implementation Activities 85
5.3. Follow-up Monitoring 87
5.4. Financial Resources for Implementation 89
6. References 91
Appendix: Case Studies A-1
Case Study 1: Calleguas Creek, California A-3
Case Study 2: Gauley River TMDLs, West Virginia A-13
Case Study 3: Long Island Sound Nitrogen TMDL A-21
Case Study 4: Lower Fox River Basin and Green Bay, Wisconsin A-27
Case Study 5: Tidal Potomac and Anacostia PCB TMDL A-37
Case Study 6: St. Marys River TMDLs, Indiana A-47
Case Study 7: Tualatin River TMDLs, Oregon A-53
Case Study 8: Virgin River TMDLs, Utah A-61
iv DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Figures
Figure 1-1. Number of TMDLs approved by fiscal year since October 1, 1995 2
Figure 1-2. Complexity of TMDLs developed, as estimated by EPA Regions 3
Figure 3-1. Process for identifying the scope of a watershed TMDL 24
Figure 4-1. General steps in developing a TMDL 38
Figure 4-2. Example map showing average TDS concentrations throughout a watershed 44
Figure 4-3. Example graph showing summary statistics in measured TDS concentrations
along the length of an impaired river 44
Figure 4-4. Example graph showing relationship between upstream and downstream data
to evaluate potential impact of an expected source 45
Figure 4-5. Example graph showing data collected upstream and downstream of an
expected source 45
Figure 4-6. Examples of different data representations to evaluate the relationship
between flow and water quality 47
Figure 4-7. Considerations for selecting a TMDL development approach 52
Figure 4-8. Model output illustrating multiple allocation scenarios that meet water quality
standards 72
Figure 4-9. Example outline for a watershed TMDL report organized by impaired
segment 76
Figure 4-10. Example outline for a watershed TMDL report organized by impairment
type 77
Figure 4-11. Example outline for a watershed TMDL report organized by impairment
type 78
Tables
Table 3-1. Summary of Common Pollutants and Sources 26
Table 4-1. Summary of Technical Considerations for Selecting a TMDL Development
Approach 52
Table 4-2. Watershed Models Commonly Used for TMDL Development 61
Table 4-3. Example Watershed TMDLs Developed with Commonly Used Watershed
Models 62
Table 4-4. Examples of Different Scenarios to Meet the Same Load Target 72
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
vi DRAFT
-------
, Introduction
Section 303(d) of the 1972 Clean Water Act (CWA)1 requires , KesoMceg.
states, territories, and authorized tribes to develop lists of impaired !
waters. These impaired waters do not meet water quality standards I For ™rf, in?.r,nf'5," °n ™°Ls
^ i j : visit EPA sTMDL Web site at
www.epa.qov/owow/tmdl.
that states, territories, and authorized tribes have set for them,
even after point sources of pollution have installed the minimum
required levels of pollution control technology. The law requires
that these jurisdictions establish priority rankings and develop total maximum daily loads
(TMDLs) for waters on the lists. A TMDL specifies the maximum amount of a pollutant
that a waterbody can receive and still meet water quality standards and allocates pollutant
loadings among point and nonpoint pollutant sources. A TMDL is the sum of the
individual wasteload allocations (WLAs) for point sources and load allocations (LAs) for
nonpoint sources and natural background (40 CFR 130.2) with a margin of safety (CWA
section 303(d)(l)(c)). The TMDL can be generically described by the following equation:
TMDL = LC = £WLA + £LA + MOS
where:
LC = loading capacity or the greatest loading a waterbody can receive without
exceeding water quality standards;
WLA = wasteload allocation, or the portion of the TMDL allocated to existing or
future point sources;
LA = load allocation, or the portion of the TMDL allocated to existing or future
nonpoint sources and natural background; and
MOS = margin of safety, or an accounting of uncertainty about the relationship
between pollutant loads and receiving water quality. The margin of safety
can be provided implicitly through analytical assumptions or explicitly by
reserving a portion of loading capacity.
The process of calculating and documenting a TMDL typically involves a number of
tasks, including characterizing the impaired waterbody and its watershed, identifying
sources, setting targets, calculating the loading capacity using some analysis to link
loading to water quality, identifying source allocations, preparing TMDL reports and
coordinating with stakeholders. EPA, states, tribes, and other TMDL practitioners have
found that it can be advantageous to simultaneously complete the TMDL process and the
associated tasks for multiple impaired waterbodies in a watershed. This process of
developing watershed TMDLs—those that establish allocations for multiple impaired
waterbodies in a watershed—is the focus of this document.
1 EPA's regulations for implementing section 303(d) are codified in the Water Quality Planning
and Management Regulations at 40 CFR Part 130, specifically at sections 130.2, 130.7, and
130.10. The regulations define terms used in section 303(d) and otherwise interpret and expand
upon the statutory requirements.
DRAFT
-------
Handbook for Developing Watershed TMDLs
December 2008
1.1. Rationale
Since 1995, more than 34,300 approved TMDLs have been developed by the states or
established by U.S. Environmental Protection Agency (EPA), as shown in Figure 1-1,
addressing more than 36,000 listed impairments. While this represents a significant
number of TMDLs and a large effort by the states and EPA, states have identified nearly
70,000 TMDLs still to be developed in the next 8-13 years (one or more impairments
affecting more than 41,500 waterbodies)2. This represents a pace of 5,300 to 8,700
TMDLs per year, nearly twice the average number of TMDLs developed per year during
the last 10 years. As many states work to refine and implement more comprehensive
monitoring and assessment strategies, it is likely that the number of additional impaired
waterbodies requiring TMDLs will continue to increase in future 303(d) lists.
In a 2001 draft study conducted by EPA (USEPA 200 la, 200 Ib), it was estimated that
"the national total undiscounted cost to develop 36,225 TMDLs for the pollutant causes
of impairment identified in the 1998 303(d) lists is estimated to be about $1 billion, with
a likely range of $0.97 to $1.06 billion." Using these figures, a conservative projection of
the necessary funding to complete the remaining 70,000 TMDLs is between $1.86 and
$2.04 billion.
8000
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Rscal Year
Figure 1-1. Number of TMDLs approved by fiscal year since October 1,1995.
Adding to the expected demand on state resources, a recent survey of EPA regional
TMDL programs indicated that efforts during previous years focused on many of the less
complex TMDLs (Figure 1-2), leaving the more complex TMDLs in terms of waterbody
type and pollutant kinetics still to be completed. Much of the TMDL efforts of states in
the late 1990s focused on single-segment TMDLs, many representing single WLAs for
point source-impaired waters. Into the early 2000s, some TMDL efforts began
using/applying a watershed framework for developing multiple TMDLs; however, a
majority of TMDLs were still being developed as single-segment TMDLs, guided by
2 Based on EPA's National Section 303(d) Fact Sheet,
http://iaspub.epa.gov/waters/national rept.control as of August 5, 2008.
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
pressures from court orders, limitations in available data, and activities to accommodate
localized NPDES activities.
O)
c
73
o
"c
FYOO
FY05
FY06
FY10
I Easy TMDLs Mid-level TMDLs • Difficult TMDLs
Figure 1-2. Complexity of TMDLs developed, as estimated by EPA Regions.
Based on these trends, states are facing a significant workload and associated financial
responsibilities to develop the 65,000 TMDLs on current 303(d) lists. It is also expected
that many of these TMDLs will require more complex analyses than the 34,300 TMDLs
developed to date. While the workload and funding needs are increasing for TMDL
development, the level of resources available to the states and to EPA are either declining
or remaining unchanged. For all of these reasons, it is necessary for states to expedite
TMDL development using an approach that will
efficiently address the maximum number of
impairments in a scientifically defensible manner. One
strategy for doing this is to use a watershed framework
for developing TMDLs. Watershed TMDLs can help
states to reduce their per-TMDL costs and address more
pollutant-waterbody combinations with the given
resources while recognizing a number of environmental
and programmatic benefits.
1.2. Purpose
The purpose of this document is to promote the
development of watershed TMDLs and to provide
TMDL practitioners the information to move from
concept to practice, making watershed TMDLs a
standard approach in their programs. While the goal of
this document is to support states in integrating a
watershed approach to TMDL development into their
overall programs, it focuses on the technical and
programmatic information and considerations for
developing and implementing watershed TMDLs
primarily at the project level. A number of reference
Definition:
Watershed TMDL vs. Single-segment TMDL
I A watershed TMDL is the result of a holistic
approach to the simultaneous development of
I multiple TMDLs for hydrologically linked
impaired segments. A watershed TMDL will
address one of the following:
• Same pollutant in multiple segments (e.g.,
fecal coliform bacteria)
• Different but similar pollutants in multiple
segments (e.g., nitrogen and phosphorus)
• Different and unrelated pollutants in multiple
segments (e.g., chromium and bacteria)
A single-segment TMDL is one that addresses
only a single impaired waterbody, whether for a
single or multiple impairments.
While a watershed TMDL evaluates multiple
segments and impairments in an integrated
analysis, a TMDL calculation with associated
LAs and WLAs is needed for each waterbody-
pollutant combination included on the 303(d) list.
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
documents already exist to help states establish a program that uses the watershed as an
operating framework for all of their statewide water programs.
This document outlines the basic technical issues related to watershed TMDLs,
identifying the issues for practitioners to consider and tools and resources that can help
them when planning for and developing watershed TMDLs. It also identifies the benefits
of developing watershed TMDLs as well as the obstacles and ways to address them.
Finally, this document highlights the connections between watershed TMDLs and other
water programs, identifying opportunities for integrating watershed TMDLs into other
watershed management efforts, such as watershed planning, permitting and water quality
trading.
The overall message of this document is that TMDL practitioners should approach
TMDL planning and development with the understanding that the watershed approach is
a framework for coordination and development that can apply to a majority of TMDLs.
During TMDL planning and scheduling, practitioners might find it necessary to identify
those cases that do not fit or are not appropriate within the watershed framework and
address those TMDLs separately.
1.3. Audience
The information in this document targets TMDL il _ M t • ~\
practitioners with an understanding of and experience || v attt »• j
with developing TMDLs. TMDL practitioners include || What Will You Get from this Document? |
those organizations actively developing TMDLs, II _.._. _ .... ... . . .. I
. ,_••.. jr-ji- ° i • ll • TMDL Practitioners—get the information i
including state and federal environmental agencies, j j you need to dedde wheyn and how to deve|op j
third-party TMDL developers, and private consultants. il watershed TMDLs i
To successfully develop TMDLs on a watershed basis, {j . Program Managers—find out how |
support and buy-in must occur at all levels of the TMDL j j watershed TMDLs will benefit your program !
program. This document presents information important j j and the issues you need to consider to j
for program managers as they outline the process and i I successfully integrate a watershed approach j
, c. r, , , j r™,™ 11 ll into your TMDL program I
benefits or developing watershed 1 MDLs and make II I
decisions about integrating the approach into their i! ' ™DL Implementation Partners— j
„_ ., , • , , ,. 1-1 ll understand how watershed TMDLs can more I
programs. While this document does discuss technical {j effectively support and coordinate |
aspects of TMDL development for watershed TMDLs, it || implementation activities and can potentially |
does not discuss in detail the overall TMDL || support and reduce your ongoing efforts j
development process or required TMDL elements. In || j
this document, practitioners will find practical
information related to commonly encountered technical and programmatic issues,
available tools, and data and information sources to specifically support the development
of watershed TMDLs.
In addition to TMDL practitioners, the information included in this document will be
helpful to TMDL implementation partners. Implementation partners include National
Pollutant Discharge Elimination System (NPDES) permitting authorities, municipalities
responsible for implementing controls or subject to municipal separate storm sewer
system (MS4) regulations, and watershed organizations working to coordinate and
implement management efforts in the watershed.
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
1.4. Organization
To present the information necessary to support development of watershed TMDLs, the
handbook is organized with the following sections:
• Section 2: Understanding Watershed TMDLs. This section defines a watershed
TMDL and discusses the environmental, financial and programmatic benefits
associated with watershed TMDLs.
• Section 3: Identifying Candidates for Watershed TMDL Development. This
section discusses a series of criteria or factors that can be considered to define the
scope of a watershed TMDL, whether on a project-specific basis or for a broader
planning exercise (e.g., statewide).
• Section 4: Developing Watershed TMDLs. This section discusses the issues and
considerations relevant to watershed TMDLs at each step of the TMDL development
process, including stakeholder and public involvement, watershed characterization,
linkage analysis, allocation analysis, and development and submittal of the TMDL
report.
• Section 5: Supporting Implementation of Watershed TMDLs. This section
discusses a number of topics related to implementation, including coordination with
related watershed programs (e.g., watershed-based permitting), follow-up
monitoring, and financial resources for implementation.
There are a series of text boxes included throughout the document to highlight certain
topics or provide specialized information related to development of watershed TMDLs.
These boxes fall into the following categories:
• Definition: Defines key terms or concepts for better understanding of the respective
discussion.
• Example: Provides real-world or hypothetical examples of watershed TMDL
development and related programs to illustrate the benefits, technical issues or
considerations associated with watershed TMDLs.
• Process Tip: Provides tips to focus efforts and successfully plan for and complete the
steps to develop watershed TMDLs.
• Program Notes: Provides background on the TMDL or other watershed programs on
topics relevant to watershed TMDL development.
• Resources: Identifies available resources to support completion of watershed
TMDLs or to find further information on related topics.
1.5. Background
1.5.1. History of the TMDL Program
While the CWA has required TMDLs since 1972, the early focus was on developing
WLAs for individual point sources. In the late 1990s, environmental groups began
bringing legal actions against EPA seeking more accurate listing of waters and
development of TMDLs. The litigation resulted in court-ordered schedules or consent
decrees outlining TMDL development schedules in a number of states and requiring EPA
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
to develop the TMDLs if the states do not. A significant effect of the litigation is that
EPA and states began focusing TMDL efforts on waters impaired by not only point
sources, but also by nonpoint sources as well as by a combination of point and nonpoint
sources.
Since then states have strived to meet TMDL commitments in an effective and efficient
manner. EPA has generated a number of guidance documents related to TMDL
development, many providing recommendations for developing TMDLs on a watershed
basis to provide both environmental and programmatic benefits. For example, EPA's first
TMDL guidance document, Guidance for Water-Quality-based Decisions: The TMDL
Process, states:
Many water pollution concerns are area-wide phenomena that are caused by multiple
dischargers, multiple pollutants (with potential synergistic and additive effects), or
nonpoint sources. Atmospheric deposition and ground water discharge may also result in
significant pollutant loadings to surface waters. As a result, EPA recommends that States
develop TMDLs on a geographical basis (e.g., by watershed) in order to efficiently and
effectively manage the quality of surface waters. (USEPA 1991)
At that time, the program began to move from a foundation of technology-based controls
to one of water quality-based controls evaluated and implemented on a watershed scale.
However, because of the workload demand and court-ordered schedules, many states
continued to focus on developing "segment-by-segment" TMDLs. In following years,
EPA policy memos reaffirmed their strategy to develop and implement TMDLs on a
watershed basis. For example, in 1997, the EPA Assistant Administrator for Water issued
a memo as a follow-up to the previously released Healthy Watershed Strategy, noting that
a key component of the strategy was developing and implementing TMDLs to manage
water quality on a watershed scale.
1.5.2. History of the Watershed Approach
The process for addressing water quality problems on a | Ptoarcun Notes'-
watershed basis is not a new idea. EPA has been I ....
,..,_, , ,.,.,,. , • i Although ten years of effort have resulted in
promoting this idea for decades, solidifying their support j general awareness of the watershed approach
in the early 1990s and continuing today through j within the Agency and at the State and local
guidance, policy and training. EPA's Office of Water \ level, recent evaluations show substantial gaps
started The Watershed Academy in 1994 to provide ! '" actual implementation. EPA believes that
. • • , , .• , . • , ., , • i the watershed approach should not be seen as
training courses and educational materials on the basics j merely g speciaffnitiative, targeted atjust a
of the watershed approach. The Watershed Academy has j se/ecf number of places or involving a
developed into one of the agency's key tools for j relatively small group of EPA or State staff.
educating and supporting agencies, stakeholders, and ! Rather, it should be the fulcrum of Federal and
citizens working to protect water and watershed ! fate ^oration ™* P"*ection efforts, and
T ^rx-V.T.A i i rrr , , r, > those of our many stakeholders, both private
resources. In 1995, EPA released Watershed Protection: j and puMc " (USEPA 2002 2005a)
A Project Focus and Watershed Protection: A Statewide \
Approach to further the premise that many water quality
and ecosystem problems are best solved at the watershed level rather than at the
individual waterbody or discharger level. In June 1996, EPA's Office of Wetlands,
Oceans and Watersheds (OWOW) released The Watershed Approach Framework,
establishing guiding principals for watershed management. EPA has continued its efforts
to promote implementation of clean water programs on a watershed basis with the
development of additional guidance, watershed information tools (e.g., Watershed
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Information Network, Surf Your Watershed), grants for watershed-based programs and
projects, and a variety of training courses.
In 2002, the EPA Assistant Administrator for Water issued a memorandum "to reaffirm
the Office of Water's commitment to advancing the watershed approach" (USEPA 2002).
Subsequently, the Office of Water's National Water Program Guidance for Fiscal Years
2005, 2006, 2007, 2008, and 2009 all included goals to "restore and improve water
quality on a watershed basis." As part of this goal, EPA encourages states "to organize
schedules for TMDLs to address all pollutants on an impaired segment and to organize
efforts so that segment level restorations are clustered together to provide improvements
on a watershed basis" (www.epa.gov/water/waterplan/). To further define this goal, the
FY2006 program guidance included a Subobjective Implementation Plan to describe
strategies and performance targets for using the watershed approach for the
implementation and integration of a number of core water programs, including water
quality standards, monitoring, TMDLs, nonpoint source control, NPDES permitting, and
wastewater infrastructure (USEPA 2005). Recognizing the potential benefits of managing
water quality using a place based approach, the current EPA administration is continuing
to promote and address issues on a watershed level and anticipates that these watershed-
based strategies and performance targets will continue in 2009 and beyond.
1.5.3. Integration of Watershed Approach into Core Water Programs
Recently, EPA's Office of Water has supported their program goals by developing
guidance for further integrating the watershed approach into their core water programs.
For example, in July 2007, the Office of
Wastewater Management (OWM) released the
Watershed-Based National Pollutant Discharge
Elimination System (NPDES) Permitting Technical
Guidance (USEPA 2007a) as a follow up to the
2003 Watershed-based National Pollutant
Discharge Elimination System (NPDES) Permitting
Implementation Guidance. Watershed-based
permitting emphasizes addressing all regulated
stressors within a hydrologically defined drainage
basin, rather than addressing individual pollutant
sources on a discharge-by-discharge basis. The
2007 guidance provides greater detail on permit
development and issuance and focuses on helping
NPDES authorities develop and issue NPDES
permits that fit into an overall watershed planning
and management approach. Watershed-based
permitting can encompass a variety of activities
ranging from synchronizing permits within a basin
to developing water quality-based effluent limits
using a multiple discharger modeling analysis. The
type of permitting activity will vary depending on
the unique characteristics of the watershed and the
sources of pollution.
In March 2008, OWOW released the Handbook for
Developing Watershed Plans to Restore and
Protect Our Waters (USEPA 2008) to help
Resources:
Recent EPA Watershed-based Initiatives
OWOW's Nonpoint Source Control Branch recently
released the "Watershed Handbook" to support
watershed managers in developing comprehensive
watershed-based plans to restore and maintain water
quality standards. The handbook focuses on
developing plans that meet 319 requirements and
focus on impaired and threatened waters.
http://www.epa.qov/owow/nps/watershed handbook/
OWM's Water Permits Division has developed a
number of policy memos, programmatic and technical
guidance, and case studies to further encourage the
practice of watershed-based permitting. The
ultimate goal of this effort is to develop and issue
NPDES permits that better protect entire watersheds.
http://cfpub.epa.qov/npdes/wqbasedpermittinq/wsper
mittinq.cfm
OWM's Water Permits Division recently released a
toolkit to support permit writers' use of water quality
trading in their watersheds. The toolkit builds on
earlier policy and guidance that highlights the
innovative approach that allows sources within a
watershed to "trade" pollution reductions with other
sources that can achieve the reductions at a lower
cost. The approach encourages achieving water
quality goals in the most cost effective and least
burdensome manner.
http://www.epa.qov/owow/watershed/tradinq.htm
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
communities, watershed organizations, and local, state, tribal, and federal environmental
agencies develop and implement watershed plans to meet water quality standards and
protect water resources. The handbook is designed to aid in the development of a
watershed plan that meets the minimum elements of a watershed plan as outlined in
EPA's Supplemental Guidelines for the Award of Section 319 Nonpoint Source Grants to
States and Territories in FY 2003. These guidelines called for an expansion of efforts to
manage nonpoint pollution on a watershed basis through the development and
implementation of watershed plans, with special emphasis on restoring impaired waters
on a watershed basis.
Building on their National Water Quality Trading Policy issued in January 2003, OWM
released the Water Quality Trading Assessment Handbook (USEPA 2004) in November
2004 to help water quality managers and watershed stakeholders determine if trading is a
potential tool for use in their watershed to make cost-effective pollutant reductions that
achieve water quality standards. The handbook helps users decide whether water quality
trading will work in their watershed and when and where trading is likely to be the
appropriate tool for achieving water quality goals. A framework for water quality trading
is best built on a watershed approach to both permitting and TMDL development. In
August 2007, OWM released The Water Quality Trading Toolkit for Permit Writers
(USEPA 2007b) as the next step in EPA's support for trading. The Toolkit provides
NPDES permitting authorities with the tools they need to incorporate trading provisions
into permits. The Toolkit is the first "how-to" manual for designing and implementing
trading programs consistent with EPA's 2003 National Water Quality Trading Policy.
As those documents have served their respective programs, this guidance document is
intended to serve as a programmatic and technical resource for TMDL practitioners to
develop watershed TMDLs, providing both environmental and programmatic benefits
over single-segment TMDLs. Developing TMDLs on a watershed basis will also be
consistent with and support the efforts of the other watershed-based initiatives in related
water programs, such as those in permitting, monitoring and nonpoint source control. The
programs overlap and can serve to support and enhance one another with the goal of
restoring and protecting water quality watershed by watershed.
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Program Notes:
Setting the Stage for Watershed TMDLs through Statewide Watershed Management
Can you organize your operations to do watershed-based TMDLs? Many states are doing just that through their
statewide watershed management programs which coordinate efforts of their monitoring, assessment, NPDES
permitting, habitat restoration, pollution prevention, TMDL development, nonpoint source grant allocation, and
other resource management programs. States have found that benefits associated with watershed-based
TMDLs—e.g. greater effectiveness and efficiency- can accrue at a higher yield when implementing a TMDL
program within a statewide watershed management framework.
If your state does not have a statewide watershed management framework, you can still begin developing
TMDLs on a watershed basis, scheduling the TMDLs during your priority ranking and TMDL development
scheduling. In addition, TMDL development can become a motivating factor for creating a statewide watershed
management framework. EPA and its partners have written numerous documents on how to develop a
statewide approach to watershed management (e.g., Watershed Protection: A Statewide Approach, USEPA
1995; Framework for a Watershed Management Program, Water Environment Research Foundation, 1996;
Statewide Watershed Management Facilitation, USEPA, 1997). Based on states that have successfully
organized statewide watershed management programs, there are four stages for developing a watershed
management framework (WERF, 1996):
Stage 1: Organizing Statewide Framework Development
Stage 2: Tailoring Statewide Framework Elements
Stage 3: Making the Transition
Stage 4: Operating Under a Statewide Approach
Stage 1. Organizing Statewide Framework Development
This stage involves establishing leadership and vision for developing a statewide framework. A champion is
needed to recruit partners from other programs and agencies, help achieve consensus on common goals and
objectives, as well as establish ground rules and a workplan for developing the framework. This "champion" is
typically from a state-level water quality program. Some states start small, with a framework built off the state's
NPDES permitting, water quality assessment, TMDL, and nonpoint source programs, while others reach out to
more than 20 partners representing federal, state, and local agencies. To ensure the statewide framework
addresses TMDL issues, a representative from the TMDL program should participate in framework development
and serve as a champion of the process.
Stage 2. Tailoring Statewide Framework Elements
This stage is essential to ensure that the watershed framework supports the daily duties and leverages the
resources of the programs being coordinated. Beginning with the end in mind, the first step is to decide the
purpose of the framework—which programs will be coordinated within the framework and what are the resulting
products (e.g., watershed plans, watershed TMDLs, rotating basin monitoring schedules). Next, and very key, is
delineating watershed management units and subunits—these become the focus of monitoring, assessment,
management and reporting. Next establish a watershed management cycle, and an associated gantt chart that
shows how program activities can be synchronized.
Stage 3: Making the Transition
This stage involves establishing organizational structures and protocols for operating under a statewide
watershed framework and synchronizing activities within the watershed management cycle. It also is the stage
for addressing resource and technical support needs and staff training.
Stage 4. Operating Under a Statewide Approach
This includes monitoring progress of framework implementation, evaluating effectiveness, and modifying the
framework as needed.
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
10 DRAFT
-------
2. Understanding Watershed TMDLs
Definition:
Watershed: Geographic Boundary vs.
Conceptual Framework
Most TMDL analyses, whether single-segment
or watershed, are defined by a watershed,
evaluating all the sources within the drainage
area of the impaired segment. However, for
watershed TMDLs, watershed goes beyond the
definition of a geographic boundary and
represents also the conceptual scope of the
TMDL—what segments and impairments are
addressed. To be a watershed TMDL, the
analyses would include and establish allocations
for multiple impaired segments and possibly
impairments within a common hydrological
network and drainage area. A watershed TMDL
can vary widely in size and does not have
inherent size minimums or maximums to define
it as a watershed TMDL. The geographic size of
the watershed will be determined on a case-by-
case basis, depending on such things as
source/impairment type, data availability,
available resources, and scheduling priorities.
A watershed TMDL will vary in geographic scale and
technical scope but it will generally involve a network
consisting of multiple hydrologically connected
waterbody segments, whether streams, rivers, lakes,
estuaries or beaches. (Section 5 further discusses
considerations and guidance for defining your
watershed.) Within that watershed, a watershed TMDL
will holistically consider multiple sources and
impairments, evaluating the watershed as a whole rather
than treating each waterbody and its impairment as
separate analyses.
Generally, the two types of watershed TMDLs are:
• Intrastate Watershed TMDLs. These are the most
common type and represent the watershed TMDLs
developed by a state (or EPA on behalf of the state)
to address multiple impaired segments at varying
geographic scales, but entirely within the
jurisdiction of a single state. Many states currently
develop TMDLs on a watershed basis, at least
occasionally if not for the majority of their TMDLs. U
States might develop watershed TMDLs because of
an established statewide watershed approach to TMDL development, or, states might
still develop single-segment TMDLs the majority of the time and use watershed
TMDLs to address special circumstances.
• Multijurisdictional TMDLs. Multijurisdictional TMDLs represent large-scale
efforts that cross major jurisdictional borders and often encompass the entire drainage
of a major regional waterbody (e.g., Chesapeake Bay, Ohio River, Klamath River).
Whether led by EPA, an individual state or an interstate organization, these TMDLs
typically involve coordinated participation from multiple states and tribes and
sometimes EPA Regions to develop a comprehensive TMDL that addresses impaired
segments throughout the multijurisdictional watershed. These TMDLs are often a
larger geographic scope than the typical watershed TMDL.
Regardless of the type of watershed TMDL, the general approach and many of the
benefits are the same. The overall benefit of watershed TMDLs is they provide the
opportunity to use the TMDL as a tool for cost effectively identifying options for
reducing point and nonpoint source loads to restore impaired waterbodies to water quality
standards. By considering all sources impacting the watershed, watershed TMDLs
provide the state with the greatest level of flexibility in allocating and subsequently
controlling loads. For the watershed TMDL to be most effective, it is important to
integrate all of the scientific, programmatic and social aspects of the TMDL within the
watershed approach. Developing single-segment TMDLs might make it more difficult in
DRAFT
11
-------
Handbook for Developing Watershed TMDLs
December 2008
many cases to efficiently use a program's resources as
well as organize and target projects, engage related
program representatives and maximize stakeholder
participation.
2.1. Environmental Benefits
Many of the environmental benefits of watershed TMDLs
are a result of evaluating and managing the watershed
holistically, allowing for an integrated and
comprehensive analysis of sources, impairments, and
management options. The environmental benefits are also
closely tied to a variety of programmatic benefits
(discussed in Section 2.2). For example, most of the
environmental benefits that result from watershed
TMDLs also result in cost savings when compared to
developing the same TMDLs as single-segment TMDLs.
The following are the primary environmental benefits
resulting from watershed TMDLs.
Definition:
Third-party TMDLs
A third-party TMDL is one developed by an
organization other than the one with TMDL
development responsibility (i.e., state water
quality agency or EPA). A third-party can be a
watershed group volunteering to lead the TMDL
to integrate it into their watershed management
efforts. It can also be another government
agency, whether local municipality, county, city
or other federal entity (U.S. Geological Survey,
U.S. Forest Service). Third-party TMDLs are
emerging as an alternative approach in
watersheds where complex issues (e.g.,
stormwater management, CSO control, interstate
systems) are prevalent. The technical process to
develop third-party TMDLs will not differ from
that for traditional TMDLs, and they are subject
to the same requirements. Therefore, third-party
TMDLs can also use the watershed framework
and can realize many of the same benefits of
traditional watershed TMDLs. Regardless of how
they are developed, third-party TMDLs must be
submitted by states and approved by EPA.
Includes a Broader Source Assessment. Developing
TMDLs on a watershed basis allows the practitioner to
evaluate all the sources in a watershed and identify their
relative impacts on water quality. By focusing only on single-segment TMDLs, a
practitioner might miss a source that is outside the immediate drainage area but
contributes a significant pollutant load that influences the segment's water quality. In
addition, a watershed TMDL can optimize allocations and target those sources that will
most efficiently and effectively result in attainment of standards. When addressing
multiple pollutants that might be contributed by common sources, a watershed TMDL
can target a source that will result in improvements for multiple pollutants, maximizing
the potential for source controls. For example, some agricultural sources (e.g., livestock-
related sources) and residential sources (e.g., septic systems) can be significant sources of
both bacteria and nutrients. By simultaneously developing TMDLs for both pollutants,
the allocations can target those sources that control both bacteria and nutrients. If the
TMDLs were developed separately, unnecessary and duplicative reductions might be
allocated to watershed sources.
Captures the Interaction Between Upstream and
Downstream Sources and Impacts. Closely related to
the benefit of more broadly evaluating sources, this
benefit of watershed TMDLs allows a practitioner to
fully understand the interactions of watershed sources
and their cumulative effects on impaired waterbodies.
Water quality is the result of a variety of load inputs, in-
stream conditions, and physical, chemical and
biological processes. Load inputs can come from
sources in the immediate drainage of a waterbody or
from upstream areas through tributary inflows, both of
which are accounted for in a watershed TMDL.
Alternatively, a watershed TMDL also considers the
impact of reductions in the loads from upstream
sources. Considering the cumulative impact of all
Ptogtom Notes:
Benefits vs. Challenges of Watershed TMDLs
While watershed TMDLs provide the potential for
a number of benefits, they are not without their
challenges. Challenges related to developing
watershed TMDLs and how to address them are
discussed throughout the document in the
relevant sections. For example, challenges in
data analysis are discussed in Section 4.2.1,
Data Analysis for Problem Identification. Some
of the typical challenges include the potential for
increased complexity in the analysis; dealing
with more data, sources and stakeholders; and
the difficulty of evaluating the multiple
impairments at an appropriate scale.
12
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
sources allows TMDL practitioners to optimize and maximize source reductions in the
allocation process. Downstream impaired segments might require less stringent
reductions to meet water quality standards because of reductions required for upstream
impaired segments. Developing single-segment TMDLs might ignore the cumulative
impacts and interactions of watershed sources and processes, potentially resulting in
overly stringent and inappropriate load reductions and subsequent ineffective use of
resources.
Reduces the Potential Need for Future TMDLs. Because the watershed approach to
TMDL development evaluates source inputs and water quality impacts throughout the
entire system, it can protect threatened or unimpaired segments in the watershed. A goal
of the watershed approach is to holistically evaluate sources and target controls to protect
the watershed as a whole, rather than narrowly focusing on portions of the watershed.
This provides water quality benefits throughout the watershed rather than just in the
impaired segments or only for the listed pollutants. This can prevent the need for future
TMDLs, and the associated expenditures, by protecting waters that might have otherwise
Watershed TMDL Development Identifies Main Source Impairing Coosa Chain of Lakes, Alabama
A watershed TMDL was developed to address organic enrichment, low dissolved oxygen and nutrients for four
reservoirs located on the Coosa River in Alabama as well as a number of tributaries to the reservoirs. In
downstream order the lakes are Lake Neely Henry, Logan Martin Lake, Lay Lake and Lake Mitchell. A watershed
model (LSPC) for the Coosa River Basin and a 3-dimensional hydrodynamic (EFDC) and water quality (WASP)
model for each lake in the Coosa Chain of Lakes provided a framework to evaluate multiple sources, pollutants
and waterbodies. During the TMDL scenario runs, it was determined that the primary source of nutrients, namely
phosphorus, was coming from Weiss Lake, located upstream on the Georgia-Alabama border. Therefore, it was
determined that if the phosphorus were reduced from the various point and non-point sources contributing loads
to Weiss Lake that the four downstream reservoirs would be able to meet established nutrient criteria.
The Coosa Chain of Lakes TMDL can be found at:
http://www.adem.state.al.usAA/aterDivisionAA/Quality/TMDLAA/QTMDLInfo.htm.
DRAFT
13
-------
Handbook for Developing Watershed TMDLs December 2008
been listed in future cycles. Developing single-segment TMDLs does not capitalize on
the added benefit of protecting good quality waters while also restoring impaired waters.
This also represents a financial and programmatic benefit by potentially reducing future
workloads associated with additional TMDLs.
2.2. Financial and Programmatic Benefits
The financial and programmatic benefits of developing TMDLs on a watershed basis
rather than as single-segment TMDLs are primarily recognized through decreased total
time spent on developing TMDLs and the resulting cost savings. Advantages also
include a variety of programmatic benefits resulting from coordinated, more efficient
completion of TMDL tasks. The following are the primary financial and programmatic
benefits resulting from watershed TMDLs.
Result in Lower Per-TMDL Development Costs. Because a program's ability to
develop TMDLs is often dictated by the amount of funding and staff resources available,
it is important to consider the potential savings from developing watershed TMDLs.
While single-segment TMDLs might require a lower initial cost to complete the project,
the cumulative savings from developing TMDLs on a watershed basis can be significant.
While a watershed TMDL might take longer to develop than a single-segment TMDL,
the per-TMDL cost and level of effort is typically lower for watershed TMDLs. The
demand for TMDL development is still substantial for many states as they work to meet
TMDL schedules. However, funding for TMDL development is remaining fairly constant
or even decreasing. By using the watershed approach to TMDL development, states can
maximize their funding and staff resources to more efficiently meet their TMDL
commitments. i
| Program Notes:
Address Greater Number of TMDL Pollutant-Waterbody j TMDL Accounting
Combinations. Because the per-TMDL cost is lower when |
developing TMDLs on a watershed basis, they provide the I While a watershed TMDL evaluates multiple
opportunity to address a greater number of pollutant-waterbody I se9ments a"d^P3"5 in ^integrated
^,..J , , -^ , , , i analysis, a TMDL calculation with
combinations at a lower cost than if they were done separately, j associated LAs and WLAs is still needed for
providing an economy of scale to TMDL development. This j each waterbody-pollutant combination
allows practitioners to more expeditiously meet their TMDL ! included on the 303(d) list. Therefore a
development schedules, especially useful when working under a ! "watershed TMDL" can count as multiple
. j . i i i i , I TMDLs toward a state's commitments.
consent decree or court-ordered schedule. I
Encourages Efficient Use of Resources and Completion of Tasks. Developing
TMDLs on a watershed basis maximizes resources and avoids duplication of effort in
completing a number of activities related to TMDL development. Efficiencies are most
easily illustrated in the public participation activities because these activities are
relatively similar across TMDLs. For a watershed that contains six impaired waterbody
segments, the state would traditionally conduct six separate public meetings and address
feedback from six separate comment periods when using single-segment TMDLs. Under
a watershed TMDL approach, the state can conduct one public meeting and have one
public comment period for the comprehensive watershed TMDL. The separate public
involvement activities would not only result in an inefficient use of resources, but would
likely ask similar sets of stakeholders to participate in multiple activities - not an
effective way to engage stakeholders.
14 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Program Notes:
What Are the Potential Cost Savings of Developing Watershed-based TMDLs?
In 2001, EPA completed a draft study to estimate the cost to develop all of the required TMDLs associated with
states' 1998 303(d) lists (USEPA2001a, 2001 b). In doing so, the study also estimated potential cost savings
associated with different approaches to TMDL development and implementation. One of the key findings is that
there can be significant cost savings in TMDL development and implementation when "clustering" TMDLs through
a watershed approach. The following table summarizes the efficiencies estimated for the major TMDL tasks,
recognizing that developing multiple TMDLs within a watershed can result in efficiencies for every step of the
TMDL process (USEPA 2001 b):
Tasks
1. Watershed characterization
2. Modeling and analysis
3. Allocation analysis
4. Develop TMDL document
5. Public outreach
6. Formal public participation
7. Tracking, planning, legal support
8. Implementation plan
% of Unit Effort of Initial Full-cost TMDL
Type A TMDL (no
efficiencies)
100%
100%
100%
100%
100%
100%
100%
100%
Type B TMDL
25%
100%
100%
25%
25%
25%
25%
50%
Type C TMDL
25%
25%
25%
25%
25%
25%
25%
15%
Type A TMDLs = Full cost; address single waterbody and single pollutant
Type B TMDLs = Standard efficiencies; subsequent pollutants in single waterbody or clustered waters
Type C TMDLs = Additional modeling-related efficiencies; multiple clustered waterbodies; one or more related pollutants
Translating these efficiencies to level of effort (LOE) and costs to develop TMDLs, the draft study identified the
following average hour and cost burdens for the different levels of efficiencies:
Type of
Efficiencies
Type A TMDLs
Type B TMDLs
Type C TMDLs
Average Unit Burden to Develop TMDL by Level of Analysis
Level 1
LOE (hours)
933
308
226
Cost (2000 $)
$36,284
$11,978
$8,789
Level 2
LOE (hours)
1,798
740
436
Cost (2000 $)
$69,924
$28,779
$16,956
Level 3
LOE (hours)
3,175
1,459
774
Cost (2000 $)
$123,476
$56,741
$30,101
Type A TMDLs = Full cost; address single waterbody and single pollutant
Type B TMDLs = Standard efficiencies; subsequent pollutants in single waterbody or clustered waters
Type C TMDLs = Additional modeling-related efficiencies; multiple clustered waterbodies; one or more related pollutants
Level 1 analysis = Simplified TMDL analysis using spreadsheet calculations
Level 2 analysis = Mid-range analysis (e.g., QUAL2E, GWLF, BATHTUB)
Level 3 analysis = Detailed analysis (e.g., HSPF, SWMM, WASP)
Similarly, technical analysis tasks such as data analysis and modeling also can experience
more streamlined efforts and avoid duplication when conducted for an entire watershed
rather than a series of single-segment TMDLs. For example, when compiling and
analyzing data for the watershed TMDL, a practitioner will conduct related activities only
once, rather than multiple times—activities such as extracting data from the applicable
databases, compiling water quality data, performing statistical summaries, and extracting
and compiling GIS data. For single-segment TMDLs, the practitioner would often
compile and analyze some (but not all) of the same data for each of the projects. So, for
each project there would be an overlap in data sets that would cumulatively result in
DRAFT
15
-------
Handbook for Developing Watershed TMDLs December 2008
duplication of efforts for data analysis and inefficient use of staff time and resources.
Similarly, in modeling activities there are clear advantages to completing TMDLs on a
watershed basis rather than as single segments. When setting up a model for the
watershed, the modeler would use much of the same data and parameters if setting up a
model for each individual segment, but develop the model once rather than six times.
This can also ensure consistent assumptions when representing the waterbodies and
sources in the analysis.
Facilitates More Effective Use of Public Participation and Stakeholder Involvement.
Developing watershed TMDLs is more likely to engage stakeholders. Similar to the
example above, multiple public meetings held for single-segment TMDLs will likely
generate less participation from stakeholders versus one meeting representing a more
comprehensive watershed TMDL. It is likely that some of the TMDLs would share
common stakeholders, and engaging stakeholders watershed-wide would streamline
participation for stakeholders with an interest in several impaired waterbodies within the
watershed. As a result, this approach has a better likelihood of avoiding stakeholder
participation "burn-out" and of gaining meaningful input. In addition, stakeholders
participating on a watershed-basis might reveal useful monitoring data (e.g., local
jurisdiction, volunteer monitoring, discharger data) that are not available from the state or
public databases to support the TMDL analysis. In addition, coordinating the public
participation activities on a broader watershed level will likely provide benefits when
implementing the TMDL as well. When the stakeholders are involved in the process, and
all of the stakeholders have been introduced to the problem and each other, they are more
likely to actively participate in development and support implementation of the TMDL.
Understanding Sources and Educating Stakeholders in the Beaver Creek and Grand Lake St. Marys Watershed
The Beaver Creek and Grand Lake St. Marys Watershed drains approximately 171 mi2 in west-central Ohio near
the Ohio-Indiana border. Many of the streams in the watershed are impaired due to bacteria and nutrients, as is
Grand Lake St. Marys, an important local recreational resource and drinking water supply. The watershed's
primary land cover is row crops and pasture/hay (more than 80 percent of the total watershed area), and it contains
an estimated 4 million animals at numerous animal feeding operations. Several wastewater treatment plants and
unsewered residential areas in the watershed are additional sources of nutrients and pathogens.
Opinions on major causes and sources varied among the stakeholders, including farmers, lakeshore residents and
recreational users. Development of the watershed TMDL, including modeling the relative magnitude and impact of
watershed sources, helped to reduce uncertainty as to the most significant pollutant sources, allowing the local
stakeholders to focus their attention on solutions. Modeling scenarios were also run to illustrate the potential impact
to Grand Lake St. Marys of implementing the TMDL.
The TMDL public meeting was attended by more than 50 stakeholders including representatives from EPA, Ohio
EPA, Ohio Department of Natural Resources, U.S. Fish and Wildlife Service, The Grand Lake/Wabash Watershed
Alliance, and The Lake Improvement Association and local politicians, numerous local farmers and sportsmen, and
members of the community. Ohio EPA facilitated discussion among these diverse and passionate stakeholders to
communicate the findings of the TMDL and to discuss options for implementation. Current efforts in the watershed
are now focused on implementing projects to reduce nutrient and pathogen loading (e.g., manure dewatering,
septic system replacement) and continuing to monitor water quality in the watershed.
In February 2008, Congressman John Boehner (R-West Chester) and Senator Sherrod Brown (D-Ohio) together
with the NRCS announced that $1 million will be available through the USDA to help improve the water quality of
Ohio watersheds. The bulk of the money will be used for the creation of buffer strips along tributaries to Grand Lake
St. Marys and for the planting of cover crops. The government will provide incentive payments to cover a portion of
the planting costs, with the rest covered by participating farmers, who must meet eligibility requirements established
by the NRCS's Environmental Quality Incentives Program (EQIP).
16 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Avoids Potential for Having to "Redo" TMDLs. Without using a watershed approach,
a practitioner might face the need to reopen and recalculate previously completed
TMDLs. For example, TMDLs completed for smaller headwater streams might contain
allocations that are not stringent enough to support TMDLs for downstream impaired
segments. A practitioner might also have to recalculate upstream TMDLs upon
completion of downstream TMDLs, resulting in unexpected costs as well as potentially
undermining the defensibility of the TMDLs.
Encourages More Comprehensive and Targeted Monitoring Programs. Through an
evaluation of all sources in a watershed and consideration of the water quality throughout
the entire system, the results of a watershed TMDL can support more effective
monitoring program design and implementation. For example, through the watershed
TMDL process, the practitioner can identify data gaps and areas to monitor for better
evaluating source impacts, sites to monitor to gauge progress in achieving load reductions
and water quality improvements, and recommended schedules considering the critical
conditions and potential implementation schedules. In addition, because a watershed
TMDL encourages the participation of multiple stakeholders there is the potential for
leveraging on the resources and involvement of other entities (e.g., local jurisdictions,
watershed groups) to extend monitoring resources.
Provides Opportunity to Integrate TMDL with Other Watershed Programs. Another
benefit of watershed TMDLs is the potential to couple the TMDL with other watershed-
based programs. Integrating watershed TMDLs with other water programs, such as
monitoring, permitting, and nonpoint source control projects (i.e., Section 319 funded
projects), provides the opportunity to more efficiently prioritize and target areas for
monitoring and restoration. Coordinating with programs that rely on similar analyses can
avoid duplicative efforts while also resulting in a more effective plan. For example,
watershed planning projects are beginning to conduct more quantitative analyses of
existing and acceptable levels of pollutant loading, similar to analyses conducted for a
TMDL. Coordinating the related efforts maximizes the cross-program resources to
coordinate and target related, yet often separate, activities. (This concept as related to
implementation of the TMDL is further discussed in Section 5.)
DRAFT 17
-------
Handbook for Developing Watershed TMDLs
December 2008
Evolution of a Successful Watershed-based TMDL Program
West Virginia provides an example of a state TMDL program that embraced the watershed TMDL concept
and integrated it successfully into their standard procedures and related water programs. When faced with
daunting consent decree schedules, West Virginia used the watershed TMDL approach to not only more
efficiently meet their commitments but also to build their TMDL program into an effective program that goes
beyond legal requirements and uses the watershed approach to identify sources, monitor water quality,
develop TMDLs, and issue consistent permits.
West Virginia Department of Environmental Protection (WVDEP) is committed to implementing a
comprehensive watershed based TMDL process that reflects the requirements of the TMDL regulations,
provides for the achievement of water quality standards, and ensures that ample stakeholder participation is
achieved in the development and implementation of TMDLs.
From 1997 through September 2003, EPA Region 3 developed West Virginia TMDLs, underthe settlement
of a 1995 lawsuit, Ohio Valley Environmental Coalition, Inc., West Virginia Highlands et al. v. Browner et al.
The lawsuit resulted in a consent decree between the plaintiffs and EPA that established a rigorous schedule
for TMDL development for the impaired waters on West Virginia's 1996 Section 303(d) list. While EPA was
working on developing TMDLs, WVDEP concentrated on building its own TMDL program. With the help of a
TMDL stakeholder committee, the agency secured funding from the state legislature and created the TMDL
section within the Division of Water and Waste Management.
The TMDL stakeholder committee consisted of 22 members with balanced interests among extractive and
manufacturing industry, environmental advocates, agriculture, forestry, state and federal government,
sportsmen associations, and municipalities. The committee made recommendations for WVDEP TMDL
development and supported general revenue funding.
Since October 2003, West Virginia's TMDLs were and continue to be developed by WVDEP. While
accommodating the remaining TMDLs required by the consent decree, WVDEP's TMDL program generates
numerous other TMDLs under a comprehensive watershed based approach. WVDEP develops TMDLs
according to the Watershed Management Framework cycle. The framework divides the state into five
hydrologic groups (groups A - E) with 32 major watersheds and operates on a 5-year rotation process.
Watershed Groupings
CH A
WEST
VIRGINIA
Watershed Groupings
8-digit(Oinfrontofnun
Hydrologic Unit Code
and Watershed Name
and Groupings
i— ED50005 G
§1 20/1
A 20/1
V ED5I
ED31
L-E021
IB
*-*3
El
LI
18
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Prior to the existence of the TMDL Program, WVDEP stream monitoring and NPDES permit reissuance activities
were organized in accordance with the Framework. The TMDL program was then designed to be synchronized
with the monitoring and implementation schedule of the Framework creating a fully integrated watershed based
program.
The TMDL development process begins with pre-TMDL water quality monitoring and source identification and
characterization. Informational public meetings are held in the affected watersheds. Data obtained from pre-TMDL
efforts are compiled, and the impaired waters are modeled to determine baseline conditions and the gross
pollutant reductions needed to achieve water quality standards. WVDEP then presents its allocation strategies in
a second public meeting, after which Final TMDL reports are developed. The draft TMDL is advertised for public
review and comment, and a third informational meeting is held during the public comment period. Public
comments are addressed, and the draft TMDL is submitted to EPA for approval.
WVDEP's 48-month development process enables the agency to carry out an extensive data generation and
gathering effort to produce scientifically defensible TMDLs. WVDEP strategically plans water quality monitoring
prior to TMDL development where numerous monitoring locations are established and a comprehensive suite of
analytes are sampled. This fine scale monitoring resolution coupled with identification and characterization of
problematic sources through field-based source tracking activities provides a sound basis for assessment and
TMDL development for all streams and impairments within the watershed.
In addition, WVDEP has created unique ways to integrate large-scale, watershed-based TMDLs with fine-scale,
highly technical methodologies that produce "implementable" TMDLs in a cost-effective manner. The
comprehensive watershed based approach typically includes all known impairments in the watershed and
involves a multi-faceted modeling approach to address total recoverable metals, dissolved metals, acidity (pH),
bacteria, and biological impairments. This watershed-based approach allows WVDEP to maximize efficiency
throughout all phases of TMDL development and thereby minimizing funding requirements of their TMDL
program. Since 2003, WVDEP's TMDL program has completed more than 1,300 EPA approved TMDLs (428
streams) with another 675 (250 streams) currently under development.
WVDEP also has designed a "TMDL on CD" concept where all relevant TMDL information (TMDL Reports and
Appendices, Technical documentation, and supporting data) is included on a CD-ROM. To further improve the
usability of the TMDLs, WVDEP developed a series of interactive tools to provide TMDL implementation
guidance. These tools are designed to simplify and assist "implementers" (nonpoint source staff and permit
writers) in using the TMDLs to develop watershed plans and issue/renew permits. An interactive ArcExplorer
geographic information system (GIS) project allows the user to explore the spatial relationships of the source
assessment data, as well as further details related to the data. Users are also able to "zoom in" on streams and
other features of interest. In addition, spreadsheet tools (in Microsoft Excel format) were developed to provide the
data used during the TMDL development process and the detailed source allocations associated with successful
TMDL scenarios. These tools provide guidance for selection of implementation projects as well as for permit
issuance and are also included on the TMDL Project CD.
ArcExplorer GIS Viewer
PS and NPS Allocation
Spreadsheets
DRAFT
19
-------
Handbook for Developing Watershed TMDLs December 2008
2.3. Implementation Benefits
While many of the benefits of watershed TMDLs are realized during the development
stages, there are several additional benefits realized during TMDL implementation.
Many of these benefits also present financial, programmatic and environmental benefits
but are included here to highlight the separate step of implementation, which often
involves its own set of unique issues and considerations compared to TMDL
development. The following are the primary benefits of watershed TMDLs as related to
TMDL implementation.
Provides a Framework for More Effective Implementation. Building on the
organizational and environmental benefits from the TMDL development stage (e.g.,
source assessment, upstream-downstream impacts), watershed TMDLs can more
effectively target source controls and allow for better implementation planning. The
implementation plans developed based on watershed TMDL analyses and allocations are
likely to be more equitable and effective than those developed for single-segment
TMDLs because they consider the relative magnitude and influence of all sources. A
watershed TMDL can also provide the necessary information to meet Section 319
requirements to receive funding for implementation of watershed management projects.
For example, TMDLs quantify source loads and identify the necessary load reductions
needed to meet water quality standards, important elements of a watershed plan required
by Section 319 supplemental guidance. In addition, a watershed TMDL is more likely to
parallel the geographic scope of a watershed management planning effort than is a single-
segment TMDL.
Facilitates Watershed-wide Planning. Applying a watershed approach to TMDLs
allows for easier integration with the overall watershed management approach. As
recently promoted in EPA's Handbook for Developing Watershed-based Plans to Protect
and Restore our Waters (USEPA 2008), TMDLs developed within a watershed
management framework can provide the quantitative link between on-the-ground actions
and the attainment of water quality standards. The TMDLs can serve as a tool to quantify
necessary load reductions and guide nonpoint source controls within the watershed
planning process. Similarly, as discussed below, a watershed TMDL can provide a
framework for implementing other watershed-based source controls, such as watershed-
based permitting and water quality trading.
Promotes "Equal Representation" Among Sources. Because watershed TMDLs
evaluate source, whether nonpoint or point, impacts across an entire watershed, they
evaluate the sources at the same time and with comparable analyses and approaches.
Sources can be represented using the same assumptions and within the same context. This
can make stakeholders representing pollutant sources feel more comfortable that they are
receiving the same treatment and representation as other sources in the watershed. This
process can help to avoid "finger-pointing" at sources that would have otherwise been
excluded from the analysis. It can also avoid the worries of stakeholders concerning new
or additional allocations that would have been developed in the future in TMDLs
developed for nearby waters or for similar pollutants.
Facilitates Use of Innovative Implementation Options. Watershed TMDLs provide an
organizational and quantitative framework for innovative approaches to point source
control such as watershed-based permitting and water quality trading. WLAs in
watershed TMDLs are developed with consideration of the cumulative effects of all the
20 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
nonpoint and point sources in the watershed; this facilitates the effective use of
watershed-based permitting and water quality trading. Because a watershed permitting
approach needs to consider all sources in the watershed when determining permit
provisions, an existing TMDL can save a great deal of time and effort if it has already
considered all of the sources to identify necessary WLAs. Similarly, developing TMDLs
on a watershed basis maximizes opportunities for implementing WLAs through water
quality trading programs. The larger geographic area for which WLAs are established in
a watershed TMDL combined with the more extensive pollutant fate and transport
analyses provide some of the needed information and drivers for water quality trading
programs to be successful. Water quality trading programs allow for those point sources
with higher pollutant control costs to achieve their WLA in a more cost effective manner
by paying another pollutant source, whose control costs are lower, to make pollutant
reductions needed to meet the WLA or the new water quality-based effluent limitation
(WQBEL) that is derived from the WLA.
More Easily Addresses Non-traditional Point Sources. A watershed TMDL provides
a tool for more effectively integrating and addressing non-traditional point sources, such
as CAFOs and Phase II stormwater communities, in the allocation process. Non-
traditional point sources are those sources that are regulated and subject to NPDES
permits yet behave like a nonpoint source, primarily delivering land-based loads in
response to precipitation runoff events. These sources present a challenge to TMDL
practitioners and regulators because of the difficulty of identifying appropriate WLAs for
sources that do not have discrete discharge points or operate under a controlled set of
conditions as do traditional permitted facilities. The watershed TMDL provides an
opportunity to evaluate these sources in the context of the larger watershed and to most
appropriately account for their behavior and response. Particularly with permitted
stormwater, a watershed TMDL can better evaluate the stormwater inputs and influence
on in-stream water quality. Because the watershed TMDL considers the entire watershed,
likely evaluating various subwatershed and land use types, it can provide flexibility in
how to express stormwater WLAs (e.g., for entire MS4 area, by subarea, by impervious
land uses) and construction stormwater future growth provisions.
DRAFT 21
-------
Handbook for Developing Watershed TMDLs December 2008
22 DRAFT
-------
3. Identifying Candidates for
Watershed TMDL Development
The first step in developing watershed TMDLs is _ _,
identifying and evaluating the opportunities for and ^'
defining the scope of the TMDLs. Ideally, a state can What is the Scope of a Watershed TMDL?
identify those waters and impairments that are appropriate
., . , „, ,^.T ., . . /I- This section provides a series of criteria to
for a watershed TMDL framework as they are developing eva|uate 303Jd) |istings jn a defined
their TMDL schedules based on current 303(d) lists. While geographic area (e.g., statewide, specific
it would be ideal for all state TMDL programs to operate basin) to narrow the larger area into multiple
on a watershed framework, that is not necessarily the "bundles" of waterbodies and impairments to
reality AssuMng tha, the majority of a^ Us, can be J^^S^JKS^M,
addressed through the use of watershed TMDLs, one of the watershed TMDLs—both the geographic
goals of this document is to encourage states to coordinate extent of the watershed and also the
their TMDL development on a watershed basis to more individual pollutant-waterbody combinations
efficiently and effectively tackle their TMDL development that wi" be addressed in the analysis.
commitments. Planning for watershed TMDLs should be
approached with the idea that most waterbody-impairment combinations can be
addressed as part of a larger watershed TMDL; however, there will be some instances
that are more appropriate as single-segment TMDLs.
The decision to develop a watershed TMDL will be guided by multiple factors that will
usually serve to determine the actual scope of the TMDL—how large is the geographic
area and what range of pollutants and sources should be addressed? If a waterbody
segment is impaired by one pollutant originating from a unique, localized source, a
watershed TMDL might not be necessary because the impairment and source activity
might not impact other segments within the larger watershed. However, many
impairments and impaired segments within a watershed can be "bundled" and addressed
through a broader watershed TMDL. There is no set size for the geographic area of a
watershed TMDL and no rules on the number of waterbodies or impairments addressed.
Watershed TMDLs have addressed areas ranging in size from a few square miles to
thousands of square miles and scopes of two impaired segments for the same pollutant to
thousands of waterbody-pollutant combinations. Ultimately, the scope of the TMDL
depends on a variety of project- and watershed-specific factors that must be considered in
the planning stages. This section addresses the various factors to consider in determining
whether a watershed TMDL is appropriate and the scope of the TMDL or whether a
single-segment analysis is best.
Figure 3-1 illustrates the typical screening criteria for identifying candidate groupings for,
and ultimately defining the scope of, watershed TMDLs. Starting with the 303(d) listings
within the area of concern, whether statewide or for a specific region or basin, evaluating
these criteria should help identify candidate groupings or "bundles" of waterbodies and
related impairments for watershed TMDL development. This process should also identify
those listings that are appropriate for single-segment TMDL development.
DRAFT 23
-------
Handbook for Developing Watershed TMDLs
December 2008
As shown in Figure 3-1, the first, and most important, screening criterion, is related to the
listing—the impairment, waterbody type, and sources. The next screening level evaluates
program commitments such as consent decrees and available resources to further narrow
the scope or prioritize the watershed candidates. The final screening is based on
evaluating the existence of ongoing watershed-based efforts that can define or guide the
scope of a watershed TMDL.
Screening Criterion 1:
Identify Opportunities based on
Impairments and Related Issues
Type/location/similarity of
impairments
Waterbody type
Expected sources and
pathways
Priority waters
Necessary level of detail or
preferred approach, if known
Screening Criterion 2:
Account for Other Programmatic
Commitments and Priorities
Consent decrees
Priority rankings
Public concerns
Available resources
Screening Criterion 3:
Identify Potential for Leveraging
Existing Watershed-based
grams or Efforts
State planning boundaries
USGS boundaries
Watersheds from related
programs (e.g., 319)
Active stakeholder groups
Identify Po
on Existii
Proc
Figure 3-1. Process for identifying the scope of a watershed TMDL.
24
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
While these screening criteria are presented in a
priority order for evaluation, they are not steps that are
dependent on one another. Practitioners can
independently evaluate each criterion and, in some
areas, not all of the considerations will be relevant or
necessary for identifying groupings for watershed
TMDLs. For example, the first criterion is applicable
to any exercise to identify candidates for watershed
TMDLs, using information contained on a state's
303(d) list. Alternatively, the final screening criterion
based on existing watershed efforts might not be
relevant to many watersheds. Either there are no
existing watershed efforts with which to coordinate or
information on their existence is not readily available
Process 7ip:
Where to Start in Identifying Watersheds for
TMDL Development
Even if a state doesn't currently schedule TMDLs
as watershed groupings, impaired segments are
typically identified and organized on a watershed
basis. State 303(d) lists might organize the listed
segments according to 8-digit hydrologic unit,
even if not by smaller subbasins. These basins
are a logical place to start in identifying the
watershed groupings for TMDLs. By looking at the
listings within an 8-digit basin, it will be easier to
identify the logical combinations of segments and
impairments for smaller watersheds.
or known to the state. However, a few quick inquiries
to representatives from related programs (e.g., Section 319 nonpoint source program,
permitting, monitoring) can hopefully identify any efforts worth considering in
identifying watershed TMDL candidates. All of the screening criteria discussed in this
section serve to work in concert to collectively narrow the starting scale (e.g., state,
basin) down to several manageable and logical groupings for watershed TMDL
development, as well as identify those listings to address in single-segment TMDLs.
3.1. Identifying Opportunities based on Impairment and Related
Issues
Process Tip:
Impairment-related Issues to Consider
when Grouping Watershed TMDLS
• Impairment type or pollutant
• Source dynamics (pollutants delivered,
delivery mechanisms, characteristics)
• Source type (point, nonpoint)
• Waterbody type and characteristics
• Critical conditions
• Required level of detail or preferred
approach, if known
The first screening criterion for identifying opportunities for
developing watershed TMDLs is based on evaluation of the
303(d) listings, including the impairment type, waterbody
type, and the associated sources, as well as what implications
these factors might have on the TMDL approach. Of these,
the impairment and waterbody type are readily available from
all 303(d) lists. They provide the most basic of filters in
identifying potential groupings for watershed TMDL
development. Source type is helpful for further grouping
waters and impairments; however, expected sources are not
included on all 303(d) lists and also are not always known.
The final factor in this category is considerations related to
the eventual TMDL technical approach. This can include
critical conditions or a preferred approach. For example,
some programs have standard TMDL development
approaches or tools. Using this as a screening tool for grouping watershed TMDLs can
identify those waterbody-impairment combinations for which that approach is
appropriate. As with all of the screening criteria discussed in this section, all of these
factors might not be relevant or even known for the area of concern.
3.1.1. Impairment and Source Type
Impairment(s) or the source(s) of the impairment can determine the groupings addressed
through a watershed TMDL. Bundling together multiple segments with the same or
related impairments for TMDL analysis can create economies of scale in data collection,
technical analysis, and report development. Impairments due to "eutrophication" or
"nutrients," for example, often require analysis of multiple processes and parameters
DRAFT
25
-------
Handbook for Developing Watershed TMDLs
December 2008
including dissolved oxygen levels, nutrient cycling, pH, and even sedimentation. For
watersheds with multiple segments listed for the same impairments, it is more efficient to
evaluate the interrelated processes and sources on the watershed scale to most efficiently
develop allocations. Grouping impairments that are "similar"—meaning those that are
likely to originate from common sources, are delivered through common pathways, or
impair the waterbody under similar conditions—provides the same benefits.
Similarly, the expected sources or delivery pathways of the impairing pollutants can help
to define the scope of a watershed TMDL. One of the greatest benefits of a watershed
TMDL is the opportunity to evaluate multiple sources impacting a watershed, thereby
evaluating their relative magnitude and maximizing the load reductions to get the most
water quality improvement with the smallest, but targeted, load controls. This can impact
multiple sources and multiple impairments. A watershed might have a number of sources
contributing to multiple impairments. While some of those sources might exclusively
contribute to single impairments, there is likely a subset of sources that generate a variety
of pollutants and contribute to more than one impairment. Considering the sources
occurring in a watershed, whether known or expected based on the impairments and
general knowledge of the area, can help to identify combinations of waterbodies and
impairments amenable to watershed TMDL development.
Table 3-1 provides the most commonly listed impairments and some of their typically
associated sources. This illustrates the potential overlap in sources and related
impairments. For example, agricultural areas are typically sources of pollutants such as
nutrients, sediment, pesticides, and bacteria; these areas are not usually associated with
metals or toxic pollutants. In addition, sediment and nutrients are often associated with
shared sources because nutrients can be sorbed to sediments and delivered to receiving
waters through erosion and runoff processes. While this type of information can provide a
general understanding of the common impairments and their associated sources,
practitioners should base candidates for watershed TMDL development on site-specific
knowledge of expected sources of the pollutants of concern whenever possible.
Table 3-1. Summary of Common Pollutants and Sources
Pollutant
Potential Point Sources
Potential Nonpoint Sources
Pathogens
' WW IPs
' CSOs/SSOs
1 Permitted CAFOs
1 Discharges from meat processing
facilities
' Lan dfills
Animals (domestic, wildlife, livestock)
Malfuncti oning septic systems
Pastures
Boat pumpout facilities
Land application of manure
Land application of wastewater
Metals
1 Urban runoff/permitted
stormwater
' WW IPs
' CSO/SSOs
1 Lan dfills
1 Industrie I facilities
1 Mine discharges
Abandoned mine drainage
Hazardous waste sites (unknown or partially treated
sources)
Marinas
Atmospher ic deposition
Nutrients
' WW IPs
' CSOs/SSOs
'CAFOs
1 Discharge from food-processing
facilities
1 Lan dfills
1 Cropland (fertilizer application)
1 Landscaped spaces in developed areas (e.g., lawns, golf
courses)
1 Animals (domestic, wildlife, livestock)
1 Malfuncti oning septic systems
1 Pastures
1 Boat pumpout
1 Land application of manure or wastewater
1 Atmospher ic deposition
26
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Pollutant
Sediment
Temperature
Potential Point Sources
• WW IPs
• Urban stormwater systems
• Construct! on
• CAFOs
• WW IPs
• Cool ing water discharges (power
plants and other industrial
sources)
• Urban stormwater systems
Potential Nonpoint Sources
Agriculture (cropland and pastureland erosion)
Silviculture and timber harvesting
Livestock (rangeland erosion, streambank erosion)
Excessive streambank erosion
Construct! on
Roa ds
Urba n runoff
Lan dslides
Abandoned mine drainage
Stream channel modification
Lack of riparian shading
Shallow or wide channels (due to hydrologic modification)
Hy droelectric dams
Urban runoff (warmer runoff from impervious surfaces)
Sediment (cloudy water absorbs more heat than clear water)
Abandoned mine drainage
The types of sources (point source, nonpoint source or mixed) can also be a factor in
evaluating how to define the scope of a watershed TMDL. The types of sources can
influence what approaches are feasible and how much detail will be necessary for the
TMDL analyses. For example, an analysis accounting for both point and nonpoint
sources might require more data and a more detailed analysis to evaluate the varying
critical conditions and types of loading. However, a watershed scale analysis for areas
with mixed sources (point and nonpoint) helps to ensure an adequate assessment of the
problems and maximize the potential for controls. It is necessary to consider the benefits
of evaluating multiple types of sources versus the need for a more rigorous analysis and
therefore a likely increase in time and resources needed for the analysis. In addition, the
coverage or extent of some sources might guide the definition of a watershed for TMDL
development. For example, the regulated areas of Phase I or Phase II MS4s might fall
within targeted watersheds and addressing associated TMDLs within the larger watershed
context can facilitate better representation and inclusion of the MS4 loadings.
3.1.2. Waterbody Type
Evaluating multiple types of waterbodies (e.g., lakes and rivers) within a watershed
TMDL sometimes adds a layer of complexity depending on how detailed an analysis is
required. In the context of a watershed TMDL, the goal is to include upstream tributaries
contributing loads to the multiple waterbodies of concern. Although the analysis might
take all waterbodies into consideration, the resulting allocations might solely focus on
one type of waterbody. For example, when dealing with a large lake that exhibits
impairments impacted by internal lake processes and conditions in addition to the amount
and timing of incoming watershed loads, a detailed analysis of the lake is typically used
to fully understand the lake dynamics and capture spatial differences. The lake-specific
analysis coupled with a comparably detailed watershed analysis to evaluate the loads
originating throughout the watershed can provide the ability to also develop TMDLs for
impaired tributaries. However, the analysis can focus only on identifying the TMDL for
the lake with the watershed loads more simply represented in the analysis as a cumulative
load entering the lake. The location and interaction of impaired tributaries in a lake's
watershed can influence whether the lake's TMDL is developed as a single-segment or
watershed TMDL. Similar situations can exist when dealing with combinations of other
types of waterbodies, such as rivers, controlled impoundments, and estuaries.
DRAFT
27
-------
Handbook for Developing Watershed TMDLs _ December 2008
3.1.3. Technical Approach Considerations
Although not always known prior to TMDL development, the preferred technical
approach or the level of detail required for a TMDL might affect how practitioners
choose to group waterbodies and impairments for watershed TMDL development. There
are instances where it is generally understood how much detail or the types or approaches
that are appropriate given the waterbodies of concern and the associated impairments and
sources. For example, in a watershed where sources are fairly well understood, a more
diagnostic approach that evaluates source- or site-level impacts is not likely necessary. In
this case, it might be feasible to develop TMDLs for a larger watershed area, using a
simpler approach and encompassing more segments given the available resources and
time. Alternatively, a complex system that has a number of sources that are not well-
defined might require a more detailed analysis involving additional monitoring and field
surveys and comprehensive modeling to understand the conditions and identify necessary
allocations. In this case, it would likely be more effective to focus the watershed at a
smaller scale around that system to allow for detailed analyses with the available
resources.
In addition, states sometimes have preferred approaches or tools that they apply for
TMDL development in an effort to promote consistency among projects and manage the
resources and effort required to develop TMDLs. In this case, practitioners can consider
the approach when identifying candidate groupings for watershed TMDL development. It
would be necessary to evaluate the listings and identify those that are appropriate for the
approach and evaluate whether grouping them together for a larger analysis makes sense.
Grouping Waters and Impairments for TMDL Development to Maximize
Use of a Proven, Consistent Approach in Southern California
EPA and California worked together to identify listings that could be grouped together to use a uniform approach
for calculating TMDLs. EPA and the state assessed listings in the region and identified many of the coastal
lagoons that are impaired by nutrients, bacteria, and sedimentation as candidates. Although the lagoons are not a
traditional watershed in the sense that they are hydrologically connected, grouping them utilizes the same
principals of a watershed TMDL — grouping similar listings and waterbodies to more effectively develop TMDLs.
Using a consistent and proven approach for multiple listings and watersheds has resulted in a significant
reduction in the costs to develop the TMDLs and buy-in from stakeholders. Facilitated by the use of a consistent
approach across the region, stakeholders have led an effort to develop a comprehensive regional monitoring
strategy to both inform the technical analyses and support tracking progress in meeting water quality standards.
3.2. Accounting for Programmatic Commitments and Priorities
The second screening criterion for determining the scope of watershed TMDLs is
programmatic commitments and priorities, as shown previously in Figure 3-1. Most
TMDL programs have existing commitments or priorities that can affect TMDL
grouping, prioritization, and development. These commitments and priorities can include
existing consent decrees that establish TMDL schedules, priority rankings identified in
303(d) lists or TMDL schedules, public concerns or priorities and the availability of
program resources to support TMDL development.
28 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
3.2.1. Consent Decrees and Settlement Agreements
Program priorities set by consent decrees or settlement agreements can often drive the
schedule of TMDL development. Depending on the type of priorities influencing the
program, they can either facilitate or hinder development of TMDLs on a watershed
basis. Many consent decrees establish TMDL schedules or commitments for states, but
do not specifically identify the waters that the state must address in the timeframe. In this
situation, the states can establish their own schedules and target specific waters as long as
they meet the court-ordered deadlines and commitments for developing a minimum
number of TMDLs. Alternatively, some states do have specific waters that the TMDL
program must address in a set timeframe, forcing the state to prioritize those waters over
others. Because those waters might not relate in terms of pollutants or sources, or share
the same geographic boundaries, practitioners might find it difficult to plan for a
watershed TMDL when operating under this type of court-mandated schedule.
For states that can establish their own TMDL schedules, using watershed TMDLs might
help to more efficiently meet TMDL development commitments. When dealing with
commitments for specific waters that are not identified based on geographic region,
impairment type or other connecting factor this might be more difficult. However,
specific waters identified in consent decrees can still be grouped with other waters (not
included in the consent decree) as part of a watershed TMDL. States can integrate the
required TMDLs into broader TMDL projects to meet their commitments while
implementing a watershed approach and more holistically and efficiently addressing their
water quality impairments.
3.2.2. Priority Rankings
States identify the priority rankings for waters included j Process Tip: I
on their 303 (d) lists. Similar to having a consent decree i . ... . „ . . ... i
.,, T,™ , , .,, , . . .. i Low vs. High Priority Waters i
specify TMDL development for certain waters, priority j a j I
rankings might dictate development schedules that j Because waters have different priority rankings on |
hamper addressing the prioritized waters as part of a j a state's 303(d) list does not mean they cannot be j
comprehensive watershed TMDL. Watersheds that ! addressed at the same time as part of a |
contain waters with varying priority levels might lead I watershed TMDL. If waters of varying priority are j
, _ . ,,,";„ . . , ., i hydrologically connected, experience similar i
the state to focus on the high priority waters before | impairments, and have similar sources, a state !
addressing the "low" priority waters in the same j should consider grouping the waters for a |
watershed - resulting in a missed opportunity for a | watershed TMDL, despite the varying priorities. j
comprehensive, integrated watershed TMDL project. If i i
the irrmaireH waters in a watershed have varied l This aPProach mi9ht add waters to their current i
the impaired waters in a watershed have varied , work,oad since they wiN be addressed sooner I
rankings, practitioners might consider using the first j than they are scheduled. However, in the long |
screening criterion to evaluate whether these waters | run, developing the watershed TMDL will provide |
have other shared factors that are appropriate for a i cost savings over addressing the waters j
watershed TMDL process (e.g., have similar ! individually and will result in a more scientifically |
, ^ . \° ' , , i defensible and implementable TMDL. i
impairments and sources). It so, the state could | |
proceed with grouping the waters as part of a
watershed TMDL, regardless of the "low" rankings of some of the segments. If the lower
rankings are appropriate to address separately from the high-priority waters (e.g., unique
impairment, localized sources, not impacted by upstream sources), it makes sense for the
state to continue with scheduling the TMDLs separately according to their original
priority rankings.
DRAFT 29
-------
Handbook for Developing Watershed TMDLs December 2008
3.2.3. Availability of Resources
While technical considerations can determine the scope of a TMDL, practical
considerations such as available budget and schedule can also be important factors in
deciding how to approach a watershed TMDL. Whether single-segment or watershed, the
cost and time associated with developing TMDLs can vary widely. Therefore, it is often
necessary to consider the expected resources needed to complete a TMDL when deciding
on its scope. The available timeframe and budget for developing TMDLs can be
considered along with the expected complexity of the analysis to find the most beneficial
scale and scope—one at which the analysis can be completed within the available
resources while still meeting the project goals in an appropriate and scientifically
defensible manner.
The perception of the funding or technical resources necessary to develop watershed
TMDLs might discourage practitioners and serve as an obstacle to initiating a watershed
TMDL process. However, as discussed in Section 3, a watershed TMDL can result in
cost savings to a TMDL program by more efficiently using the available resources and
reducing the per-TMDL costs. Due to the potential time-savings associated with a
watershed TMDL process, the project might have a shorter lifespan with a larger upfront
budget—a budget that will likely be less than the combined budget of the corresponding
single-segment TMDLs. Developing TMDLs on a watershed basis ultimately saves
money and maximizes resources. They also provide the state with maximum flexibility in
setting allocations, thereby providing more opportunities for optimizing the cost of
implementation.
The availability or experience of staff to complete TMDLs can also influence the
decisions regarding scope of a TMDL. Watershed TMDLs might require more complex
approaches (e.g., dynamic modeling) and can present difficulties for inexperienced
practitioners. However, most TMDL practitioners have access to technical support staff,
technical training or qualified consultants to support more detailed analyses, if necessary.
In addition, having limited technical expertise or resources should not preclude
practitioners from choosing to develop watershed TMDLs. Watershed TMDLs can
accommodate a variety of technical approaches and still accomplish many of the benefits
of the most detailed, complex watershed TMDLs. For example, a watershed TMDL
might use a load duration or other statistically based TMDL approach and apply that
approach at various locations throughout the watershed to address impaired segments.
Although this approach does not quantitatively account for upstream-to-downstream
effects and the cumulative impact of watershed sources, it still evaluates the watershed as
a whole. It allows the practitioner to identify major sources throughout the watershed,
evaluate spatial variations in water quality and source activity, and target controls more
efficiently.
Planning for TMDLs should consider the expected level of resources and time needed to
address a particular waterbody or impairment when deciding how to group watershed
TMDLs. Waterbodies with impairments and sources that are fairly well understood will
likely have more straightforward approaches and predictable resource needs. However,
some TMDLs might appear to require more analyses, resources and time to adequately
address, such as impairments that are impacted by a number of varying pollutants and
sources; loads contributed by multiple permitted and unpermitted sources; and complex
physical, biological and chemical dynamics and processes. Anticipated costs and
complex technical considerations might indicate the need to consider a more focused
single-segment TMDL rather than a watershed TMDL.
30 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Example Analysis to Identify Groupings for Watershed TMDL Development
This example identifies potential watershed TMDL groupings based on listed impairments and related issues (i.e.,
Screening Criterion 1 in Figure 3-1). Because most of this information is readily available on a state's 303(d) list
or the supporting assessment documentation, it will be the first and most commonly applied criterion for
identifying groupings. The remaining criteria (i.e., programmatic commitments, existing watershed-based efforts)
are likely to be more site-specific in their availability and usefulness. They should be considered when identifying
groupings, when available and relevant, but are not explicitly illustrated in this example. The goal of this grouping
exercise is to review available listing and assessment information for similarities in type/location/behavior of
impairments, waterbody type, expected sources and pathways, priority waters, level of detail/approach.
The evaluation for identifying watershed TMDL groupings can be conducted at any geographic scale, whether
statewide or within a specific watershed of concern. If dealing with the listings for an entire state it will be useful to
approach the listings for smaller divisions. For example, most state lists are organized according to 8-digit
subbasins or other state planning watersheds. This can be a logical scale at which to begin. In some states,
watershed TMDLs are developed at the 8-digit watershed scale. However, available resources, schedule and
priorities cannot always accommodate that approach. For this example, listings are evaluated for an 8-digit
watershed, highlighted in the following figure, to identify smaller groupings appropriate for watershed TMDLs, as
well as those impairments that are better addressed using an individual TMDL analysis.
/\/ Impaired segments
i"^ 6-digit basin
| | 8-digitsubbasin
The following figure presents the hypothetical 303(d) listings for the fictional 8-digit Oak River subbasin. The
impairments listed in the table are presented on a map of the subbasin to illustrate the distribution and location of
similar impairments. As shown in the table and map, there are multiple impaired waters in the subbasin, many
with multiple impairments. The 303(d) list also identifies expected sources of the impairments, including
agriculture and urban runoff. These sources have been labeled on the map to illustrate the distribution of sources
in relation to impaired segments. Also included on the map are the boundaries of the state's existing planning
level subwatersheds (12-digit) that can help to identify the associated watersheds of the identified groupings.
DRAFT
31
-------
Handbook for Developing Watershed TMDLs
December 2008
Example Analysis to Identify Groupings for Watershed TMDL Development (continued)
A/Nutrents
'Kt Fecal cdifcrm
Ben thic impairment
nature
Other (metals; PCBs, sediment, turbidity, DO;
8-digit subbasin
12-dgtsubwatershed
As shown in the above table and map, there are a number of waterbodies that share common impairments and/or
sources. Visually evaluating the location and connection of these impairments and sources can help in identifying
potential groupings for watershed TMDL development. For example, many of the segments are listed as impaired
by fecal coliform and nutrients from either agriculture or urban runoff. The shared pollutants can provide a basis
for grouping, but different sources can sometimes require different analyses to account for varying delivery
pathways and source behavior. Many of the segments in the middle of the watershed (e.g., Oak River, Meadow
Creek, Dobbs Run) seem to be impacted by urban runoff while those in the northern part of the watershed (Turtle
Creek) are impaired by agriculture. However, the watershed seems to transition from agriculture uses to urban
uses as evidenced by the most upstream segment of Oak River, which is identified as impaired by both
agricultural and urban runoff. This apparent transitioning of land uses provides some guidance for how to group
many of these segments for analysis to evaluate the separate and also cumulative effect of the different sources
on the connected segments. Within this grouping Turtle Creek also is impaired for low dissolved oxygen and
elevated turbidity. While these listings are not the same as the other similar fecal coliform, nutrients and benthic
impairments, they will likely involve many of the same analyses. Dissolved oxygen is likely linked to the nutrient
impairments and the turbidity impairment can be caused by many of the same sources (e.g., runoff from
agriculture or algal growth from excessive nutrients).
Alternatively, there are some impairments or waters that can merit smaller groupings/watersheds or even
individual analysis either because of the nature of their impairment or the location of the waterbodies. For
example, within the Oak River grouping, there are some impairments listed for temperature and PCBs. The
temperature listings might share a source (urban runoff) with the nutrient and fecal coliform impairments, but the
dynamics and behavior of the source and resulting impairment are often different. Because of the unique nature
of temperature impacts and impairments, the data analysis, source assessment, TMDL calculation approach and
even the identification and expression of allocations will likely be very different from those conducted for other
pollutants. For this reason, it would be appropriate to address the temperature listings separately from other
impairments; however the temperature-impaired segments can be grouped together for analysis to evaluate the
common sources and causes, more efficiently analyze the data and apply a common technical approach. Also
within this grouping, PCBs are listed as an impairment for only one segment in the subbasin. PCBs are typically
related to legacy sources and might be the result of highly localized issues. These issues and other unique
characteristics (environmental persistence, unknown sources) provide an ideal justification for addressing the
PCS impairment as an individual TMDL.
32
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Example Analysis to Identify Groupings for Watershed TMDL Development (continued)
In addition, while Dry Run, Daniels Branch and Jacks Branch experience similar impairments as the Oak River
grouping, they are not hydrologically connected to those impaired segments, and any water quality improvement
or impairment would not have an impact on the Oak River segments and vice versa. Therefore, it makes sense to
group the three segments in the eastern portion (Jacks Branch, etc.) together for analysis. Their impairments of
nutrients and benthic impacts can feasibly be analyzed together because of likely common sources (although
unknown at listing) as well as shared data analyses. Benthic impairments are usually linked to sources of
excessive sediment or nutrients, and the analysis to identify the stressors often involves evaluation of much of the
same data as would be analyzed for a nutrient impairment. In addition, sources targeted for control of nutrients
loads would often be the same as those impacting the benthic community.
Island Creek also represents isolated segments that are not hydrologically impacted by other impaired segments.
In addition, they are listed for metals, an impairment that is not associated with any other segment in the
subbasin. The list indicates the impairment is due to historical mining activities. Because of the location and the
localized impairment, these segments and their associated metals impairments can be grouped for TMDL
development.
While the upstream segments in the northwestern portion of the watershed (Wards Creek, Blacks Run, Bird
Creek) are hydrologically connected to downstream impaired segments, they are located upstream of a controlled
reservoir that can act as a divide for the watershed. The reservoir can serve as an upstream boundary for the
watershed of the Oak River grouping and allow for the upstream segments to be analyzed separately from
downstream segements. These upstream segments experience a variety of impairments, including sediment,
temperature and nutrients. While these impairments are not the same and they might not share sources
(temperature and sediment impacts from construction vs. nutrients from agriculture), the stream segments (Wards
Creek, Bird Creek, Blacks Run) could still be grouped together for a larger watershed TMDL. Sediment and
nutrients often require similar analyses and although impairments are not the same, there will likely be efficiencies
in data analysis and other TMDL activities. On the other hand, it would also be appropriate to break these
upstream segments into groupings based on their impairment and associated sources. For example, Wards
Creek and its impaired tributary would be grouped to address their multiple impairments for temperature and
sediment, while Bird Creek and Blacks Run would be grouped together to address their nutrient impairments. This
decision would likely be based on program-specific factors such as available resources (funding and staff time)
and existing studies or activities for the area.
The following table and map illustrate the resulting groupings based on the above evaluation, with each color
grouping representing a separate watershed TMDL and the corresponding impaired segments and impairments.
/\/ Impaired s egnents
TMDL Groupings
Island Creek, UT- Iron, Alum
OakR, OakRSF, MeadowCr, DobbsRun,
Bennett Cr, Turtle Cr- Fecal cdiform, benthic,
nutrients, DO, turbidity
Jacks B~, Daniels B~, Dry Run - Nutrients, benthic
^ Wards Cr, UT- Sediment, temperature
^ Brd Cr, Blacks Run - Nutrients
| |OakR-PCB
| [Oak R, MeadowCr, DobbsRun- Temperatu
DRAFT
33
-------
Handbook for Developing Watershed TMDLs December 2008
3.2.4. Public Concerns
Sometimes the level of detail used in an analysis is influenced by the political or social
concern over the waterbody. For example, a waterbody that is a "special resource" to the
region, perhaps providing important recreational or economic opportunities, might merit
a more detailed analysis because of the increased public interest and also the potential
economic impact to the area if the water is not restored. When deciding how to group
watershed TMDLs, it might be important to define the scope of a TMDL to allow the
necessary level of detail and focus for these special waters.
3.3. Leveraging Existing Watershed-based Programs or Efforts
In some cases, ongoing and related watershed-based programs might facilitate identifying
waterbodies and impairments to group in a watershed TMDL. Considering these existing
watershed-based efforts represents the third screening criterion for determining the scope
of watershed TMDLs, as shown previously in Figure 3-1. Many states coordinate some, if
not all, of their water programs on a rotating basin or watershed framework. Whether or
not this is the case for TMDLs, the existing planning structure can guide watershed
TMDL planning. For example, planning basins used to coordinate monitoring schedules
often serve as a good template for defining the geographic boundary of the TMDL.
Established planning basins and associated schedules can also help to schedule the timing
of watershed TMDL development. For example, if a basin is scheduled to undergo its 5-
year cycle of monitoring in the next year, it might be helpful to evaluate the TMDLs that
fall within that basin and schedule the TMDLs for a year following the monitoring and in
groupings that correspond to monitoring schedules and station locations. Similarly,
NPDES permits are sometimes reviewed and issued by basin on a set interval (e.g., every
5 years). Where this is the case, a practitioner can consider grouping TMDLs using a
similar geographic organization and schedule TMDL development to precede permitting
activities to ensure permit writers can incorporate relevant WLAs during the issuance or
reissuance process.
Another consideration in identifying and grouping potential watershed TMDLs is the
activities of existing stakeholder groups. Stakeholder groups have the potential to play an
important role in TMDL development by providing data and information, identifying
potential sources, providing historical and local perspective, and identifying opportunities
for implementation activities. Some watersheds have active watershed groups that
conduct a range of activities, from promotion of basic education and awareness to large-
scale volunteer monitoring and cleanup activities to development of comprehensive
watershed management plans. Large watersheds that have active watershed or
stakeholder groups can be targeted for TMDL development to capitalize on the ongoing
efforts and participation of the groups that already have a defined geographic focus.
Related to existing watershed groups and activities, ongoing watershed planning efforts
under the Section 319 nonpoint source control program might also serve to guide
identification of watershed TMDLs and their scope. While some Section 319 efforts
focus on small, site-scale projects, others focus on wider watershed efforts for
implementing BMPs or conducting monitoring. An existing 319 project can support
watershed TMDL planning and development by helping to define the geographic scope
of the analysis and providing a basis for the TMDL analysis. Recently EPA developed
Handbook for Developing Watershed Plans to Restore and Protect or Waters (USEPA
2008) to encourage the development of comprehensive watershed-based plans to meet
319 requirements. These plans are typically more quantitative than historical 319 plans in
34 DRAFT
-------
December 2008 _ Handbook for Developing Watershed TMDLs
Guiding Watershed TMDL Development based on Related Previous, Ongoing or Planned Efforts
Sometimes the scope of a watershed TMDL can be defined to either build upon related previous or ongoing
watershed-based efforts or similarly to facilitate planned efforts. The following are examples of how watershed
TMDLs were chosen for development based on the goals or efforts of related watershed-based programs:
• Simpson Northwest Timberlands Sediment and Temperature TMDLs: The TMDLs for waters included in
Simpson Timber Company land were developed concurrently with an aquatic Habitat Conservation Plan (HCP)
developed under the Endangered Species Act (ESA) for the same geographic area. The HCP describes a suite
of management prescriptions, assessment, and monitoring actions, with Simpson's conservation program
emphasizing the protection of riparian forests and erosion control to satisfy the requirements of ESA §10. The
TMDL provided the analytical framework to support the necessary riparian and sediment management practices
identified in the HCP and provided needed technical information to guide monitoring efforts that satisfy the
adaptive management aspects of the HCP. This unique coordination of two federal programs relied on
cooperative input from the core group of Simpson, National Marine Fisheries Service, U.S. Fish and Wildlife
Service, Washington Department of Ecology, and EPA. To read the Simpson TMDLs, please visit
www.ecv.wa.gov/pubs/9956.pdf and www.ecv.wa.qov/biblio/0010047.html.
• Long Island Sound Nitrogen TMDL: Beginning in 1985, New York, Connecticut and the EPA formed the Long
Island Sound Study (LISS) to promote measurable improvements to the water quality of the Sound. In 1 994, the
LISS completed a Comprehensive Conservation and Management Plan (CCMP) under EPA's National Estuary
Program. The CCMP identified seven priority issues, the highest of which was low dissolved oxygen levels in the
Sound. By 1998, the LISS CCMP adopted a 58.5 percent reduction target for nitrogen loads and specified in its
implementation plan that a TMDL for nitrogen be adopted with LAs and WLAs for all sources in the watershed.
As a result, the states of New York and Connecticut jointly developed and submitted a TMDL for nitrogen, which
was approved by EPA in 2001.
their analysis of sources and linkage to water quality. Any existing or planned 319
watershed planning efforts provide an ideal basis for integrating and developing
watershed TMDLs.
In addition, it is helpful to consider other existing state and federal programs that might
impact TMDL development either through the provision of data for developing the
TMDL or through their potential for leveraging implementation funds. How many
NPDES facilities are contributing to the impaired segments? Are there areas in the
contributing drainage that are covered by Phase II stormwater permits? What pollutant
parameters are being discharged and are they related to the TMDL?
The NPDES program and permittees can provide data for the source characterization and
analysis. In addition, these key stakeholders also serve as critical implementation
partners for the TMDL. Any NPDES watershed-based permitting effort involves a
TMDL-like analysis of pollutant fate and transport in a watershed; therefore, TMDL
practitioners should investigate if such an effort is underway in the watershed and what
type of information is available from the permitting activity for use in the watershed
TMDL.
DRAFT 35
-------
Handbook for Developing Watershed TMDLs December 2008
36 DRAFT
-------
4. Developing Watershed TMDLs
The general process for developing watershed
TMDLs is much the same as that for
developing single-segment TMDLs. Figure 4-1
illustrates the typical steps for developing a
TMDL, including:
• Stakeholder involvement and public
participation to engage affected parties and
solicit input, feedback and buy-in for a
successful TMDL. This process should
occur throughout the TMDL development
(and implementation) process.
• Watershed characterization to identify the
watershed, waterbody and impairment
conditions; TMDL targets; and potential
sources.
• Linkage analysis to calculate the loading
capacity.
Resources:
Guidance on Developing TMDLs
For more information on the general TMDL development
process, refer to the following EPA references:
Guidance for Water Quality-based Decisions: The TMDL
Process (USEPA 1991):
www.epa.qov/OWOW/tmdl/decisions/
Protocol for Developing Nutrient TMDLs (USEPA 1999a)
www.epa.gov/owow/tmdl/nutrient/pdf/nutrient.pdf
Protocol for Developing Sediment TMDLs (USEPA 1999b):
www.epa.gov/owow/tmdl/sediment/pdf/sediment.pdf
Protocol for Developing Pathogen TMDLs (USEPA 2000):
www.epa.gov/owow/tmdl/pathogen all.pdf
More technical and policy support documents are available
on EPA's TMDL Program Web site at
www.epa.gov/owow/tmdl.
• Allocation analysis to evaluate and assign WLAs to point sources and LAs to
nonpoint sources.
• Development of the TMDL report and administrative record for submittal to EPA.
While these steps are common to all TMDL development projects, there are a number of
considerations for each step when developing multiple TMDLs within a watershed
framework. This section highlights the critical issues for each step as related to
developing watershed TMDLs.
4.1. Stakeholder and Public
Involvement
In any TMDL process, practitioners should engage stakeholders early with activities such
as project scoping and data collection and continue throughout the allocation and
implementation phases. Stakeholders to involve in TMDL development can include
partner state and federal agencies; pollutant sources, such as permitted facilities or
landowners that are likely to receive WLAs or LAs; and citizen groups, watershed
organizations, and other interested parties in the watershed that might provide assistance
in TMDL implementation. Depending on the complexity of the TMDL, stakeholders can
participate in a variety of capacities ranging from attending public information meetings
to supporting selection of technical approaches and contributing data to participating in
implementation decisions. For a watershed TMDL, stakeholder and public involvement is
likely to expand due to the larger geographic area and increased number of sources
addressed through the watershed TMDL process. Practitioners should anticipate more
DRAFT
37
-------
Handbook for Developing Watershed TMDLs December 2008
stakeholders involved in the development process and anticipate stakeholders to bring to
the table watershed concerns and issues that transcend the scope of the watershed TMDL.
The stakeholder engagement phase is critical because there are likely to be more potential
stakeholders than those involved in a single-segment TMDL.
Watershed Characterization
Compile and analyze watershed and waterbody data (e.g., CIS,
in-stream monitoring, weather).
Characterize in-stream conditions and impairments.
Gain basic understanding of waterbody and watershed
characteristics affecting impairment.
Identify WQS and other TMDL targets.
Identify potential sources.
c
.o
'
'•£ Linkage Analysis to Calculate Loading Capacity
£
o • Select and apply approach to establish a link between pollutant
IS loading and water quality.
3
°- • Estimate existing source loads.
~o
|g • Calculate allowable loading capacity.
>
§
— Allocation Analysis
0)
^
•g • Select appropriate level (geographic, temporal and source) for
•= allocations for successful implementation.
^
Q • Evaluate allocation scenarios representing different combinations
w of load reductions (WLAs and LAs).
• Select most appropriate and feasible allocation scenario.
TMDL Report and Submittal
• Prepare TMDL report.
• Document all required elements of a TMDL.
• Compile administrative record.
Figure 4-1. General steps in developing a TMDL.
38 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
In watersheds, engaging stakeholders early in the TMDL development process often
results in better integration of stakeholder data, more buy-in and acceptance of the
process and resulting allocations, and fewer "surprises" during the later public comment
stages. A watershed TMDL provides an opportunity to engage all of the stakeholders in
the same process at the same time, using the TMDL as a unifying goal and focal point for
discussion and involvement. This helps to raise awareness about the stakeholders'
priorities and concerns and avoid potential delays caused by having to revisit decisions
and previously conducted analyses because of new information or issues introduced by
stakeholders who were not engaged in the early stages of the process. Stakeholders can
also be important sources of data (e.g., volunteer monitoring data, knowledge of key
watershed characteristics, facility discharge data) and can provide information on the
existence and locations of critical sources that might otherwise have been unaccounted
for (e.g., historical land uses). Broad-based stakeholder involvement also feeds into
implementation phases of the TMDL since many stakeholders are also key
implementation partners that will perform the on-the-ground work necessary to reduce
loading by upgrading treatment processes, installing BMPs, and acquiring funding for
necessary implementation activities.
Some watershed TMDL projects cross jurisdictional boundaries, including county, city
and sometimes even state boundaries and federal jurisdictions, introducing a number of
regulatory agencies into the TMDL development process, as well as implementation
efforts. Watershed TMDLs that move into the category of multijurisdictional TMDLs
(those involving multiple states or tribes) can present a unique set of challenges to TMDL
development, including varying water quality standards, TMDL schedules and
implementation goals. The approach for planning and completing a multijurisdictional
watershed TMDL is generally the same as any other TMDL. However, the process
should include increased attention to and coordination with the other affected groups so
that all parties involved are included in all communications and at critical decision points
and have a thorough understanding of the expectations, legal requirements, priorities and
needs of each entity. The TMDL should be developed with an active stakeholder process
to ensure all agencies are informed and that any issues requiring decisions (e.g., approach
to maintaining varying water quality standards) are introduced early in the process. Open
communication and consensus-based decision-making are essential for the successful
completion of multijurisdictional TMDLs.
DRAFT 39
-------
Handbook for Developing Watershed TMDLs _ December 2008
Maximizing the Benefits of Stakeholder Involvement in the San Jacinto Watershed, California
The Santa Ana Watershed Project Authority (SAWPA) developed TMDLs for Lake Elsinore and Canyon Lake as
part of development of a nutrient management plan for the San Jacinto Watershed. During the effort, SAWPA
engaged the active stakeholder community and relied heavily on its participation to provide data, previously
developed tools, historical knowledge, and review of potential management options.
The effort involved application of a watershed and lake modeling system to evaluate nutrient sources and transport
in the 780-mi2 watershed under a range of hydrologic regimes. The modeling system was also used, through
consultation with the stakeholder group, to identify and test potential BMP strategies to meet lake water quality
goals. The results of an analysis of relative impacts of alternative management solutions were used as decision
support for development of the final nutrient management plan and placement of an appropriate suite of BMPs.
Following adoption of the nutrient TMDL, stakeholders were tasked with developing a monitoring plan to address
data gaps and provide data for evaluating compliance to TMDLs. SAWPA and the TMDL Task Force evaluated the
ongoing and planned monitoring programs of multiple stakeholders to develop a single monitoring plan. The
resulting plan provided significant cost savings (phased monitoring approach with reduced costs ranging from
$20,000 to over $200,000 per year) to stakeholders. SAWPA worked closely with stakeholders to reach an
agreement regarding priorities, schedule, data gaps, and requirements for measuring compliance to the TMDL.
The increased stakeholder participation fostered by SAWPA led to greater acceptance of the management
strategies and increased the likelihood of implementation and success. Because the stakeholders were actively
involved in the decision-making process they were more confident in the resulting allocations and therefore more
willing to expend resources to implement them. The TMDL allocations not only provided a framework for
distributing load reductions but also associated funding responsibilities among the stakeholders.
The San Jacinto watershed provides an example of a watershed TMDL facilitating increased stakeholder
participation, which then led to benefits from greater accessibility to watershed data and tools, consensus-based
decisions, increased acceptance and ownership of allocations, and maximizing of resources and coordination for
implementation and follow-up monitoring.
To read the plans developed for the San Jacinto watershed, please visit
www.sawpa.org/tmdl/Lake elsinore Canyon lake.html.
4.2. Watershed Characterization
The TMDL process requires a thorough
understanding of the watershed characteristics, available data, causes of impairment,
sources, water quality standards, and potential targets. Some of this information will be
available through a state's 303(d) list and waterbody assessment documentation, but
much of the information will have to be gathered and summarized while completing the
TMDL. Collectively, this is referred to here as the watershed characterization
component of the TMDL. Watershed characterization serves as the foundation of the
TMDL analysis, providing a basic understanding of the impairments of concern, the
desired levels for restoration (e.g., water quality standards and TMDL targets) and the
likely sources contributing to the impairment. Characterizing the watershed, as well as
the waterbody and the associated impairments and sources, provides the necessary
background information to support decisions regarding the approach used for calculating
the TMDL, the level of detail or focus of the analysis and ultimately TMDL
implementation. The following sections describe the major elements of the watershed
characterization, including:
• Data analysis for problem identification
• Identification of TMDL targets
"40 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
• Identification and assessment of potential sources
4.2.1. Data Analysis for Problem Identification
The objective of problem identification in the TMDL
process is to identify the nature of the impairment(s)
being addressed by the TMDL. In many cases, the
listing itself is not enough to fully inform the problem
identification process. A state or tribe's 303(d) list
identifies the basic information regarding the impaired
waterbody and the observed impairment, usually
including the waterbody characteristics (e.g., name,
location, size), the water quality standard that was
violated, the pollutant of concern, and the suspected
causes and sources of impairment. Additional
information, however, is often desired for the TMDL
process. For example, a TMDL practitioner might
want to know:
Process Tip:
What is the Problem?
The problem identification step is likely the most
important in the TMDL process. This step
determines how subsequent steps are conducted
and focused. This is especially true with
watershed TMDLs. The problem identification
identifies the areas/segments, pollutants,
sources and conditions of concern, which in turn
help to focus source characterization, determine
technical approaches, and identify allocations.
• How does the available water quality data vary over space and time?
• How are the water quality data and impairments in a watershed related?
• Did the 303(d) process correctly identify the causes and sources of impairment in
the waterbody?
• Have new data been collected since the initial 303(d) listing, and what additional
information do those data provide?
The data analysis activity of a TMDL serves to answer these questions and support
problem identification. It involves a review of the 303(d) listings, a thorough inventory of
watershed conditions and systems, and the mapping of s
the spatial distribution of pollutant sources as they
relate to the water quality impairment. Critical issues
are identified by developing a preliminary description
of water quality problems and basic interactions (when,
where, under what conditions is problem evident?)
through the analysis of in-stream monitoring data (e.g.,
flow, water quality, bioassessment). The answers to
these questions help to define many of the technical
aspects of the TMDL, including what sources are
quantified, what approaches can be used, how
allocations are determined, and on what time and spatial
scale the analysis is conducted.
Because a watershed TMDL can include multiple
subwatersheds, impaired segments, pollutants of
concern, expected sources and critical conditions
affecting impairment, the problem identification stage is
critical to effectively focus the analysis to identify and
address the critical issues and also maximize the
resources available to develop a watershed TMDL. For
this reason, it is important to carefully analyze available
monitoring data to identify any patterns or trends that
can highlight important sources, connect multiple
Online Sources of Physical, Chemical and
Biological Monitoring Data
STORET is EPA's database for the storage and
retrieval of ground water and surface water
quality data. In addition to holding chemical and
physical data, STORET supports a variety of
types of biomonitoring data on fish, benthic
macroinvertebrates, and habitats. STORET data
can be downloaded from
www.epa.qov/STORET/index.html.
The National Water Information System Web
site (NWISWeb) is the USGS's online database
for surface water and ground water flow and
water quality data. The NWISWeb database
provides access to water resources data
collected by USGS at approximately 1.5 million
sites in all 50 states, the District of Columbia,
and Puerto Rico. Data can be downloaded at
http://waterdata.usqs.gov/nwis.
State and local environmental agencies might
have additional data that are not yet included in
STORET or other available databases.
DRAFT
41
-------
Handbook for Developing Watershed TMDLs
December 2008
impairments, and describe critical conditions or important watershed processes affecting
impairment. Important aspects of the data analysis to support problem identification for
the watershed TMDL include:
• Spatial analysis to identify spatial variations in waterbody and watershed conditions
to identify sources and understand the relationship among impaired segments
• Temporal analysis to evaluate the timing of impairment and potential source loading
or other conditions contributing to impairment
• Analysis of the relationships among multiple parameters or in-stream measures (e.g.,
pollutant concentration and flow) to understand critical conditions and identify
potential sources
Because watershed TMDLs are often developed for a broader geographic scale than
single-segment TMDLs, it is desirable to have sufficient data to characterize the water
quality and sources throughout the watershed and at key locations (e.g., tributary
confluences, upstream/downstream of major sources). With the larger area, there is more
potential for variations in the amount, type, and quality of data throughout the watershed.
Incomparable data can create difficulties in conducting meaningful statistical or modeling
analyses. However, data limitations and variations among datasets can be an issue with
any TMDL, whether single-segment or watershed. TMDLs are sometimes developed
with less than desirable amount, period and spatial distribution of data. Existing TMDL
program guidance (USEPA 1991) suggests that having limited data is not a sufficient
reason for not developing a TMDL and recommends that TMDLs be developed with the
best available data. If data are limited the TMDL should be developed, recognizing that
additional data and information could be used in the future to revise the TMDL as
necessary.
Alternatively, although a larger watershed can mean data from a greater number of
agencies and of varying types, it also provides more opportunity to extrapolate an
understanding of conditions throughout a watershed. When focusing on a single-segment
TMDL with insufficient data to characterize the watershed, it might be necessary to
evaluate data outside of the study area to make assumptions about the area without a real
understanding of how the areas are related or similar. However, developing a watershed
TMDL encourages evaluating data for a larger area and the relationships among that data.
It provides a stronger foundation for any assumptions about conditions in data-poor areas.
The following sections provide more details on the common types of data analyses
conducted for a watershed TMDL to understand impairment conditions and identify
sources.
Spatial Analysis
TMDL practitioners often times want to know how
watershed characteristics (e.g., land use, water quality,
soils/geology) vary throughout their watershed. A
spatial analysis of these characteristics can help to
inform multiple steps of the TMDL process. For
example, a spatial analysis of water quality data can
help to identify the location and distribution of areas
that exhibit increased pollutant levels. Through
additional analysis of land use and other waterbody
Process Tip:
Bigger Watershed = More Data
When conducting a watershed TMDL, there will
be more data and information to organize and
analyze. It is important to establish an effective
and consistent way of organizing data, including
in-stream flow and water quality data, GIS
coverages and watershed reports.
EPA's Watershed Handbook (USEPA 2008)
includes tips on gathering and organizing data
from disparate sources on a watershed scale.
42
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
information, a TMDL practitioner can begin to understand what sources might be
contributing to impairment and the conditions or processes that might be exacerbating
impairment in the watershed. Regardless of the type of TMDL (single-segment or
watershed), a spatial analysis of the available data is a critical step in understanding the
linkage between sources, causes of impairment, and waterbody response.
When completing a watershed TMDL, the spatial analysis step can be more important
because the larger land areas, multiple tributaries, and multiple sources might require
more careful evaluation to fully understand the relationship between sources and
waterbody impairment. For example, a downstream waterbody might be impaired
because of a number of upstream tributaries and sources. A spatial analysis of the
available water quality data would help to identify and locate the sources, including any
potential unknown sources, and would inform future steps in the TMDL process
including target setting, source assessments, and load allocations. Understanding the
spatial variation in water quality levels can also help to determine how to subdivide the
watershed for subsequent analyses (e.g., watershed modeling) by isolating areas of
increased loadings or unique characteristics.
When completing a watershed TMDL, spatial analyses can be used to look at trends
throughout the entire watershed or more specifically evaluate data bracketing areas of
concern or of expected source activity. For example, Figure 4-2 presents a map generated
using GIS coverages of monitoring station locations on impaired segments in a watershed
and corresponding statistics for water quality data. The map can be used to generally
identify areas exhibiting higher TDS concentrations. Similarly, Figure 4-3 shows
statistics for data collected at a number of stations located along the length of a river.
Understanding the increases and decreases in pollutant concentrations can help to identify
types and locations of sources in the watershed. Alternatively, spatial analyses can also
be used to evaluate specific sources. Figure 4-4 presents a graph showing matching data
collected upstream and downstream of an expected source to evaluate the potential
impact of the source on downstream water quality. Plotting the upstream data versus the
downstream data can examine whether upstream and downstream concentrations differ
and indicate whether a source discharging to the impaired stream between the stations is
contributing to the impairment. As shown in the figure, the downstream (Station 4) TSS
values are consistently higher than those measured at the upstream site, indicating that
there is some loading input between the stations. Figure 4-5 presents an alternative
representation of the data in Figure 4-4, simply plotting the data chronologically at the
two stations to visually determine whether upstream and downstream stations are
comparable and follow similar patterns. This type of analysis can be effective for
evaluating the influence of tributary inputs as well. Reviewing data collected upstream
and downstream of a tributary confluence can provide insight into whether the tributary
subwatershed is contributing elevated levels of pollutants and affecting downstream
water quality. This can be especially useful when developing watershed TMDLs and
evaluating multiple subwatersheds of impaired segments.
DRAFT 43
-------
Handbook for Developing Watershed TMDLs
December 2008
Average IDS
s. 0-135
•i 135 - 290
8 290-616
S
998-2013
Figure 4-2. Example map showing average IDS concentrations throughout a watershed.
10BOO
8.000 ;
6,000 •
* # ,-3" J? # 3 <¥ J" J
I I 25tti-75thPercentle * Average I Min-Max Standard
Figure 4-3. Example graph showing summary statistics in measured IDS concentrations
along the length of an impaired river.
44
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
250-
•j 200 -
E
OT
?
ro
01
In
c
I 100 -
•c
•*
§
ra
55 50 -
0 -
0
i
•
• '" "•
"•
" "5" :
mb ^
if"
• i
i
x"'
,"V"
!t!_
•
•
•
•
" f^r
X
•
•
. 1:1/
/
•
. /
x
0 50.0 100.0 150.0 200.0 250.0
Station 3 (upstream) - TSS (mg/L)
Figure 4-4. Example graph showing relationship between upstream and downstream data to
evaluate potential impact of an expected source.
1,000
1
Oct-97
Jan-98
Apr-98
Jul-98
-Station 3 - Upstream
-Station 4 -Downstream
Figure 4-5. Example graph showing data collected upstream and downstream of an
expected source.
DRAFT
45
-------
Handbook for Developing Watershed TMDLs _ December 2008
Using Spatial Analyses to Identify Sources in the Big Spring Creek Watershed, Montana
PCS sampling in the lower reach of Big Spring Creek, Montana, (Waterbody ID MT41S004_020) found elevated
levels in fish tissues, and the segment was placed on Montana's 2004 303(d) list. The suspected source of the
impairment was the "Brewery Flats" industrial area in Lewistown, Montana. Instead of focusing on a single-
segment TMDL for Big Spring Creek and the Brewery Flats area, Montana DEQ initiated a watershed TMDL for
the entire Big Spring Creek watershed. As part of this effort, DEQ collected additional PCS data and conducted a
watershed-scale spatial analysis of the concentrations. The spatial analysis showed that PCS concentrations
were highest in the headwaters of the watershed (in Waterbody Segment MT41S004_010) and then declined in a
downstream direction past the Brewery Flats industrial area. Further investigations found that the source of the
PCBs was contaminated paint from the Big Spring Creek fish hatchery's raceways and not the Brewery Flats
industrial area. Completion of the watershed scale spatial analysis identified a previously unknown source of
PCBs and identified a new waterbody segment as impaired because of PCBs.
Temporal Analysis
As with any TMDL, evaluating the temporal patterns in water quality data can lead to an
understanding of how source behavior, weather patterns, and waterbody conditions relate
to resulting impairment. Evaluating these factors in the context of a watershed TMDL can
identify important issues to be addressed in the technical approach. A temporal analysis
of water quality data can help to identify seasonal sources and associated loads (e.g.,
grazing, seasonal residents, and recreational uses), identify new sources, and compare
pre- and post-source water quality data. Understanding the different sources and their
times of loading and influence can guide the selection of a technical approach and help to
better maximize the watershed-wide benefits of load reductions during the allocation
analysis.
While temporal variations in water quality can be affected by source activity, they are
more often related to environmental conditions such as weather and resulting flow
patterns. Evaluating the relationship between water quality, flow and seasonality can be
done using a variety of techniques including simple visual comparison of graphed time-
series data, regression analyses, or the use of flow duration curves. Figure 4-6 includes
examples of each of these types of data representation using the same dataset. As shown
in the figure, all of the figures can be used to show the relationship between bacteria and
flow. While the regression plot does not show a strong correlation between flow and
bacteria, the chronological and flow duration graphs show that they do tend to follow
similar patterns, with elevated bacteria typically occurring during higher flows. Because
discharges from certain types of sources are typically observed during particular flow
conditions, evaluations of flow and corresponding water quality can be a helpful tool in
identifying potential sources of impairment in a watershed and also understanding
waterbody and impairment conditions for selection of an appropriate technical approach.
The evaluation of temporal variations in water quality for watershed TMDLs will not
likely be different than for single-segment TMDLs. However, as with many of the data
analysis activities, the analysis has the potential for evaluating multiple pollutants and
impairment types and the effects of multiple sources. This can result in the identification
of varying temporal trends (e.g., depending on varying source behavior or pollutant fate
or behavior) that can affect what approach is selected for TMDL development.
46 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Regression - Flow vs. Fecal Coliform
1200
1000
o
^ 600 -
o
8 :
T! 400 -
iZ :
200
<
o
0
0 0.
£B&c
' «
* 0
»
»
X
<
•
•
0
«
/
*
R2 =
/
7
= 0.6796]
/
>
0 100 200 300 400 500 600 700
Row(cfs)
Time-series of Row and Fecal Coliform Data
10,000
1,000 -.
Fecal coliform (#/100 m L) Flow (cfs)
Row Duration Curve and Corresponding Feca\ Coliform
* «
k&£
3,000 —
20% 30% 40% 50% 60% 70% 80% 90% 100%
Row Duration Interval
Figure 4-6. Examples of different data representations to evaluate the relationship between
flow and water quality
DRAFT
47
-------
Handbook for Developing Watershed TMDLs
December 2008
Evaluation of Multiple Parameters
Many pollutants causing impairments throughout a watershed can originate from a
common source(s). For example, nutrients and bacteria often are associated with the same
types of sources, such as failing septic systems, wildlife populations, and agricultural
livestock uses. Evaluating the correlation among multiple pollutants can help to
understand the types of sources in the watershed and better focus subsequent assessment
and allocation. In addition, some pollutants might in fact be dependent on other
pollutants. For example, some pollutants (e.g., nutrients, metals) can be delivered to
receiving waters adsorbed to sediment particles. In agricultural areas that have
experienced fertilizer application, areas of increased erosion (e.g., degraded cropland,
overgrazed areas, areas experience streambank erosion) can deliver significant amounts
of nutrients that have attached to watershed soils. Similarly, some waterbodies
experienced historical accumulation of pollutants, such as metals or pesticides, that
adsorb to sediments but do not die-off. These pollutants can remain in bottom sediments
but can also be resuspended during increased flows or other disturbances. Identifying a
relationship between increased sediment concentrations and other pollutants can help to
identify these situations to better target source assessment.
4.2.2. TMDL Target Identification
All TMDLs require a target or indicator that can be used to evaluate attainment of water
quality standards in the listed waterbody. Often, the numeric target value for the TMDL
pollutant will be the numeric water quality criterion for the pollutant of concern. In some
cases, however, TMDLs must be developed for pollutants that do not have numeric water
quality criteria. When numeric water quality criteria do not exist, impairment is
determined on the basis of narrative water quality criteria or identifiable degradation of
designated uses (e.g., impaired fishery). The narrative criterion is then interpreted to
develop a numeric TMDL target that represents attainment of the water quality standards.
One of the benefits of completing a watershed TMDL
is that the TMDL targets can be set based on an
understanding of upstream and downstream
conditions. For example, when interpreting narrative
criteria for single-segment TMDLs, it is possible to
set a target that is not protective of the upstream or
downstream beneficial uses. The single-segment
target (and the associated TMDL) could be set, only to
be revisited later when another segment's TMDL
requires more stringent targets. The watershed TMDL
process allows the practitioner to consider multiple
upstream and downstream segments when setting
TMDL targets, thereby insuring that targets are
protective for multiple segments.
Target identification for watershed TMDLs can be
complicated when the watershed is multi-
jurisdictional. Waterbodies that cross state or tribal
boundaries might be subject to multiple or differing
water quality criteria for the same pollutant. When
dealing with numeric criteria, the watershed TMDL
would typically apply the most stringent criteria.
Setting Variable Targets to Support the Tualatin
Watershed TMDL, Oregon
Oregon Department of Environmental Quality
developed a watershed-based TMDL to address
impairments related to temperature, bacteria,
dissolved oxygen, pH and chlorophyll a in the
Tualatin River and several tributaries. The pH
impairment is tied to excess algal growth, which is
measured by chlorophyll a and driven by total
phosphorus concentrations. Therefore, total
phosphorus was used as the indicator for the
combined pH and chlorophyll a TMDL. Natural
conditions were chosen to represent target levels
for phosphorus. Because the TMDL addressed
several impaired segments, data analysis and
mass balance evaluations were conducted for the
impaired mainstem and tributaries to identify
background levels of phosphorus for development
of segment-specific total phosphorus targets. To
read the Tualatin TMDL, please visit:
www.deq.state.or.us/wq/TMDLs/willamette.htm#t.
48
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
However, because this might result in more stringent allocations in one jurisdiction to
support criteria in the downstream jurisdiction, it is important to involve all jurisdictions
and agencies in the process. When dealing with narrative criteria in multi-jurisdictional
TMDLs, it is important to understand how each jurisdiction applies the criteria for
impairment determination and listing. These assessment methodologies can support
identification of an appropriate TMDL target, and when they vary by jurisdiction, all
parties should be involved in identifying an acceptable TMDL target.
Varying water quality criteria or TMDL targets can also be an issue when dealing with
multiple types of receiving waters (e.g., lakes, bays, rivers) or waterbodies that have site-
specific targets that differ from the generally applicable criteria. Practitioners should
consider this when selecting the most appropriate approach for TMDL development. The
approach should be at a level of detail sufficient to evaluate the potential relationship
among the targets and allow for evaluation of impacts of source loads on all of the
waterbodies of concern to ensure that allocations meet all targets.
For watershed TMDLs, the process of identifying numeric targets can be more in-depth
than for single-segment TMDLs because of multiple waterbodies and possible pollutants.
Conclusions drawn from the data analysis and problem identification step should guide
the development of targets for the impaired segments and pollutants of concern. It is
possible that the analysis might indicate a need for different targets for different
segments. For example, if a TMDL addressing sediments aims to develop in-stream or
loading targets based on background conditions, the specific target value could vary
throughout the watershed. In areas where pollutants subject to narrative criteria are the
primary impairment, developing TMDLs on a watershed basis ensures that the targets
work together to ensure use attainment throughout the watershed. If the TMDLs were
developed individually, the variable targets might not be set at levels to ensure use
support in downstream waters.
4.2.3. Source Assessment
A source assessment can significantly influence TMDL development, associated
allocations, and subsequent implementation. The source assessment should be an
extension of the analyses conducted during the problem identification step to further
characterize the important sources and better define their location, behavior, magnitude
and influence. The source assessment should result in an understanding of what major
sources are contributing to impairment, which sources are contributing which pollutants
and which sources have an impact on which segments. Again, this can affect what
approach is selected and how it is applied for TMDL development and help to focus the
allocation analysis as well as future implementation.
While the pollutant loads originating with each source are typically quantified during the
linkage analysis (Section 4.3), the information necessary to understand their location and
discharge behavior and characteristics is compiled and reviewed during this step. In
general, the methods that are used to complete a source assessment do not differ between
single-segment and watershed TMDLs, and they involve identification and
characterization of point sources (e.g., wastewater treatment plants, industrial facilities)
and nonpoint sources (e.g., grazing, timber harvest, septic systems). The methods for
completing a source assessment vary with the type of watershed, pollutants, and sources
but typically rely on information from state or national databases, literature reviews, and
local knowledge from state or local contacts. It is important to correlate the assessment of
both point and nonpoint sources the data analysis to characterize source impacts and
DRAFT 49
-------
Handbook for Developing Watershed TMDLs
December 2008
behavior. For example, land use, locations of facility
discharges and other source information should be
evaluated along with water quality data analyses (e.g.,
spatial analysis) to understand potential impacts from
the various sources or explore unknown sources.
Point sources are generally easier than nonpoint sources
to identify and quantify as they are usually permitted
and tracked under EPA's NPDES. Point sources might
include surface water discharges from permitted
facilities (e.g., WWTPs, industrial discharges),
stormwater discharges, groundwater discharges, and
construction discharges (depending on the watershed
and the state or tribe's permitting program). Identifying
the number, type and location of NPDES permitted
point sources in a watershed can start with searching
available databases (e.g., PCS) but might also require
coordination with relevant state or tribal permitting staff
to obtain further information about the source and its discharge. Typical information used
to characterize a point source includes facility type, design flow, permit limits, number
and location of permitted outfalls, and available discharge monitoring data. TMDL
developers should coordinate with the state staff to identify and obtain additional
information that might be available from the facilities. For example, facilities might
collect more water quality data than are required for or reported in the discharge
monitoring reports (DMRs), and these data could be helpful in calculating more accurate
loads and in addressing seasonality.
Resources:
Identifying Point Sources
Information on point sources permitted through
EPA's NPDES can be obtained from EPA's
Permit Compliance System (PCS) database,
which allows the user to query and obtain
information on permitted facilities through an
online interface. PCS is available at
http://www.epa.gov/enviro/html/pcs/index.html.
TMDL developers should also contact state
permitting staff to obtain information. PCS might
not include all available data or information.
Some sources, such as permitted stormwater,
might require additional information (e.g., MS4
drainage maps) than what is included in PCS to
fully characterize.
Identification and assessment of nonpoint sources are
usually based on review of aerial photos, satellite
imagery, GIS coverages (e.g., land use/cover, soils),
windshield surveys, and other maps and available data.
TMDL practitioners can use these data to determine the
location and extent of nonpoint sources at the watershed
scale. Sometimes more information is available from
literature reviews and previous studies of nonpoint
sources in a given watershed. Conservation districts,
county planning or environmental agencies, and state
water quality programs might also have supplemental or
site-specific information on land uses and sources.
Resources:
Compiling Land Use and Cover Information to
Identify Nonpoint Sources
Nationally available sources of land use data
include satellite data from the Multi-Resolution
Land Characteristics Consortium's (MRLC's)
National Land Cover Database (NLCD)
(available at http://www.epa.gov/mrlc/) and aerial
photos available from the National Agriculture
Imagery Program (NAIP) (described at
http://165.221.201.14/NAIP.html). Local land use
data might also be available through county
agencies (e.g., planning offices, environmental
offices).
Because of the larger area addressed in a watershed
TMDL, there will likely be more sources to evaluate than when dealing with a single-
segment TMDL. Because of this it is often important to use the source information to
build off the data analysis, focusing the more detailed characterization to sources
expected to be contributing to impairments. It is also important to decide the appropriate
scale for source evaluation within a watershed TMDL. For example, at what level of
detail will land uses be identified and evaluated, or how much effort will be expended for
site-scale identification of sources (e.g., through watershed surveys)? The TMDL
developer should balance the level of detail with the goals, priorities, and available data
and resources for the project. The scale should capture the major sources without
overburdening the analysis with little added benefit.
50
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Although watershed TMDLs can expand the scale of and effort required for the source
assessment in comparison to single-segment TMDLs, practitioners will realize the
benefits of a watershed TMDL through this activity. Evaluating the sources at the
watershed scale can focus the analysis to those sources that have the greatest impact on
water quality. It also facilitates more targeted allocations by evaluating the relative
magnitude and impact of sources. This is especially important when dealing with a
watershed affected by both point and nonpoint sources. Watershed TMDLs can provide a
platform for more effectively evaluating all sources and various combinations of source
controls that can represent attainment of water quality standards. This is important in
watersheds where water quality trading might be used as a tool for implementation.
4.3. Linkage Analysis
Watershed
Characterization
Allocation
Analysis
TMDL Report
Submittal ,
Stakeholder Involvement and Public Participation
For all TMDLs, the linkage analysis establishes
the cause-and-effect relationship between pollutant sources and the waterbody response.
Selecting what approach is used for this analysis is often guided by a number of technical
and practical factors, such as waterbody type, pollutant type and behavior, source type
and behavior, data availability, spatial and temporal needs, and user considerations (e.g.,
experience needed, anticipated level of effort). When dealing with watershed TMDLs,
there are a number of specific technical considerations that can affect what approaches
can be used in TMDL development and how they are applied. This section identifies the
factors unique to watershed TMDL development that can affect selection of a TMDL
development approach. It then discusses commonly used TMDL approaches and some
practical considerations for their application for development of watershed TMDLs.
4.3.1. Factors Affecting Selection of Technical Approach for Watershed
TMDL Development
When selecting an approach for TMDL development, a number of factors are often
considered. As shown in Figure 4-7, these can include user needs or requirements,
programmatic considerations, and technical needs. While user needs and programmatic
considerations will often guide the general type of approach (e.g., simple vs. complex,
modeling vs. non-modeling), the technical considerations will weigh heavily in the
selection of a specific approach or methodology. The technical considerations define the
following three needs for the TMDL analysis:
• Spatial scale/resolution
• Temporal resolution/time scale
• Processes or features that need to be included (e.g., pollutant type, dynamic
waterbody conditions, in-stream transport)
The watershed characterization step of TMDL development (Section 4.2) should generate
the necessary information to define these needs by providing an understanding of the
impaired waterbodies, the surrounding watershed and the associated impairments.
Specifically, the major considerations or questions that were addressed during the
watershed characterization that can support selection of an appropriate approach for
TMDL development include:
• What are the applicable water quality criteria?
• What are the impairments and associated critical conditions?
DRAFT
51
-------
Handbook for Developing Watershed TMDLs
December 2008
• What are the sources?
• How are the multiple sources, impaired waterbodies or impairments related?
Table 4-1 summarizes the considerations related to each of the three technical needs for
these defining topics of water quality standards, impairment, and sources. The answers to
the questions outlined in Figure 4-7 and more specifically in Table 4-1 will guide
approach selection for TMDL development. The following sections discuss these
questions and how they relate to the selection of an appropriate approach for watershed
TMDL development.
User and Application
Considerations
What experience or training is
required to apply the approach?
What level of effort is needed for
application?
What are the data needs?
What is the expected cost (of
necessary software and of time
and labor for application)?
Programmatic
Considerations
What is the schedule?
Are there existing tools available for
the waterbody/watershed?
Are there any planned future uses for
the approach (e.g., linkage to other
analyses, implementation planning)?
Are there proven and accepted
methods applied for similar projects?
Technical
Considerations
What are the applicable water quality
criteria?
What are the impairments and critical
conditions?
What are the sources and their
behavior and characteristics?
How are the multiple sources,
impaired waterbodies or impairments
related?
Figure 4-7. Considerations for selecting a TMDL development approach.
Table 4-1. Summary of Technical Considerations for Selecting a TMDL Development Approach.
Technical Needs
of Approach
Spatial Needs
Technical Considerations for Approach Selection
Water Quality Criteria
• Are different criteria
applicable in different
locations within the
watershed?
Impairments and Critical
Conditions
• What is the location and
distribution of impaired
segments?
Sources
• What type of sources/land
uses exist in the watershed?
• What is the location and
distribution of sources?
• At what level do the sources
need to be isolated (e.g.,
gross loading vs. land use
specific loading)?
52
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Technical Needs
of Approach
Time-scale Needs
Processes to
Include
Technical Considerations for Approach Selection
Water Quality Criteria
• What are the duration
and frequency of
applicable criteria?
• Is criterion based on
pollutant level (e.g.,
concentration) or a
measure of response or
condition (e.g., DO,
eutrophication)?
• What are the pollutants?
Impairments and Critical
Conditions
• What is the timing associated
with impairment (e.g.,
instantaneous vs. chronic or
cumulative effects)?
• Are there any temporal trends
to capture (e.g., seasonality in
waterbody conditions)?
• Is meeting WQC dependent on
or affected by other waterbody
measures (e.g., nutrient levels,
temperature, pH)?
• What are the in-stream critical
conditions for loading
response (e.g., dynamic, flow
variable vs. steady-state)?
• If dealing with multiple
pollutants, how are they
related?
Sources
• Are the impacts due to
cumulative or acute loading
conditions?
• Are there temporal variations
in source loading (e.g., due to
weather patterns, seasonal
activities)?
• At what temporal scale do the
sources need to be
estimated?
• What is the source loading
behavior (e.g., precipitation-
driven, direct discharge)?
• Do sources impact multiple
impaired segments (i.e., need
for in-stream routing and
transport)?
• Does the analysis need to
evaluate individual and/or
cumulative impact of
sources?
What Are the Applicable Water Quality Criteria?
Section 4.2 discussed the activity of identifying TMDL
targets based on numeric water quality criteria or an
interpretation of narrative water quality criteria. The type
and expression of the TMDL target(s) is a major influence
on approach selection. The most basic factor is the pollutant
or other indicator for which the target is established because
some approaches might be more or less appropriate for
certain types of pollutants.
Process Tip:
Selecting an Approach: Questions Related
to Water Quality Criteria or Targets
• For what indicators (e.g., pollutant,
parameter) are the targets set?
• How are the targets expressed (e.g.,
average, maximum, concentration, load)?
• How are the multiple targets similar or
related?
The applicability of an approach is also affected by its
ability to simulate at a time-scale necessary for comparison j
to the TMDL target's magnitude, duration and frequency.
While many established water quality criteria are based on daily maximums or daily
averages, TMDLs are also commonly developed for water quality targets based on longer
timeframes. For example, TMDLs developed to meet narrative criteria for sediment or
nutrients can be developed to meet a target monthly loading rate based on loading
conditions in a reference watershed. Endpoints designed to address acute (short-term)
impairments are typically based on instantaneous maximums or daily averages while
chronic (long-term) problems (e.g., eutrophication, sediment loading and deposition) can
be represented by endpoints with longer durations (e.g., monthly average concentration,
annual loading).
While these issues can affect approach selection for any TMDL, unique considerations
for watershed TMDLs arise when dealing with multiple targets. A watershed TMDL can
include a variety of targets, whether because of multiple types of waterbodies and
impairments and associated pollutants or because of criteria that vary depending on
waterbody-specific conditions (e.g., metals criteria based on site-specific hardness
DRAFT
53
-------
Handbook for Developing Watershed TMDLs December 2008
values). Criteria or targets with different time scales can impact selection of the TMDL
approach and how it is applied because TMDL development approaches should be able to
evaluate the water quality at a time scale sufficient to compare to targets. The similarity
or differences in targets addressed in watershed TMDLs should be considered to select an
approach or combination of approaches that can accommodate the multiple targets while
still efficiently and effectively addressing all of the segments and impairments.
What Are the Impairments and Associated Critical Conditions?
EPA regulations require that TMDLs consider critical j _ _, f ]
conditions to ensure that the established allocations will j ^' |
result in water quality standards. Understanding the critical | Selecting an Approach: Questions Related j
conditions builds upon the previous analyses of spatial and j t° Critical Conditions j
temporal trends and relationships among pollutants and i ,.„,.. „ I
.,. , . -. . . „. ?., ._ , i • When do impairments occur? \
processes (discussed in Section 4.2) and identifies the i !
environmental conditions under which impairment occurs. ! " Where do imPairments occur? |
As with all the other analyses discussed, understanding j' What pollutants are causing impairment? |
critical conditions can provide clues about the location, | • What processes or conditions can affect j
timing and type of sources affecting impairment. In | the occurrence or magnitude of j
addition, understanding critical conditions, particularly with I |
watershed TMDLs, is crucial in supporting selection and j !
subsequent application of a TMDL technical approach.
Evaluating the impairment of concern and associated critical conditions helps to identify
the watershed and waterbody processes, spatial scale and temporal scale necessary for the
approach to capture the impairment and effectively evaluate the source impacts and
identify allocations. Many impairments are based on levels of a specific pollutant (e.g.,
violation of aluminum criterion), while some are based instead on the resulting
waterbody conditions (e.g., impaired biological community, eutrophication caused by
elevated nutrients). Some impairments also depend on or are affected by a variety of in-
stream measures and processes. For example, critical conditions related to low dissolved
oxygen often relies on the timing and availability of nutrient loads but also on other
factors such as resulting algal growth, flow and temperature.
With watershed TMDLs, the potentially greater number of sources, pollutant types, and
waterbody types can lead to varying impairment and critical conditions. Sometimes this
can require the selection of a more complex approach or combination of approaches that
can simulate a number of processes with spatial and temporal variation. It is important to
weigh the necessity of representing certain processes and variables for representing
impairment conditions to effectively evaluate source loads and identify allocations.
What Are the Sources?
The most basic distinction in sources and how they affect selection of a TMDL technical
approach is whether they are delivered through surface runoff (e.g., precipitation-driven)
or discharged directly to the waterbody at a discrete location. This traditionally defined
the differences in nonpoint (i.e., diffuse, unregulated) and point (i.e., discrete, regulated)
sources; however, these distinctions are not as clear-cut when dealing with permitted
stormwater or areas that experience unregulated direct discharges such as illicit
discharges or watering livestock. Whether categorized as nonpoint or point sources, the
representation of sources in a TMDL approach typically falls into the "precipitation-
54 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
driven" or "direct-discharge" categories, and the number i
and distribution of these types of sources can significantly ^' |
influence TMDL approach selection. Selecting an Approach: Questions Related j
to Source Characteristics j
The consideration of precipitation-driven sources .. . , , .. ,. i
, ,, , , • How are source loads delivered to i
sometimes necessitates the use or an approach that can impaired waterbody? I
simulate time-variable weather and resulting runoff and ,, , ., , , . . .. , i
,. ... ° , • How do the sources characteristics and i
pollutant loading. However, it is important to carefully |oads impact the waterbody? I
evaluate the level of detail in representing the variable . How are source£ distributed and |ocated j
loading in relation to the waterbody response when in re|ation to impaired segments? !
selecting the approach. For example, while some j
impairments are primarily influenced by precipitation-
driven sources, evaluation of their impact on designated uses might be more appropriate
on a longer-term, average basis. Consider a situation where sediment is causing
impairment to a stream, stormwater runoff from agricultural and urban areas represent the
major sources of sediment loading. While loads can change frequently and rapidly in
response to rain events, they do not typically cause acute, instantaneous effects on
designated uses. Their impacts are more chronic in nature, resulting from the continued
delivery and accumulation of sediment. In this case, using an approach that evaluates
longer-term loading (e.g., monthly, annual) and waterbody response might be appropriate
(e.g., mass balance, simple watershed model), and a model simulating daily loading in
response to precipitation and runoff would not be necessary.
Another consideration for approach selection regarding sources is how well they and their
impacts are understood. The watershed characterization activities discussed in Section 4.2
should provide a general understanding of what and how sources are affecting
impairment. This can help to identify the type of information that the technical approach
will need to include and also produce, thereby narrowing the range of approach options.
If sources and their impacts (to both immediate and downstream impaired segments) are
well understood based on available data and local knowledge, it might not be necessary
to use an approach that evaluates individual sources, provides the ability to predict effects
from existing and future source inputs, or simulates the routing of source loads into
upstream waterbody segments and through to downstream segments. Some approaches,
such as receiving water modeling, can potentially provide a great deal of information on
how known sources will affect receiving water quality but will not provide much
information on unknown sources. Watershed modeling, on the other hand, can help to
quantify the relative significance of various sources such as urban runoff compared to
point source discharges and often include in-stream routing to evaluate the effect of
upstream sources on downstream segments. Non-modeling approaches, such as load
duration curves, that rely on evaluation of in-stream loads based on monitoring data do
not typically support direct calculation of loads originating from individual sources.
However, these types of approaches can still be meaningful and useful applied within a
watershed context when sources are well understood or their application captures the
conditions at key locations in the watershed.
In addition to understanding the individual sources, it is important to evaluate the
distribution and location of all the major sources. Understanding the relationship among
sources in regard to their individual and cumulative impact on impairments and impaired
segments can support selection of the TMDL approach. For example, if sources
throughout the watershed are fairly homogenous it might not be likely that single sources
are affecting a great impact throughout multiple segments. In this case it might not be
necessary to use an approach that relies of detailed and quantitative evaluation of source
DRAFT 55
-------
Handbook for Developing Watershed TMDLs December 2008
loads and assessing upstream-downstream effects. However, if there are a few sources
that are expected to have a significant impact on not only the impaired segments to which
they immediately drain but also on those impaired segments located downstream, it might
be warranted to use a detailed watershed model that can simulate the upstream-to-
downstream impacts of sources loads.
How Are the Sources, Multiple Impaired Waterbodies or Impairments
Related?
This consideration builds on previous questions, evaluating j
the comprehensive effects and relationships among the ^' j
multiple sources, impaired segments and impairments. Selecting an Approach: Questions Related j
to Overall System Behavior j
Similar to the relative impact of sources and how to ., , , . , ., , I
. . f , „, ,^T . . . • At what scale should sources be i
represent that in a watershed TMDL approach, it is evaluated and allocations be established? I
importanttoconsiderhowtocapturetherelatonship . What sources affectwhich impairments? |
among sources, impaired waterbodies and the observed I
impairments when selecting an approach. This can affect " Which imPairments are similar or related? |
both the necessary spatial scale of the approach and its j
application and also the processes that should be considered. Typically, watershed
TMDLs involve identification of key "assessment locations" throughout the watershed to
allow for calculation of allocations for each impaired segment. This can include division
of the watershed into subwatersheds for modeling or individual application of non-
modeling approaches such as load duration curves at multiple locations. The need for and
location of the assessment locations can also be affected by the location of watershed
sources. For example, if multiple sources impact an impaired segment it might be helpful
to establish multiple assessment points to isolate areas dominated by certain sources.
Evaluating multiple impaired segments and possibly multiple pollutants and related
parameters for watershed TMDLs might depend on the ability to consider the relationship
among and impacts of individual segments and pollutants on other segments and
waterbody conditions. For example, in watersheds where it is expected that there exists a
common source(s) that is contributing to the impairment in multiple impaired segments
or by multiple pollutants, it is useful to select an approach that can address the impact of
and routing of loads from one segment to another and also the fate and transport of
multiple inter-related pollutants.
Similarly, in cases where multiple impairments are related or even dependent on one
another it is appropriate to select an approach that can consider all of the necessary
pollutants and resulting waterbody impacts. For example, if the impaired segments are
impaired by a variety of metals and pH due to mining activities, it would be helpful to
select an approach that can account for the metals loadings and resulting waterbody
impacts to concentrations and pH but also the waterbody conditions that can affect how
metals behave and impact the waterbody (e.g., temperature, hardness).
4.3.2. Practical Applications of Various Approaches for Watershed TMDL
Development
Most approaches used for single-segment TMDL development can also be applied for
watershed TMDLs. However, the extent to which benefits or efficiencies are recognized
can depend on the type of approach and how it is applied. For example, some approaches,
such as watershed modeling, are able to simulate the relative magnitudes and relationship
56 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
among sources and their impact on immediate and
downstream impaired waters, allowing for more flexibility
in setting and prioritizing allocations. Alternatively, non-
modeling approaches that do not directly simulate the
hydrologic network or relative impacts of sources still
benefit from the comprehensive watershed approach for such
things as stakeholder involvement, data analysis, source
evaluations and implementation planning.
Process 7ip:
Documenting Selection and Application of
TMDL Development Approach
Regardless of the approach(es) selected to
develop watershed TMDLs, it is important
to thoroughly document the rationale for
using the approach and all assumptions
and supporting data for its application.
Documenting decisions throughout the
process will facilitate preparation of the
TMDL report and administrative record and
will provide a foundation for explaining the
process to stakeholders.
This section introduces some commonly used approaches for
TMDL development and describes the considerations for
their use in watershed TMDL development. The approaches
are generally categorized as modeling approaches and non-
modeling approaches. Modeling approaches include
watershed modeling and receiving water modeling, while non-modeling approaches
include a variety of approaches that depend on calculation of loading capacity using
monitoring data, empirical approaches or literature values.
Modeling Approaches
Directly simulating the upstream-downstream effects of source loading and to support a
top-down analysis of loading assessments is typically accomplished through dynamic
modeling. With dynamic models, practitioners can track the fate of pollutant loads
transported downstream from subwatershed to subwatershed. This section introduces
some of the commonly used modeling tools applied to watershed TMDL development
and provides information on where to obtain additional details about their use.
Watershed Models
Many TMDLs use watershed models to evaluate existing and allowable pollutant loads to
identify allocations, load reductions and management scenarios. A primary advantage of
developing a watershed TMDL using a watershed model is the ability to consider the
entire watershed and use a "top-down" method of assessing loading and determining
allocations. This maximizes the allocations and fully considers the relative impact of the
various sources.
Definition:
The Top-Down Approach
Watershed models emphasize description of watershed hydrology and water quality,
including runoff, erosion, and washoff of sediment and pollutants. Some models
simulate only the land-based processes while some can
also include linked river segments and simulate in-stream
transport and water quality processes. Watershed models
vary in the level of detail, including what processes they
simulate and the simulation timestep (e.g., daily vs.
monthly). The complexity of watershed models can range
from the use of loading functions—empirically based
estimates of load based on generalized meteorologic
factors (e.g., precipitation, temperature)—to physically
based simulations—scientifically based equations to
represent the physical, chemical, and biological processes
associated with runoff, pollutant accumulation and
washoff, and sediment detachment and transport.
In developing watershed TMDLs, the top-
down approach consists of evaluating sources
from upstream to downstream in watersheds
with multiple impaired segments. Doing so can
maximize the water quality benefits throughout
the watershed of strategically applied
reductions at upstream locations that impact
multiple downstream impaired segments. The
approach provides flexibility in targeting
source controls and avoids the potential for
redundant or overly stringent reductions
resulting from individual analyses.
DRAFT
57
-------
Handbook for Developing Watershed TMDLs
December 2008
The primary reason, and benefit, for applying a
watershed model for TMDL development typically
is the ability to predict pollutant generation from
varying land uses and from multiple subwatersheds.
The level of detail for predicting spatial loading and
then the fate and in-stream transport of the loads
depends on the type of model used and the way in
which it is setup and applied to the watershed.
Several considerations for selecting a watershed
model for use in TMDL development and in how to
apply it include:
Resources:
Watershed Water Modeling
For more information on watershed models used in
TMDL development and their applicability to different
waterbody types, pollutants and sources, refer to the
following EPA references:
Compendium of Tools for Watershed Assessment
and TMDL Development—
www.epa.qov/owow/tmdl/comptool.html
TMDL Model Evaluation and Research Needs—
www.epa.gov/nrmrl/pubs/600r05149/600r05149.htm
• Multiple impaired segments: Watershed
models with an in-stream routing component (e.g., HSPF, LSPC, SWAT) allow for
the evaluation of pollutant loading and transport through a segmented network of
connected waterbodies. This allows for the evaluation of multiple impaired segments
and the impact of upstream controls on downstream segments, allowing for more
efficient reductions throughout the watershed to address multiple segments.
• Variable sources and land uses: Watershed models can include land uses or
individual sources as separate inputs with unique characteristics that define their
related pollutant generation and delivery. This allows for the evaluation of the
relative magnitude of various sources in a watershed.
• Multiple subwatersheds: Related to the evaluation of multiple waterbody segments
and multiple land uses, watershed models can be set up to simulate multiple
connected subwatersheds to capture the spatial variations in source inputs. This can
target evaluations to specific impaired segments or areas of increased/concentrated
source activity while still evaluating sources and water quality on a watershed-wide
scale. However, those without in-stream routing do not simulate the transport of
loads through a system, from one subwatershed to the next.
• Evaluation of dynamic processes and relationships: An important capability of
watershed models is the ability to simulate dynamic relationships and processes that
affect receiving water quality. For example, some models have the capability to
simulate dissolved oxygen and the related processes and parameters, including
available nutrient loads, algal processes, temperature, and sediment and biological
oxygen demand. Additional examples include simulation of delivery and transport of
pollutants adsorbed to sediment particles and simulation of die-off, settling,
resuspension and other waterbody inputs and losses. The type and level of detail in
these capabilities vary among watershed models.
• Time-variable processes and conditions: Some watershed models provide the
capability to simulate time-variable watershed conditions affecting hydrology and
pollutant transport. For example, some more detailed models include parameters that
represent pollutant accumulation and can vary monthly. Another example are models
that include modules to simulate snowfall and snowmelt.
These capabilities also present some considerations for setting up and applying the
watershed model for TMDL development. The main consideration is the scale of the
analysis, including the number of waterbody segments, number and size of
subwatersheds, and the type and number of land uses. Using a watershed model for
TMDL development typically involves representing the watershed as a system of
58
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
connected waterbodies and their associated subwatersheds. A TMDL practitioner has
flexibility in how to establish the network, whether using a coarser representation
focusing on major river segments and tributaries or a more detailed division capturing
major segments as well as smaller tributaries. However, the network should represent a
balance between detail and usefulness. More detail (i.e., more segments and
subwatersheds) will require more effort in the model setup and the processing and
analysis of modeling results. The system configuration should be detailed enough to
capture the major segments and isolate areas representing spatially diverse sources, while
providing a scale that doesn't overburden the analysis. The system should also be
designed to isolate the individual impaired segments within the network to facilitate
identification of allocations for each impaired segment. The number of subwatersheds
will be determined by the network configuration, with each waterbody segment having an
associated subwatershed.
These level-of-detail considerations also extend to the representation of land uses and
sources in the watershed model. Depending on the type of land use information that is
being incorporated into the model, a watershed might have dozens of land use types.
However, many of these land use types are similar, particularly when considering their
related characteristics for pollutant generation and delivery. For example, a land use
coverage might include multiple categories for residential land uses (e.g., single-family,
low-density, high-density). The sources of pollutants for these categories are likely the
same, with the main difference impacting the pollutant generation and delivery being the
amount of impervious cover associated with each category. For a modeling analysis, it
might be appropriate to consolidate all of the residential categories into one category and
use an average value of imperviousness for its representation in the model. Eliminating
multiple, similar land uses will promote more efficient model application and analysis.
The pollutant of concern can also affect how the land uses in a watershed are evaluated
within a model. The analysis should isolate land uses that represent different sources of
the particular pollutant. For example, agriculture land uses might be represented by
several subcategories (e.g., cropland, rangeland, pastures). For a watershed TMDL
addressing metals loading, it would not be necessary to represent these subcategories, but
rather group them into a general "agriculture" category. However, if the TMDL is
addressing bacteria or nutrients, agricultural areas likely represent the primary sources
and it might be meaningful to evaluate the subcategories because they represent distinct
types of sources and delivery pathways.
Also affecting how a model is set up and applied for
watershed TMDL development are the important
processes affecting the pollutants, impairments and
sources of concern. If the data analysis has indicated
that sediment and nutrient impairments could be
related, it might be appropriate to choose and apply a
model that can simulate the adsorption of pollutants
to sediment. Another example is the need to simulate
and predict dissolved oxygen levels resulting from a
combination of processes and conditions.
Process Tip:
Focusing and Maximizing Modeling Efforts
Using a watershed model can provide significant
benefits to a watershed TMDL analysis but can also
add a level of complexity and potentially time and
effort. As with other tasks, it is important to use the
available data to focus the efforts and maximize the
use of resources. Rely on the conclusions of the
Watershed Characterization to identify the level of
detail (both spatially and temporally) necessary for
the modeling analysis to represent sources,
waterbody segments and in-stream conditions for
effective TMDL development.
DRAFT
59
-------
Handbook for Developing Watershed TMDLs December 2008
While these considerations are relevant to most watershed modeling applications, some
watershed models commonly used for TMDL development have unique considerations,
advantages and disadvantages, as summarized in Table 4-2. As discussed in the box
below, EPA's TMDL Model Evaluation and Research Needs (USEPA 2005b) rates
models in a number of categories that can affect their use in developing watershed
TMDLs (e.g., pollutants, waterbody types, land uses, in-stream routing). This provides a
consistent rating system for all models to allow a TMDL practitioner to compare the
range of simulation needs across any number of models. Alternatively, Table 4-2
provides more qualitative descriptions of a number of watershed models. The table
identifies considerations for the models, such as their ability or inability to route flow and
loads from one subwatershed to the next, the pollutants they simulate or the land uses and
source processes they include, that can determine their applicability to developing a
watershed TMDL, both in general and on a project-by-project basis. For example, most
watershed models include some level of in-stream routing for connecting the waterbody
segments for each simulated subwatershed. However, simpler models such as GWLF do
not include in-stream routing although it can be applied for multiple subwatersheds and
land uses. Alternatively, detailed models such as HSPF and LSPC not only include in-
stream routing but also include a full water quality component that can simulate chemical
and biological processes within receiving water segments and simulate the fate and
transport of flow and loads from one segment to the next. Table 4-3 identifies some
example watershed TMDLs developed using the models discussed in Table 4-2.
Which Model is the Right Model?
EPA and other agencies have developed a number of resources discussing the applicability of models for water
resource and environmental management. There are two documents that specifically discuss the applicability of
models to TMDL development and watershed assessment. Those documents include
• Compendium of Tools for Watershed Assessment and TMDL Development (USEPA 1997a,
www.epa.qov/owow/tmdl/comptool.html)
• TMDL Model Evaluation and Research Needs (USEPA 2005b,
www.epa.gov/nrmrl/pubs/600r05149/600r05149.htm
Both of these documents provide detailed information on the capabilities, limitations, appropriate uses, and input
and resource requirements of a wide-variety of models. The information contained in these documents can
provide useful insight into the appropriateness of a given model for watershed TMDL development. For example,
TMDL Model Evaluation and Research Needs provides a series of tables rating the capabilities or applicability of
more than 60 available watershed and receiving water models in the following categories:
• TMDL Endpoints. Considers the model's ability to simulate typical TMDL target pollutants (e.g., nutrients,
toxics, bacteria) and expressions (e.g., load vs. concentration). Characterizes the models depending on the
timestep of the simulation for the target—steady state, storm event, annual, daily or hourly.
• General Land and Water Features. Rates models according to their ability to simulate general land uses
(e.g., urban, agricultural) and waterbody types (e.g., river, lake, estuary).
• Special Land Processes. Rates models on their ability to simulate more than 15 special land processes such
as wetlands, hydrologic modification, urban BMPs (e.g., street sweeping, detention ponds) and rural BMPs
(e.g., nutrient control practices, irrigation practices).
• Special Water Processes. Rates models on their ability to simulate special processes occurring in receiving
waterbodies such as air deposition, stream bank erosion, algae and fish.
• Application Considerations. Rates models on the following practical considerations affecting their
application—experience required, time needed for application, data needs, support available, software tools
and cost.
60 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Table 4-2. Watershed Models Commonly Used for TMDL Development
Model
Considerations
AGNPS/
AnnAGNPS
Is a distributed model, providing information on impacts at various locations in a watershed
Originally designed to evaluate agricultural management practices but can also be applied to
mixed-land-use watersheds.
Simulates only sediment, nutrients and pesticides
Includes simplified stream routing to allow for multiple drainage areas with runoff and pollutants
routed to downstream areas
Includes special components to handle concentrated sources of nutrients (e.g., feedlots, point
sources), sediment (e.g., gullies) and added water (e.g., irrigation)
Can be used to simulate effect of agricultural BMPs (e.g., ponds, irrigation, tile drainage, vegetative
filter strips, riparian buffers)
Is appropriate for use on watersheds up to 200 mi2
Can simulate precipitation-driven sources and direct discharge sources (i.e., point sources)
GWLF
Designed to simulate mixed land use watersheds
Simulates sediment, nitrogen and phosphorus
Typically used to evaluate long-term loading; provides monthly output
Does not include stream routing in the original version
Can simulate precipitation-driven sources, direct discharge sources (i.e., point sources) and inputs
from septic systems
Requires a low level of expertise
Appropriate for situations of limited data for calibration
Requires combination with a receiving water model to directly evaluate water quality impacts
HSPF
Simulates watershed runoff and hydrology, pollutant buildup and washoff processes, and in-stream
water quality and routing
Appropriate for all types of land uses; distinguishes pervious and impervious areas
Simulates wide range of conventional and toxic organic pollutants and sediment
Simulates on a daily, hourly, or sub-hourly timestep, capturing variable flow conditions
Can simulate snowfall/snowmelt processes
Includes agricultural components for land-based nutrient and pesticide processes
Includes a special actions block for simulating management activities
Has a number of modules representing different processes, allowing flexibility in how simple or
complex the model is set up
Can simulate precipitation-driven sources and direct discharge sources (i.e., point sources)
Requires extensive calibration
Requires experience or training to setup and apply and understand model assumptions
Can be applied for streams/rivers and well-mixed reservoirs
LSPC
Similar considerations as HSPF, but with the following additions or enhancements:
• Simulates multiple modules simultaneously at each timestep, allowing for simulating dynamic
interaction between land and stream modules for representing processes such as irrigation,
BMP/impoundment interaction with the landscape, and other more complex water routing
configurations.
• Has no inherent limitations in terms of modeling size or model operations; is capable of simulating
thousands of subwatersheds in a single file, as needed
• Is designed to accommodate large-scale model development in areas with a high degree of spatial
resolution (especially beneficial in mountainous areas)
• Provides post-processing and analytical tools designed specifically to support TMDL development
and reporting requirements
• Has data management tools to facilitate evaluation of multiple watersheds simultaneously
• Includes specialized MDAS (Mining Data Analysis System) module for simulating pH associated
with mining activities
• Simulates time-variable land use change (e.g., urbanization of a watershed, forest fires, harvesting
and regrowth)
DRAFT
61
-------
Handbook for Developing Watershed TMDLs
December 2008
Model
Considerations
SWAT
Simulates watershed hydrology, sediment and water quality
Can network multiple subwatersheds and representative streams
Division within subwatersheds is based on Hydrologic Response Units (HRUs), unique
combinations of soil and vegetation type, rather than land use
Appropriate for streams
Is suitable for watersheds from small to very large in size
Simulates agricultural practices (e.g., planting, tillage, irrigation, fertilization, pesticide management,
grazing and harvesting)
Simulates in-stream biological and nutrient processes, including algal growth, death and settling
SWMM
Is primarily applied to urban areas with impervious drainage
Was originally developed for the analysis of surface runoff and flow routing through complex urban
sewer systems
Simulates watershed hydrology and water quality
Can network multiple subwatersheds and representative streams
Can be applied for single-event or long-term simulations
Provides flexibility in simulation of quality through buildup/washoff, rating curves or regression
techniques, providing varying levels of data input needs
Simulates pollutants typically associated with urban areas (nutrients, metals, sediment, pathogens)
Can simulate storage, treatment and other BMPs
Table 4-3. Example Watershed TMDLs Developed with Commonly Used Watershed Models
Model
AGNPS/
AnnAGNPS
GWLF
HSPF
LSPC
SWAT
SWMM
Example Watershed TMDLs
• Zorinsky Lake, NE (siltation, nutrients and organic enrichment/low dissolved oxygen):
www.deq. state. ne.us/Publica.nsf/0/113b45ab14a8f68206256be5005874ff/$FILE/zorinskv-tmdl. PDF
• Carbury Dam, ND (nutrients, sediment and dissolved oxygen):
http://www.health.state.nd.us/WQ/SW/Z2 TMDL/TMDLs Completed/Carburv%20Dam%20Final%20
TMDL%202006111 3.pdf
• Con odoguinet Creek Watershed, PA (sediment and phosphorus):
www.dep. state. pa. us/dep/deputate/watermat/wap/wastandards/tmdl/Conodoauinet TMDL.pdf
• Wissahickon Creek, PA (sediment and nutrients):
www.epa.qov/req3wapd/tmdl/pa tmdl/wissahickon/index.htm
• Little Beaver Creek watershed, OH (phosphorus, ammonia, dissolved oxygen, fecal coliform,
siltation): www.epa.state.oh.us/dsw/tmdl/LittleBeaverCreekTMDL.html
• Suwannee River Basin, GA (dissolved oxygen):
www.qaepd.orq/Files PDF/techquide/wpb/TMDL/Suwannee/FinalSuwanneeDOTMDLs.pdf
• Ballona Creek, CA (metals):
www.waterboards.ca.qov/losanqeles/board decisions/basin plan amendments/technical document
s/bpa_28_2005-007_td.shtml
• Cahaba River watershed, AL (nutrients):
www.adem.state.al.usAA/aterDivision/WQualitv/TMDL/CRNutTMDL.pdf
• Los Angeles River, CA (metals):
www.waterboards.ca.qov/losanqeles/water issues/proqrams/tmdl/tmdl list.shtml
• Mahoninq River, OH (fecal coliform): www.epa. state. oh. us/dsw/tmdl/MahoninqRiverTMDL. html
• Huron River, OH (sediment, nutrients): www.epa. state. oh. us/dsw/tmdl/HuronRiverTMDL. html
• Pajaro River, CA (sediment):
www.waterboards.ca.qov/centralcoast/TMDL/303dandTMDLproiects.htm
• Winter Haven Southern Chain of Lakes, FL (nutrients):
www.epa.qov/reqion04/water/tmdl/florida/documentsAA/Haven chainlakes Nutrient TMDL 001.pdf
62
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Receiving Water Models
In some cases, receiving water models are used to i j „ f ]
support TMDL development, either alone or in | j ' j
combination with a watershed model. Receiving 11 Receiving Water Modeling j
water models differ from watershed models in that II . f ±. , , , , ,. i
j , ^ ,.i. .j . . . II For more information on watershed models used in i
they only represent conditons within a receiving j | TMDL deve|opment and their applicability to different !
water, such as a stream or reservoir. Receiving water |j waterbody types, pollutants and sources, refer to the |
models simulate conditions within a receiving || following EPA references: j
waterbody (e.g., lake, stream, estuary) based on II I
representation of physical, chemical and biological || <%ffiff^^""**"1 A88e88mmt \
processes. Inputs to the waterbody are often defined j j www.epa.aov/owow/tmdl/comptool.html |
as boundary conditions or developed using linked j j j
dynamic output from a watershed model. Receiving i! TMDL Model Evaluation and Research Needs— I
water models are typically either steady-state or (| www.epa.qov/nrmrl/pubs/600r05149/600r05149.htm |
dynamic models. Steady-state models operate under " j
a single nonvariable flow condition with constant inputs, typically used to evaluate
conditions for a design or critical flow. Dynamic models allow for variations in both flow
and meteorologic conditions on a small timestep, typically shorter than daily. Level of
complexity in receiving water models is also determined by spatial detail described as
one, two or three dimensions.
While the capabilities and applicability of receiving water models can vary widely, they
generally provide the capability of simulating a number of water column processes
unavailable in many watershed models, such as the effects of nutrient cycles on dissolved
oxygen. For TMDLs addressing waterbodies with complex biological and chemical
processes impacting water quality and associated impairments, receiving water models
can more accurately evaluate the allowable pollutant loads. Because they can represent a
stream, lake, or estuary using numerous analytical elements - in some cases tens of
thousands - model predictions can be very accurate at many locations along the length of
a receiving water. With this capability, receiving water models can be used to most
accurately determine the specific amount of a contaminant that can enter a receiving
water at different locations while still achieving water quality criteria throughout. Thus,
allocations can potentially be made at the very detailed spatial level. However, they do
not explicitly represent land-based contributions, which are typically addressed through
designation of boundary conditions (often based on monitoring data) or through
development of a separate watershed model. Therefore, when applied alone, the
allowable loads calculated at different points represent a cumulative load entering the
waterbody at that point. Because of this, it is important to understand the watershed and
related sources and impairments when using a receiving water model for supporting
watershed TMDLs independently from a watershed model. When applied alone, they do
not explicitly represent land-based sources and a separate analysis might be needed to
allocate the allowable load among land-based sources within the drainage area.
Non-modeling Approaches
Sometimes there is a misperception that developing watershed TMDLs requires the use
of complex models. However, it is not always feasible or necessary to use watershed or
receiving water models. Using a watershed framework to develop TMDLs accommodates
various non-modeling approaches while still providing many of the same benefits. Non-
modeling approaches to TMDL development are typically based on statistical analysis of
ambient data or on an empirical calculation representing land-based processes. As
DRAFT 63
-------
Handbook for Developing Watershed TMDLs December 2008
compared to modeling approaches they typically include a _ _, i
more simplified representation of watershed and receiving ^' j
water processes. However, they typically require less effort, Considerations When Selecting and Using j
time and experience to apply and can often be more easily Non-Modeling Approaches for Watershed j
communicated to the public. While non-modeling ™DL Development j
approaches, such as load duration curves, statistical . Ab t() eva|uate and quantjfy re|atjve |
approaches or mass balance analyses, might not magnitude and impact of multiple !
quantitatively track the transport of loads as a model can, the sources |
TMDL still involves a thorough data analysis and source . Ability to evaluate the impact of and |
evaluation to identify critical loading conditions for relationship among multiple pollutants j
significant sources in the watershed and helps to identify key . Ability to link hydrologic segments or !
areas for management. Even though the analysis does not consider in-stream fate and transport j
employ a quantitative link between various sources and • Ability to capture spatial variability in !
segments, by understanding the contributions and impacts of waterbody conditions and source loading j
all sources in the watershed, the analysis is still holistic. • Ability to evaluate time-variable |
processes and conditions j
When applying non-modeling approaches for watershed j
TMDL development, many of the same considerations are
important as when using a modeling approach. The most important of these are
representation of the multiple pollutants and sources, level of spatial and temporal detail,
and the considerations of important processes affecting impairment. For example,
regardless of the approach used it is important to evaluate the appropriate temporal and
spatial scale for application to evaluate the important sources, capture the conditions of
impairment and allow for comparison to applicable water quality criteria or TMDL
targets. For example, when using a load duration approach, the in-stream analysis of flow
and water quality can be conducted at any point in the watershed with sufficient data.
Evaluating such characteristics as the location of impaired segments, locations of key
sources, land use distribution, and data availability to capture variable in-stream
conditions can support selection of assessment points for conducting the analysis and
ultimately setting allocations.
Within the non-modeling categories, there are various types of approaches, all of which
can be further characterized based on the type of simulation or calculation they perform.
The approaches either calculate land-based loads or the resulting waterbody loads. The
"land-based" approaches, such as using export coefficients or the Simple Method,
calculate loading from land-based runoff processes assuming some measure of
precipitation and characteristics representative of the watershed (e.g., soils,
imperviousness). The "waterbody-based" approaches, such as load duration curves or
mass-balance analyses, calculate the "delivered" load in the waterbody based on
waterbody conditions, either using observed monitoring data (i.e., concentration and
flow) or assuming some user-defined load inputs and outputs. Many of these approaches
are applied in combination to represent both source loading and waterbody response to
develop TMDLs. Some examples of non-modeling approaches are discussed below.
Load Duration
The load duration methodology (USEPA 2007c) relies on using observed flows and water
quality criteria to establish a curve of loading capacities for various flow conditions. This
builds on using flow duration curves, which use hydrologic data to evaluate the
cumulative frequency of historic flow data over a specified period. A water quality
criterion or other target concentration can then be multiplied by the observed flows to
create a curve representing the distribution of allowable loads as a function of daily flow,
64 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
representing the loading capacity of the stream. The entire curve can be used to represent
flow-variable loading capacities or allowable loads can be identified for specified flow
intervals, used as a general indicator of hydrologic condition (e.g., wet versus dry and to
what degree).
Because the method identifies the allowable and existing loads for all flow conditions, it
provides insight into the critical conditions and inherently accounts for the natural
variations in loading and in-stream conditions. However, because the methodology is
based on observed in-stream conditions, it provides limited information regarding the
relative magnitude of source loads and requires a supplemental analysis to distribute the
calculated loading capacity into source-based allocations. This does not typically present
a problem in watersheds where the sources are well understood. When using the approach
for watershed TMDLs, one consideration is the need for robust and consistent records of
flow and in-stream water quality data. Because it is necessary to identify allocations for
each impaired segment, it would be likely that the load duration approach be applied for
each impaired segment. However, sometimes data, especially flow, are not sufficient to
perform the analysis. In this case, assumptions are usually made to extrapolate flow
and/or water quality conditions from nearby segments to support TMDL calculation for
the impaired segments. In addition, load duration curves for TMDL development are
applicable only to non-tidal streams or rivers and might not be appropriate or would
require combination with other approaches in watersheds with different types of impaired
waterbodies (e.g., lakes and streams).
Using a Load Duration Approach and Field Work to Focus Resources
in the Duchesne River Watershed TMDL, Utah
Several segments in the Duchesne River watershed, Utah, are listed as impaired by TDS due to elevated in-
stream salinity, the result of natural geology and years of hydromodification for irrigation. TMDLs for the
Duchesne River watershed were calculated using the load duration approach, a statistical method relating
pollutant loads to the frequency of observed in-stream flows. The load duration approach uses flow with observed
TDS data to estimate existing loads and with the TDS TMDL target to estimate allowable loads over a range of
flow percentiles. The approach was applied at representative monitoring stations for each impaired segment and
provided information on water quality and flow conditions to support TMDL calculation.
To supplement the analysis, a series of watershed surveys and a variety of statistical analyses on historical data
were conducted to better understand sources and water quality conditions. During the watershed surveys the
types, locations, and severity of sources expected to contribute to water quality impairment were noted. Two
visual, screening level assessments of TDS sources were conducted throughout the watershed and included
photo documentation, global positioning system (GPS) locational indexing, and narrative descriptions of current
and potential sources of water quality impairment in listed segments. Each of the listed segments was surveyed
from available access points and road networks, and relevant features were documented. Obvious water quality
impairments associated with the identified sources were noted (e.g., streambank erosion and destabilization,
dewatered stream channels, natural sources).
Statistic analyses were also performed to evaluate the water quality conditions in the streams, particularly
evaluation of background and natural conditions. Evaluation of historical data and local historical observations
and anecdotal information suggested that natural conditions in two of the listed segments were in exceedance of
water quality criteria. For these segments, site-specific TDS criteria were calculated to reflect the naturally
elevated TDS.
Because the multijurisidictional watershed includes federal, state and tribal lands, an extensive stakeholder
coordination process supported the TMDL process. Stakeholder involvement facilitated increased access to
streams during the field surveys, improved historical and local knowledge of sources and conditions, and
consensus on decisions regarding allocations and site-specific criteria.
DRAFT 65
-------
Handbook for Developing Watershed TMDLs December 2008
Steady-State or Mass-Balance Analysis
The steady-state and mass-balance type approaches rely on the assumption of
conservation of mass into a waterbody. The analysis might calculate loads entering a
waterbody using export coefficients or observed data and calculate the resulting
waterbody concentrations on the basis of simple representation of pollutant fate and
transport, typically including estimated losses (e.g., settling, decay) and inputs. The
approach relies on identifying the necessary loads entering a waterbody that will meet the
desired waterbody target after the consideration of all inputs and losses.
Applying a mass balance or steady-state analysis to calculate a loading capacity relies on
some calculation of the incoming existing pollutant load. This might be done by "back-
calculating" an existing load based on observed concentrations and stream flow (or
volume for lakes and reservoirs) and accounting for any expected losses (e.g., die-off,
settling). If this is the case, the existing load represents a cumulative load from all of the
sources contributing the pollutant of concern to that segment. In other cases source load
inputs can be calculated using observed data (e.g., monitoring data for tributaries, DMR
data for point sources). Assuming there are no data available to directly calculate loads
from individual sources, the most likely approach for distributing the total load into
source allocations will be to use some measure of source area (e.g., percent land use
area), not accounting for natural variations in loading from different sources. Another
option for calculating the existing load for a mass-balance analysis is the use of export
coefficients. Similar to the Percent Reduction method, this would produce source-specific
loads that could be applied a reduction to calculate allocations that meet the overall
loading capacity. However, the loads are typically based on literature values representing
longer-term loading (e.g., monthly, annual).
In the context of a watershed TMDL, the steady-state or mass-balance analysis can
accommodate inputs of multiple sources. There is also the capability, although simplified,
to route output of flow and load from one segment to the next, assuming some constant
input and losses. However, the analysis is typically a static calculation, evaluating either a
critical condition (e.g., design flow) or a longer-term average condition (e.g., average
monthly loading). This can limit the ability to evaluate different types of sources and
variability in source behavior (especially precipitation-driven sources) and impairment
conditions. As with all approaches, the technical needs of the watershed TMDL analysis
should be weighed against the necessary level of detail to develop effective allocations.
Percent Reduction
The percent reduction method assumes a 1:1 relationship between surface water
concentrations and pollutant loading to calculate a loading capacity. The existing
pollutant concentrations are compared to applicable water quality criteria to calculate a
necessary reduction. This reduction is then applied to an estimate of "existing" loading to
calculate the loading capacity. Existing loads are often calculated using ambient
monitoring data (e.g., concentration and flow) or some estimation of land-based loading
(e.g., export coefficients). While this methodology is easily applied and requires little
effort, it also provides little to no information on watershed sources and does not provide
the ability to evaluate the relationship among multiple impairments and impaired
segments. The approach uses a static calculation, focusing on a critical condition (e.g.,
maximum concentration) or a long-term average condition, not allowing for evaluation of
variability in loading or waterbody response.
66 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Using a Mass Balance Analysis to Develop PCB TMDLs in the Shenandoah River Watershed, VA/WV
Several segments of the Shenandoah River were identified on Virginia's and West Virginia's 1998 303(d) lists as
impaired due to fish consumption advisories for PCBs. A TMDL was developed to address the PCB impairments
using a mass balance approach. Based on the data availability for the river, a one-dimensional, steady-state,
plug-flow system was developed to represent the linkage between PCB sources in the Shenandoah watershed
and the in-stream response. The Shenandoah River was segmented into a series of plug-flow reactors (defined
along the entire length of the impaired segment) to simulate a steady-state distribution of PCBs. This approach
accounted for the water balance between each segment and the impact of point sources and tributaries to the
main stem of the Shenandoah River. Each of the plug-flow reactors defined a mass balance for PCBs for the
sediment-water system. PCBs in the water column and sediment layers were computed as concentration profiles
with respect to distance.
^xjUFDik
RJVET
\fj-in S-PI-I-I
SbELLi
Ri'.ier
USGSDlSlMi
HzpgyQwk-
Wimc'!:; _ r •
Rt-t
s = length.
C =
£p = len*tli
I :-EIL^'ir
c, = c-^areatratioa: a ±e pluj flmv
Tr.':oar--
The Shenandoah River Watershed PCB TMDL is available online at
http://www.epa.gov/req3wapd/tmdl/wv tmdl/Shenandoah/index.htm
Export Coefficients/Pollutant Budgets
This category encompasses a number of approaches built on empirical relationships
among watershed processes and pollutant loading as well as the use of literature values of
typical watershed loading rates. Examples include using monthly load rates from various
land uses to calculate allowable loading from an impaired watershed. Another example is
using an empirical relationship that allows a user to calculate an allowable load
depending on desirable conditions (e.g., target runoff/waterbody concentration or
indicator levels). An example of such an analysis is using the Vollenweider approach for
identifying an allowable phosphorus loading rate for a lake to meet a desired trophic level
and based on such lake characteristics as depth.
Export coefficients are measures of typical loading rates from certain land uses or
sources. Used within a TMDL analysis they would typically be used to calculate existing
loads based on the land use distribution in a watershed and would often be combined with
a supplementary approach that calculates an allowable load based on waterbody targets
(e.g., percent reduction, mass balance). Export coefficients can be obtained for literature
DRAFT
67
-------
Handbook for Developing Watershed TMDLs December 2008
values from regional or national studies (e.g., EPA's Nationwide Urban Runoff Program
[NURP] study [USEPA 1983]) or based on site-specific sampling of runoff from
individual land uses. Issues or cautions in using export coefficients in watershed TMDLs
are the same as when using them in all TMDLs. The TMDL practitioner should evaluate
the applicability of the coefficients to the watershed of the impaired waterbodies and
decide whether they are representative and appropriate. Because export coefficients only
represent loading rates from designated land uses, direct-input sources or those that are
not precipitation-driven will not be accounted for in the analysis. In addition, export
coefficients do not evaluate or consider waterbody conditions. To provide even a simple
evaluation of waterbody response or in-stream routing, the export coefficients would be
used with some type of analysis of waterbody conditions to support calculation of a
loading capacity.
For development of a watershed TMDL, this approach provides little utility in the
evaluation of relative magnitude or impact of watershed sources, and the utility of any
evaluation of waterbody response to individual or multiple sources would depend on
what supplementary waterbody-based approach were used.
Simple Method
The Simple Method (CWP 2005, Schueler 1987) is an empirical equation used to
calculate pollutant loading based on drainage area, pollutant concentrations, a runoff
coefficient and precipitation. In the Simple Method, the amount of rainfall runoff is
assumed to be a function of the imperviousness of the contributing drainage area. When
using the Simple Method, the TMDL loading capacity would typically be calculated
using a combination approach with the percent reduction method or similar approach. As
with the other non-modeling approaches, using the Simple Method for a watershed
TMDL would require its application at key locations in the watershed to calculate a
loading capacity for each impaired segment. Because the method assumes all loading
originates on impervious surfaces during storm events, it does not account for runoff
from impervious areas or subsurface inputs and baseflow loading. Therefore it would not
be appropriate for watersheds with sources such as agricultural runoff, failing septic
systems, or direct inputs. The method also uses a static runoff concentration (e.g., event
mean concentration), not accounting for variability in loading or in-stream levels.
Because the method was originally developed to evaluate site-scale stormwater loading in
urban areas, it is not typically appropriate for large watersheds (>1 mi2) or non-urban
areas.
Integration of Different Technical and Modeling Approaches
Using a watershed approach also lends itself well to the use of multiple technical and
analytical tools to evaluate water quality conditions and loading, and to considering
equitability in making allocations. Due to the complexities and heterogeneities of systems
addressed by watershed TMDL analyses, it is sometimes helpful to combine multiple
technical and modeling approaches to develop the TMDL. Various tools can be utilized
to assist with loading estimations for different pollutant parameters. For example, while
an overall TMDL for nutrients might be developed using the Soil and Water Assessment
Tool (SWAT) model, site-specific nutrient export studies using a field scale model can be
utilized to refine the parameterization of the SWAT model. While using combination
approaches might make the analysis more difficult, it is sometimes necessary to represent
the various sources and their unique features or impacts. A watershed TMDL allows for
integrating multiple approaches for the most benefit. Because the framework allows for
68 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
maximum flexibility in applying approaches and facilitates a more comprehensive and
integrated evaluation of the watershed, it avoids added burden without added value to the
analysis.
Watershed
Characterization
Linkage
Analysis
Stakeholder Involvement and Public Participation
4.4. Allocation Analysis
As with the other tasks, developing and
evaluating allocations for a watershed TMDL follows much the same process as that of a
single-segment TMDL, but it is applied at a larger scale and likely involves more
potential allocation combinations. Generally, a TMDL's allocation analysis involves
applying load reductions to combinations of watershed sources to identify various
scenarios that meet water quality standards. The final scenario is then selected and used
to establish LAs for nonpoint sources and WLAs for permitted point sources. Developing
TMDLs on a watershed basis often provides greater flexibility in developing source
allocations. Although the level of detail might vary depending on the approach used, a
watershed TMDL considers all the sources in a watershed
and evaluates their relative magnitude to some extent.
Because of this, there will likely be a number of
allocation scenarios that will result in meeting the overall
water quality goals. Deciding which scenario to choose
can sometimes be a difficult task. Considerations to
support the decision include:
TMDL Report
' Submittal,
Scale or resolution of source allocations
Equitability or feasibility of allocations
Stakeholder priorities and implementation plans
Process 7ip:
Factors Affecting Allocation Decisions
Location and relative magnitude of sources
Pollutants of concern
Feasibility of necessary load reductions
Equitability among sources
Ongoing or planning controls
Stakeho Ider priorities
4.4.1. Scale or Resolution of Source Allocations
The scale at which an analysis considers pollutant loading from sources can affect the
identification of allocations. The spatial scale of allocations can range from a gross load
for an entire watershed to loads by land use to loads by land use by subwatershed. The
watershed TMDL approach allows for the spatial evaluation of sources and their impacts
and that should be captured in the allocations. Establishing a gross allocation for an entire
watershed defeats the purpose of conducting a watershed TMDL analysis; it ignores the
spatial variability of sources throughout the watershed and their relative magnitude and
influence. Establishing allocations at a smaller scale will likely be more meaningful and
effective.
An analysis that evaluates sources at a more detailed scale provides the greatest
flexibility in identifying and targeting source reductions to meet water quality goals. The
allocation analysis should establish loads and necessary reductions at a scale that
maximizes the benefits from the established allocations, especially when addressing
multiple impaired segments and multiple pollutants, without overburdening the analysis.
For example, when evaluating a watershed with multiple impaired segments, source
reductions can be targeted to upstream subwatersheds first because they will have an
impact on water quality in upstream and downstream segments. Evaluating the upstream -
to-downstream effects of pollutant loading is a primary benefit of developing TMDLs on
a watershed scale, particularly when using a watershed model. After identifying
necessary load reductions to meet targets in upstream impaired segments, the model can
evaluate the effect on downstream water quality. Upstream reductions can significantly
DRAFT
69
-------
Handbook for Developing Watershed TMDLs December 2008
': Program Notes:
t
," TMDL Allocations to Sources Upstream
t
In the context of a watershed TMDL, developers might wonder about possible implications of LAs or WLAs
developed for sources that are not directly discharging to a listed waterbody. To answer this question it helps to
distinguish the different types of waterbodies relative to the impaired segments for which the TMDL is to be
developed. As such, there are essentially four kinds of waterbodies for which source allocations might be
developed in a watershed TMDL:
1. Waterbodies that are included in the state's Section 303(d) list
2. Waterbodies that are not on the state's Section 303(d) list but are shown to cause and contribute to the
impairment based on data and/or observations
3. Waterbodies that are not on the state's Section 303(d) list and are not proven to cause or contribute to the
impairment
4. Waterbodies that are unassessed or not listed but for which modeling or analysis indicates that there might be
an impairment
Allocations might be required for sources that are contributing either directly or indirectly to impaired segments to
meet water quality standards in downstream impaired segments. Allocations can also be developed for sources
on unimpaired and or non-contributing waterbodies that, for implementation or other programmatic reasons, a
state might wish to identify reasonable allocations. These allocations would not be formally approved by EPA but
would be developed as part of the overall assessment of loading in the watershed. In essence, they can be
considered "for your information" allocations.
decrease or even eliminate the need for reductions in the lower subwatersheds, thereby
optimizing the necessary reductions and associated source controls.
Similarly, depending on the approach chosen, the TMDL might consider pollutant
loading down to the land use or source level. When addressing multiple impairments,
certain land uses or sources might impact the levels of multiple pollutants while others
only impact single pollutants. For example, when developing TMDLs for bacteria and
nutrients, many of the sources are the same, such as failing septic systems or livestock
related areas or activities, and targeting single sources will have benefits in reducing both
pollutants. However, some sources might only impact one of the pollutants, such as
cropland runoff that contains elevated nutrients but minimal bacteria, and therefore
targeting that source might not provide the maximum benefits in addressing both
impairments. While watershed TMDLs allow for identifying these efficiencies when
targeting source controls, the allocation analysis should evaluate the benefit of such
reductions in the context of the state's overall priorities. For example, a source might
contribute multiple pollutants but is already subject to a number of controls, making
further load reductions infeasible.
70 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Process Tips:
How to Integrate Pre-existing Approved TMDLs
The TMDL practitioner might find in some situations that TMDLs already exist (for one or more segments) within
the watershed addressed by the watershed TMDL. The practitioner has at least two options for dealing with these
pre-existing TMDLs:
• Incorporate the existing TMDL and analysis into the ongoing watershed TMDL and let the existing allocations
remain in effect, or
• Apply the new analysis to calculate new allocations that reflect the interactions of all segment within the
hydrologically linked watershed.
If it is decided to use the existing TMDL allocations, it is a good idea to make certain that the current land use,
loadings and other key assumptions used to develop the original TMDL are still representative. In some cases,
states might want to consider selecting the most stringent allocations by comparing the new calculation with the
allocations contained in the existing TMDL. If the state decides to replace the existing TMDLs, they are in effect
withdrawing and redoing the in-place TMDLs. This should be done in accordance with any state requirements and
should be clearly explained in the watershed TMDL report distributed for public review and comment.
4.4.2. Priority and Feasibility of Source Allocations
Another consideration for the development of effective allocations in a watershed TMDL
is the relative priority and feasibility of potential allocations scenarios. Whether or not
developed on a watershed basis, TMDLs typically include multiple sources, often
including both point and nonpoint sources. When establishing allocations and necessary
load reductions among various sources, the issues of equitability and feasibility often
arise. The allocation analysis can be used to evaluate a variety of possible allocation
schemes to target or prioritize source controls as appropriate. For example, a goal might
be to strike a balance among allocations and distribute necessary load reductions
equitably among sources. Alternatively, the allocations might target those sources that
represent the majority of the load input or those that are more feasible to control. For
example, some sources that already contribute a small portion of the overall load might
not be able to reduce the load any further. Other sources that represent a larger percentage
of the total load and have a greater opportunity for reductions (e.g., more land area and
delivery pathways to apply BMPs) can be targeted with larger reductions. Table 4-4
illustrates this concept, presenting two scenarios that meet the same overall allowable
load with different approaches to source-level allocations. Scenario 1 represents an
equitable distribution of load reduction among sources.
Reductions are applied so that the resulting loads are the
same percentage of the total as under existing
conditions. Alternatively, Scenario 2 represents a
scenario in which more controllable sources (e.g., roads,
cropland, pasture) are targeted to meet the load reduction _ ,., ... ... , ...
^ ^ ' ° One of the greatest benefits of evaluating
target.
Even when trying to target those sources that have the
most impact on receiving water quality and that are more
. 11 i i ., -111-1 11 1^-1 • ...i ... allocations and allows for flexibility in
controllable, there will likely be multiple scenarios that estab|ishing the fina, allocations. In addition,
result in meeting the water quality target (Figure 4-8),
and there will be case-specific challenges, priorities and
issues to consider in choosing the final TMDL
allocations. This is the case with all TMDLs, regardless
Process 7ip:
Capitalizing on the Use of a Watershed-based
Framework to Establish Accepted Allocations
sources and engaging stakeholders within a
watershed TMDL framework is realized during
the allocation stage. Dealing with multiple
sources provides increased opportunities for
stakeholders gain valuable perspective on their
allocations in the context of the watershed and
other sources when they are involved in
evaluating TMDL allocations scenarios.
of whether developed as single-segment or watershed
DRAFT 71
-------
Handbook for Developing Watershed TMDLs
December 2008
TMDLs. However, watershed TMDLs provide
more flexibility in identifying and evaluating
potential allocation schemes because there are
more sources to consider when identifying the
ultimate cumulative impact. Because the
watershed analysis framework allows for relative
comparison of sources and an evaluation of the
cumulative affect of source inputs, it can more
effectively be used to identify an appropriate
allocation scenario.
Program Notes:
Reasonable Assurance in Watershed TMDLs
Based on EPA guidance (USEPA 1991, 1997b) all
TMDLs that include a mix of point sources and nonpoint
sources, whether single-segment or watershed, should
include reasonable assurances that nonpoint source
control measures will achieve expected load reductions.
The watershed TMDL framework typically provides
greater flexibility in targeting source controls and
potential for increased stakeholder participation, thereby
facilitating identification of feasible allocation options and
supporting reasonable assurance.
Table 4-4. Examples of Different Scenarios to Meet the Same Load Target
Source
Roads
Pasture/Hay
Cropland
Forest
Landfill
Residential
Groundwater
Total
Existing
Phosphorus
Loading
78
21
218
97
7
6
111
539
Scenario 1
% Load
Reduction
26%
26%
26%
26%
26%
26%
26%
26%
Allowable Load
58
16
162
72
5
5
83
400
Scenario 2
% Load
Reduction
20%
10%
55%
0%
0%
0%
0%
26%
Allowable Load
62
19
98
97
7
6
111
400
1.400
dp dp
'
TMDL Target - Baseline - Scenario A - Scenario B - Scenario C
Percent of total TMDL load
Septics WWTP Scenario A
Septics WWTP Scenarios
4% 4%
Septics WWTP Scenario C
2% 2%
11%
Figure 4-8. Model output illustrating multiple allocation scenarios that meet water quality standards.
72
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
4.4.3. Stakeholder Priorities and Implementation Goals
Stakeholder concerns can also be a significant factor in establishing allocations for a
watershed TMDL. Ideally, major stakeholders have been involved throughout the
watershed TMDL development process and were involved in decisions regarding the
scale of the analysis and the representation of the sources in the analysis. Stakeholders
can provide valuable information in establishing and evaluating the appropriateness of
allocation scenarios, such as information on planned or expected implementation
activities and the likelihood of meeting necessary reductions (i.e., feasibility). For
example, a major source of bacteria in a watershed might be outdated and failing septic
systems. Through coordinating with stakeholders to identify and understand sources, the
TMDL practitioner learned that the city sewer is being extended to the area of the
watershed currently served by septic systems, thereby eliminating a major source over the
next few years. Because of this, a TMDL practitioner can target significant reductions to
the septic system load knowing that it will be met and therefore avoid applying
unnecessary reductions to other sources.
In addition, when stakeholders are involved throughout the TMDL process, it is more
likely that they will be involved in implementation planning. Having an understanding of
which stakeholders are more interested, willing, and able to implement control efforts as
a result of the TMDL process allows a TMDL practitioner to identify management
scenarios that are more likely to be implemented. Having stakeholders involved in the
decision-making process for establishing allocations will result in more realistic
allocations and will also open a dialogue among stakeholders that can support such
efforts as watershed-based permitting and water quality trading.
Using Stakeholder Input to Identify, Understand and Evaluate Potential Management Strategies in the Lower
Fox River Watershed, Wisconsin
Many of the BMPs that will be necessary to achieve the phosphorus and sediment reduction goals for the Lower
Fox River Basin TMDL will require voluntary cooperation from landowners. If landowners are expected to
voluntarily implement the elements of a TMDL implementation plan, the plan must not only address ecological
functions in the basin, but also consider issues that directly impact the individual. To account for this, the TMDL
for the Lower Fox River Basin includes extensive stakeholder involvement, including development of an Outreach
Committee to implement a communication and outreach strategy for TMDL development and implementation and
meet with key stakeholder groups. To facilitate more successful BMP implementation, the Lower Fox River TMDL
process also includes an evaluation of social indicators to gain a better understanding of the potential
effectiveness of BMPs that rely on outreach and behavior change. The process has included source-specific
stakeholder meetings or "focus groups" (agriculture, stormwater, dischargers) to identify BMP scenarios to
evaluate in the TMDL and optimization studies. A watershed survey of area farmers was also conducted to
identify what BMPs are currently used and why.
4.5. Development and Submittal of
TMDL Report
Watershed
Gh.im tf n/.ition
Analysis
Allocation
Analysis
lvement and Public Participation
Preparing the TMDL report and assembling the submittal package for EPA's review and
approval can be one of the most important steps in the TMDL process. No matter how in-
depth, accurate or meaningful the analysis, if it is not documented and represented in the
report and associated administrative record, EPA and the public might not understand the
TMDL and associated allocations. When dealing with a watershed TMDL, there is likely
DRAFT
73
-------
Handbook for Developing Watershed TMDLs December 2008
more information and more intricacies to the approach to document, making it important
to organize the information in a useful and understandable manner.
Special considerations for preparing for submittal of a watershed TMDL mostly relate to
the organization of the report. Often a comprehensive report for a watershed TMDL can
result in more efficient report preparation when developing supporting information for
the TMDL report that is common to all listed segments/impairments, such as maps, data
analysis graphics, and language on watershed characteristics and sources. On the other
hand, when dealing with multiple impaired segments, multiple pollutants, and a variety of
source allocations, a TMDL report can become cumbersome and confusing.
Redundancies can occur across sections that discuss similar or common information for
different areas of the watershed. It is important to decide how to present allocations to not
only meet regulatory requirements and accurately represent specific waterbody-pollutant
combinations being addressed by the TMDLs but also to support public consumption.
There are a number of ways to present individual TMDLs developed within the
watershed TMDL—by impaired segment, by pollutant, by subwatershed or planning
basin, etc. The TMDL writers should evaluate what format and structure is most
meaningful for the specific watershed.
Some of the considerations for deciding how to structure and prepare a watershed TMDL
report include:
• Organization of Multiple Waterbodies and Impairments. A watershed TMDL
might address only two waterbody-pollutant combinations or it could address
hundreds. When dealing with multiple TMDLs in one report, deciding how to
organize the TMDL report can be difficult and affect how readily the TMDL is
understood and accepted by the public. Many times the background information such
as watershed characteristics, land use, and water quality standards can be described
for the entire watershed and all pollutants, rather than addressing each waterbody or
impairment in separate sections. Usually the more difficult decisions arise when
deciding how to present allocations. Regardless of how many pollutant-waterbody
combinations are addressed in the watershed TMDL, TMDL allocations (WLAs and
LAs) need to be documented for each combination. When dealing with multiple
impaired segments and multiple impairments, there will be a number of options for
how to present allocations. Some reports provide sections for each impaired
waterbody and document allocations for the various pollutants within the segment-
specific sections (Figure 4-9). Others group all of the impairment-specific TMDLs
together to provide allocations for multiple segments within sections dedicated to
individual pollutants (Figure 4-10 and Figure 4-11). Regardless of how the
allocations are presented, it can also be helpful to present them in an alternate manner
in an appendix. For example, if the main report presents allocations by subwatershed
to target areas of concern and specific sources, it might also include the allocations
grouped by pollutant in an appendix to illustrate the distribution and relative
magnitude of pollutant-specific load reductions throughout the watershed.
• Audience. For any TMDL report, it is important to consider the audience when
developing the outline and preparing the document. This is especially true for
watershed TMDLs that might have a greater number of involved stakeholders and
affected sources. The organization and content of a report can be critical to the
comprehension and acceptance of the TMDL by the affected parties. Many sources
(e.g., MS4s, private farmers, WWTPs) might not be familiar with the location, or
even concept, of watersheds and subwatersheds. Depending on how the report is
written and the allocations are presented, sources might not understand where they
~74 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
are located in relation to impaired segments and related allocations. To facilitate an
understanding by the public and affected parties of the relative location of segments,
sources and areas targeted for controls, it is useful for watershed TMDL reports to
include descriptive maps with such things as roads, boundaries (counties, towns,
parks, federal lands) and other recognizable landmarks or points of reference.
• Level of Detail. Watershed TMDLs typically generate more information to present in
the TMDL report when compared to reports addressing single-segment TMDLs.
Because of the larger area evaluated and the multiple waterbodies and impairments,
there will likely be more information to document for such topics as watershed
characteristics, impairment information, data analysis, sources and technical
analyses. The level of detail included for each topic can help or hinder the readability
of the document. A high level of detail, particularly on the data analysis and technical
approaches, can result in a cumbersome document that is daunting to stakeholders
and members of the public and difficult to navigate for EPA reviewers. The use of
appendices can alleviate this problem by moving some of the background details out
of the main body of the report. For example, watershed TMDLs that involve complex
modeling sometimes include a short summary of the technical modeling approach in
the main report with additional details on model development, calibration and
application and associated assumptions and data inputs in a technical appendix. This
way, the analysis information is available to those interested (and also for the
administrative record) but does not overwhelm the main report and bury the
allocations.
DRAFT 75
-------
Handbook for Developing Watershed TMDLs
December 2008
sIS
1 "5 -5 3 T5 S 3 t; 1 :1 5 ? £ £ J f -3 -1
>IJIJ""IIII j. --*•--'•'•
«««»»=«««,«! 1^ ggsgle
S t Illfll
l
l
11
I 2
i^'ilfi^Il^lllllil il 11. •
S ^=5
X 5C 3C X.
Figure 4-9. Example outline for a watershed TMDL report organized by impaired segment.
76
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
s
o
u
1
JJ
Jranch of UK Middle Fort I.I
£
: I : : :
: | : : :
: v~- '. '. '.
: | : : :
, ""
;l, : ;
; JS 6 ; ;
:« « : :
Bavtr Creek, Uslie Kun-E
v Creek Kcu Middltion .Vs.
'ALLOCATIONS
I s g t- ^ £
: : : 5 :
: : : ^ :
: : : J :
: : : | :
. H
: : : j, :
: : :-! S
: .- : 1 S
.\RIAT10N"
•VnoS AND MONITORINf
.-\ssuranccs
mum moments of an Approv
1 Reasonable As*tii'.tnci:si Sui
ij j- 1 -J ~
1 M i
•j - : :
•5 .£ : :
II ; :
J^ ; ;
^5 g '
£ 'S ' ':
il : :
= J ; ;
jii
•^11 ;
111 *
6 - = s
I'll g
0.^ S S3
^^^ t £
- 'i£ % t
5 ' -
1 i i
i - -
^ :
C : :
52 : -z
s ; I
•5 -a ^
1? * S*
1 6 1
uiti-Rcsolulion Land Cliarat:1
lilc Beaver Creek Ailatnmcn
;\'e1opmctrt of Watershed Lo
?.. '^ £
.£
3ft
1
J
:j
^
c:
'-
§
i
"3
5J
1
S
!
1
1
(S
= 3
.
v S .a — : : a : Is : : :
SS:£2 ; : j* :'s ; • ;
-* >-
— •— > C .CQ,.y . > . C
iHl^iiMils
IIIIiilj'liH!?
iiti ;iil|;u||
™ — X
XXX
i i 1:1;
: : S :.' :
2 '-g-
>||
ll^
f* i -
° 2 if r
- ^
.^:lM:iii||2
Figure 4-10. Example outline for a watershed TMDL report organized by impairment type.
DRAFT
11
-------
Handbook for Developing Watershed TMDLs
December 2008
o 2 ~_
pi £
tf « _
5 5 2 £ , >
I liiil
Jf !2 ^ s s «
Figure 4-11. Example outline for a watershed TMDL report organized by impairment type.
78
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Watershed TMDL Demonstration Projects Sponsored by EPA
EPA is currently sponsoring three pilot projects to demonstrate the efficiencies (both in pollutant reductions and
cost) of using a watershed approach to TMDL development. The following are summaries of the projects:
Lower Fox River, Wisconsin
A watershed TMDL is being developed for the Lower Fox River (LFR) Basin, including the Green Bay Area of
Concern (GB AOC) to address 13 waterbodies impaired due to excessive sediment and phosphorus. Highlights of
the project include:
• Pre-TMDL Analysis of the Cost-effectiveness of Agricultural BMPs. Used SWAT in conjunction with an
optimization model to compare 416 BMP scenarios along with implementation costs to identify optimal BMPs
for reducing phosphorus.
• Extensive Stakeholder Involvement. Developed a multi-agency Outreach Committee to work with key
stakeholders (agriculture, stormwater, industrial and municipal dischargers, etc.) to determine and analyze
BMP scenarios. Conducted targeted stakeholder meetings (agriculture, stormwater). Conducted basin-wide
survey of 300 dairy farmers to raise awareness about water quality issues, identify existing agricultural
practices and gain information to support modeling scenarios and TMDL implementation planning.
• Application of Multiple Technical Tools for TMDL Development. Will apply SWAT to estimate existing
loads and calculate necessary load reductions to meet the numeric targets. Will include evaluation of
phosphorus and sediment loading MS4 urban areas covered under WPDES Municipal General Permits. Will
incorporate data from recent urban stormwater modeling using the Source Loading and Management Model
(SLAMM) conducted for MS4s within the LFR Basin. Will use load duration curves to illustrate the water quality
targets and modeling results for each impaired segment.
Lower Silver Creek, Utah
A TMDL was developed previously to address cadmium and zinc impairments in Lower Silver Creek. Estimates
for the cost of implementation were greater than $100 million dollar with no incentive to begin cleanup. The initial
TMDL assessment included gross (watershed-scale) allocations but provided an insufficient level of detail
necessary to justify the expense of specific source reduction and remediation efforts. This demonstration project
is designed to address the ongoing need for additional analysis of the pollutant source reduction options to better
understand how to optimize cost and pollutant reduction effectiveness at the watershed scale and to inform
regulators on the best mix of NPDES, nonpoint source, and other cleanup options. Because Lower Silver Creek
watershed includes CERCLIS NPL "Alternative" Sites as well as 303(d) impaired waters, the project uses a cross-
programmatic (Water and Waste Programs) approach to assessment and cleanup. Activities include:
• Watershed reconnaissance survey to fill the data gaps and characterize the nonpoint and point sources for
model development.
• Conceptual model development to identify known and suspected pathways of pollutant transport, known and
suspected sources of pollutants, and known or potential human and environmental receptors.
• Detailed water quality and stream sediment sampling throughout the upper and lower watershed to support
model calibration and validation.
• Development of reactive transport models for both the Upper and Lower Silver Creek watersheds to quantify
significant sources of metals loadings, addressing the geochemical behavior of the source materials as well as
the fate and transport of metals via surface and subsurface pathways, and evaluate various remedial
scenarios.
• Identification and evaluation of load reduction options and remedial action objectives for the identified sources
of pollutants associated with mining activities.
• Quantitative evaluation of watershed-scale options and development and optimization of alternatives for
cleanup that consider their effectiveness, implementability, and cost.
DRAFT 79
-------
Handbook for Developing Watershed TMDLs December 2008
Watershed TMDL Demonstration Projects Sponsored by EPA (continued) j
Ballona Creek and San Gabriel River, Greater Los Angeles Area, CA j
: The purpose of this pilot project is to develop a cost-benefit framework for optimizing BMP selection in watershed j
; TMDLs being developed in the Los Angeles Region. The project will compare alternative pollutant reduction and j
associated cost scenarios that result in the most efficient and timely attainment of standards in two watersheds j
(Ballona Creek and San Gabriel River), using the following 4-step process: j
: 1. Select a suite of BMPs for testing of various attainment scenarios. j
; 2. Apply a watershed model to evaluate the effectiveness of various BMP scenarios in the two watersheds to j
; meet water quality standards. j
3. Evaluate the relative costs of the BMP scenarios against the expected reductions in pollutant loadings and j
water quality concentrations. j
: 4. Design a monitoring framework to demonstrate attainment with water quality standards. j
This BMP analysis and optimization tool will be integrated with the allocation scenario and selection process to be j
implemented over the next 2 years as part of TMDL development for Dominguez Channel and the Ports of Los j
Angeles and Long Beach (all part of the same integrated watershed-scale TMDL project). This pilot project will j
also be used to assist development and revision of several additional TMDLs in the Los Angeles Region, j
including the Los Angeles River bacteria TMDL. The optimization tool can provide a framework for pollutant j
trading scenarios within a TMDL and can be applied in other Southern California counties facing similar problems, j
80 DRAFT
-------
5. Supporting Implementation of
Watershed TMDLs
This section focuses on activities and issues related to watershed TMDL implementation.
As noted in Section 2.3, watershed TMDLs can generate a variety of implementation
benefits. Through a watershed TMDL approach, implementation has the potential to
connect to a broader variety of watershed-based tools and activities. Watershed TMDLs
support and inform other watershed-focused efforts, such as watershed planning,
permitting and trading. TMDLs often provide the core quantitative analysis for these
programs and implementation activities. As a result, these programs and activities
represent major vehicles for watershed TMDL implementation. Watershed TMDLs
establish comprehensive stakeholder involvement processes to garner support from all
watershed interests. As the watershed TMDL process moves from the development
phase to the implementation phase, watershed stakeholders are more likely to identify as
a watershed group and collectively identify and implement strategies to achieve LAs.
5.1. Multiple Program Involvement/Coordination
Watershed TMDLs set the stage for TMDL practitioners and stakeholders to consider
innovative approaches to implementation that focus on broad watershed issues and
require broad stakeholder involvement. These innovative implementation approaches,
such as watershed-based NPDES permitting, water quality trading, and watershed-based
planning, often work in concert, resulting in a coordinated effort across multiple
programs to achieve water quality goals. Provided below are brief descriptions of
watershed-focused implementation programs and activities that can support and be
coordinated with watershed TMDL implementation.
5.1.1. Watershed-based Permitting
The NPDES program is an important part of an integrated |
watershed approach to achieving water quality goals. To j
integrate NPDES permitting into a watershed approach, | Watershed-based Permitting
NPDES permit writers need to develop or use an existing i
, ,, 11- ri • • I For more information on watershed-based
watershed-based analysis as part of the permitting j NRDES permitting] including watershed-based
process. A watershed-based analysis, referred to as a j permitting implementation and technical
watershed permitting analytical approach in the context of j guidance, as well as case studies, visit EPA's
watershed-based permitting, considers watershed goals, I watershed permitting web site at
identifies and evaluates multiple pollutant sources and ! http://cfpub.epa.qov/npdes/wqbasedpermittinq/
. ^ ^ i wspermittmq.cfm.
stressors, including nonpomt source contributions, and I
constructs an NPDES watershed framework that
identifies NPDES permitting implementation options and other water quality program
tools to achieve water quality goals. Where TMDL practitioners develop watershed
TMDLs, the TMDL development process is likely to serve as a basis for the watershed
permitting analytical approach necessary to support the development and implementation
DRAFT 81
-------
Handbook for Developing Watershed TMDLs December 2008
of an NPDES watershed framework. Implementation options available under an NPDES
watershed framework can include a wide range of activities, such as:
• NPDES permit development and issuance on a watershed basis
• Wet-weather integration
• Indicator development for watershed-based stormwater management
• Monitoring consortium development
• Permit synchronization
• Statewide rotating basin planning approach (USEPA 2007a)
Through the watershed TMDL development process, TMDL practitioners might have
obtained the information necessary to demonstrate to NPDES permit writers that NPDES
permit development and issuance on a watershed-basis is a technically feasible option for
permittees in the watershed. For example, the watershed TMDL analysis might show
that the conditions in the watershed are well understood, that there are common pollutants
of concern among sources in the watershed, and that point sources have a significant
impact in the watershed. In this case, the NPDES permit writer, along with point source
dischargers and other watershed stakeholders, might determine that developing and
issuing NPDES permits on a watershed basis is an appropriate approach for addressing
point source loads of one or more pollutants. The specific conditions and types of point
source dischargers located in the watershed will influence the type of NPDES permit that
is appropriate for the watershed. Information generated through the watershed TMDL
will help NPDES permit writers evaluate the applicability of the following permit types:
coordinated individual permits, integrated municipal permits, and multisource watershed-
based permits. It is important to keep in mind that NPDES permit writers might want to
include other pollutants in NPDES permits developed for point source dischargers within
a watershed to address other local water quality goals and concerns. As a result, NPDES
permit writers might find it necessary to supplement the watershed TMDL analysis with a
watershed permitting analytical approach to address pollutants outside the scope of the
watershed TMDL.
I Definition: \
| Types of NPDES Permits Issued on a Watershed-basis to Facilitate Watershed TMDL Implementation j
j Coordinated Individual Permits: Individual permits issued to each point source discharger in the watershed j
j that contains water quality-based effluent limits and other conditions based on the watershed TMDL analysis (i.e., j
| holistic analysis of watershed conditions) rather than on a permit-by-permit basis. j
j Integrated Municipal Permits: A single individual permit issued to a municipality that bundles a number of point j
j source permit requirements (e.g., publicly-owned treatment works, stormwater, combined sewer overflows, j
j pretreatment) that take place within a watershed's boundaries and contain effluent limitations and permit j
| conditions designed to specifically address watershed goals, including WLAs under a watershed TMDL. j
j Multisource Watershed-based Permits: A single permit, either individual or general, issued to a group of point j
j source dischargers located within a watershed either to address a single pollutant or to address multiple j
| pollutants from similar types of discharges that contain requirements specific to achieving watershed goals, such j
| as water quality-based effluent limits generated using a holistic watershed analysis based on the watershed j
j TMDL, watershed monitoring requirements, and other watershed provisions (e.g., trading, stakeholder j
j involvement, reporting). j
| For more information on developing NPDES permits on a watershed-basis refer to EPA's Watershed-based j
j NPDES Permitting Technical Guidance at www.epa.gov/npdes/pubs/watershed techquidance entire.pdf j
82 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
5.1.2. Water Quality Trading
Water quality trading most often takes place on a | D \
watershed-basis under a pollutant cap established | ' j
through a TMDL analysis, allowing point sources j Water Quality Trading j
that face higher pollutant control costs to achieve i i
11 . . i . • 11 i • 11 . . I For more information on the type of information i
pollutant reduction goals by purchasing pollutant j necessary to develop and impy|ement water quality j
reduction credits from other sources, both point and | trading activities to achieve WLAs under watershed |
nonpoint, that can generate reductions at a lower j TMDLs, see EPA's Toolkit available at j
cost. EPA's Water Quality Trading Toolkit for j www.epa.qov/waterqualitvtradinq/WQTToolkit.html. j
Permit Writers (Toolkit) provides the fundamental ' j
information necessary to develop and conduct water quality trading through NPDES
permitting as a tool for TMDL implementation. Information generated through
watershed TMDLs is essential to support key components of water quality trading
programs, including defining the appropriate geographic area, identifying potential
trading program participants within the watershed, establishing baselines (i.e., WLAs and
LAs) for all potential participants, and conducting the necessary technical analysis to
support the development of trade ratios based on watershed conditions and
characteristics.
Defining the appropriate geographic scope is an essential first step to ensure water quality
trading will effectively address water quality concerns. According to EPA's Toolkit, the
geographic scope of water quality trading should occur only within a hydrologic unit that
is defined to ensure that trades will maintain water quality standards within that unit, as
well as within downstream and contiguous waters. Because the purpose of water quality
trading is to more cost effectively improve water quality, buyers and sellers should
discharge directly or indirectly (e.g., to a tributary within the impaired watershed) to the
same waterbody where pollutant load reductions are necessary to achieve water quality
standards or sellers could discharge to an upstream tributary within the watershed. The
analyses done for watershed TMDLs are likely to provide more information on potential
buyers and sellers of pollutant credits and on pollutant fate and transport in a watershed,
which will better support a water quality trading feasibility analysis, as opposed to single-
segment TMDLs that might have a more limited geographic scope and would require
additional pollutant analyses beyond this limited geographic area to determine if a trading
program could be feasible.
Watershed TMDLs comprehensively and simultaneously address the development of
WLAs and LAs for sources contributing to the water quality impairment within the
watershed. In the context of water quality trading, WLAs and LAs serve as the baseline
for point sources and nonpoint sources potentially involved in trading. It is imperative
that the WLAs and LAs are set consistently among potential trading partners and are not
likely to change as a result of having to "redo" upstream TMDLs based on pollutant load
reductions necessary to achieve downstream TMDLs. Unexpected changes to WLAs and
LAs for trading partners can create uncertainty in the trading market and result in
decreased willingness to participate.
Water quality trading might require the use of one of two types of fate and transport trade
ratios - location and delivery ratios. Location and delivery ratios account for the unique
watershed conditions and characteristics that affect pollutant fate and transport between
upstream and downstream locations. Models used to support watershed TMDL
development are likely to address these fate and transport issues when calculating loading
capacity and allocations. This information can directly support the development of
DRAFT 83
-------
Handbook for Developing Watershed TMDLs December 2008
Using Water Quality Trading to Implement the Long Island Sound Nitrogen TMDL
The states of New York and Connecticut jointly submitted a TMDL for nitrogen which was approved by EPA in
2001. The TMDL includes reductions to in-basin (phase III) and out-of-basin (phase IV) point and nonpoint
sources. In-basin sources refer to those sources within the Connecticut and New York portions of the drainage
basin, while out-of-basin reductions refer to those nutrients from all other sources beyond the in-basin boundaries,
including tributary transport from north of Connecticut and oceanic transport. The first phase of reductions (phase
III) focus on in-basin reductions, since the majority of nitrogen loading is attributed to those sources. Phase V of
TMDL implementation actually addresses non-treatment alternatives such as outfall relocation, seaweed farms
and oxygen injection techniques.
One of the primary mechanisms for implementation of the TMDL was the option for nitrogen trading among
sources. Because 90 percent of the reductions were expected to be from point sources, the State of Connecticut
developed a trading program to address its 84 municipal treatment facilities. Geographically, the in-basin areas
were divided into 12 management zones, one of which comprised the Sound itself. The terrestrial zones
generally follow river basin boundaries in Connecticut and political boundaries in New York. To accommodate the
development of Connecticut's trading program, the larger management zones in Connecticut were also
subdivided into tiers. The tiers were designed to account for differences in attenuation and transport of nitrogen
from one tier to the next and were used in assigning attenuation factors. Attenuation factors are used in
quantifying the relationship between a discharge point and actual delivery of nitrogen to the Sound and help to
determine allowable trades between facilities.
location and delivery ratios necessary to understand the downstream impact of a pound of
pollutant on the waterbody of concern discharged by an upstream source.
5.1.3. Watershed-based Planning
While TMDL implementation for point sources occurs
through NPDES permits, the primary mechanism for
addressing nonpoint source pollutant load reductions is
Resources:
Watershed-based Planning
Plans to Restore and Protect Our Waters
(USEPA2008), available at
www.epa.gov/owow/nps/watershed handbook/.
provides a step-by-step approach for developing
a watershed plan that addresses each of the
nine elements.
often through watershed-based planning. More
._ „ & i . i i i j r? j i ! EPA s Handbook for Developing Watershed
specifically, watershed stakeholders often develop
watershed-based plans eligible for funding through
EPA's section 319 nonpoint source pollution control
grant program or other funding sources. The process for
watershed-based planning includes characterizing the
watershed, identifying watershed-specific problems,
setting goals, identifying solutions, establishing
partnerships, and measuring progress. EPA recently identified nine elements that are
critical for developing and implementing effective watershed-based plans to improve
water quality (USEPA 2002b, 2008). EPA now requires these nine minimum elements
for watershed plans funded through section 319 grants and strongly recommends the use
of these elements in all watershed plans intended to address impaired watersheds.
Development of watershed TMDLs address many of the nine minimum elements
necessary to include in watershed plans that stakeholders will use to secure section 319
grant funds to implement nonpoint source pollution control measures. Specifically,
watershed TMDLs will produce a watershed pollutant source characterization that
identifies major sources and causes of impairment, as well as water quality standards and
other watershed goals for addressing the impairments (Element a). Watershed TMDLs
also quantify pollutant loads for sources within the watershed and identify the reductions
necessary to achieve water quality standards (Element b). Of all the nine minimum
elements, quantifying the pollutant loads and pollutant load reductions necessary to meet
water quality standards is likely the most significant contribution that a watershed TMDL
can make to the development of a watershed-based plan intended to satisfy EPA's
"84 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Definition:
Nine Minimum Elements of a Watershed-based Plan
a. An identification of the causes and sources or groups of similar sources that will need to be controlled to
achieve the load reductions estimated in this watershed-based plan (and to achieve any other watershed goals
identified in the watershed-based plan), as discussed in item (b) immediately below. ...
b. An estimate of the load reductions expected for the management measures described under paragraph (c)
below (recognizing the natural variability and the difficulty in precisely predicting the performance of management
measures overtime). ...
c. A description of the NPS management measures that will need to be implemented to achieve the load
reductions estimated under paragraph (b) above (as well as to achieve other watershed goals identified in this
watershed-based plan), and an identification (using a map or a description) of the critical areas in which those
measures will be needed to implement this plan.
d. An estimate of the amounts of technical and financial assistance needed, associated costs, and/or the
sources and authorities that will be relied upon, to implement this plan. ...
e. An information/education component that will be used to enhance public understanding of the project and
encourage their early and continued participation in selecting, designing, and implementing the NPS management
measures that will be implemented.
f. A schedule for implementing the NPS management measures identified in this plan that is reasonably
expeditious.
g. A description of interim, measurable milestones for determining whether NPS management measures or
other control actions are being implemented.
h. A set of criteria that can be used to determine whether loading reductions are being achieved over time and
substantial progress is being made towards attaining water quality standards and, if not, the criteria for
determining whether this watershedbased plan needs to be revised or, if a NPS TMDL has been established,
whether the NPS TMDL needs to be revised.
i. A monitoring component to evaluate the effectiveness of the implementation efforts overtime, measured
against the criteria established under item (h) immediately above.
requirements. Watershed TMDLs can also help with the development of progress
indicators to assess achievement of pollutant reduction targets (Elements g and h) and the
development of a monitoring component to assess implementation efforts over time
(Element i).
5.2. Structuring the TMDL to Support Implementation Activities
In addition to organizing the TMDL process and analytical activities to support the needs
of other CWA programs, both the TMDL process and document itself need to be
structured in such a way that they facilitate implementation.
Building on the early stakeholder involvement efforts of the TMDL process, as discussed
in Section 4.1, can facilitate implementation activities. Continual stakeholder support and
involvement ensures consideration of important analytical considerations for
implementation. This might include stakeholder input regarding the relative amount of
reductions or allocations and what pollutant parameters should receive greater focus for
reductions. For example, modeling might indicate that several NPDES dischargers in a
watershed need to reduce BOD, TP, and TN discharge concentrations to specific levels to
reduce periphyton and nuisance algal growth in a waterbody. Given the ability to
participate in the reduction scenario analysis, the dischargers' input might indicate that
greater relative controls on BOD discharges from the facilities can be more effective than
equal reductions on all three.
DRAFT 85
-------
Handbook for Developing Watershed TMDLs December 2008
Program Notes:
How Are WLAs Translated into NPDES Effluent Limitations?
TMDL WLAs are assigned for both continuous discharging facilities and for diffuse discharges such as stormwater
outfalls that are covered by a Phase II NPDES stormwater permit. NPDES regulations at § 122.44(d)(1)(vii)(B)
require that any NPDES permits subject to a TMDL WLA be updated to include effluent limitations that are
consistent with the assumptions and requirements of the WLA. To facilitate implementation, it can be helpful for
TMDL practitioners to understand the basic procedures used by permit writers in translating TMDL WLAs into
permit limits. EPA guidance is available to assist permit writers in developing effluent limits from TMDL WLAs.
Separate techniques are applied for aquatic life WLA's and human health WLA's.
Implementing Aquatic Life WLAs
For continuous discharges, permit writers often use a statistical procedure such as the one discussed in Section
5.4 of the EPA's Technical Support Document for Water Quality-based Toxics Control (Technical Support
Document), or similar procedures adopted by the permitting authority, to translate WLAs expressed as a monthly
average or shorter time period (e.g., 4-day average) into effluent limits.
The statistical process consists of the following steps:
• Calculate a long-term average (LTA) pollutant concentration for each WLA (generally an acute and chronic
WLA)
• Select the lowest (most protective) LTA as the performance basis for the permitted discharge
Calculate an average monthly limitation (AML) and maximum daily limitation (MDL) from the most protective LTA
Implementing Human Health WLAs
The exposure period of concern for human health water quality standards can be up to 70 years and the average
exposure rather than the maximum exposure is usually of concern. Because compliance with effluent limitations is
normally determined on a daily or monthly basis, it is necessary to set human health effluent limitations for
continuous discharges so that they meet a human health WLA every month. In Section 5.4 of the Technical
Support Document, EPA recommends a procedure for establishing effluent limitations for protection of human
health:
• Set the AML equal to the human health WLA
• Calculate the MDEL for human health by multiplying the AMEL by the ratio MDELAMEL ratio provided in the
Technical Support Document
For both human health and aquatic life WLAs, the Technical Support Document provides formulae and tables to
assist with calculating effluent limitations for human health protection.
Long-Term Effluent Limitations
In some cases, the long-term nature of impacts from pollutants such as nutrients, especially in downstream
waterbodies such as lakes and estuaries, combined with the difficulty of predicting seasonal fluctuations in
treatment efficiency and loading of such pollutants have led to development of annual WLAs. If a WLA is
expressed as an annual requirement (e.g., annual average concentration or annual loading), and fluctuations in
treatment efficiency and loading make establishing shorter-term (e.g., monthly, weekly, or daily) effluent
limitations impracticable, the NPDES permit could include an annual effluent limitation. In such cases, the
NPDES permit should clearly state the method for determining compliance with the annual limitation.
Concentration-based and Mass-based Effluent Limitations
Where a WLA is expressed in terms of concentration, effluent limitations calculated from the WLA should be
expressed in terms of concentration. Likewise, if a WLA is expressed in terms of mass, the corresponding effluent
limitations should be expressed in terms of mass. If there are WLAs that apply to a discharger expressed in terms
of both concentration (e.g., WLAs calculated directly from concentration-based standards to prevent near-field
impacts) and mass (e.g., WLAs from a TMDL) it might be necessary to calculate the corresponding effluent
limitations in more than one form to allow comparison and selection of the most protective limitations for the
permit.
Water Quality-based Effluent Limitations for Permitted Stormwater
EPA recommends that for NPDES-regulated municipal and small construction stormwater discharges effluent
limits should be expressed as BMPs or other similar requirements, rather than as numeric effluent limits, to be
consistent with the WLA. (USEPA 2002c)
86 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
In preparing the final document and deciding how to present allocations, practitioners
should consider the ultimate implementation goals of the TMDL. For example, are there
going to be both LAs and WLAs? What types of facilities will receive WLAs and how
do permit writers include WLAs in their permits? This can facilitate adaptation of the
required WLAs by NPDES permits, as well as LAs through necessary BMP
implementation to reach land-use based loading goals. Developers and stakeholders
might work together to determine what format the TMDL document should use to present
LAs and WLAs to ensure the allocations are understandable and readily translated to
promote implementation. Especially for MS4 permits, where TMDLs are sometimes
referenced in or attached to the implementing permits, it is especially important for
TMDL practitioners to provide adequate information in the text and analysis so that
permittees can identify appropriate BMPs to implement WLAs. TMDL practitioners
might consider presenting MS4 allocations at various levels of detail to assist permittees.
For example, MS4 WLAs might be provided for the MS4 area as a whole, but also
describe the associated reductions identified for subbasins and individual land uses in the
boundary area. This additional detail and context for the WLAs, (e.g., 80 percent of
loading reductions must come from urban areas as compared to residential areas)
permittees will have more information to support them in meeting their allocations.
| Program Notes:
\ Approved vs. FYI WLAs in Watershed TMDLs
j When applying the watershed approach in the development of TMDLs, pollutant sources are identified and
j quantified throughout an entire waterbody and multiple impaired segments. The process might identify the need to
I establish allocations in segments not formally defined as impaired (on the state's 303(d) list of waters requiring a
| TMDL).
j As stated in 40 CFR 130.7(d)(1), "all WLAs/LAs established ...for water quality limited segments [any segment
j where it is known that water quality does not meet applicable water quality standards] shall continue to be
| submitted to EPA for review and approval." Clearly, then, WLAs developed in TMDLs for waters on the State's
| 303(d) list are formally approved by EPA, and must be incorporated into NPDES permits (see CFR section 144).
j However, WLAs developed for un-listed segments that are an integral part of the watershed's hydrologic network
j might not need to be formally approved by EPA. WLAs developed in the TMDL for unlisted segments fall into two
j categories, WLAs for segments "causing and contributing" to a listed segment's impairment and WLAs for
| segments not "causing or contributing" to a listed segment's impairment.
j The WLAs for segments determined to be ""causing or contributing" to an impairment will be acted upon by EPA
j in an identical manner as those for the formally listed segments. The allocations to these segments will be
j formally approved by EPA when the TMDL is submitted, and the WLAs must be incorporated into permits (see
| CFR section 144). The WLAs for segments determined not to be "causing or contributing" to an impairment will
j not be formally approved acted upon by EPA, and are for informational purposes only.
tee
5.3. Follow-up Monitoring
Post-implementation monitoring is critical to determining the success of any
implementation effort and this is no different for watershed TMDLs. States should
ideally plan for both pre- and post-TMDL monitoring needs as a regular part of their
monitoring strategies to effectively allocate resources to conduct both activities. The
TMDL modeling analysis can be used to assist in the development of the post-
implementation monitoring plan to identify the location of assessment points and the
frequency of assessment in the follow-up period. If detailed enough, the modeling can
even be used to help measure compliance with the TMDL by developing rating curves of
allowable loads against which post-implementation monitoring results can be compared.
DRAFT 87
-------
Handbook for Developing Watershed TMDLs December 2008
By comparing monitoring results at key locations against specific metrics obtained with
model results at those locations states can have a real measuring stick to guide them in
assessing the success of their reduction efforts.
Often with large area TMDLs there are significant areas or categories of information for
which additional detail or monitoring can help to more fully inform the analysis. For
situations such as these, the TMDL analysis is useful in identifying what information
(either geographic locations or specific parameters) needs to be addressed in the post-
implementation monitoring. Upon further investigation, results can be incorporated into
an updated TMDL in an adaptive implementation process.
Watershed TMDL Development in the DuPage River and Salt Creek Watersheds, Illinois: Promoting
Development of a Stakeholder Workgroup and Monitoring to Guide Post-TMDL Implementation Efforts
TMDL development in Illinois is conducted on a watershed basis using a three stage process that emphasizes the
importance of monitoring and implementation. Stage 1 consists of watershed characterization and data analysis.
Stage 2 involves follow-up data collection for situations where additional information is deemed needed. Stage 3
is focused on modeling, TMDL development, and preparation of implementation plans. In the case of the DuPage
River/ Salt Creek, development of the watershed TMDL and implementation plans led to the formation of local
stakeholder group that is taking a distinctive approach to address the findings of the TMDL reports.
The DuPage River and Salt Creek, significant tributaries of the Des Plaines River, flow through rapidly urbanizing
watersheds in the Chicago metropolitan area. The two watersheds combined encompass an area of
approximately 360 mi2. The watersheds are home to 55 municipal entities and 25 POTWs, and at least 21 dams
have been constructed in the watersheds to address issues such as flooding and recreational needs. Illinois EPA
developed watershed TMDLs to address impairments to several segments by chloride and low dissolved oxygen.
Representatives from municipalities affected by the TMDL discussed the need for a framework to collect data and
carry out other technical activities to support implementing the TMDLs. This led to creation of the DuPage River
Salt Creek Workgroup (DRSCW, www.drscw.org), a collaborative effort by sanitary districts, municipalities,
counties, forest preserve districts, state and federal agencies, and private environmental organizations.
Stakeholders affected by the TMDL allocations wanted an opportunity to "substantiate" implementation strategies
and determine whether there were other cost-effective options. It was also envisioned that the DRSCW could help
stakeholders establish a solid foundation for future TMDLs, contribute to the development of nutrient criteria, and
address other water quality or regulatory issues in the watersheds.
At the outset, DRSCW members acknowledged the need for better data to make informed decisions. As a result,
establishing and implementing a monitoring program have been the DRSCW's highest priorities and have helped
to unify the group. Better monitoring data allows the DRSCW to understand the sources of impairment, identify
priority restoration activities and track implementation effectiveness, calibrate water quality and watershed
models, determine progress toward achieving water quality standards, and assess the overall health of the
watersheds. To augment routine fixed-station monitoring by the Illinois EPA, Metropolitan Water Reclamation
District (MWRD) of Greater Chicago and USGS, the DRSCW established a network of continuous monitoring
probes throughout the watersheds. To date, the monitoring network includes 10 submerged probes that measure
DO and collect hourly data on pH, conductivity, and temperature. Agency members of the DRSCW also contribute
data from an additional five probes. The Workgroup also initiated an extensive bioassessment program across
DuPage County.
The collaborative, locally led approach initiated by DRSCW members, while still in its beginning phases, has
several early indicators of success, including support from state and federal regulatory agencies, financial support
from all levels of watershed stakeholders, membership that continues to grow, and more watershed monitoring
data to facilitate science-based collaborative decision-making.
88 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
5.4. Financial Resources for Implementation
The early engagement of stakeholders in a watershed TMDL is not only an opportunity
for better coordination of activities, data integration and increased buy-in, but it is also an
opportunity for the identification and coordination of funding resources.
Many sources of funding require a "match." For every dollar a funder provides, the
recipient organization or agency must "match" a certain percentage of the grant with
funds obtained from other entities. By requiring a match, the funder guarantees that they
are not the sole supporter of a project and that partnering with other funders or
stakeholders will occur. Stakeholders can be source of matching funds when
implementing a watershed TMDL. Matching funds can be in the form of cash raised or in
the form of in-kind contributions for allowable costs under the grant. For example, if a
professional donates their time to conduct a stream assessment, their hourly rate for the
time they spent conducting the assessment might be used as a match. Other examples of
in-kind match include donated tools, the rental value of meeting space used for the
project, mileage, postage and materials. Every match, whether in-kind or cash, should
have clear, supporting documentation as federal grants are subject to audits.
Watershed Funding
EPA's Section 319 Grant program provides states with funding to address nonpoint source pollution. Within this
funding program, "incremental funds" are available for the development and implementation of watershed-based
plans and TMDLs. State agencies might offer subawards of this money to public and private entities. Section 319
Grant guidance can be found in Applying for and Administering CWA Section 319 Grants: A Guide for State
Nonpoint Source Agencies (http://www.epa.qov/owow/nps/319/319statequide-revised.pdf).
The Watershed Funding Web site (www.epa.qov/owow/fundinq.html) includes tools, databases, and information
about sources of funding to watershed practitioners. The Catalog of Federal Funding for Watershed Protection
can be accessed through this site and it includes a searchable database of more than 80 federal programs
offering financial assistance. Also accessible through this site is the Plan2Fund Objective Prioritization Tool
(Plan2Fund OPT). This web-based decision model generates comparative reports based on the user's objectives
and decisions related to the project and was designed to help an organization develop and implement a long-term
financial strategy to meet strategic goals.
DRAFT 89
-------
Handbook for Developing Watershed TMDLs December 2008
90 DRAFT
-------
6. References
CTDEP and NYSDEC. 2000. A Total Maximum Daily Load Analysis to Achieve Water
Quality
Standards for Dissolved Oxygen in Long Island Sound. Connecticut Department of
Environmental Protection and New York State Department of Environmental
Conservation. December 2000.
CWP. 2005. Simple Method to Calculate Urban Stormwater Loads. Center for Watershed
Protection, Stormwater Manager's Resource Center (SMRC), Ellicott City, MD.
Schueler, T.R. 1987. Controlling Urban Runoff: A Practical Manual for Planning and
Designing Urban BMPs. Metropolitan Council of Governments, Washington, DC.
USEPA. 1983. Results of the Nationwide Urban Runoff Program: Volume 1 - Final
Report. Water Planning Division. U.S. Environmental Protection Agency, Water
Planning Division, Washington, DC.
http://www.epa.gov/npdes/pubs/sw nurp vol 1 finalreport.pdf
USEPA. 1991. Guidance for Water Quality-based Decisions: The TMDL Process. EPA
440/4-91-001. U.S. Environmental Protection Agency, Office of Water, Office of
Wetlands, Oceans, and Watersheds, Washington, DC.
http://www.epa.gov/OWOW/tmdl/decisions/
USEPA. 1997a. Compendium of Tools for Watershed Assessment and TMDL
Development. EPA841-B-97-006. U.S. Environmental Protection Agency, Office of
Water, Office of Wetlands, Oceans, and Watersheds, Washington, DC.
USEPA. 1997b. New Policies for Establishing and Implementing Total Maximum Daily
Loads (TMDLs). Memorandum from Robert Perciasepe, Assistant Administrator, Office
of Water, to Regional Administrators and Regional Water Division Directors. August 8,
1997.
USEPA. 1999a. Protocol for Developing Nutrient TMDLs. EPA 841-B-99-007. U.S.
Environmental Protection Agency, Office of Water (4503F), Office of Wetlands, Oceans,
and Watersheds, Washington, DC.
http://www.epa.gov/owow/tmdl/nutrient/pdf/nutrient.pdf
USEPA. 1999b. Protocol for Developing Sediment TMDLs. EPA 841-B-99-004. U.S.
Environmental Protection Agency, Office of Water (4503F), Office of Wetlands, Oceans,
and Watersheds, Washington, DC.
http://www.epa.gov/owow/tmdl/sediment/pdf/sediment.pdf
USEPA. 2000. Protocol for Developing Pathogen TMDLs. EPA 841-R-00-002. U.S.
Environmental Protection Agency, Office of Water (4503F), Office of Wetlands, Oceans,
and Watersheds, Washington, DC. http://www.epa.gov/owow/tmdl/pathogen all.pdf
DRAFT 91
-------
Handbook for Developing Watershed TMDLs December 2008
USEPA. 2001a. The National Costs of the Total Maximum Daily Load Program (Draft
Report). EPA 841-D-01-003. U.S. Environmental Protection Agency, Office of Water,
Washington, DC. August 1, 2001.
USEPA. 200Ib. The National Costs to Develop TMDLs (Draft Report): Support
Document #1 for "The National Costs of the Total Maximum Daily Load Program (Draft
Report). EPA 841-D-01-004. U.S. Environmental Protection Agency, Office of Water,
Washington, DC.
USEPA. 2002. Committing EPA's Water Program to Advancing the Watershed
Approach. Memorandum from G. Tracy Mehan, III, Assistant Administrator, Office of
Water, to Office Directors and Regional Water Division Directors. December 3, 2002.
USEPA. 2002b. Supplemental Guidelines for the Award of Section 319 Nonpoint Source
Grants to States and Territories in FY 2003. U.S. Environmental Protection Agency,
Office of Water, Washington, DC. August 26, 2002.
http://www.epa.gov/owow/nps/Section319/319guide03.html
USEPA. 2002c. Establishing Total Maximum Daily Load (TMDL) Wasteload
Allocations (WLAs) for Storm Water Sources and NPDES Permit Requirements Based on
Those WLAs. Memorandum from Robert H. Wayland, III, Director, Office of Wetlands,
Oceans and Watersheds, and James A. Hanlon, Director, Office of Wastewater
Management, U.S. Environmental Protection Agency, Washington, DC.
USEPA. 2003. Nonpoint Source Program and Grants Guidelines for States and
Territories (FY2004). Federal Register/Vol. 68, No. 205. U.S. Environmental Protection
Agency, Office of Water (4503F), Office of Wetlands, Oceans, and Watersheds,
Washington, DC. http://frwebgate.access.gpo.gov/cgi-
bin/getdoc.cgi?dbname=2003 register&docid=fr23oc03-39.pdf
USEPA. 2004. Water Quality Trading Assessment Handbook. EPA 841-B-04-001. U.S
Environmental Protection Agency, Office of Water, Office of Wastewater Management,
Washington, DC.
USEPA. 2005a. Improve Water Quality on a Watershed Basis: Implementation Plan for
Subobjective 2.2.1. SIP Draft FY06 National Program Guidance. U.S. Environmental
Protection Agency, Office of Water, Washington, DC. April 28, 2005.
USEPA. 2005b. TMDL Model Evaluation and Research Needs. EPA/600/R-05/149. U.S.
Environmental Protection Agency, Office of Research and Development, National Risk
Management Research Laboratory, Cincinnati, OH.
USEPA. 2007a. Watershed-based National Pollutant Discharge Elimination System
(NPDES) Permitting Technical Guidance. EPA 833-B-07-004. U.S. Environmental
Protection Agency, Office of Water, Office of Wastewater Management, Water Permits
Division, Washington, DC.
USEPA. 2007b. Water Quality Trading Toolkit for Permit Writers. EPA-833-R-07-004.
U.S. Environmental Protection Agency, Office of Water, Office of Wastewater
Management, Washington, DC.
92 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
USEPA. 2007c. An Approach for Using Load Duration Curves in the Development of
TMDLs. EPA 841-B-07-006. U.S. Environmental Protection Agency, Office of Water,
Office of Wetlands, Oceans and Watersheds, Washington, DC.
USEPA. 2008. Handbook for Developing Watershed Plans to Restore and Protect Our
Waters. EPA 841-B-08-002. U.S. Environmental Protection Agency, Office of Water,
Office of Wetlands, Oceans and Watersheds, Washington, DC.
DRAFT 93
-------
Handbook for Developing Watershed TMDLs December 2008
94 DRAFT
-------
Appendix: Case Studies
This appendix highlights a number of watershed TMDL efforts in various stages of completion, from some
that are just beginning to those that are years into implementation. To illustrate a greater range of issues,
the cases presented include both state TMDL efforts as well as regional TMDLs. While some technical
issues are discussed, the case studies try to focus on issues such as the development process and other
topics related to the implications of using a watershed approach to develop TMDLs, including insights
from individuals involved in the actual development process and ensuing implementation activities. To
the extent information was available, various "lessons learned" are presented highlighting difficulties
encountered, realized benefits of the watershed process, and implementation highlights.
Note that most of the case studies represent fairly large-scale efforts, allowing for the illustration of
multiple issues, sometimes on an exaggerated scale, related to watershed TMDL development. For
example, large watershed TMDLs can address hundreds of waterbody-pollutant combinations and
include multiple states and numerous stakeholders. The increased scale provides more opportunity for
unique technical and programmatic issues to arise and therefore illustrate in these case studies.
However, this is in no way meant to imply that "watershed TMDLs" must be, by definition, large in area.
Small-scale watershed TMDLs can still result in the same types of efficiencies and benefits as those
developed for larger geographic areas.
The following table summarizes the case studies included in this appendix.
Case Study
Callegas Creek, CA
Gauley River, WV
Long Island Sound, NY & CT
Lower Fox River, Wl
Potomac River PCB, VA, WV, MD
St. Mary's River, IN
Tualatin River, OR
Virgin River, Utah
Type
3
£
X
X
X
X
X
X
Regional
X
X
Pollutants
Phosphorus
X
X
X
Nitrogen
X
X
X
Sediment
X
X
X
Bacteria
X
X
(0
;o
o
w
n
>
o
(0
(0
b
15
X
X
c
Q)
U)
>•,
X
Dissolved C
X
X
X
X
c
E
ns
Q.
P
Biological 1
X
Q)
Temperatu r
X
X
Aluminum
X
c
2
X
Manganese
X
u>
&
ns
W
X
00
o
Q_
X
X
(0
ns
•s
5
ns
O
X
Program Factor
U)
^P
P
NPDES Per
X
X
U)
c
K
X
X
•<*
w
X
X
0
w
o
X
X
o
LL.
o
X
c
o
•s
•5
n
o)
Source Wat
X
ns
c
O
+^
.:£
Multi-Jurisd
X
X
o
?
o
Consent De
X
DRAFT
A-1
-------
Handbook for Developing Watershed TMDLs December 2008
A-2 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 1: Calleguas Creek, California
Watershed TMDL at a Glance
Waterbody:
Calleguas Creek
Drainage Area:
343 square miles
Parameters:
Organochlorine (OC) Pesticides and PCBs
Trash
Nutrients (ammonia, oxidized nitrogen, and
algae/dissolved oxygen (DO))
Salts (boron, chloride, sulfate, total dissolved
solids (IDS))
Additional TMDLs were developed separately for
siltation, metals, and toxicity.
Development Status:
The TMDLs were developed from 2001 to 2007.
Developed by:
Larry Walker Associates on behalf of the
Calleguas Creek Watershed Management Plan for
the Los Angeles Regional Quality Control Board
Jurisdictions:
Ventura County and a small portion of western
Los Angeles County.
# Impaired Segments Listed:
OC pesticides and PCBs - 11 segments
Trash - 2 segments
Nutrients -12 segments
Salts - 11 segments
Technical Approach:
The OC pesticides and PCB loads into and out
of Callegas Creek subwatersheds were
calculated using DDE as a representative
constituent. A numerical model was developed
for this purpose.
A conceptual mass balance spreadsheet model
forms the basis of the information used to
develop the nutrient and salts TMDLs.
Source Types:
NPS and PS
Background
Calleguas Creek and its tributaries are located in southeast Ventura County and a small portion of
western Los Angeles County. Calleguas Creek drains an area of approximately 343 square miles from
the Santa Susana Pass in the east to Mugu Lagoon in the southwest. The main surface water system
drains from the mountains in the northeast part of the watershed toward the southwest where it flows
through the Oxnard Plain before emptying into the Pacific Ocean through Mugu Lagoon. The watershed,
which is elongated along an east-west axis, is about thirty miles long and fourteen miles wide. The Santa
Susana Mountains, South Mountain, and Oak Ridge form the northern boundary of the watershed; the
southern boundary is formed by the Simi Hills and Santa Monica Mountains.
Land uses in the Calleguas Creek watershed (CCW) include agriculture, high and low density residential,
commercial, industrial, open space, and a Naval Air Base located around Mugu Lagoon. The watershed
includes the cities of Simi Valley, Moorpark, Thousand Oaks, and Camarillo. Most of the agriculture is
located in the middle and lower watershed with the major urban areas (Thousand Oaks and Simi Valley)
located in the upper watershed. The current land use in the watershed is approximately 26% agriculture,
24% urban, and 50% open space. Patches of high quality riparian habitat are present along the length of
Calleguas Creek and its tributaries.
The watershed is generally characterized by three major subwatersheds: Revolon Slough in the west,
Conejo Creek in the south, and Arroyo Simi/Las Posas in the north. Additionally, the lower watershed is
also drained by several minor agricultural drains in the Oxnard plain. Figure 1 depicts the CCW with reach
names and designations used in the OC Pesticides and PCB TMDL, the three major subwatersheds, and
six smaller subwatersheds which are defined for analysis and modeling in this TMDL (Mugu, Revolon,
Calleguas, Conejo, Arroyo Las Posas, and Arroyo Simi).
DRAFT
A-3
-------
Handbook for Developing Watershed TMDLs
December 2008
The main surface water system drains from the mountains toward the southwest, where it flows through
the Oxnard Plain before emptying to the Pacific Ocean through Mugu Lagoon. Dry weather surface water
flow in the Calleguas Creek watershed is primarily composed of groundwater, municipal wastewater,
urban non-stormwater discharges, and agricultural runoff. In the upper reaches of the watershed,
upstream of any wastewater discharges, groundwater discharge from shallow surface aquifers provides a
constant base flow. Additionally, urban non-stormwater runoff and groundwater extraction for construction
dewatering or remediation of contaminated aquifers contribute to the base flow. Stream flow in the upper
portion of the watershed is minimal, except during and immediately after rainfall. Flow in Calleguas Creek
is described as storm peaking and is typical of smaller watersheds in coastal southern California.
Impairment Listing Information
There are 11 waterbodies included on the 303(d) impaired list for OC pesticides and PCBs, 2 for trash, 12
for nutrients, and 11 for salts (Table 1).
Table 1. 303(d) Impaired Waterbodies
Waterbody
Mugu Lagoon, Lower
Calleguas Creek, Upper
Calleguas Creek, Lower
Camrosa Diversion
Revolon Slough
Beardsley Channel
Arroyo Las Posas
Arroyo Simi, Upper
Tribs to Arroyo Simi
Dry Calleguas
Arroyo Conejo, Upper
Arroyo Conejo, Lower
Conejo Creek
Conejo Creek, Mainstem
Conejo Creek, Hill Canyon
Arroyo Santa Rosa
Conejo Creek, North Fork
Conejo Creek, South Fork
Pollutant
OC Pesticides, PCB, Nutrients
Nutrients, Chloride, IDS
OC Pesticides, PCB, Nutrients
Sulfates, IDS
OC Pesticides, PCB, Trash, Nutrients, Boron, Sulfates, TDS
OC Pesticides, PCB, Trash
OC Pesticides, PCB, Nutrients, Chlorides, Sulfates,
TDS
Nutrients, Boron, Chlorides, Sulfates, TDS
Boron, Chlorides, Sulfates, TDS
Nutrients
Nutrients
Nutrients
OC Pesticides, PCB
OC Pesticides, PCB, Chlorides, Sulfates, TDS
OC Pesticides, PCB, Chlorides, Sulfates, TDS
OC Pesticides, PCB, Nutrients, Sulfates, TDS
OC Pesticides, PCB, Nutrients, Sulfates, TDS
OC Pesticides, PCB, Nutrients, Chlorides, Sulfates,
TDS
A-4
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Watershed Features
Figure 7. Map of Calleguas Creek Watershed, showing 303d listed reaches for OC pesticides and PCBs, (Source: Larry
Walker Associates, 2005)
Applicable Standards and Water Quality Targets
California implements the federal water quality standard requirements by providing for the reasonable
protection of designated beneficial uses through the adoption of water quality objectives (CA Water Code
§13241)which may be numeric values or narrative statements. For inland surface waters in the Los
Angeles Region, beneficial uses and numeric/narrative objectives are identified in the Basin Plan and
additional numeric objectives for toxic pollutants are contained in the California Toxics Rule as adopted
by the U.S. EPA (40 CFR 131.38).
OC Pesticides and PCBs
Numeric targets identify specific goals for the OC Pesticides and PCBs TMDL. Multiple numeric targets
are often considered when there is uncertainty that a single numeric target is sufficient to ensure
protection of designated beneficial uses. The 2002-303(d) list for the Calleguas Creek Watershed
contains listings for OC pesticides and PCBs (OCs) in the water column, fish tissue, and sediment. In
order to address these listings, water criteria, and fish tissue and sediment guidelines were selected as
numeric targets (Table 2).
DRAFT
A-8
-------
Handbook for Developing Watershed TMDLs
December 2008
Table 2. Numeric targets for water, fish tissue, and sediment for PC Pesticides and PCBs (Larry Walker Associates 2005)
Constituent
Aldrin
Chlordane
Dacthal
ODD
DDE
DDT
Dieldrin
Endosulfan 1
Endosulfan II
Endrin
HCH (alpha-BHC)
HCH (beta-BHC)
HCH (delta-BHC)
HCH (gamma-BHC)
Heptachlor
Heptachlor Epoxide
PCBs
Toxaphene
Water Quality Targets1 (ug/L)
Freshwater
3.0"
0.0043
3500b
NA
NA
0.001
0.056
0.056
0.056
0.036
NA
NA
NA
0.954
0.0038
0.0038
0.014B
0.00020
Marine
1.3"
0.0040
NAb
NA
NA
0.001
0.0019
0.0087
0.0087
0.0023
NA
NA
NA
0.164
0.0036
0.0036
0.030B
0.00020
Fish Tissue
Targets2
(ug/Kg)
0.050
8.3
NAb
45
32
32
0.65
65,000
65,000
3200
1.7
6.0
NA
8.2
2.4
1.2
5.3
9.8
Sediment Targets'* (ug/dryKg)
Freshwater,
TEL
NA
4.5
NA
3.5
1.4
NA
2.9
NA
NA
2.7
NA
NA
NA
0.94
NA
0.6
34'
NA
Marine, ERL
NA
0.5
NA
2.0
2.2
1.0
0.02
NA
NA
NA
NA
NA
NA
NA
NA
NA
23'
NA
[1] USEPA. 2000. Water Quality Standards; Establishment of Numeric Criteria for Priority Toxic Pollutants for the State of
California; Rule.
May 18, 2000. The human health criteria listed are "For the Consumption of Organisms Only".
[2] Obtained from the USEPA 1980 Ambient Water Quality Criteria Documents for each constituent.
[3] Included in the list of "Chem A" pesticides.
[4] The numeric target for aldrin was derived from a combination of aldrin and dieldrin risk factors and BCFs as recommended in
"Ambient Water Quality Criteria for Aldrin/Dieldrin" (USEPA 1980, 1990).
[5] Applies to the sum of all congener or isomer or homolog or Aroclor analyses.
[6] PCBs in water are measured as sum of seven Aroclors.
[7] PCBs in fish tissue and sediment are measured as sum of all congeners.
"NA" indicates that no applicable target exists for the constituent.
Trash
The target for trash is zero.
Nutrients
Multiple, nutrient related numeric targets are applicable to the CCW. Ammonia and oxidized nitrogen
targets apply to all reaches of the watershed. The dissolved oxygen target applies only to the Conejo
Creek system, which is the only reach listed on the 303(d) list for low dissolved oxygen. Algal biomass
targets only apply to the Conejo Creek system, Revolon Slough, and Beardsley Channel, because these
are the only reaches for which algae is listed on the 303(d) list. Table 3 summarizes the nutrient targets
for the watershed.
Table 3. Nutrient targets for the Callequas Creek watershed (Larry Walker Associates 2001)
Reach
Arroyo Simi, Upper
Arroyo Simi/ Las Posas
Dry Calleguas
Arroyo Conejo Upper
Arroyo Conejon, Lower
Arroyo Santa Rosa
Conejo Creek, Upper
Conejo Creek, Lower
Calleguas Creek, Upper
Calleguas Creek, Lower
Revolon Slough and Ag drains
30-Day
average
ammonia
target (mg/L)1
1.8xWER
2.7 x WER
1. Ox WER
1.7xWER
3.1 xWER
2.4 x WER
3.3 x WER
3.5 x WER
2.7 x WER
2.2 x WER
2.5 x WER
1-Hour
maximum
ammonia
target (mg/L) 1
3.9 x WER
8.4 x WER
5.7 x WER
3.2 x WER
8.4 x WER
5.7 x WER
5.7 x WER
4.1 xWER
2.7 x WER
2.7 x WER
4.9 x WER
Nitrate-N
+ nitrite-N
target
(mg/L)
10
10
10
10
10
10
10
10
10
10
10
Dissolved
oxygen target
(mg/L)
Minimum 5.0
Minimum 5.0
Minimum 5.0
Minimum 5.0
Algal biomass
target (mg/m2
chl-a)
150
150
150
150
150
A-6
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Mugu Lagoon
2.7xWER
2.7xWER
10
1 A Water Effects Ratio (WER) is a mechanism for adjusting national criteria to reflect site-specific conditions in the Calleguas Creek
watershed based on monitoring conducted in the watershed. In the event a site specific WER is not developed for a given reach, the
WER for that reach will be set equal to 1.0.
2 This algal biomass target has been selected from literature and is not based on local consensus as to what constitutes nuisance
conditions. This target can be adjusted based on further studies and public input.
Salts
The Basin Plan numeric water quality objectives were selected as numeric targets for the salts TMDL
covering boron, chloride, sulfate and IDS (Table 4). These numeric targets were applied at the base of
each of the subwatersheds defined in the TMDL for allocations.
Table 4. Salts numeric targets (Larry Walker Associates 2007)
Subwatershed
Simi
Las Posas
Conejo
Camarillo
Pleasant Valley
(Calleguas Creek
Reach 3)2
Pleasant Valley
(Reaches 4 and 5)3
Boron target (mg/L)1
1.0
1.0
Chloride target (mg/L)
150
150
150
150
150
150
Sulfate target (mg/L)
250
250
250
250
250
250
IDS target (mg/L)
850
850
850
850
850
850
1. The Boron target only applies to the subwatersheds containing listed reaches. The other subwatersheds do not exceed the boron
objective.
2. The targets apply upstream of Potrero Road. Downstream of Potrero Road, the creek is tidally influenced and the salt objectives
do not apply.
3. The targets apply upstream of Laguna Road. Downstream of Laguna Road, the creek is tidally influenced and the salt objectives
do not apply.
Stakeholder Process: A Third Party TMDL
The Calleguas Creek Watershed Management Plan Committee has been active since 1996, with its initial
purpose being to develop a comprehensive watershed management plan for Calleguas Creek that
addresses a broad range of land use, environmental, resource management, economic, public
infrastructure, recreation, and other issues. Participants on the Committee included representatives of
public agencies as well as private stakeholders. In 2001, the group began discussions with the Regional
Board and USEPA to provide assistance in the development of the TMDLs for the watershed, making the
Calleguas Creek TMDLs some of the first third-party TMDLs to be developed and approved. In December
2002, the group developed TMDL work plans for most constituents on the 2002-303(d) list. A Watershed
Management Plan subcommittee assisted with the development of the TMDLs to ensure input from local
expertise and reach a broad group of stakeholders. The subcommittee also participated in development
of implementation plans to resolve the water quality problems within the watershed. Stakeholders
included representatives of cities, counties, water districts, sanitation districts, private property owners,
agricultural organizations, and environmental groups with interests in the watershed.
As a third party TMDL, a high level of stakeholder involvement has occurred throughout the TMDL
development process. There have been no interventions from outside groups, and much of the work has
been performed, or paid for, by members of local government agencies with partial USEPA grant funding.
Pollutant Sources
OC Pesticides and PCBs
Most sources of pesticides and PCBs to surface waters in the CCW are related to historical uses. .
Agricultural runoff is likely responsible for the majority of OC pesticides introduced into the watershed
overtime. Past use of PCBs as coolants and lubricants in transformers, capacitors, and other electrical
equipment is suspected as the primary source of PCB residues. Available evidence suggests that
POTWs, groundwater, atmospheric deposition, and imported water are not responsible for major
contributions to current loading of OCs in the watershed (Larry Walker Associates 2005).
DRAFT
A-7
-------
Handbook for Developing Watershed TMDLs December 2008
Trash
Sources of trash in Revolon Slough and Beardsley Wash include: storm drains, wind action, and direct
disposal. It is estimated that 80% is due to nonpoint sources and 20% is due to point sources (CRWQCB
2007).
Nutrients
POTWs were identified as the most significant source of ammonia and TKN in the watershed. Nonpoint
sources, groundwater, and atmospheric deposition did not contribute significant loads of ammonia or TKN
to the creek system. For oxidized nitrogen compounds (nitrate+nitrite), POTWs are only a significant
source if they have implemented nitrification treatment processes without also denitrifying. Agriculture
contributes significant oxidized nitrogen loadings to the watershed, especially in Revolon Slough. Other
nonpoint sources, groundwater, and atmospheric deposition contribute significantly smaller loadings of
oxidized nitrogen (Larry Walker Associates 2001).
Salts
Conceptually, six possible sources of salts to the watershed exist: water supply (water imported from the
State Water Project or Freeman Diversion and deep aquifer groundwater pumping), water softeners,
POTW treatment chemicals, atmospheric deposition, pesticides and fertilizers, and indoor water use
(chemicals, cleansers, food, etc.). These salts are then transported through POTW discharges and dry
weather runoff to three possible endpoints: surface water, shallow groundwater, and/or stranded on the
watershed in the soils. The salts stranded in the soils are eventually transported to surface water when
precipitation mobilizes them and carries them to the creek system. Groundwater pumping and exfiltration
move salts from groundwater to surface water and surface water infiltration transports salts from the
surface water to groundwater. Additionally, groundwater saturation of historic marine sediments can
mobilize existing background salts from previously dry soil and transport them to the groundwater.
However, none of these transport mechanisms add salts, they just move salts from one endpointto
another. Salts transported in the surface water to the ocean are currently the only salts that are exported
from the watershed (Larry Walker Associates 2007).
Available Data
Since the mid-1990s various studies have been conducted to assess water, sediment, and fish tissue
quality in the CCW. Portions of the data collected through these studies were incorporated into the
1996,1998, and 2002 LARWQCB Water Quality Assessments to identify exceedances of water quality
objectives.
Technical Approach
OC Pesticides and PCBs
Concentrations of OCs in water are primarily the result of loads from nonpoint sources and discharges
from point sources. The TMDL analysis considers OC loads into and out of CCWsubwatersheds, using
DDE as a representative constituent. The numerical model developed for this purpose is characterized
as:
• Empirical - Based on the statistics of the available data;
• Static- Simulating conditions as annual averages;
• Stream reach based - Simulating conditions in representative stream reaches; and
• Water quality based - Focused on the physical and chemical conditions in the modeled stream reaches
that determine concentrations and loads of OCs in water.
Load (mass per time, L) is calculated as the product of concentration (C) and flow rate (Q):
L = C*Q
Flow rates for each land use in each subwatershed were calculated from daily mean values or water
years 1990-2003 estimated by the Dynamic Calleguas Creek Modeling System, or DCCMS (LWA,
2004b). DDE concentrations from major sources to water were estimated based on the land-use runoff
and discharge data.
A^8 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Loads were calculated according to two different methods. First, DDE sources from land-use runoff and
point source discharges to receiving water were quantified. The sum of those loads presumably
represents the total load of DDE to water. Second, actual DDE loads in water from representative
reaches were quantified using receiving water data presented in the Current Conditions section. These
loads should mirror the loads to water, but could vary depending on the effects of in-stream processes
(Larry Walker Associates 2005).
Trash
Final WLA and LA are zero trash (CRWQCB 2007).
Nutrients
Conceptual models of nutrient inputs and interactions were developed The assumptions and analyses
used to develop the conceptual models formed the basis of the information used to develop the nutrient
TMDLs. The conceptual models were used to assess potential sources of nutrients to the watershed and
identify processes that impact surface water concentrations of these constituents. A spreadsheet model
was used to obtain a general idea of the impacts on surface water concentrations resulting from
implementation of BMPs in the watershed (Larry Walker Associates 2001).
Salts
The framework for the salts modeling effort was a numerical mass balance water quality model originally
developed for use in the Calleguas Creek Nutrient TMDL effort and accepted by State and Federal
regulatory authorities for use in the Nutrient TMDL process for the CCW.
The water quality simulation component of the CCMS (acronym) is built on a spreadsheet mass balance
model. To model the CCW, the entire watershed was divided into 15 subwatersheds based on drainages
to sampling locations and significant tributaries. A computational element was assigned to each
subwatershed for calculating the changes in stream flow and water quality due to processes present
along stream reaches circumscribed by the subwatersheds. The model was expanded to accommodate
stochastic input, which allows calculation of the likely distribution of in-stream salts concentrations (Larry
Walker Associates 2007).
Allocations
OC Pesticides and PCBs
Table 5. DDE Allocations
Subwatershed
Mugu Lagoon
Calleguas Creek
Revolon Slough
Arroyo Las Posas
Arroyo Simi
Conejo Creek
Total (lb/yr)=
Percent of Total
Average Annual DDE Load (Ibs/yr)
Urban
0.07
0.31
0.16
0.09
0.93
0.8
2.35
7.1%
Native
-
-
-
-
-
-
-
-
Agric
2.79
5.48
11.3
6.84
2.05
2.09
30.55
92.9%
POTW
-
0.21
-
0.09
0.67
0.79
1.76
5.3%
GW
-
0.03
-
-
0.02
0.01
0.06
0.2%
Total
2.9
5.8
11.5
6.9
3
2.9
32.9
100.0%
Trash
Both point sources and nonpoint sources were identified as sources of trash in Revolon Slough and
Beardsley Wash. For point sources, the strategy for attaining water quality standards focuses on
assigning WLAs to the California Department of Transportation (Caltrans) Permittees and Co-Permittees
of the Ventura County Municipal Separate Storm Sewer System (MS4) Permit, including the Ventura
County Watershed Protection District, the City of Camarillo, and the City of Oxnard. The WLAs will be
implemented through permit requirements. For nonpoint sources, the strategy for attaining water quality
standards focuses on assigning LAs to land owners and agencies in the vicinity of Revolon Slough and
Beardsley Wash, including the County of Ventura, City of Camarillo, City of Oxnard, and Agricultural
DRAFT
A-9
-------
Handbook for Developing Watershed TMDLs
December 2008
entities in the Revolon Slough and Beardsley Wash subwatersheds. The LAs will be implemented through
regulatory mechanisms that implement the State Board's 2004 Nonpoint Source Policy such as
conditional waivers. Final WLA and LA are zero trash (CRWQCB 2007).
Nutrients
Both WLAs and LAs were expressed as concentrations, not loads. WLAs for oxidized nitrogen and
ammonia were set for five of the six POTWs in the watershed. Wasteload allocations were not determined
for the Olsen Road treatment plant because it was in the process of being shut down. WLAs for TKN
were developed only for the two treatment plants that discharge to the Conejo Creek system, the Hill
Canyon and Camarillo plants. The Camrosa plant is able to achieve all proposed nutrient WLAs with
existing facilities and operations. Moorpark is currently constructing facilities that will allow compliance
with the WLAs by September 2001. The Camarillo plant is able to achieve the proposed ammonia WLA
with existing facilities and operations but would have to reduce oxidized nitrogen levels by about 70% to
achieve the proposed WLAs for these pollutants. The Hill Canyon plant is able to achieve the proposed
oxidized nitrogen and ammonia WLAs with existing facilities and operations but would have to reduce
TKN levels by about 60% to achieve the TKN WLA. The Simi Valley plant would have to reduce ammonia
levels by between 70% and 90% (depending on the site-specific adjustment to the ammonia criteria) to
achieve the ammonia WLA and would have to reduce nitrate levels by between 40% and 70% (depending
on the degree of nitrification needed to achieve the ammonia WLA) to achieve the oxidized nitrogen WLA.
Load allocations were set only for agricultural discharges of oxidized nitrogen. In the Calleguas Creek
watershed, agriculture is assigned a load allocation of 10 mg/L of nitrate-N + nitrite-N. All other non-point
sources of nutrients were sufficiently below the numeric target and comprised such a small portion of the
total pollutant loading that they were not considered to be significant loadings and, consequently, were
not assigned load allocations. Load allocations were assigned to agriculture as a category, rather than to
individual dischargers. Agricultural loadings of oxidized nitrogen would require an average reduction of
about 70% to meet the assigned load allocations.
Salts
The loading capacity was calculated using the average of the critical condition dry weather flow rates. The
following table (Table 6) represents the current loading capacity of the stream with the percent reductions
in current average loads to achieve loading capacity. However, the loading capacity will increase over
time as the POTW flows increase to design flow. The loading capacity shown in the table represents all of
the flow discharged to the stream. Some of this flow is removed from the stream through groundwater
recharge and diversions. However, the flow is available for carrying load prior to its removal from the
stream and is therefore considered in the loading capacity.
Table 6. Current salt loading capacity with the percent reductions in current average loads to achieve loading capacity.
Subwatershed
Simi
Las Posas
Conejo
Camarillo
Pleasant Valley
(Calleguas)
Pleasant Valley
(Revolon)
Boron loading
capacity (Ibs/day)
117(0%)
131 (0%)
127(0%)
152(0%)
182(0%)
50 (39%)
Chloride loading
capacity (Ibs/day)
17,593(0%)
19,721 (0%)
19,073(7%)
22,756(13%)
27,247(12%)
7,552 (4%)
Sulfate loading
capacity (Ibs/day)
29,322 (28%)
32,869 (28%)
31,788(0%)
37,927 (0%)
45,411 (4%)
12,586(38%)
TDS loading
capacity (Ibs/day)
99,695 (14%)
111,754(14%)
108,080(0%)
128,953(2%)
154,398(3%)
42,793(15%)
Implementation
Each TMDL report includes an Implementation Plan. Implementation of the TMDLs operates within an
adaptive management framework where compliance monitoring, special studies, and stakeholder
interaction guide the process as it develops through time. Compliance monitoring is envisioned to
generate information critical for measuring progress toward achievement of WLAs and LAs, and may
suggest the need for revision of those allocations in some instances. Additionally, data from ongoing
A-10
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
monitoring could reveal necessary adjustments to the implementation timeline and may serve to initiate
reevaluation when appropriate. Special studies will increase understanding of specific
conditions/processes in the watershed, allowing for more accurate prediction of results expected from
various implementation efforts. Thus, adaptive management allows this TMDL to be an ongoing and
dynamic process, rather than a static document.
OC Pesticides and PCBs
The Implementation Plan includes the following elements:
• Source control activities to reduce any active sources of OC pesticides and PCBs in the watershed;
• Implementation and evaluation of agricultural best management practices (BMPs) in the watershed;
• Special studies to identify sediment transport and OC content and areas where BMP implementation
may be more effective.
• Monitoring for OC pesticides and PCBs in water, fish tissue, and sediment throughout the watershed.
Trash
Compliance with the TMDL is based on the Numeric Target and the Waste Load and Load Allocations,
which are defined as zero trash in Revolon Slough, Beardsley Wash and their tributaries. Consequently,
compliance is based on implementing a program for trash assessment and collection, or alternatively for
point source dischargers, full capture devices, to attain a progressive reduction in the amount of trash in
Revolon Slough, Beardsley Wash and their tributaries. Dischargers who do not implement full capture
devices shall propose a program for a Minimum Frequency of Assessment and Collection (MFAC). The
MFAC program is required to attain a progressive reduction in the amount of trash collected from Revolon
Slough, Beardsley Wash and their tributaries through implementation of BMPs. Dischargers may
implement structural or nonstructural BMPs as required to attain a progressive reduction in the amount of
trash in Revolon Slough, Beardsley Wash and their tributaries (CRWQCB 2007).
Nutrients
Implementation of the TMDLs is phased to allow implementation of the ammonia TMDL first and the
oxidized nitrogen and algae/dissolved oxygen TMDLs at a later date. The implementation plan includes a
schedule with the dates by which wasteload allocations are to be incorporated into NPDES permits and
load allocations go into effect. The implementation plan also outlines special studies that may be
conducted during the implementation period to address the uncertainties in the TMDLs.
Salts
The goal of the TMDL implementation plan is to achieve a salts balance within the CCW, attain water
quality standards, and protect salt-sensitive beneficial uses. Through achieving a salts balance, water
quality is expected to improve and allow achievement of water quality standards.
Through achievement of a salts balance, surface water and groundwater quality within the CCW should
improve because salt will no longer build-up in surface soils and groundwater basins. In addition, the
implementation actions include elements to ensure the protection of sensitive beneficial uses in the CCW.
A salt balance will be achieved through the implementation of actions to:
1. Reduce the amount of salts imported into the CCW.
2. Reduce the amount of salts added to water in the CCW.
3. Transport salts down gradient and export them out of the watershed.
4. Provide protection to sensitive beneficial uses.
5. Monitor and track achievement of the salt balance and the associated impacts on water quality.
The implementation actions described in this plan represent a range of activities that could be conducted
to achieve a salts balance in the watershed. The implementation plan has been developed as a phased
plan to allow for a review of implemented actions to assess the impacts on the salt balance and water
quality. The specific actions taken to achieve the salt balance may vary to some degree from the
elements presented here based on this evaluation and future analyses of the most cost effective and
DRAFT A-11
-------
Handbook for Developing Watershed TMDLs December 2008
beneficial mechanisms for achieving the salt balance. To the extent possible, all ideas being considered
as mechanisms for implementing the TMDL have been included in the plan. Future considerations may
result in other actions being implemented rather than the options presented. However, any proposed
actions will be reviewed using the salt balance model to ensure the action does not adversely impact
other implementation actions in the watershed or the salt balance of a downstream subwatershed.
References
California Regional Water Quality Control Board (CRWQCB). 2007. Trash Total Maximum Daily Load
For Revolon Slough and Beardsley Wash in the Calleguas Creek Watershed.
Larry Walker Associates. 2007. Calleguas Creek Watershed Boron, Chloride, TDS, and Sulfate TMDL,
Public Review Technical Report.
Larry Walker Associates. 2005. Calleguas Creek Watershed OC Pesticides and PCBs TMDL Technical
Report.
Larry Walker Associates. 2001. Calleguas Creek Nutrient TMDLs.
A-12 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 2: Gauley River TMDLs, West Virginia
Jurisdictions:
Watershed incorporates all or portions of 8
counties
# Impaired Segments Listed:
2006 303(d) List included 104 impaired streams
Technical Approach:
Selenium - criterion x flow
All other pollutants: Continuous Simulation -
Mining Data Analysis System (MDAS)
Source Types:
NPS and PS
Watershed TMDL at a Glance
Waterbody:
Gauley River
Drainage Area:
8-digit HUC, 1,419 square miles
Parameters:
Total Iron
Dissolved Aluminum
Fecal coliform bacteria
Total selenium
PH
Biological Impairments addressed as well
Development Status:
Submitted to EPA Sept. 2007
Developed by:
West Virginia DEP
Background
The Gauley River watershed, in southern West Virginia encompasses approximately 1,419 square miles
of mountainous, mainly forested (85.7%) lands. Other important landuses include grasslands (7%),
mining (2.4%), abandoned mine lands (AMI) (1.5%), and urban/residential (1.3%). The TMDLs
developed for streams in the Gauley River watershed were completed in the context of significant
historical TMDL activity in West Virginia. From 1997 to September 2003, the USEPA developed a series
of large scale waershed TMDLs in West Virginia under the settlement of a lawsuit which resulted in a
consent decree between EPA and the plaintiffs1. The consent decree established a rigorous schedule
under which TMDLs were to be completed for the waters included on West Virginia's 1996 303(d) list.
The schedule included TMDL development dates extending into March 2008. Beginning in October 2003,
West Virginia took over development of the consent agreement TMDLS. Since then, all TMDLs
developed for West Virginia streams have been developed by the West Virginia Department of
Environmental Protection (WVDEP). The Gauley River watershed TMDLs account for a portion of the
TMDLs required under the consent decree plus additional TMDLs for segments identified as impaired
since the 1996 listing cycle.
Prior to October 2003, while WVDEP was assisting EPA in the development of the consent decree
TMDLs, WVDEP was also working to build its own TMDL program. With the help of a TMDL stakeholder
committee, the agency secured funding from the state legislature and created the TMDL section within
the Division of Water and Waste Management.
The TMDL stakeholder committee consisted of 22 members with balanced interests among extractive and
manufacturing industry, environmental advocates, agriculture, forestry, state and federal government,
sportsmen associations, and municipalities. The committee made recommendations for WVDEP TMDL
development and supported general revenue funding. For additional details please see the highlight box
in Section 2 of the main document.
1 Ohio Valley Environmental Coalition, Inc., West Virginia Highlands, et al., v. Browner et al.
DRAFT
A-13
-------
Handbook for Developing Watershed TMDLs December 2008
Since WVDEP assumed responsibility for development of its consent decree related and other TMDLs,
they have worked to enhance and broaden the applicability of the watershed TMDL development tools
initially developed by EPA as a matter of necessity to facilitate development of large scale TMDL in a
timely manner. They have also integrated the TMDL program into the overall watershed management
program, utilizing a rotating basin monitoring framework and a 48 month TMDL development cycle. The
process used to develop the TMDLs for the Gauley River watershed are typcial of all TMDLs developed
for West Virginia waterbodies.
Impairment Listing Information
Based on extensive pre-TMDL monitoring conducted from July 2003 to June 2004, VW DEP refined the
impairments of previous listing cycles and identified other impaired waterbodies that were not previously
listed. After the 2006 listing cycle, 104 streams were included on VW's 303(d) list for impairments related
to numeric criteria for fecal coliform bacteria, dissolved aluminum, total iron, total selenium, and pH. In
addition, waters were also included for biological impairments.
For assessment and modeling purposes, the Gauley River watershed was divided into TMDL watersheds,
major contributing streams draining directly to the Gauley mainstem for which TMDLs were to be
developed. TMDL watersheds were further subdivided into subwatersheds, a more detailed delineation
of smaller catchments where pollutant sources, allocations and reductions are better incorporated into
permitting activities and TMDL implementation (Figure 2). The entire Gauley watershed contains 520
subwatersheds. The 15 TMDL watersheds were delineated into 447 subwatersheds containing impaired
waters or contributed to impaired waters.
Applicable Standards and Water Quality Targets
Because multiple listings and parameters of concern were involved in the TMDLs for the Gauley River
watershed, the array of uses, standards and applicable targets is wide. VWDEP issued a single TMDL
report document that discussed the individual TMDLs separately. However, the analysis performed to
develop the TMDLs for parameters and impairments that are related (e.g., iron, dissolved aluminum, pH,
fecal coliform bacteria, and biological) was not done separately. For example, to account for all potential
stressors affecting benthic organisms (biological impairment), it was necessary to evaluate multiple
factors such as sedimentation, acidity, dissolved metals toxicity, and organic enrichment. Once the
potential stressors for each biologically impaired stream were identified, one or more pollutant specific
TMDLs were developed to address the biological impairment. Table 7 summarizes the applicable water
quality standards and targets of each TMDL analysis.
A-14 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
S g Beaver Cree
MuddletyC
Liele Laurel Creek
Tirtey Creek
Peters Cree
Twentymile Creel
Scrabble Creeic
Ricfi Creek
I I ~MDL Watersnecs
•33 jley River Watersneds
~] Big Beaver Creek
^ Big Rjn
™ Cherry River
Cranberry River
Hominy Creek
Little Laurel Creek
Meadow River
Mudd;e:y Creek
Panther Creek
_] Peters Creek
~" Rich Creek
| Scrabble Creek
I "urkey Creek
: Twentymile Creek
Williams River
•Vllans River
Cranberry Rwe'
Cherry Rive-
'antrer Creek
Horrirry Creek
Mesdwv River
15
25 Mies
Figure 2, Gauley River watershed subbasin delineation
Table 7, Applicable Standards and Water Quality Targets
POLLUTANT
Aluminum
dissolved
(|jg/i-)
Iron, total
(mg/L)
Selenium, total
(|jg/i-)
PH
Fecal coiform
bacteria
USE DESIGNATION
Aquatic Life
Wamwater Fisheries
Acute3 Chronic"
750
—
20
6.0 to 9.0
750
1.5
5
6.0 to 9.0
Troutwaters
Acute3 Chronic"
750
—
20
6.0 to 9.0
87
0.5
5
6.0 to 9.0
Human Health
Contact
Recreation/Public
Water Supply
1.5
10
6.0 to 9.0
Human Health Criteria Maximum allowable level of fecal coliform content for Primary
Contact Recreation (MPN or MF) shall not exceed 200/100 mL as a monthly geometric
mean based on not less than 5 samples per month; nor to exceed 400/100 mL in more
than 1 0 percent of all samples taken during the month.
One hour average concentration not to be exceeded more than once every 3 years on the average
Four-day average concentration not to be exceeded more than once every 3 years on the average
DRAFT
A-15
-------
Handbook for Developing Watershed TMDLs December 2008
Pollutant Sources
As one might expect for such a large watershed, with the exception of selenium, pollutant sources include
both point and nonpoint sources. Hughes Fork of Bells Creek in the Twentymile Creek TMDL watershed,
was the only impaired reach listed for selenium. The listing was based on WVDEP collected data during
the pre-TMDL monitoring period for two violations of the chronic aquatic life criterion. Given the high
selenium content of coals in the region, the prevalence of mining activity and the association of selenium
mobilization from surface disturbance activities, it was deemed that the only source of selenium is from
mining activities in the watershed. Therefore the TMDL assigned WLAs to all mining permits in the
watershed, applying the water quality criteria as end of pipe concentrations. For the other impairments,
bacteria, metals, pH and biological impairments, source assessments found that both point and nonpoint
sources attributed to impairments. Generally, pollutant sources included mine drainage from active sites
and abandoned mine lands, untreated sewage and sediment.
To assess the biological impaired streams, a stressor identification process was applied. Sources of
biological stressors are often analogous to those already described: mine drainage, untreated sewage,
and sediment. The general stressor identification process involved reviewing available information,
forming and analyzing possible stressor scenarios, and implicating causative stressors. Candidate
causes of biological stresses included metals contamination, acidity (low pH), high sulfates and increased
ionic strength, increased TSS and erosion, altered hydrology, algal growth (food supply shifts), ammonia
toxicity, and chemical spills. Ultimately, metals toxicity, pH toxicity, ionic toxicity, sedimentation and
organic enrichment were implicated as stressors responsible for the biological impairment listings. TMDLs
were developed for all except the ionic toxicity listings, for which adequate data were not available for
TMDL development. Those listings were retained on VW's 303(d) list for future assessment.
Waters identified with metals and pH toxicity also demonstrated exceedences of iron, aluminum or pH
criteria. VWDEP determined that implementation of those TMDLs would address the biological
impairments. Waters identified with organic enrichment stressors also demonstrated exceedences of
fecal coliform criteria; WVDEP determined that implementation of the bacteria TMDLs would address
those impairments.
The stressor identification process indicated sedimentation as a causative stressor for four biologically
impaired streams. WVDEP initially pursued the development of TMDLs directly for sediment for those
streams. The approach involved selection of a reference stream with an unimpaired biological condition,
prediction of the sediment loading present in the reference stream, and use of the area-normalized
sediment loading of the reference stream as the TMDL endpoint for sediment-impaired waters.
Additionally, all of the sediment-impaired waters also were impaired pursuant to total iron water quality
criteria and the TMDL assessment for iron included representation and allocation of iron loadings
associated with sediment. In each stream, the sediment loading reduction necessary for attainment of
water quality criteria for iron exceeds that which would be necessary under the reference approach. As
such, the iron TMDLs were deemed an acceptable surrogate for biological impairments from
sedimentation.
Technical Approach
To address the selenium impairment, WVDEP used a simple calculation of the assimilative capacity for
selenium available at the mouth of Hughes Fork (the only stream impaired by selenium) at 7Q10 flow.
WLAs for contributing sources are based on achieving the chronic aquatic life criterion in the discharge.
Additional impairments for which TMDLs were developed were addressed using a modeling application,
the Mining Data Analysis System (MDAS) that was originally developed by EPA Region III to complete
several early consent decree TMDLs that necessitated analysis of large scale, data intensive analyses.
Since WVDEP began developing their own TMDLs in 2003, they have continued to apply the MDAS
system, supporting development of additional capabilities as the needs of the TMDL program have
required. MDAS allowed for a performing a continuous simulation of loading in the watershed, predicting
both loads and in-stream concentrations, and representation of all major pollutant sources. With
A-16 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
customization, the watershed model was used to represent critical processes and factors associated with
evaluating compliance with water quality critieria such as frequency and duration of exceedences. Key
technical factors associated with developing TMDLs for the Gauley River and accommodated by the
model include:
• Large scale analysis,
• Point and nonpoint sources,
• Temporally and flow variable metals and bacteria impairments,
• pH related impairments,
• Time-variable landuse practices and their impacts on water quality,
• Variable and weather-dependant transport mechanisms for metals and bacteria.
From the data analysis, it was determined that separate processes may be contributing to pH
impairments in the watershed, either historical mining related metals discharges or acid deposition (wet
and dry) in conjunction with low watershed buffering capacity. MDAS was enhanced with new simulation
modules to more explicitly simulate the processes associated with both sources of pH impairments. For
details on the model and specific processes represented, please see the Gauley River Watershed TMDL
Report and its supporting technical report and appendices (West Virginia DEP, 2007).
Allocations
Generally, WV's water quality critieria and an explicit margin of safety were used to identify endpoints for
the TMDLs and associated allocations. An implicit MOS was included in the selenium TMDL, where
WLAs were prescribed for the surface mining point sources at water quality criteria at the end-of-pipe.
Except for selenium, the TMDLs are presented as average annual loads because they were developed to
meet TMDL endpoints under a range of conditions observed throughout the year. Analysis of available
data indicated that critical conditions occur during both high- and low-flow events; the TMDLs were
therefore developed using continuous simulation (modeling over a period of several years that captured
precipitation extremes), which inherently considers seasonal hydrologic and source loading variability.
Equivalent, daily average TMDLs are also presented.
The selenium TMDL is presented as an equation for the maximum daily load that is variable with
receiving stream flow. For pH impairments associated with atmospheric deposition, TMDLs are
presented as the annual net acidity load associated with maintenance of the pH TMDL endpoint of 6.02.
The allocation process applied a "top-down' methodology, where headwaters were first analyzed because
of their impacts on downstream water quality. Load contributions were reduced from applicable sources
and the TMDLs were identified. The loading contributions of unimpaired headwaters and the reduced
loadings for impaired headwaters were then routed through downstream waterbodies. Using this method,
contributions from all sources were weighted equitably. Reductions in sources affecting impaired
headwaters ultimately led to improvements downstream and effectively decreased necessary loading
reductions from downstream sources. Nonpoint source reductions did not result in loadings less than
natural conditions, and point source allocations were not more stringent than numeric water quality
criteria.
Critical source categories receiving allocations for each TMDL are summarized in Table 8. Due to the
level of detail at which source allocations were determined, WVDEP presented summary allocation tables
for LAs and WLAs in the TMDL document. For detailed, source-specific allocations by subbasin,
WVDEP also developed spreadsheet tables in Microsoft Excel to assist in implementation efforts.
DRAFT A-17
-------
Handbook for Developing Watershed TMDLs
December 2008
Table 8. Allocated Sources for the Gau/ey River Watershed TMDLs
TMDL
Dissolved Aluminum
Total Iron
Fecal coliform
PH
Selenium
Sources receiving LAs
Abandoned mine lands (disturbed
land, highwalls, deep mine discharges
and seeps)
Bond forfeiture sites (unreclaimed
mine areas)
Sediment associated landuses (barren
land, harvested forest, oil and gas well
operations, unpaved roads,
streambank erosion)
Pasture/Grass areas
Onsite Sewage Treatment Systems a
Residential areas
Background and Other NPS (wildlife) b
Atmospheric deposition
NA
Sources receiving WLAs
Active mining operations
Non-mining point sources (e.g., industrial
stormwater)
Construction stormwater
Future growth construction stormwater
NPDES permitted facilities (e.g., STPs)
NA
Active mining operations
illegal discharges such as those from failing septic systems and straight pipes as well as sanitary sewer overflows received zero
allocations
received no reductions
Implementation
Because of the prevalence of mining related permits in the watershed, the WLAs assigned to those
facilities represent the results of what is in fact a watershed-based permitting analysis. Individual WLAs
were presented in terms of both load and concentration. The concentration-based WLAs were prescribed
as the operable term since the TMDL allocations reflected pollutant loadings that are necessary to
achieve water quality criteria at distinct locations (i.e., the pour points of delineated subwatersheds). In
contrast, effluent limitation development in the permitting process is based on the
achievement/maintenance of water quality criteria at the point of discharge. Future permits are not
precluded in impaired watersheds and will be assigned end-of-pipe limits at applicable water quality
criteria. For aluminum, where the criterion relates to the dissolved portion of the metal, an appropriate
total to dissolved translator must be applied. In many cases, the implementation of the TMDLs for fecal
coliform will consist of providing public sewer service to unsewered areas.
With regard to water quality trading to implement TMDL allocations, the TMDL neither prohibits nor
authorizes trading in the watersheds addressed in the document. According to the TMDL document:
"WVDEP generally endorses the concept of trading and recognizes that it might become an
effective tool for TMDL implementation. However, significant regulatory framework development
is necessary before large-scale trading in West Virginia can be realized. Furthermore, WVDEP
supports program development assisted by a consensus-based stakeholder process. Before the
development of a formal trading program, it is conceivable that the regulation of specific point
source-to-point source trading might be feasible under the framework."
As part of the permit review process, permit writers must incorporate the required TMDL WLAs into new
or reissued permits. Both the permitting and TMDL development processes have been synchronized with
the Watershed Management Framework cycle, such that TMDLs are completed just before the permit
expiration/reissuance timeframes. Existing permit reissuance in the Gauley watershed is scheduled to
begin in July 2007 for non-mining facilities and in January 2008 for mining facilities. Therefore, the WLAs
for existing activities will be promptly implemented. New facilities will be permitted in accordance with
future growth provisions.
WVDEP uses a geographically based approach to manage water resources - the Watershed
Management Framework. WVDEP uses the Framework to organize and prioritize activities associated
with monitoring and TMDL development and implementation. Each year, TMDLs are developed in
specific geographic areas according to the schedule and priorities mapped out in the Framework. In
A-18
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
addition to scheduling and developing TMDLs, the Framework includes a schedule forTMDL
implementation.
The Framework provides a six-step process for developing integrated management strategies
(monitoring, TMDLs, implementation through permitting, grants, etc.), and action plans for achieving the
state's water quality goals. Step 3 of the process involves identification of strategies needed to meet
needed pollutant reductions (TMDLS). Steps 5 and 6 provide for preparation, finalization and
implementation of a Watershed Based Plan to improve water quality. These plans are based on the
efforts of locally based watershed teams composed of members of the West Virginia Watershed Network.
The West Virginia Watershed Network is an informal association of state and federal agencies, and
nonprofit groups promoting watershed management in the state. On an annual basis, the Network
evaluates WVDEP's Watershed Management Framework to coordinate and prioritize its existing
programs, local watershed associations and limited resources. The evaluation includes a review of TMDL
recommendations for watersheds under consideration. Formation of local teams and development of
Watershed Based Plans is assisted by WVDEP's Nonpoint Source Program.
Lessons Learned: Developing Watershed TMDLs in the Real World2
WV has chosen to continue the precedent set by the early consent decree TMDLs of large-scale analysis
because it allows for a more coordinated response to water quality issues both in terms of collecting data,
analyzing pollutant sources (modeling), and implementing solutions. Critical lessons learned by the
program are summarized below.
• Pre-TMDL monitoring and pollutant source tracking is important not only for adequately
representing sources but for achieving programmatic efficiency. WVDEP strategically plans
water quality monitoring prior to TMDL development where numerous monitoring locations are
established and a comprehensive suite of analytes are sampled. This fine scale monitoring resolution
coupled with identification and characterization of problematic sources through field-based source
tracking activities provides a sound basis for assessment and TMDL development for all streams and
impairments within the watershed. This watershed based approach allows WVDEP to maximize
efficiency throughout all phases of TMDL development and thereby minimizing funding requirements
of their TMDL program.
• Large scale, highly detailed modeling facilitates efficiency and permit coordination. As noted
above, the comprehensive watershed based approach employed by WVDEP, typically includes all
known impairments in the watershed. This involves a multi-faceted modeling approach to address
total recoverable metals, dissolved metals, acidity (pH), bacteria, and biological impairments. WVDEP
has created unique ways to integrate large-scale, watershed based TMDLs with fine-scale, highly
technical methodologies that produce "implementable" TMDLs in cost-effective manner. Overtime,
this has led to WVDEP utilizing predicted model output in absence of water quality data to identify
waters as threatened and present "modeled impairment TMDLs".
• Highly detailed, source specific allocations must be verified in reality. Typically when large
scale TMDLs are developed, allocations are presented across broad categories that can often "hide"
pollutant sources allowing for significant interpretation for implementation. However, presenting highly
detailed, source specific allocations forces the need to assess the practicality of implementation and
management objectives - a process that can be very complex and labor intensive.
• Large scale modeling facilitates permit coordination. Based on the history of TMDLs in WV,
where large watersheds have been involved, with literally hundreds of point source discharges, WV
has realized the importance of synchronizing the permitting and TMDL development processes. In
effect, all TMDLs in the state are essentially, watershed-based permits. The permitting mechanism
relies on issuance of individual permits to facilities but provisions of applicable TMDLs are
incorporated.
2 Special thanks to Dave Montali, TMDL Program Manager, WVDEP
DRAFT A-19
-------
Handbook for Developing Watershed TMDLs December 2008
• Developing user-friendly TMDL end products provides sound basis for implementation
guidance. WVDEP has designed a "TMDL on CD" concept where all relevant TMDL information
(TMDL Reports and Appendices, Technical documentation, and supporting data) is included on a CD-
ROM. To further improve the "usability" of the TMDLs, WVDEP developed a series of interactive
tools to provide TMDL implementation guidance. These tools are designed to simplify and assist
"implementers" (nonpoint source staff and permit writers) in using the TMDLs to develop watershed
plans and issue/renew permits. An interactive ArcExplorer geographic information system (CIS)
project allows the user to explore the spatial relationships of the source assessment data, as well as
further details related to the data. Users are also able to "zoom in" on streams and other features of
interest. In addition, spreadsheet tools (in Microsoft Excel format) were developed to provide the data
used during the TMDL development process, and the detailed source allocations associated with
successful TMDL scenarios. These tools provide guidance for selection of implementation projects
as well as for permit issuance and are also included on the TMDL Project CD.
• Continual improvement - a key to success. As VWDEP's TMDL program continues to evolve,
focus is placed on refining all facets of TMDL development. Over the past 5 years, significant
improvements have been made in pollutant source tracking and data transfer from WVDEP to their
TMDL development contractor, where spreadsheet tools and databases have been developed to
specifically address issues that have been previously encountered. Furthermore, WVDEP continually
attempts to improve upon programmatic issues as feedback is provided from stakeholders and
interagency personnel.
References
West Virginia DEP, 2007. Total Maximum Daily Loads for Selected Streams in the Gauley River
Watershed, West Virginia. September 2007. Prepared by Tetra Tech, Inc., Charleston, WV.
West Virginia DEP, 2007. Total Maximum Daily Loads for Streams in the Gauley River Watershed, West
Virginia. FINAL TECHNICAL REPORT, September 2007. Prepared by Tetra Tech, Inc.,
Charleston, WV.
A-20 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 3: Long Island Sound Nitrogen TMDL
Watershed TMDL at a Glance
Waterbody:
Long Island Sound
Area:
16,000 square miles
Parameters:
Nitrogen
Year Developed/Approved:
2001
Developed by:
Connecticut Department of Environmental
Protection (CDEP) and New York State Department
of Environmental Conservation (NYDEC)
# Impaired Segments Addressed:
Long Island Sound in its entirety; TMDL addresses
all sources of Nitrogen to the sound
Jurisdictions:
Watershed includes most of Connecticut, portions
of Massachusetts, New
Hampshire, Vermont, a small area in Canada, as
well as portions of New York City, and
Westchester, Nassau, and Suffolk Counties in
New York.
Technical Approach:
Developed and applied a linked circulation and
water quality model (LIS 3.0) to simulate hypoxia
in Long Island Sound
Incoming loads were represented based on a
variety of data.
Nonpoint source loads: atmospheric monitoring
data, landuse specific export coefficients based
on literature and calibrated to monitoring data
Point source loads: point source monitoring data
Source Types:
Point and Nonpoint
Background
The watershed of Long Island Sound (LIS) drains an area over 16,000 square miles, covering most of
Connecticut, portions of Massachusetts, New Hampshire, Vermont, a small area in Canada, as well as
portions of New York City, and Westchester, Nassau, and Suffolk Counties in New York. The Sound
itself is about 1300 square miles, measuring about 100 miles long and 21 miles at its widest point.
Depths in the middle of the Sound range from 60-120 feet. It is one of the most highly urbanized and
suburbanized areas on the eastern seaboard.
Hypoxia has for sometime been a common occurrence in the bottom waters of Long Island Sound,
usually occurring in the late summer months of July through September. It is linked to excess nitrogen
and is exacerbated by the naturally occurring density stratification of the water column. Resulting algal
blooms and anoxic conditions contribute to insufficient habitat for submerged aquatic vegetation and
overall impairment of the health and functioning of the Sound. Due to the complexity of the problem and
the cost magnitude of potential remedies, it was understood from an early point that new, flexible
approaches would be needed to address and implement solutions to the hypoxia issue.
Applicable Standards
In neither New York nor Connecticut, were there numeric criteria for nitrogen. The LISS determined
however that reducing nitrogen loads necessary to achieve water quality standards for dissolved oxygen
(DO) would protect and maintain designated uses in the Sound. In both cases, DO standards for saline
(New York) and coastal/marine waters (CT) were used (Table 9).
DRAFT
A-21
-------
Handbook for Developing Watershed TMDLs
December 2008
Table 9. DO criteria for Long Is/and Sound
Class
SA
SB
SC
1
SD
SA
SB
Description
Marketable shellfishing, recreation, fishing
Recreation and fishing
Fishing (must be suitable for fish propagation and survival)
Secondary contact recreation and fishing
Fishing (must be suitable for fish survival)
Marine fish, shellfish and wildlife habitat, shellfish harvesting
for human consumption, recreation
Marine fish, shellfish and wildlife habitat, shellfish harvesting
for transfer to depuration, recreation, industrial and other such
as navigation
DO Criterion
Not less than 5.0
mg/L
Not less than 5.0
mg/L
Not less than 5.0
mg/L
Not less than 4.0
mg/L
Not less than 3.0
mg/L
Not less than 6.0
mg/L
Not less than 5.0
mg/L
State
NY
NY
NY
NY
NY
CT
CT
Pollutant Sources
Nitrogen is provided by multiple sources, which enter through tributary inputs, the two ocean boundaries
and by atmospheric deposition. Nonpoint sources include traditional land-based runoff such as from
agricultural and developed land areas. Point sources include over 80 treatment plants. Additionally,
CSOs and stormwater discharges also provide significant loading to the Sound. Given the geographic
scale of the Long Island Sound TMDL and the land use based approach used to estimate loadings, it was
not feasible to meaningfully separate loadings from point source stormwater runoff and CSOs from the
general nonpoint source categories, with the exception of the New York City CSO loads. As a result,
these loads were considered nonpoint sources for the TMDL.
Available Data
Beginning in 1985, New York, Connecticut and the EPA formed the Long Island Sound Study (LISS) to
promote measurable improvements to the water quality of the Sound. Intensive water quality monitoring
of the Sound was conducted from April 1988 to September 1989. Over 25 constituents were monitored,
including transparency, salinity, temperature, nutrients and their forms, chlorophyll a, DO, BOD, total
organic carbon, and suspended solids. CTDEP expanded on this monitoring to include monthly sampling
at 18 stations with expansion to 30 additional stations in the summer months. Data from NYCDEP's
Harbor Survey (16 stations) were used as well. Additionally, the Interstate Sanitation Commission has
collected weekly surveys in the Narrows and western basin during summer months (21 stations) for which
temperature, salinity, and DO at multiple levels is measured. Several citizens monitoring programs also
made data available to the effort.
Watershed loads were estimated from various data sources. Categories of loading data evaluated
included point source related data and nonpoint source data. The bulk of nitrogen loading was
determined to be due to point source contributions. CTDEP and NYSDEC collected NPDES discharge
monitoring data (flow and effluent quality) from the major municipal and industrial dischargers in the
drainage area. CSO data (for New York City) were used in determining incoming loads; while CSO
related loads (and other stormwater and landuse related loads) for CT and the rest of the contributing
area were estimated using landuse specific export coefficients. Additionally, attenuation in rivers was
estimated based on flow data and other information. Atmospheric deposition was estimated based on wet
deposition monitoring data.
TMDL Development Approach
In 1994, the LISS completed what is known as a Comprehensive Conservation and Management Plan
(CCMP) under the EPA's National Estuary Program. The CCMP for Long Island Sound identified seven
A-22
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
priority issues for the Sound, the highest of which was low dissolved oxygen levels in the Sound. By
1998, the LISS CCMP adopted a 58.5 percent reduction target for nitrogen loads to the Sound and
specified in its implementation plan that a TMDL for nitrogen be adopted with load and waste load
allocations for all sources in the watershed. As a result, the states of New York and Connecticut jointly
submitted a TMDL for nitrogen which was approved by EPA in 2001.
Partner Coordination and Stakeholder Involvement
The states were primarily responsible for writing and submitting the TMDL for EPA approval. However, in
this case, the LISS partnership provided the bulk of the technical support, including water quality
modeling, which provided the scientific underpinning for the TMDL. Although the TMDL is clearly
identified as a CT and NY product, there was a lot of effort on the part of EPA Long Island Sound Office
and a very comprehensive public process providing input and comment that went well beyond the LISS
partnership.
Technical Approach - Linking Sources and Water Quality
A three-dimensional, time-variable hydrodynamic/water quality model (LIS 3.0) of Long Island Sound was
developed to simulate hypoxia conditions in the Sound. The model incorporates advanced physical,
biological and chemical processes that in turn relate nutrients and carbon-based pollutants to
phytoplankton dynamics and DO. The model was calibrated with extensive monitoring data collected as
part of the LISS. Boundary conditions for the model were established based on analysis of various types
of data related to nutrient loading from the contributing area.
The LIS 3.0 model was calibrated using 18-months of ambient monitoring data (April 1988 through
September 1989). Tributary loadings and CSO contributions to the model were determined using time-
variable rainfall and flow data. The calibrated model was reviewed by the LISS Modeling Evaluation
Group and approved as being appropriate for use as a predictive tool. It was then applied to TMDL
development.
Based on model results, it was demonstrated that nitrogen loadings throughout the year contribute to the
pool of nitrogen available for phytoplankton uptake and therefore, long-term annual loading controls
rather than seasonal or shorter term limits are warranted.
Allocations
For purposes of implementation, the LIS TMDL distinguishes between in-basin and out-of-basin sources.
In basin sources are those originating from within the CT and NY portions of the drainage basin including
those directly deposited to the Sound's surface. Out-of-basin sources are all other sources beyond the
in-basin boundaries. Both the in- and out-of-basin sources were also further subdivided into a pre-
colonial load and a human induced load. The pre-colonial condition estimates what natural loading may
have been. Both the in-basin and out-of-basin loads are made up of various sources such as wastewater
treatment facilities, nonpoint runoff, CSOs and atmospheric deposition. Atmospheric deposition was
accounted for in the geographic category in which it is deposited. Table 10 provides a summary of what
sources were considered under either the Load orWasteload Allocation portion of the TMDL. Because
CT and NY cannot enforce nitrogen reductions from point and atmospheric sources in other states,
specific WLAs for facilities contributing to out-of-basin loads were not identified in the TMDL.
DRAFT A-23
-------
Handbook for Developing Watershed TMDLs
December 2008
Table JO. Sources considered under WLA and LA portions of the TMDL
In-basin
WLAs
LAs
WWTPs
CSOs (NY)
Pre-colonial
Terrestrial3
Atmospheric
Out-of-basin
WLAs
LAs
None There was
a gross 25% WLA
recommended for
out of basin
sources. EPA
Region 1 is
treating that as a
binding WLA.
Pre-colonial
Tributary Loads D
Boundary Loads0
a includes CSOs in CT as well as stormwater contributions
b Include contributions from groundwater, overland runoff, atmospheric deposition, CSOs and VWVTP discharges in Massachusetts,
Vermont and New Hampshire
0 from Atlantic Ocean and New York Harbor
Note that with the exception of New York City CSOs, for which monitoring data were available to
characterize loading quantities, nitrogen loads from stormwater runoff and CSOs (in CT) were given load
allocations. The TMDL states that "development of the phase II stormwater permitting program over the
next few years will provide opportunities for the states to elucidate the load from stormwater sources and,
building on the phase II regulations, identify appropriate wasteload allocations."
Implementation
The TMDL is designed to be implemented in phases. Phase III represents achieving a 58.5% reduction
from in-basin sources within a 15-year time frame (beginning August 1999) with five-year incremental
targets. 100% of the 58.5% reduction is to be met by 2014. However, Phase III reductions are not
enough to meet the necessary reductions for attaining water quality criteria. The entire TMDL is the sum
of the Phase III reductions (58.5% of in-basin sources), reductions from outside the basin, and application
on non-treatment alternatives necessary to meet water quality standards for DO with an implicit margin of
safety.
TMDL = 58.5 % in-basin reduction + out-of-basin reductions + non-treatment alternatives + margin of
safety
Phases IV and V address action necessary to meet out-of-basin reductions and non-treatment
alternatives.
Significant implementation activities under Phase III have involved the development of a general permit
for nitrogen discharges in CT covering all 79 of the state's WWTPs as well as a nitrogen credit exchange
program. Each year the permit specifies an annual statewide aggregate target for nitrogen removal as
new upgrades are brought online. It also sets annual end-of-pipe limits in pounds per day of TN,
apportioned by plant discharge volume to meet the aggregate statewide target. Under the permit,
facilities can purchase or sell nitrogen credits annually based on the facilities' performance with respect to
their annual limit. According to state program staff, development of the trading program was a
considerable effort, but has paid dividends in terms of rate of implementation, and cost efficiency.
CT has also seen some cross-program implementation successes in focusing stormwater permitting
towards nitrogen control, and emphasizing nitrogen controls in Sec. 319 nonpoint source program. In
terms of impacting atmospheric sources, the TMDL has resulted in pushing atmospheric controls forward,
particularly through NOx reduction goals established within the New England Governors and Eastern
Canadian Premiers "Acid Rain Action Plan".
Monitoring
A-24
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Significant resources have been devoted to post TMDL monitoring including additional atmospheric
deposition monitoring to better characterize the relative importance of that source to overall loading to the
Sound. Land based contribution estimates were also refined through the development of an HSPF
watershed model of the drainage basin in Connecticut. Watershed modeling was used to improve
understanding of how nitrogen is attenuated in the watershed and is especially relevant to implementation
efforts related to BMPs as well as to understanding acceptable trading ratios between point source
trading partners.
Lessons Learned: Developing Watershed TMDLs in the Real World3
Ultimately, the desirability of developing a watershed TMDL for LIS included technical, environmental,
political and programmatic reasons. Technical and environmental, as discussed above, but also political
because the impairment cannot be corrected unilaterally by one state. And, each state recognized that
their efforts alone wouldn't be sufficient to do the job so there had to be assurance that both CT and NY
would proceed equally. Other contributing state reductions are currently under more scrutiny and will be
subject of next year's (expected 2009) revised TMDL to provide a more complete and comprehensive
plan, building on the recommendations of the current TMDL. Among the reasons for revising the TMDL
are a new Systemwide Eutrophication Model (SWEM) as well as new state DO criteria
(http://www.ct.gov/dep/lib/dep/water/water quality standardsl/wqs.pdf).
Facilitation of complex technical evaluation
One of the primary technical advantages of having developed the LIS TMDL on a watershed scale was
having the support of EPA and the LISS to assist with the technical evaluations. According to the state
TMDL program, the complex technical evaluations necessary for the LIS TMDL would not likely have
been accomplished as well if the states were working independently on the problem. Programmatically,
having both states on board and implementing the same level of reduction in both states minimized
"finger pointing".
Dealing with obstacles
While there were advantages associated with developing the TMDL on the watershed scale, difficulties
and frustrations were also encountered during the process including the sheer magnitude of work
necessary to understand the problem and the length of time required to work through the process as well
as difficulties associated with navigation of interstate and multimedia issues.
According to state program staff familiar with the LIS TMDL, the magnitude of the required scientific
understanding required to understand what needs to be done to restore LIS was probably the biggest
obstacle. Costs were high in terms of both time and money. The LISS highlighted hypoxia as a primary
water quality issue, supported monitoring and research, and was instrumental in the modeling required to
develop a credible and defensible management plan, i.e., the TMDL. It took from 1985 until 2001 for the
TMDL to be developed and adopted. Interim plans were developed during that time, but the fact that it
took 15 years of work from discovery of hypoxia to a final management plan reflects the complexity of the
problem as well as the implementation intricacies of all the negotiations over who does what and to what
degree. In addition, input from the regulated community and the public had to be considered. Connecticut
began engaging the regulated community, especially municipalities with sewage treatment plants, as
early as the late 1980's, and kept them informed about their potential role in solving the problem.
Implementing multimedia and multi-jurisdictional controls
Another significant issue has been trying to impact air emissions through a TMDL, which has no binding
on air programs. Better multimedia coordination among EPA's air and water programs has been cited as
being needed to fully implement the air deposition reductions that are identified in the TMDL. Additionally,
strong EPA support in engaging upstream states with respect to multijurisdictional issues is also
necessary for ensuring progress toward the TMDL. For example, a strong EPA voice in the TMDL
process can facilitate getting monitoring requirements for out of state sources.
3 Special thanks to Paul Stacey, Director of Planning and Standards in the Bureau of Water Protection
and Land Reuse of CT DEP
DRAFT A-25
-------
Handbook for Developing Watershed TMDLs December 2008
References
CTDEP and NYDEC, 2000. A Total Maximum Daily Load Analysis to Achieve Water Quality Standards
for Dissolved Oxygen in Long Island Sound.
EPA, 2003. Watershed-Based Permitting Case Study: Final Permit, General Permit for Nitrogen
Discharges (Long Island Sound). Fact Sheet #1.
(http://www.epa.gov/npdes/pubs/wq casestudy factshtl .pdf)
EPA, 2007. Water Quality Trading Toolkit for Permit Writers: Appendix A, Water Quality Trading
Program Fact Sheets, pp. A-13 - A-26.
A-26 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 4: Lower Fox River Basin and Green Bay,
Wisconsin
Watershed TMDL at a Glance
Waterbody:
Lower Fox River (LFR Basin)
Drainage Area:
638 square miles
Parameters:
Phosphorus
Total Suspended Solids (TSS)
Development Status:
Currently in the workplanning and stakeholder
engagement process. Next phase will be TMDL
development; Draft TMDL due Spring 2009
Developed by:
Wisconsin Department of Natural Resources
(TMDL)
Oneida Nation (Watershed Management Plan)
Background
Jurisdictions:
Portions of Brown, Calumet, Outagamie, and
Winnebago Counties as well as most of the
Oneida Nation Reservation.
# Impaired Segments Listed:
24 Segments listed as impaired for phosphorus
or sediment
Technical Approach:
Undetermined but will likely utilize the existing
technical tools already developed for the area
such as the Lower Fox River SWAT watershed
model, and an optimization model used to
evaluate the cost effectiveness of multiple
implementation scenarios.
Source Types:
NPS and PS
The Lower Fox River Basin and Green Bay are important environmental and economic resources for the
State of Wisconsin and the local community. People have long used the river and bay for transportation,
commerce, energy, food, and recreation. Situated on one of the major bird migration routes in North
America, the Mississippi flyway, the Lower Fox River and Green Bay environment provides essential
habitat for breeding and migratory birds. The terrestrial, wetland, and aquatic habitats in the basin support
a wide diversity of songbirds, shorebirds, waterfowl, and birds of prey.
The river and bay also support a nationally known fishery. Green Bay is the largest freshwater estuary in
the world; the bay itself is an inflow to Lake Michigan. The wetlands along Green Bay's west shore, as
well as the wetlands lining the Lower Fox River that flows into Green Bay, provide critical fish spawning
habitat for perch, northern, walleye and the elusive spotted musky. The natural resources of the Lower
Fox River Basin and Green Bay are critical to tourism and the local economy. The river and bay support
popular recreational activities such as hiking, boating, fishing, snowmobiling, cross-country skiing, and ice
fishing. Many local and state parks and wildlife areas are scattered throughout the basin. These areas
provide opportunities to enjoy camping, trails, and hunting, and passive nature observation.
The Lower Fox River Basin encompasses approximately 638 square miles in northeastern Wisconsin
(Figure 3). Portions of several counties and most of the Onieda Nation Reservation lie within the
drainage area. The lower portion of the watershed is designated as an Area of Concern (AOC) under the
Great Lakes Water Quality Agreement between the United States and Canada. The AOC consists of the
lower 6.9 miles of the Fox River below DePere Dam and a 21 square mile area of southern Green Bay
out to Point au Sable and Long Tail Point. In addition to the AOC, the rest of the drainage area of the
Lower Fox River Basin contains streams that have been identified on Wisconsin's 303(d) list as impaired
by phosphorus and/or sediment due to various sources. As a result of the phosphorus and sediment
loading to the River and Lower Green Bay, excess aquatic algal growth contributes to severe depletion of
oxygen levels in the waterbodies of the Basin and the Bay, impacting fish and other aquatic life. Sediment
also decreases the availability of light to support submerged vegetation which in turn provides habitat and
DRAFT
A-27
-------
Handbook for Developing Watershed TMDLs
December 2008
food sources for fish, birds, frogs, turtles, insects and other wildlife. Submerged vegetation also plays an
important role in producing oxygen, stabilizing bottom and shoreline sediments, and taking up nutrients
that would otherwise support nuisance algae.
i II'I Mi \\lll <'()
Lake
Winnehago
^^_ Waters Impaired for
Phosphorus or Sediment
River or Stream
j County Boundary
Lake or River
Oneida Reservation
City
| | Lower Fox Watershed
Figure 3. Lower Fox River Basin (Source: CADMUS 2007)
Impairment Listing Information
The approach will be to establish a Phosphorus and Sediment Watershed TMDL for the LFR and Green
Bay AOC, as well as a Watershed Management Plan (WMP) for the impaired waters within the boundary
of the Oneida Nation Reservation. There are currently 24 impaired segments in the LFR Basin, including
A-28
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
the Green Bay AOC, which will be addressed by the TMDL and WMP (Table 11). Section 303(d) of the
Clean Water Act requires a TMDL be developed for each pollutant for each listed waterbody. The LFR
Basin and Green Bay AOC Watershed TMDL will address impairments through 41 individual TMDLs.
Table 7 7. Impaired Segments Covered Under the Watershed TMDL for the Lower Fox River and Green Bay AOC
Waterbody
Apple Creek Segment 1
Apple Creek Segment 2
Ashwaubenon Creek
Baird Creek Segment 1
Baird Creek Segment 2
Bower Creek Segment 1
Bower Creek Segment 2
Duck Creek Segment 1
Duck Creek Segment 2
Dutchman Creek
East River
East River
Fox R. Lower Segment 1
(D
Fox R. Lower Segment 2
(D
Fox R. Lower Segment 3
(D
Green Bay AOC (inner bay)
(D
Kankapot Creek Segment 1
Kankapot Creek Segment 2
Mud Creek Segment 1
Mud Creek Segment 2
Neenah Slough
Plum Creek Segment 1
Plum Creek Segment 2
Plum Creek Segment 3
County
Brown
Outagamie
Brown
Brown
Brown
Brown
Brown
Brown
Outagamie
Brown
Brown
Brown
Outagamie
Brown
Brown
Brown
Outagamie
Outagamie
Outagamie
Outagamie
Winnebago
Outagamie
Outagamie
Outagamie
Pollutants
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus
Phosphorus
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Phosphorus, Sediment
Sediment
Phosphorus
Phosphorus, Sediment
Sediment
Sediment
Impairments
Degraded Habitat, Dissolved Oxygen,
Temperature
Dissolved Oxygen, Sediment
Degraded Habitat, Dissolved Oxygen
Degraded Habitat, Dissolved Oxygen,
Temperature
Degraded Habitat, Dissolved Oxygen
Degraded Habitat
Degraded Habitat
Dissolved Oxygen, Sediment
Dissolved Oxygen, Sediment
Dissolved Oxygen
Degraded Habitat, Dissolved Oxygen,
Sediment
Degraded Habitat, Dissolved Oxygen,
Sediment
Degraded Habitat, Dissolved Oxygen
Degraded Habitat, Dissolved Oxygen
Degraded Habitat, Dissolved Oxygen
Degraded Habitat, Dissolved Oxygen
Degraded Habitat
Degraded Habitat
Degraded Habitat
Degraded Habitat
Dissolved Oxygen
Degraded Habitat & Temperature
Degraded Habitat & Temperature
Degraded Habitat & Temperature
Note: Bold indicates proposed additions based on impending 2008 Impaired Waters List
Pollutant Sources
Sources of phosphorus and sediment loading to the LFR Basin and Green Bay AOC include polluted
runoff from nonpoint sources, such as the abundance of dairy farms in the area, pastures and crop land,
stormwater, rural and urban land, and construction sites, as well as treated effluent from permitted
municipal and industrial point source dischargers. Point source facilities have already begun to reduce
their discharge of phosphorus as part of their permit requirements established by WDNR. While
additional reductions from point source facilities may be needed to restore water quality in the river and
bay, reducing phosphorus and sediment loading to the LFR Basin and Green Bay AOC will require
significant reductions in polluted runoff from nonpoint sources.
Applicable Standards and Water Quality Targets
Neither Wisconsin nor the Oneida Nation has numeric criteria for phosphorus or TSS. However, a
numeric target is needed for the TMDL and WMP in order to calculate reductions in phosphorus and
sediment loading necessary to meet water quality objectives and protect designated uses. WDNR has
DRAFT
A-29
-------
Handbook for Developing Watershed TMDLs December 2008
proposed the following numeric targets for the LFR and Green Bay AOC TMDL: mean summer total
phosphorus (TP) concentration of 120 ug/L (0.12 mg/L) and mean summer total suspended solids (TSS)
concentration of 16 mg/L. The targets chosen will reflect what is both feasible and reasonable to meet
water quality restoration goals. Numeric targets for phosphorus and TSS for the impaired tributary
streams will be determined by the WDNR and Ad-Hoc Science Team for the LFR Basin by the TMDL
Development Phase.
Available Data
There are 25 years of data in this watershed, next draft to include a brief summary.
Stakeholder Process
A number of factors, including the impairment status of the Lower Fox River Basin, the high priority status
of the Lower Green Bay AOC, and the availability of significant monitoring data to characterize loading in
the watershed, led to area stakeholders to recognize that the Basin offers an excellent opportunity for
developing a TMDL and subsequent implementation plan. The effort that is now underway represents a
unique approach to TMDL development compared to how many are traditionally done. The process is
following a four-phased approach. In Phase I, which is now complete, a preliminary modeling study was
performed to develop a framework in which managers can explore optimal watershed management
options for restoring water quality in the basin. The primary goal of Phase I was to look at a combination
of optimal best management practices (BMPs) for agriculture. This optimization framework will be further
expanded upon to include additional agricultural BMPs and coasts, stormwater BMPs and costs, and
upgrade costs to point sources during the TMDL development phase of the project. Phase II, which is
underway, includes multiple stakeholder involvement efforts. Facilitated stakeholder discussions have
been held for both large and small agricultural producers, crop consultants in the basin, and local agency
staff. Discussions will continue with the stormwater stakeholders throughout the watershed, as well as
elected officials and environmental groups. A basin-wide social indicators survey was sent to 300 dairy
farmers with a 58% response rate. Another set of surveys will be sent to a randomly selected portion of
the general public in the summer of 2008. The social indicators survey will raise awareness of water
quality issues, find out what BMPs are currently being done, help plan for modeling restoration sceneries,
and assist in planning for TMDL implementation. Phase III is development of the TMDL. This phase will
include defining water quality targets, exploring different restoration scenarios, determining reductions
needed for all sources, a draft report, and a final public meeting. Phase IV will include implementation of
the TMDL by both point sources (permitted entities) and nonpoint sources (where cost share resources
are available). An Implementation Plan will be included in the final TMDL report.
Organizational Structure
Figure 4 illustrates the organizational structure and role of those involved with the TMDL for WDNR and
the WMP for the Oneida Nation Reservation. The TMDL development process will be led by WDNR, with
guidance from EPA and technical support from a contractor. Several committees have been formed to
support the development and implementation of the TMDL and WMP: The Outreach Committee, the Ad-
Hoc Science Team, and the Technical Team.
A-30 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
TMDL
Technical
Team
Restoration
Scenarios &.
Load Allocations
USEPA
Regulatory
Guidance &
TMDL Review
& Approval
WDNR
TMDL Lead
Oneida Nation
TMDL
Outreach
Committee
WMPLead
Implementation
Planning
Support
TMDL Ad-hoc
Science Team
Modeling and
Technical
Support
Figure 4. Organizational Structure for the Development of the TMDL and WMP
Outreach Committee
The Outreach Committee will play a key role in public and stakeholder outreach for both development and
implementation of the TMDL. Objectives for this committee include but are not limited to: developing key
messages, developing and implementing a communication and outreach strategy for TMDL development
and implementation, and meeting with key stakeholder groups. The committee will work closely with key
stakeholders (agriculture, stormwater, industrial and municipal dischargers, etc.) to determine and
analyze Best Management Practice (BMP) scenarios as part of the pollutant load reduction optimization
modeling.
Members of the Outreach Committee will be asked to share any information gathered through outreach
tools (media, stakeholder meetings, social indicators) with the TMDL Technical Team when considering
the feasibility of the allocation and restoration scenarios.
The Outreach Committee includes representatives from WDNR, EPA, UW-Green Bay (UW-GB), UW-
Extension, UW-Sea Grant, Oneida Nation, Brown County Land Conservation Department (LCD), Green
Bay Metropolitan Sewerage District (GBMSD) and Fox Wolf Watershed Alliance (FWWA).
Established in fall of 2006, the Outreach Committee meets approximately every two months and has
already held a series of stakeholder outreach meetings to inform the community about the TMDL, answer
questions, and listen to stakeholder concerns. A kick-off meeting to present the TMDL and WMP work
plan and proposed numeric targets was held January 23, 2008. The Outreach Committee will continue to
assist in organizing two additional community meetings during the TMDL development process: 1)
Restoration Scenarios Meeting to review the results of the modeling and discuss feasibility of various
BMPs in meeting TMDL targets (Fall 2008); and 2) Public Comment Period Meeting to present the final
TMDL (Spring 2009).
Ad-Hoc Science Team
The role of the Ad-Hoc Science Team is to contribute local data and scientific expertise to set the numeric
targets and restoration goals of the TMDL. The Ad-Hoc Science Team includes: staff from WDNR, UW-
GB, UW-Milwaukee Water Institute, GBSMD, UW-Sea Grant, Oneida Nation and EPA. The Ad-Hoc
Science Team has already held a series of discussions to analyze the numeric targets for the TMDL.
Data analysis and modeling is currently taking place to define numeric criteria for total suspended solids
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
(TSS) and phosphorus (P) in the tributary streams. Numeric targets for the Lower Fox River, Green Bay
and all the impaired tributaries in the basin, will be established by the beginning of the TMDL
development phase. Once the team completes the scientific analyses and determines the numeric
targets for the TMDL, this team will no longer be active, but participants may be asked to provide scientific
expertise to the Technical Team.
Technical Team
The role of the Technical Team will be to evaluate and comment on various load allocation scenarios and
restoration strategies that have been selected by WDNR. WDNR will welcome and consider comments
from the technical team when deciding on the final methodology to ensure that the allocation scenario is
feasible and will meet water quality standards.
Members were solicited for this team, from those attending various TMDL outreach meetings. However,
due to the large number of people interested in participating, WDNR will select members from each key
stakeholder group (county land and water conservation departments, crop consultants, point source
facilities, agricultural producers, municipalities, etc.). Members with various backgrounds and interests
will be chosen to ensure broad representation. Team members will be chosen by the end of July 2008.
Participation on this team will require attending a series of meetings with a contractor to discuss results of
the modeling and proposed allocation scenarios. While the load allocations and wasteload allocations will
be determined by the WDNR, the decision making process will be informed by the allocation scenario
chosen by the Technical Team. The Technical Team will also be encouraged to participate in the second
Public Outreach meeting to be held in Fall 2008.
Local Watershed Websites
The Lower Fox River Watershed Monitoring Program (LFRWMP)
(https://www.uwqb.edu/watershed/about/index.htm The LFRWMP is a multi-year water monitoring
program which will provide independent, high-quality data that can be used to make resource decisions to
improve water quality and foster habitat restoration within the Fox River Basin. Funded by a grant from
Arjo Wiggins Appleton, the program involves coordination between area high school students and
teachers, university students and researchers from the University of Wisconsin-Green Bay (UWGB) and
the University of Wisconsin-Milwaukee (UWM), the Cofrin Center for Biodiversity, the Green Bay
Metropolitan Sewerage District (GBMSD), and the US Geological Survey (USGS).
University of Wisconsin-Extension Basin Education Initiative
http://basineducation.uwex.edu/lowerfox/index.htm The UW-Extension Basin Education Initiative with the
Wisconsin Department of Natural Resources designs and delivers educational programs, assists
organizations, and builds partnerships to promote understanding and stewardship of Wisconsin's natural
resources at the watershed and landscape scale.
Wisconsin Department of Natural Resources
http://dnr.wi.qov/orq/water/wm/wqs/303d/FoxRiverTMDL/ Lower Fox River & Green Bay Area of
Concern Total Daily Maximum Load (TMDL)
A-32 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
TMDL Development Approach
The previously calibrated and validated Soil & Water Assessment Tool (SWAT) will be used to develop
the TMDL and WMP for the LFR Basin and Green Bay AOC (i.e., to estimate existing loads and calculate
necessary load reductions to meet the numeric targets). SWAT is a distributed parameter, daily time-step
model that was developed by the U.S. Department of Agriculture - Agricultural Research Service (USDA-
ARS) to assess nonpoint source pollution from watersheds and large complex river basins. SWAT
simulates hydrologic and related processes to predict the impact of land use management on water,
sediment, nutrient, and pesticide export. With SWAT, a large heterogeneous river basin can be divided
into hundreds of subwatersheds; thereby, permitting more realistic representations of the specific soil,
topography, hydrology, climate and management features of a particular area. Crop and management
components within the model permit reasonable representation of the actual cropping, tillage, and nutrient
management practices typically used in northeastern Wisconsin.
Modeled output data from SWAT can be easily input to a spreadsheet or database program, thereby
making it easier to model large complex watersheds with various management scenarios efficiently.
Major processes simulated within the SWAT model include: surface and groundwater hydrology, weather,
soil water percolation, crop growth, evapotranspiration, agricultural management, urban and rural
management, sedimentation, nutrient cycling and fate, pesticide fate, and water and constituent routing.
SWAT also utilizes the QUAL2E sub-model to simulate nutrient transport. A detailed description of
SWAT can be found on the SWAT website (http://www.brc.tamus.edu/swat/).
The SWAT model was previously calibrated and validated by Mr. Paul Baumgart (UWGB) in 2005 for use
in estimating TP and TSS loading in the LFR Basin. The SWAT model framework that Mr. Baumgart
applied to the LFR Basin in 2005 was refined as part of a recent demonstration project (by Cadmus and
Mr. Baumgart) and reapplied to estimate the load reduction associated with various combinations of
agricultural BMP scenarios for the LFR Basin. A Quality Assurance Project Plan (QAPP) has already
been developed for the recalibration and validation of the SWAT model for use in the load allocation
modeling analysis for the TMDL and WMP.
SWAT will be used to estimate existing TP and TSS loading to each of the impaired segments in the
basin. SWAT will also be used to calculate the tributary/segment specific loading targets and associated
load reduction necessary to meet the TMDL and the targets in the WMP. TMDLs will be developed for
each impaired segment.
Future efforts will refine SWAT to make use of new data sets of continuous flow and daily loads of TP and
TSS from the five Lower Fox River Watershed Monitoring Program (LFRWMP) monitoring stations. The
urban stormwater component of the model will be updated to allow for the evaluation of phosphorus and
sediment loading MS4 urban areas covered under WPDES Municipal General Permits. MS4 urban
loading and non-MS4 urban loading are tracked differently in a TMDL and subject to different regulatory
requirements, therefore, it is important to have the ability to evaluate loading from both MS4 urban areas
and non-MS4 urban areas. The urban stormwater component of SWAT will also be updated to
incorporate data from recent urban stormwater modeling using the Source Loading and Management
Model (SLAMM) conducted for MS4s within the LFR Basin. SLAMM is used to simulate pollutant loads
and demonstrate compliance with requirements stipulated in NR 151 for urban areas.
In addition, the streambank erosion sub-model within SWAT will be updated. Previously, this component
of the modeling framework had been "turned off' and effectively lumped in with upland sources because
1) county land conservation departments assessed the streambank contributions of TSS and TP during
watershed planning and estimated that they were relatively minor compared to upland sources; and 2)
actual watershed-wide measurements of streambank contributions were not available to calibrate the
model. However, urbanization along tributaries within the LFR Basin is likely to change the hydrologic
regime and potentially create unstable streambanks and beds, which could contribute significant loads of
TSS and TP to the streams. Data from an ongoing sediment source tracing study of LFR tributaries has
recently been made available, and may be used to calibrate the SWAT model. This study, which is being
conducted by the UWGB and UW-Milwaukee, utilizes radionuclide analysis of suspended sediments and
DRAFT A-33
-------
Handbook for Developing Watershed TMDLs December 2008
compares the results to sediment sources, such as streambanks, urban areas, and agricultural fields to
estimate the proportional contribution from each source. Existing local estimates of streambank
contributions will be augmented with data from similar areas to calibrate the model if we determine that
the data from the sediment source tracing study is not enough to support recalibration.
Finally, the SWAT modeling framework will be updated to allow for the simulation of potential load
reductions associated with the implementation of potentially restorable wetlands within the basin.
Potential wetland restoration sites will be identified using the same methodology used for the Rock River
Basin TMDL analysis (WDNR is in the process of documenting the methodology).
Implementation Approach
Define Restoration Goals and Identify Restoration Scenarios
The restoration goals for the TMDL should focus on biological benefits and endpoints for the river and bay
(e.g., increased submerged aquatic vegetation). EPA, WDNR, UWGB, and other creditable sources have
previously identified approaches and potential restoration goals. WDNR, the Oneida Nation Reservation,
the Outreach Committee and the Technical Team will help define the restoration goals for the TMDL and
WMP.
WDNR, the Oneida Nation Reservation, and the Outreach Committee will identify all potential restoration
scenarios to analyze with the load reduction optimization-modeling framework. Both agricultural and
urban stormwater BMPs, as well as point source facility upgrades will be considered. The agricultural
BMPs evaluated in the demonstration project will be reevaluated; additional agricultural BMPs will be
identified for inclusion in the optimization analysis. WDNR, the Oneida Nation Reservation, and the
Outreach Committee will also identify urban stormwater BMPs to include in the optimization analysis.
WDNR, the Oneida Nation Reservation, and the Outreach Committee will discuss BMP options with
stakeholders and assess the potential for implementation success based on the BMP's feasibility,
acceptability, and sustainability. Many of the BMPs necessary to achieve the phosphorus and sediment
reduction goals for the LFR Basin and Green Bay AOC will require voluntary cooperation from
landowners. The Outreach Committee has been conducting an assessment to identify socioeconomic
indicators to gain a better understanding of the social systems that influence water quality in the LFR
Basin and Green Bay AOC. The results of the social indicators work will be used to help gauge the
potential effectiveness of the various BMPs that have outreach and behavior change components.
Perform Cost Analysis of Restoration Scenarios
A final list of BMPs and other restoration scenarios will be included in the cost and load reduction
optimization analysis. Site-specific (i.e., local) total annual costs associated with implementation of each
of the agricultural and urban stormwater BMPs in the LFR Basin will be calculated. This estimate will
include all costs associated with the BMP, including implementation expenses and costs associated with
incentives (e.g., provided by the government or other agency). Costs will also take into account both
initial implementation costs and annual operation and maintenance (O&M) costs. In 1991, the Southeast
Wisconsin Regional Planning Commission (SEWRRC) prepared a technical report entitled "Costs of
Urban Nonpoint Source Water Pollution Control Measures," which includes estimated costs for urban
BMPs. The costs of some of the BMPs in this report (i.e., those selected by WDNR, the Oneida Nation
Reservation, and the Outreach Committee) will be updated to reflect current costs.
BMPs have varying lifetimes; therefore all costs will be reduced to their annual values. Annualizing BMP
costs provides a means of comparing BMPs by cost and supplying cost values that can be utilized in
conjunction with average annual TP and TSS load reduction estimates associated with the BMPs to
identify the optimal combination of BMPs. Point source facility upgrades (including O&M costs) will be
estimated on costs provided by point sources (time permitting) or calculated based on similar studies
throughout the state.
Perform Load Reduction Optimization Analysis
A watershed-level optimization-modeling framework will be used for determining the optimal combinations
of BMPs and potential point source facility upgrades for reducing TP and TSS loading in the LFR Basin
A-34 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
and Green Bay AOC. Site-specific BMP and point source facility upgrade costs will be used in
conjunction with estimated load reductions (from SWAT) associated with implementation of the BMPs and
facility upgrades to identify the ten most cost-effective combinations of implementation scenarios that
achieve the TMDL targets for TP and TSS. The optimization analysis will be conducted using SWAT in
conjunction with a Generic Algorithm Optimization Model (OptiMod), which is a refinement of the
Optimization Model previously developed for the demonstration project.
The most cost effective combinations of restoration scenarios that achieve the TMDL targets for TP and
TSS will be presented to EPA, WDNR, the Oneida Nation Reservation, the Outreach Group, and the
Technical Team who will comment on the scenarios and provide input to WDNR and Oneida
Nation. WDNR and Oneida National will choose the final scenario to serve as the basis for the
implementation plan for the TMDL and WMP.
Lessons Learned: Developing Watershed TMDLs in the Real World
Start Upstream
Since this watershed is rich with 25 years of data it was a logical place to start with a TMDL project. The
Upper Fox and Wolf Basins, above the Lower Fox River covers more land area and has less data . Since
50% of the total load to the Lower Fox River and Green Bay is entering from the Upper Fox and Wolf
Basins, TMDLs must be completed in the near future to reach the water quality goals proposed for the
Lower Fox River TMDL.
Agency Coordination is Critical
Since this watershed TMDL will include point and nonpoint sources, WDNR realizes the importance of
synchronizing the permitting, runoff management, water quality standards, and TMDL development
processes. Involving additional agency personnel, particularly at the regional level, is important to the
appropriate development and implementation of a TMDL.
Using the Local Hero Approach
WDNR has been lucky to have an active local watershed group, including university professors,
spearheading the original efforts for data analysis in the watershed. Involving stakeholders in facilitated
discussions and surveys has helped find out what BMPs are currently being implemented, plan for
modeling restoration scenarios, and will help pave the way for TMDL implementation. Active public
involvement will be a vital part of the development and implementation of the Lower Fox River Basin and
Green Bay TMDL. Accomplishing reductions in phosphorus and sediment loadings to the river and bay
will require participation from every community member. It is hoped that implementation scenarios will be
carried out by individual "local heroes" and "tributary teams" throughout the watershed.
Having lots of data helps, not having numeric criteria is problematic
Having 25 years of water quality data for the watershed has made modeling efforts for the watershed
much easier than if significant amount of data collection had been necessary. However, once a model
has given an output, not having codified numeric criteria from which to calculate a TMDL, is problematic.
WDNR has had to include the additional step of developing watershed specific numeric targets for both
phosphorus and TSS in this watershed. Fortunately the plethora of monitoring data in the basin has given
WDNR and researchers the opportunity to base site-specific numeric targets on local scientific data.
References
CADMUS, 2007. Final Report: Integrated Watershed Approach Demonstration Project, A Pollutant
Reduction Optimization Analysis for the Lower Fox River Basin and the Green Bay Area of Concern
Wisconsin Department of Natural Resources. 1993. Lower Green Bay Remedial Action Plan, 1993.
Update for the Lower Green Bay and Fox River Area of Concern. Wisconsin Department of
Natural Resources. 152 pp.
DRAFT A-35
-------
Handbook for Developing Watershed TMDLs December 2008
A-36 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 5: Tidal Potomac and Anacostia PCB TMDL
Jurisdictions:
Virginia Department of Environmental Quality
District of Columbia Department of the
Environment
Maryland Department of the Environment
# Impaired Segments Addressed:
28
Technical Approach:
Linked modeling applications based on various
loading estimation methodologies depending on
source
Source Types:
IMPS and PS
Watershed TMDL at a Glance
Waterbody:
Tidal reaches of the Potomac and Anacostia
Rivers
Drainage Area: The drainage area is divided into
(1) direct drainage and (2) non-tidal upstream
tributaries. The definition and interpretation of
tributary and direct drainage areas have been
defined by the Chesapeake Bay Program.
Parameters:
Polychlorinated Biphenyls (Total PCBs)
Development Status:
Approved on October 31, 2007
TMDL Developed by:
Interstate Commission on the Potomac River
Basin
Background and Rationale for Using a Watershed Approach
A total of 28 segments of the tidal Potomac and Anacostia Rivers were placed on the Maryland, Virginia,
and District of Columbia's 303(d) lists as being impaired by PCBs (Figure 7). The District of Columbia
Potomac PCB listings were covered by a consent decree between the Environmental Protection Agency
(EPA) and the U.S. District Court scheduled to be addressed by September 30, 2007. While Maryland
and Virginia were not bound by the same deadline, in 2004 the three jurisdictions recognized the
importance of consistent methods and agreed to coordinate their PCB TMDL development for the tidal
Potomac and Anacostia listings.
A consolidated TMDL effort was expected to minimize anticipated confusion associated with making
sense of independent TMDLs completed on different dates, using different models and assumptions, and
possibly reaching different conclusions. A single TMDL effort was also deemed to be more cost effective
as the three jurisdictions were able to share cost associated with data collection, model development, and
stakeholder participation. EPA was closely involved in this project and provided funding for consulting
services and model development. The Interstate Commission on the Potomac River Basin (ICPRB)
assisted the jurisdictions in the TMDL development effort by providing overall coordination and acting as
a technical resource, while consulting services for model development were funded by EPA.
DRAFT
A-37
-------
Handbook for Developing Watershed TMDLs
December 2008
* ImpBHtwnl R«i«r«riG8 Numoet
— Hrtrwyiii*
— Slate tie
• CC 3ftJid| PCB iirtairtd witen
Mar>iand >33ni> PCB mpars-d WWB-*
DOdtiar 2D07
Figure 5, Waters of the Tidal Potomac and Anacostia PCB TMDL
Impairment Listing Information
In all three jurisdictions, a waterbody may be listed as impaired by PCBs based on water column (i.e.,
violation of the water column criteria) or fish tissue data (i.e., fish tissue threshold exceedance). All of the
tidal Potomac and Anacostia segments (Table 12) were placed on the 303(d) list due to elevated fish
tissue concentrations.
Some of these segments are also listed for other reasons (e.g., nutrients, sediments, bacteria, and
impacts to biological communities). The tidal Potomac and Anacostia PCB TMDL addressed only the
PCB listings. The Potomac and Anacostia PCB TMDL replaces a 2003 Anacostia River PCB TMDL
developed for the DC portion of the river.
A-38
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Table 12. PCB impaired waterbodies in the tidal Potomac and Anacostia rivers
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
Impaired Waterbody
Upper Potomac
Middle Potomac
Lower Potomac
Upper Anacostia
Lower Anacostia
Accotink Bay
Aquia Creek
Belmont Bay / Occoquan Bay VA
Chopawamsic Creek
Coan River
Dogue Creek
Fourmile Run
Gunston Cove
Hooff Run & Hunting Creek
Little Hunting Creek
Monroe Creek
Neabsco Creek
Occoquan River
Pohick Creek / Pohick Bay VA
Potomac Creek
Pot. River, Fairview Beach
Powells Creek
Quantico Creek
Upper Machodoc Creek
Tidal Anacostia
*Potomac River Lower
*Potomac River Middle
*Potomac River Upper
Juris
DC
DC
DC
DC
DC
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
VA
MD
MD
MD
MD
Description
Chain Bridge to Key Bridge
Key Bridge to Hains Point
Hains Point to Wilson Bridge (DC/MD border)
DC/MD border to Pennsylvania Ave. bridge
Pennsyl. Ave. bridge to Potomac River
In each Virginia embayment, the impairment generally includes
all tidal waters within the embayment, from head-of-tide to the
Potomac river mainstem. The Potomac River, Fairview Beach,
impairment is an area on the mainstem off the beach.
See the Virginia 2006 Integrated Assessment report for
specific descriptions of the geographic extent of each
impairment.
From head of tide on NE and NW Branches of the Anacostia to
the DC/MD border
Mouth of the Potomac to Smith Point, Charles County
Smith Point to Pomonkey Point, Charles County
Pomonkey Point to DC/MD line at Wilson Bridge
Note: * Only the tidal portion of these MD 8-digit waters have been listed as impaired for PCBs.
Applicable Standards and Water Quality Targets
The designated uses of the tidal Potomac and Anacostia include primary and secondary contact
recreation, protection of aquatic life, and fish consumption. Additionally, the Upper Machodoc Creek is
covered by a shellfish water designation.
The PCB listings for all of the tidal Potomac and Anacostia Rivers were based on the exceedence of the
jurisdictional fish tissue threshold levels (i.e., human health risk assessment limits for fish consumption).
Consequently, the objective of the tidal Potomac and Anacostia PCB TMDL was to identify maximum
allowable PCB loads that would not result in "fish consumption" use impairments.
Table 13lists the specific PCB fish tissue thresholds and water quality criteria used by each jurisdiction.
The variations among these values are mainly due to different assumptions applied by each jurisdiction
with regards to acceptable risk levels, individual weight, exposure duration, and the amount offish and
drinking water intake.
Table 13. Applicable Criteria for the PCBs TMDLs in the Tidal Potomac and Anacostia Rivers
Jurisdiction
DC
MD
VA
Fish Tissue Threshold (ppb)
20
88
54
Water Quality Criteria (ng/L)
0.064
0.64
1.70
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
Stakeholder Process
The TMDL development process was initiated with the formation of a Steering Committee (SC) and
Technical Advisory Committee (TAG). The SC members coordinated activities, guided the TMDL
development process, and routinely updated TAG members on the status of the project. Additionally two
sets of public meetings were held and public comments were solicited.
Steering Committee
Steering Committee representatives included staff from the District of Columbia Department of the
Environment (DCDOE), Maryland Department of the Environment (MDE), Virginia Department of
Environmental Quality (VADEQ), the EPA, the Interstate Commission on the Potomac River Basin
(ICPRB), LimnoTech, Inc., and the Metropolitan Washington Council of Governments (MWCOG).
Specific roles of the individual Steering Committee members included:
• States: regulatory guidance and decision making, monitoring and funding, internal review and
coordination, stakeholder notification.
• EPA: regulatory guidance and decision making, funding,
• ICPRB: coordination, contract management, data analysis, scenario development and TMDL
document development, public meeting facilitation.
• MWCOG: expertise and regional perspective.
• Consultant: model development, technical expertise.
The SC members held conference calls on a regular basis (1-4 times a month). Additionally, topic
specific workgroups were convened, to provide input on relevant issues. Work groups collaborated on
such issues as monitoring, loading estimation methods, modeling, and implementation.
Technical Advisory Committee
Membership of the TAG consisted of institutional stakeholders likely to be affected by the TMDL. These
included civic, environmental, and business groups. TAG participation was solicited via broad email
notification sent out to the identified stakeholders. The participation list was continually updated.
Six publicly noticed TAG meetings were held between September 2005 and September 2007. The TAG
was briefed and asked to provide feedback related to modeling, data analysis, and policy decisions.
Public Information Meetings
Public information meetings were held in each jurisdiction in June 2006 and July 2007. The purpose of
these meetings was to provide the public an opportunity to learn about and comment on the activities
associated with the Potomac and Anacostia PCB TMDL development. Additionally, a draft TMDL report
was made available for public comment on July 17, 2007. Approximately 100 comments were received
from 17 organizations.
Available Data
An initial examination of historical PCB data collected by multiple agencies revealed a lack of consistency
among the analytical methods and geographic differences in the extent of sampling. In order to overcome
these obstacles, the SC restricted initial analysis to a set of commonly reported PCB homologs.
During the course of the project, additional water column and point source discharge samples were
collected (between 2005 and 2007) and analyzed using high resolution, congener based laboratory
techniques. Point source effluent monitoring included sampling of 15 wastewater treatment plants.
Water column samples were collected at the Potomac River at Chain Bridge and 26 other nontidal
tributaries.
Pollutant Sources
A-40 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Three principal data sources were used to develop most of the PCB load estimates: historical PCB data,
new PCB samples, and regression-derived PCB data. These sources were supplemented by additional
information from the literature.
Table 14 summarizes external PCB loads to the Potomac and Anacostia Rivers and the methodology
used to estimate these loads.
Table 14. PCB Sources Considered by the TMDL
Source
Potomac at Chain Bridge
Basin Tributaries D
Direct Drainage
WWTPs
CSOs
Atmospheric Deposition
Contaminated Sites
Bay Boundary
Estimation Method
Loads were generated by applying PCB and carbon regressions with TSS to
daily times series of TSS concentration predicted by the LOADEST
regression model with USGS flow data a
Loads were generated by applying PCB and carbon regressions
with total suspended solids (TSS) to daily times series of TSS concentration
predicted by the Chesapeake Bay Model (WM5). These loads represent a
sum of loads from atmospheric deposition, unidentified contaminated sites,
and stormwater runoff (regulated and unregulated)
Loads were generated by applying PCB and carbon regressions
with TSS to daily times series of TSS concentration predicted by the WM5.
These loads represent a sum of atmospheric deposition, loads from small
tributaries not specified in the WM5, regulated and unregulated stormwater,
unidentified contaminated sites, unspecified point source discharges
Monitoring data (PCB, BOD and flow)
Modeled flows and monitoring data
Literature values
Contaminated soil data maintained by each jurisdiction. Delivery rate to the
nearby waterbody was determined using Revised Universal Soil Loss
Equation, Version 2 (RUSLE2)
POTPCB model
Notes: a This method provided better agreement with observed data than the WM5 model.
b Specific point sources were not characterized for the upstream loads.
For additional details on the source loading estimations and data and methods used to develop them, see
Appendix A of the TMDL Report.
Technical Approach
POTPCB Model
The POTPCB model (Limnotech, Inc. 2007) is a coupled, hydrodynamic, salinity, sorbent dynamics
(sediment), and PCB mass balance model for the tidal portions of the Potomac and Anacostia Rivers. It is
closely based on a similar linked modeling platform developed for the Delaware River Estuary PCB
TMDL. Figure 6 presents a graphic representation of the modeling framework.
POTPCB provided daily water column and sediment concentrations in each of 257 model segments in
response to loading inputs generated externally of the model (see Table 14). The Hydrodynamic model is
based on a version of the Dynamic Hydrologic Model (DYNHYD). For sorbent dynamics and PCB mass
balance, a version of the Water Quality Analysis Simulation Program 5 (WASP)/Toxic Chemical (TOXI5)
was developed.
DRAFT
A-41
-------
Handbook for Developing Watershed TMDLs
December 2008
Hydrodynamic
Mode!
Flows
Tidal
Sorbent Dynamics
Mode!
Resuspertsion
Nei
PCB
Model
Water
Balance
Organic Carbon
Mass Balance
PCB
Mass Balance
Figure 6, POTPCB Model Framework (Source: Tidal Potomacand Anacostia PCB TMDL)
PCB Targets
While POTPCB model simulates water column and sediment concentrations, it does not predict the
associated fish tissue concentrations. Therefore, a method external to the POTPCB model was required
to relate water column and sediment PCB concentrations to fish tissue concentrations. For this purpose,
species-specific bioaccumulation factors (BAFs) were derived from the observed fish tissue, water
column, and surface sediment PCB concentrations. The BAFs were in turn used to establish water
quality and sediment targets for the POTPCB model used to determine loading scenarios necessary for
achieving desirable fish tissue PCB concentrations.
Table 15 summarizes the water quality targets calculated using species specific bioaccumulation factors.
For comparison purposes applicable water column criteria and fish tissue thresholds are also listed.
Table 15. BAF-based TMDL Targets
DC
Maryland
Virginia
Fish Tissue PCB
Impairment
Threshold (ppb)
20
88
54
PCB Water
Quality Criteria
(ng/L)
0.064
0.64
1.7
BAF-based Target PCB
Water Concentration
(ng/L)
0.059
0.26
0.064
BAF-based Target PCB
Sediment Concentration
(ng/g dry wt)
2.8
12.0
7.6
As the BAF-based water concentration targets were found to be more stringent that the existing water
column criteria, the BAF based targets were used to determine a set of PCB loads that would result in the
desirable fish tissue concentrations. The POTPCB model was used to predict water column and sediment
concentrations for given loading scenarios, which then were compared against the appropriate targets.
A-42
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Allocations
TMDL loads were allocated to each of the identified source categories within each impaired segment.
This was done as part of a deliberate process starting with a number of diagnostic model runs that
provided a general sense of required load reductions to achieve the PCB targets in each impaired
segment. The next step included a series of model runs with adjusted loads from specific the source
categories in order to arrive at a set of loads that provided quasi-equilibrium PCB concentrations at or
below the appropriate targets in each of the model segments.
Wasteload Allocations
For the TMDL, the jurisdictions agreed to apply a consistent approach for WWTPs. Only those WWTPs
facilities (22) with the greatest annual flows within the direct drainage of the tidal Potomac and Anacostia
Rivers were assigned a WWTP WLA. These WLAs were determined based on the facility design flow
multiplied by the applicable water quality target.
The regulated NPDES stormwater allocations within the direct drainage area were expressed as a single
stormwater WLA for each impaired waterbody. These allocations were determined by multiplying the
direct drainage PCB load for the TMDL scenario by the percent of land classified as developed (i.e.,
covered by an NPDES stormwater permit). The nontidal tributary regulated and unregulated stormwater
loads are included in the LA portion of the TMDL.
Combined Sewer Overflow WLAs were based on daily flows and modeled loads for each outfall (53 in the
District and 4 in VA).
Load Allocations
Load allocations include loads from upstream nontidal tributaries, which include upstream point sources,
direct drainage nonpoint source runoff, atmospheric deposition to water surface, and runoff from known
contaminated sites. Additionally, PCB exchanges with the Chesapeake Bay were considered in the
analysisbut isnot tracked and identified in the final TMDLs. Table 14 describes how the nonpoint source
load estimates were developed, for additional details, please see the TMDL report.
MOS
While conservative assumptions were used in developing load estimations and in setting up the POTPCB
model, an additional explicit margin of safety of 5% was applied to all source categories with an exception
of WWTPs.
Daily Load Expression
Fish tissue PCB concentrations and associated human health impacts are reflective of exposure to
elevated PCB concentrations over an extended time period. Consequently, the final TMDL allocations
were expressed on annual basis. Two additional PCB loading expressions were also provided:
a) average daily loading condition - calculated as the annual load divided by 365;
b) peak one-day loads for the TMDL year - calculated differently for different source
categories (for details see Table 16).
DRAFT A43
-------
Handbook for Developing Watershed TMDLs
December 2008
Table 16. Methods for Peak One Day Loads
Source
Tributaries
Direct drainage areas
CSOs
Blue Plains WWTP
Atmospheric deposition
Contaminated sites
WWTPs (other than Blue Plains)
Calculation Method
Annual maximum daily load in the daily
time series for the TMDL year
load
Annual load divided by 365.
1.31 x average daily load a
a A statistical procedure recommended by EPA for identifying daily loads for long term allocations, that
relates maximum daily concentrations to a long term average. The procedure is outlined in EPAs 1991
Technical Support Document for Water Quality-based Toxics Control.
Overall Reductions
Overall, the PCB TMDL constitutes a 96% reduction of PCBs from the 2005 baseline load of 37,156
grams/year (Figure 7).
Total PCB loads to the tidal
Source category
Pcr.omac fa] Chain Bridge1
Lower Basin Tributaries2
Direct drainage3
WWTP4
CSO5
Atmospheric deposition'
Contaminated si:es7
TOTAL3
1 The non-tdal Potomac River above Cha n Brdge
of-T de of fie tida Potomac R ver, or estuary.
: The lower basin is that portion of the Potorrac RK
watershed above Chain Bridge. The tributaries are
yVatersh&d Model f'/v'MS) as tributaries.
* That part of the lower basin watershed that is not
adjacent to the Potomac and Ana cost a rivers.
~ Waste water treatment plan*.
° Comb ned sewer overflow system.
° Atmospher c PCBs depos ted d rectly on the tidal
Potomac and Anacostia rivers, in g/year
Baseline TMDL
(0/year) (g/year)
16,433 329
2,657 407
10,996 413
762 63.2
3,020 61.2
3,070 217
15.1 1C. 8
37,156 1,505
Reduction
98%
86%
&6%
91%
98%
93%
28%
96%
n the D str ct of Columbia. Cha n Brdge s the approx rrate head-
er watershed that contributes to the tda waters
the 17 streams in the lower basin def ned n the
and exc udes the
Chesapeake Bay
n a WM5 defined tributary. D rect drainage areas are located
•vater surface.
7 Those sites that have been identif ed as contaminated by PCBs. some of which have been remed
B This total does not include changes in the Downstream Boundary condition for reasons exp ained
ate-d.
n Section V(5.2;>
Figure 7. Summary of Potomac and Anacostia PCB TMDLs (source Tidal Potomac PCB TMDL)
Implementation
Due to the uncertainty associated with the TMDL loading capacity and load allocations, the three
jurisdictions agreed to follow the adaptive implementation guidelines. Collection of additional data, PCB
source-tracking, and PCB source minimization and reduction measures will be the focus of the
implementation efforts rather than end-of pipe treatment measures.
Potomac and Anacostia PCB TMDL calls for large reductions from a number of upstream nontidal
tributaries. While the NPDES permitted point sources in the nontidal tributary have not been assigned a
WLA as part of this TMDL effort, it is expected that the upstream facilities will be subject to similar permit
provisions as those for facilities with an existing WLA. Additionally, the SC recommended that the
A-44
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
jurisdictions, ICPRB, and EPA Region III continue to work together in gathering additional monitoring data
to better characterize loadings from non identified sources.
Each of the jurisdictions intends to apply existing programs in the implementation of the PCB TMDL along
with follow-up monitoring. All three will utilize high resolution analytical techniques (such as Method
1668A) for sample analysis for surface water, sediment, and permit related monitoring.
Important implementation aspects in the District include implementation of the Long Term Control Plan for
CSO Discharges, implementation of its stormwater management plan for MS4 areas, and follow-up
monitoring. Authorities in Maryland intend to evaluate the impact of atmospheric deposition on point
source loadings including regulated stormwater. Additionally, a monitoring plan is being drafted to
evaluate contributions from the upstream nontidal portion of the Potomac and Anacostia watersheds. For
WLAs, permitting activities will first address monitoring, followed by minimization and reduction measures.
Also, implementation of existing nutrient and sediment TMDLs in the region is expected to result in
reductions of PCB loadings. In Virginia, implementation efforts will focus on the impacts of atmospheric
deposition to regulated source loadings. The strategy and framework for Virginia's implementation
activities is detailed in the PCB Strategy for the Commonwealth of Virginia, published in 2004.
Lessons Learned4
Participant insights:
Given the complexity associated with the type of pollutant and the interconnected nature of the tidal
Potomac subwatersheds, addressing the tidal Potomac and Anacostia PCB listings would have been
much more difficult on a jurisdiction by jurisdiction level.
Will additional listings be addressed in a coordinated manner as well?
Maryland and DC have coordinated on other TMDLs developed for the Anacostia River (sediment and
nutrient), while the three jurisdictions, EPA, and ICPRB are currently coordinating their work on the PCB
TMDL implementation measures.
Additional multi-jurisdictional efforts include: work towards addressing nutrient and sediment impairments
in the Chesapeake Bay and a bacteria TMDL for the restricted shellfish harvesting/growing areas of the
Pocomoke River.
What were particular difficulties with this approach?
While the TMDL established an allowable PCB load that the watershed can assimilate and still meet
water quality standards, the specific ongoing sources are still unknown. The wide historical use of PCBs
and the time constraint associated with the consent decree made it difficult to identify ongoing sources on
a manageable scale. Thus, the initial stages of implementation will need to focus on further source
identification rather than being able to move directly to elimination of the ongoing sources.
Applying a 5% margin of safety to all source categories created an unanticipated situation in which some
sources which otherwise would not need to reduce their loads, ended up with a 5% reduction.
References
Haywood, H. C. and C. Buchanan. 2007. Total maximum daily loads of polychlorinated biphenyls (PCBs)
for tidal portions of the Potomac and Anacostia rivers in the District of Columbia, Maryland, and
Virginia. Interstate Commission on the Potomac River Basin. ICPRB Report 07-7. Rockville, MD.
September 2007.
LimnoTech. 2007. PCB TMDL Model for Potomac River Estuary. Combined Report on
Hydrodynamic/Salinity and PCB Transport and Fate Models. Prepared for U.S. Environmental
4 Special thanks to Anna Soehl, TMDL Technical Review Coordinator MDE/Science Services
Administration
DRAFT A45
-------
Handbook for Developing Watershed TMDLs December 2008
Protection Agency, Region 3, Philadelphia, PA, through Battelle, Applied Coastal and
Environmental Services, Duxbury, MA.
USEPA, 1991. Technical Support Document for Water Quality-based Toxics Control. EPA/505/2-90-001
PB91-127415. U. S. Environmental Protection Agency, Office of Water, Washington, DC.
Available on-line at: http://www.epa.gov/npdes/pubs/owm0264.pdf
A-46 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 6: St. Marys River TMDLs, Indiana
Jurisdictions:
Watershed incorporates all or portions of
Adams, Allen, and Wells Counties, as well as
City of Fort Wayne
# Impaired Segments Listed:
TMDL addressed 42 impairments on 34
segments
Technical Approach:
Load Duration Curve
Source Types:
IMPS and PS
Watershed TMDL at a Glance
Waterbody:
St. Marys and Upper Maumee Rivers
Drainage Area:
8-digit HUC, 814 square miles
Parameters:
E. coli bacteria
Total suspended solids
Nutrients (total phosphorus)
Impairments biological communities (IBC)
Development Status:
Approved by EPA Sept. 2006
Developed by:
Indiana DEM
Background
The St. Marys River watershed is located in northeastern Indiana (Figure 8), covering more than 810
square miles. The river originates in Ohio and flows into Indiana through Adams County. The St. Marys
continues in a northwest direction into Allen County where it joins the St. Joseph River at Fort Wayne to
form the Maumee River. Land use in the watershed is predominantly agriculture, which represents over
84% of the total land cover. Corn and soybeans comprise the majority of crops produced in the
watershed. Other important landuses include urban / residential (7.3%), woodland / forest (7.1%), and
wetlands (1.2%). In addition to Fort Wayne, other major communities in the Indiana portion of the St.
Marys watershed include Decatur and Berne. Along with the mainstem St. Marys River, this watershed
TMDL covers tributaries located in Indiana including Habegger Ditch, Gates Ditch, Blue Creek, Yellow
Creek, Martz Ditch, Borum Run, Holthouse Ditch, Kohne Ditch, Gerke Ditch, and Nickelsen Creek.
One of the driving factors for the Indiana Department of Environmental Management (IDEM) to develop a
watershed TMDL for the St. Marys was to build on their efforts in preparing a Watershed Restoration
Action Strategy (WRAS). The WRAS consisted of two parts: 1) Characterization and Responsibilities,
and 2) Concerns and Recommendations. Multiple stakeholders in the watershed were involved in the
development of the WRAS. Priority issues identified during the process included data / information
collection, targeting areas of concern within the watershed, streambank erosion and stabilization, failing
septic systems, and general water quality. Because segments in the St. Marys watershed were on
Indiana's §303(d) list, there was an awareness by stakeholders of the potential utility of the TMDL
process to address the priority issues. Following publication of the WRAS in 2001, IDEM decided to take
advantage of this local interest and initiate development of a broader watershed TMDL for the St. Marys.
Several major stakeholders were also key players in the St. Joseph River Watershed Initiative (SJRWI), a
progressive effort to address water quality problems in the basin immediately north of the St. Marys.
Included were the Allen County Soil & Water Conservation District, the City of Fort Wayne, the Maumee
River Basin Commission, and the Allen County Health District. These groups were very interested in the
St. Marys TMDL process, in particular the technical approach that was being contemplated by IDEM to
evaluate water quality data. As a result of this interest, these stakeholders became actively engaged in
supporting the St. Marys watershed TMDL development effort through major assistance in ambient
monitoring and coordinating public involvement activities. For instance, the City of Fort Wayne partnered
DRAFT
-------
Handbook for Developing Watershed TMDLs
December 2008
with IDEM in a collaborative monitoring effort to collect data, which were used in preparation of the TMDL.
The Allen County Health Department also provided ambient data from their weekly sampling program
conducted over a four year period; information that proved to be quite useful in evaluating conditions and
potential sources of E. coli in the St. Marys watershed.
Figure 8, St. Marys River watershed
Impairment Listing Information
Numerous streams in the St. Marys River watershed are identified on Indiana's 2002 and 2004 §303(d)
lists as impaired for E. coli, impaired biotic communities (IBC), and nutrients. Based on data collected in
2004 by IDEM and the City of Fort Wayne, a reassessment was conducted of the St. Marys watershed.
The purpose of the reassessment was to define the extent of the 2004 impairments and in turn confirmed
the 2002 listings. The reassessment for the E. coli impairment resulted in the addition of the sixteen
segments in the St. Marys River watershed to the 2006 §303(d) list. Table 17 summarizes segments and
impairments addressed by the St. Marys Watershed TMDL.
In order to assist with the assessment and TMDL development process, the St. Marys River watershed
was divided into subwatersheds, a more detailed delineation of smaller catchments. The entire watershed
TMDL contains six subwatersheds, which include the mainstem St. Marys, Blue Creek, Yellow Creek,
Borum Run, Holthouse Ditch, and Nickelsen Creek.
Table 77. Impaired Segments Addressed by the St. Marys Watershed TMDL.
^H WaterbodyName ^H
St. Marys-Willshire
St. Marys River
Blue Creek
Blue Creek
Duer Ditch (Adams) and Other Tribs
Blue Creek Headwaters (Adams)
Habegger Ditch
Wittmer Ditch, No. 1
Farlow Ditch and Tribs
Gates Ditch
Segment ID Number
INA0434 00
INA0441 00
INA0442 T1007
INA0445_T1006
INA0445 00
INA0442 00
INA0443 T1008
INA0443 T1020
INA0443-T1019
INA0443 T1014
Length
(mi)
2.84
0.86
11.94
12.28
9.33
8.46
5.80
2.98
11.01
1.17
Impairment
E. coli
E. coli
E. coli
E. coli, IBC, ammonia,
nutrients
E. coli
E. coli
E. coli, IBC, nutrients
E. coli
E. coli
E. coli
A-48
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Little Blue Creek
Borum Run and Tribs
Yellow Creek
Martz Creek-Ruppert Ditch and Unnamed
Tributaries
Holthouse Ditch-Kohne Ditch
St. Marys River
St. Marys River
Unnamed Trib of St. Marys River
Pleasant Mills and Tribs
Decatur Tribs
Gerke/Weber Ditch and Tribs
Snyder Ditch and Other Tribs
Junk Ditch
Spy Run Creek
Unnamed Tributaries to Spy Run Creek
Lowther Neuhaus Ditch
Unnamed Tributary to Lowther Neuhaus
Ditch
St. Marys River
INA0444 00
INA0448 00
INA0447 00
INA0447_T1002
INA0452 00
INA0461 T1004
INA0463 T1003
INA0465 T1002
INA0448 T1016
INA0449 T1017
INA0453 T1018
INA0454 T1021
INA0446 T1015
INA0454 T1012
INA0446 00
INA0449 00
INA0453 00
INA0463 00
INA0465 00
INA0465 T1011
INA0466 T1012
INA0466 T1013
INA0466_T1014
INA0466 T1022
22.12
21.65
32.79
9.82
10.16
37.70
4.79
2.84
15.30
7.12
17.53
10.61
6.55
8.75
5.08
3.03
3.00
0.50
E. coli
E. coli
E. coli, IBC, nutrients
E. coli
E. coli, IBC
E. coli
E. coli
E. coli, IBC
E. coli
E. coli
E. coli
E. coli
E. coli
E. coli
E. coli
E. coli
E. coli
E. coli
Applicable Standards and Water Quality Targets
Beneficial uses and the criteria to protect those uses are identified in Part 327 of the Indiana
Administrative Code (IAC). With respect to the St. Marys watershed, full body contact recreation is a
designated use [327 IAC 2-1-3. Sec 3. (a)(1)]. In addition, all surface waters in the St. Marys watershed
must support a well-balanced, warm water aquatic community [327 IAC 2-1-3. Sec. 3. (a) (2)(A)],
For waters in Indiana within the Great Lakes system, numeric criteria have been adopted for E. coli
bacteria to protect the primary contact recreation use under IAC 2-1-6(a)(3)(d). The criteria, which apply
during the recreation season (April 1 - October 31), state that:
"E. coli bacteria, using membrane filter (MF) count, shall not exceed one hundred twenty-five
(125) per one hundred (100) milliliters as a geometric mean based on not less than five (5)
samples equally spaced over a thirty (30) day period nor exceed two hundred thirty-five (235) per
one hundred (100) milliliters in any one (1) sample in a thirty (30) day period."
Narrative criteria were applied to establish numeric targets, which address the impaired biotic
communities portion of the St. Marys watershed TMDL. In particular, the relevant narrative criteria state
that:
"All surface waters at all times and at all places, including waters within the
mixing zone, shall meet the minimum conditions of being free from substances,
materials, floating debris, oil, or scum attributable to municipal, industrial,
agricultural, and other land use practices, or other discharges that do any of the
following:...
(a)re in concentrations or combinations that will cause or contribute to the growth
of aquatic plants or algae to such degree as to create a nuisance, be unsightly,
or otherwise impair the designated uses."[327 IAC 2-1-6. Sec. 6. (a) (1)(D)]
(a)re in amounts sufficient to be acutely toxic to, or to otherwise severely injure or
kill, aquatic life, other animals, plants, or humans."[327 IAC 2-1-6. Sec. 6. (a)
DRAFT
-------
Handbook for Developing Watershed TMDLs December 2008
Indiana utilized benchmarks to evaluate conditions that may contribute to the biological impairments.
Indicators evaluated included total phosphorus, nitrate, dissolved oxygen, pH, algal conditions, and total
suspended solids (TSS). IDEM assessed the Index of Biological Integrity (IBI) scores relative to the
benchmarks, considering the effect of each indicator on the aquatic ecosystem. Based on a review of all
information, IDEM used the benchmarks for total phosphorus (0.3 mg/L), nitrates (10 mg/L), and TSS (30
mg/L) as the numeric TMDL targets, which would address the biological impairments.
Pollutant Sources
The St. Marys watershed contains a mix of both point and nonpoint sources. Source assessment
information was analyzed by major subwatershed. Point sources in the St. Marys center on five major
categories: municipal wastewater treatment facilities (WWTF), combined sewer overflows (CSO),
municipal separate storm sewer systems (MS4), industrial facilities, and confined animal feeding
operations (CAFO). Significant nonpoint sources evaluated in the St. Marys watershed TMDL included
agriculture, which is the dominant land use, and rural domestic on-site septic systems. Potential
contributions from wildlife, especially E. coli, were also evaluated.
Permit files, discharge monitoring reports, land use inventories, air photos, and ambient water quality data
were all considered as part of the source assessment process. In addition, IDEM conducted public
meetings in the watershed as part of an effort to solicit additional input from stakeholders and other
interested groups. Because a large portion of the watershed is in agricultural land use, contributions of E.
coli, sediment, and nutrients were looked at closely in the source assessment process. Aerial photos and
CIS files were evaluated for significant changes in land use, location of feeding operations relative to
receiving waters, and for the distribution of cropping patterns. The presence of CSOs in three
communities also prompted a close review of available data regarding these sources.
One concern that emerged from the public information meetings was the potential effect of failing septic
systems on streams throughout the watershed. Because of the rural nature of the watershed, the Adams
County Health Department conducted studies in 2001, which demonstrated a high likelihood that on-site
septic systems represent a major source of E. coli bacteria. Numerous systems were identified, which
discharge directly to streams or are directly connected to old agricultural field tiles. In addition, the study
identified a number of unsewered communities in the St. Marys watershed, which are neither connected
to a municipal treatment plant nor use a complete functioning on-site septic system.
Technical Approach
The St. Marys watershed TMDL used the duration curve framework to link water quality data with the
source assessment information. A collaborative monitoring effort on the part of IDEM and the City of Fort
Wayne was utilized to generate sufficient ambient data to help identify major sources of concern and
effectively target priority areas for subsequent implementation. IDEM sampled fourteen sites, once every
other week from March 2004 to October 2004. The City of Ft. Wayne sampled seven of the same sites
as IDEM on opposite weeks from July of 2004 through October of 2004
IDEM applied the duration curve framework for the TMDL in an effort to consider the general hydrologic
loading conditions of the watershed, and subsequently, to enhance the source assessment process.
Pollutant delivery mechanisms likely to exert the greatest influence on receiving waters (e.g., point source
discharges, surface runoff) could be matched with potential source areas appropriate for those conditions
(e.g., riparian zones, impervious areas, uplands). Patterns associated with certain source categories are
often apparent when visually assessing data by flow conditions.
Because the flow duration interval serves as a general indicator of hydrologic loading condition (i.e., wet
loading versus dry loading and to what degree), allocations and reduction targets were linked to source
areas, delivery mechanisms, and the appropriate set of management practices. The use of duration
curve zones (e.g., high flow, moist, mid-range, dry, and low flow) allowed the development of allocation
tables, which were used to summarize potential implementation actions that most effectively address
water quality concerns.
A-50 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
The analysis was conducted by major subwatershed. Water quality patterns were evaluated for each flow
condition. One of the advantages of conducting the subwatershed assessments under the "umbrella" of a
larger watershed TMDL effort was the ability to consolidate the overall analytical process. This was
particularly important in developing estimates of flow on ungaged streams, a critical part of the duration
curve method. Furthermore, conducting the overall analysis at the watershed scale enabled the ability to
"cluster" data sets based on land use similarities. This was particularly useful in evaluating the relative
importance of old field tile lines as a delivery mechanism for E. coli from on-site septic systems.
Allocations
Critical source categories receiving allocations for each TMDL are summarized in Table 18. Due to the
level of detail at which source allocations were determined, IDEM presented summary allocation tables
for LAs and WLAs in the TMDL document. For detailed, source-specific allocations by subbasin, IDEM
also developed spreadsheet tables in Microsoft Excel to assist in implementation efforts.
Table 18. Allocated Sources for the St. Marys River Watershed TMDLs
TMDL
E. coli
Total phosphorus
Nitrate
Total suspended solids
Sources receiving LAs a
Agricultural areas
Rural residential areas
Woodland / forest areas
Background and Other NPS (wildlife)
Sources receiving WLAs a
NPDES permitted facilities (e.g., WWTFs)
Combined Sewer Overflows
MS4 Stormwater Communities
Onsite Sewage Treatment Systems b
Sanitary Sewer Overflows b
Concentration-based TMDL and allocations
Illegal discharges such as those from failing septic systems and straight pipes as well as sanitary sewer overflows received a
wasteload allocation of zero.
Implementation
The next phase of this TMDL will be to identify and support the implementation of activities that will bring
the St. Marys River watershed in compliance with the E. coli criteria, and with the phosphorus, nitrate,
and total suspended solids targets. IDEM continues to work with its existing programs on
implementation, as well as with local stakeholder groups to pursue best management practices that will
result in improvement of the water quality in the St. Marys River watershed. Prior to the TMDL, two 319
grants were awarded to the Adams County Soil and Water Conservation District in 1999 and 2000 to
address nutrient management. The information gathered for these grants was useful in building upon
existing work in this watershed, which has led to additional 319 grants to fund needed implementation.
More recently, the Allen County Soil and Water Conservation District received a 319 grant to complete a
Watershed Management Plan. The Adams County Soil and Water Conservation District and the City of
Fort Wayne are both partners in this project. The Watershed Management Plan will be designed to
achieve the reductions in pollutant loads called for in the nonpoint source section of the St.
Marys/Maumee TMDL. An Implementation Plan will be developed describing in detail all the activities
planned during the implementation phase of this project to address the needed reductions.
The Maumee River Basin Commission (including Steuben, Dekalb, Allen, Adams Co) offers a cost-share
program for voluntary agricultural land use conversion whereby an agricultural land owner may put in
filterstrips, grass buffer zones, French Drains, Wetland Restorations, or Woodland / Reforestation in the
floodplain. Though the goal of the program is to reduce flooding, there are water quality benefits.
In addition, a Clean Water Indiana grant has led to the hiring of a local project manager to foster greater
adoption and implementation of conservation practices by working with landowners and farmers in the St.
Marys and Maumee River watersheds. The focus of this effort is to launch an intensive public education
and outreach effort, which creates an awareness of currently available local, state and federal
conservation programs. A resource inventory will be developed, which includes water quality conditions
and a geo-referenced database of existing conservation practices. The inventory effort will be supported
by the Maumee River Watershed Project (MRWP), currently underway by the Ohio Department of Natural
DRAFT
A-51
-------
Handbook for Developing Watershed TMDLs December 2008
Resources and Ohio NRCS. The inventory will guide the project's efforts by identifying priority areas
within the watersheds to maximize environmental benefit. In addition to the public education, technical
assistance, and resource inventory assessment efforts, the project will conduct focus group meetings to
help identify other local resource concerns that may not be able to be discerned from water quality or
related data, such as land-use issues and community development planning. The focus groups will
include landowners, farmers, local business owners, local and state agency personnel, local government
leaders, as well as concerned citizens. Participants in the focus groups will be encouraged to assist with
the development of Watershed Management Plans (WMP) as WMP Steering Committee members.
Lessons Learned
IDEM continues to take advantage of opportunities for large-scale analysis because it allows for a more
coordinated response to water quality issues in terms of collecting data, analyzing pollutant sources, and
implementing solutions. Critical lessons learned by the program are summarized below.
The value of utilizing local interest cannot be overstated. Incorporating any and all information that can
be gathered often builds credibility and trust with local stakeholders. It is important to ensure that
stakeholders are engaged in the process without being overwhelmed, as well as to listen to their
thoughts, ideas, and perspectives. In addition to meeting regulatory requirements, the TMDL process
should strive to fit local needs and interest.
The ability to "cluster" data sets, particularly when several counties and government agencies are
involved, maximizes use of available information and enables efficient targeting of priority source areas.
Coordinated monitoring and assessment between multiple counties and governmental agencies is time
intensive. However, good communication coupled with continued coordination helps ensure stakeholder
"buy-in" as the TMDL is developed and sets the stage for a smooth transition into implementation.
Successful implementation of the TMDL is dependant on the involvement/participation of local
stakeholders. Initiating the interest, and maintaining that interest in the project is a key element needed
to support successful TMDL implementation. Statewide, Indiana has hired five watershed specialists to
take the lead on these types of initiatives.
References
Indiana Department of Environmental Management. June 2006. Total Maximum Daily Load For E. coli
Impairment in the St. Marys River Watershed and Maumee River Adams and Allen Counties and
Total Maximum Daily Load For Impaired Biotic Community Impairment in the St. Marys River
Watershed Adams and Allen Counties. IDEM Office of Water Quality ~ TMDL Program.
Indianapolis, IN.
Indiana Department of Environmental Management. January 2001. St. Marys River Watershed
Restoration Action Strategy - Part I: Characterization and Responsibilities. IDEM Office of
Water Quality. Indianapolis, IN.
Indiana Department of Environmental Management. January 2001. St. Marys River Watershed
Restoration Action Strategy - Part II: Concerns and Recommendations. IDEM Office of Water
Quality. Indianapolis, IN.
A-52 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 7: Tualatin River TMDLs, Oregon
Watershed TMDL at a Glance
Waterbody:
Tualatin River and tributaries
Drainage Area:
Approximately 710 square miles
Parameters:
Temperature
Dissolved Oxygen
Bacteria
Phosphorus
Development Status:
Revised / Finalized 2001
Developed by:
Oregon DEQ
Jurisdictions:
Watershed incorporates all or portions of 15 cities, including
portions of Portland
# Impaired Segments Listed:
Temperature: 19
Bacteria: 26
Dissolved Oxygen: 22 plus mainstem
Phosphorus: 8 (chlorophyll a)
Technical Approach:
Temperature: Continuous simulation model (Heat Source)
Bacteria: Event based unit load model
Dissolved Oxygen and related parameters: Steady State
QUAL2E simulation (tributaries) and dynamic simulation
(mainstem) using CE-QUAL-W2
Source Types:
NPS and PS
Background
The Tualatin River is a major tributary of the Willamette in the Northwest corner of Oregon. The
watershed lies almost entirely within Washington County with small portions in Multnomah, Clackamas
and Yamhill Counties. It is approximately 83 miles in length with a flat gradient for most of its length.
Major tributaries include Scoggins, Gales, Dairy, Rock and Fanno Creeks. Summer flows are
supplemented by releases from two Reservoirs (Scoggins and Barney). The urban areas in the
watershed are rapidly growing and are served by four wastewater treatment plants (WWTPs) which are
operated by the public utility Clean Water Services (CWS). In addition, CWS also has two industrial
stormwater permits and is a co-permittee on a Municipal Separate Storm Sewer permit (MS4) permit.
The Tualatin drains urban, agricultural, and forested lands.
The primary driver for developing this TMDL on a watershed-scale was political, as it was one of the first
TMDLs developed in Oregon, and was the focus of the initial 1986 lawsuit. Although the litigation was
concerned with nuisance algal growth and phosphorus in the lower Tualatin and Lake Oswego, several
factors were clear from the start of TMDL development. The dominant nutrient sources during the
summer low flow period were two major advanced wastewater treatment facilities operated by the Unified
Sewerage Agency (USA) of Washington County (now Clean Water Services, or CWS). However,
significant nonpoint source loads from multiple sources (urban stormwater, agriculture, and forest lands)
were delivered during major rain events. Also, flows in the Tualatin River are managed by releases from
a multiple purpose reservoir (Scoggins) and diversions from the Trask Basin through Barney Reservoir,
making water quantity a significant issue. Because of the high visibility the litigation was receiving, and
because of major stakeholder interest from multiple parties, the Oregon Department of Environmental
Quality (DEQ) decided that developing the TMDL for the entire Tualatin was the only logical path to take.
Two major advisory committees were formed; in addition to DEQ sampling & analysis, CWS also
operated a large ambient WQ monitoring network in the basin and had a history of data sharing with
DEQ. The initial Tualatin experience of developing TMDLs on a watershed basis set the tone for
Oregon's statewide program as it moved forward (including the 2001 revision to the Tualatin TMDL).
DRAFT
A-53
-------
Handbook for Developing Watershed TMDLs
December 2008
Impairment Listing Information
Population growth in the watershed has contributed to water quality decline. In 1988, the Oregon
Department of Environmental Quality (DEQ) developed an ammonia TMDL to address problems with low
dissolved oxygen (DO) in the mainstem river and total phosphorus TMDL to address high pH and
nuisance algal growth occurring in the reservoir like section between river miles 24 and 3.4. The
ammonia TMDL was largely addressed through upgrades to the WWTPs in the watershed. The
phosphorus TMDL involved both point and nonpoint sources. WWTPs upgraded processes to increase
phosphorus removal and BMPs were implemented by Designated Management Agencies (DMAs)
throughout the watershed, which resulted in significant decreases in overall phosphorus concentrations in
both the mainstem and the tributaries.
While nuisance algal blooms were reduced as a result of implementation measures from the 1988
TMDLs, the watershed is not meeting phosphorus limits set by the TMDL. The Tualatin Basin Policy
Advisory Committee was appointed to develop recommendations to DEQ regarding the TMDLs.
Concurrently, Oregon updates to the 303(d) list included additional listings for temperature, bacteria,
dissolved oxygen, pH, biological criteria, arsenic, iron, and manganese.
In 2001, new TMDLs were issued for the Tualatin watershed, updating the ammonia and phosphorus
TMDL from 1988 and instituting new TMDLs for temperature, bacteria and dissolved oxygen (volatile
solids). Based on further data analysis and assessment of background conditions in the watershed,
TMDLs were deemed not necessary for arsenic, iron, manganese, and low pH. Impairment of biological
criteria was deemed to be addressed by temperature and DO TMDLs.
Applicable Standards and Water Quality Targets
Because multiple listings and parameters of concern were involved in the TMDLs for the Tualatin
watershed, the array of uses, standards and applicable targets is wide. Oregon DEQ issued a single
TMDL report document that discussed the individual TMDLs separately. However, the analysis
performed to develop the TMDLs for parameters and impairments that are related (e.g., DO, temperature,
phosphorus) was not done separately. For example, to account for all potential factors associated with
the DO listing it was necessary to evaluate multiple factors such as pH, phytoplankton growth,
temperature, oxygen demanding substances, and nutrient levels. While the DO TMDL specifically
identifies loading capacities for ammonia and sediment oxygen demand, the loading capacities for the
related parameters of temperature and phosphorus are presented in those TMDLs. Table 19 summarizes
the applicable water quality standards and targets of each TMDL analysis.
Table 19. Applicable Standards and TMDL Targets
Temperature
D0a
Phosphorus e
Bacteria
Standard
Targets
Surrogate
Standard
Targets
Standard
Targets
Standard
Targets
No measurable increase from anthropogenic activities
Multiple, determined by designated use
Effective shade (for nps loading)
Numeric, minimum levels specified forsalmonid spawning,
and warm-water aquatic life
cold-water, cool-water
salmonid spawning -11.0 mg/L
cold-water - absolute minimum = 8.0 mg/L b
cool-water - absolute minimum = 6.5 mg/Lc
warm-water- absolute minimum = 5.5 mg/Ld
Nuisance phytoplankton growth rule
Numeric pH criteria
pH 6.5 -8.5
Chlorophyll a action level = 15 ug/L or that associated with
loading conditions
Background phosphorus concentrations
Tributaries from 0.04 mg/L -0.1 9 mg/L
Mainstem from 0.04 mg/L - 0.1 1 mg/L
natural phosphorus
Numeric E. coli; Year-round
30-day log mean = 126 organisms / 100 mL (5 sample minimum)
Single sample maximum = 406 organisms /100 mL
A-54
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
a
(related to ammonia, SOD, phosphorus, and temperature)
bDEQ Discretion: 30-day mean minimum = 8 mg/L; 7-day mean minimum = 6.5 mg/L; absolute minimum = 6
mg/L
0 DEQ Discretion: 30-day mean minimum = 6.5 mg/L; 7-day mean minimum = 5 mg/L; absolute minimum = 4
mg/L
d DEQ Discretion: 30-day mean minimum = 5.5 mg/L; absolute minimum = 4 mg/L
e (related to D.O., pH and chlorophyll a)
Pollutant Sources
For each TMDL, pollutant sources include both point and nonpoint sources. Bacteria, DO-related, heat
loading and total phosphorus sources all include forestry, agriculture, transportation, rural residential,
urban, industrial discharge, and WWTPs. Different methods for developing loading estimations were
applied for each pollutant.
Temperature
According to the TMDL, the largest source of heat loading is from nonpoint sources, with anthropogenic
nonpoint source heat loading as the dominant pollutant source. NPDES point source loading is relatively
small. Nonpoint source thermal loading is attributed to lack of riparian shading (e.g, near-stream
vegetation disturbance and removal).
Bacteria
Source assessments conducted for the bacteria TMDL examined run-off related and non-runoff related
loading. This analysis indicated that while criteria exceedences occur year-round, higher concentrations
are associated with wet weather across the watershed. Based on the data, instream bacteria levels,
especially in urban areas are quite high during runoff events, implicating urban runoff as one of the most
significant sources of bacteria loading. Ultimate sources of urban loading probably include pet waste,
illegal dumping, failing septic systems and sanitary sewer cross connections and overflows.
DO
Factors affecting DO levels instreams of the Tualatin watershed include nitrification, carbonaceous
biochemical oxygen demand (CBOD), algal growth, sediment oxygen demand (SOD), and temperature.
DEQ conducted an assessment of critical DO factors in each watershed based on beneficial use
categorization to identify the most critical sources for control. In Gales Creek, ammonia was identified as
a minor contributor to the DO problem while SOD was identified as a significant contributor. Temperature
was also identified as a factor. In Fanno Creek, algae, BOD and SOD were considered critical factors to
the overall DO levels as was the case for Lower Rock and Beaver Creeks. For Scoggins Creek, the
primary source appears to be low DO waters discharged from Scoggins Dam.
The primary pollutants affecting DO levels in the tributaries are solar loading (increased water
temperatures) and SOD. Modeling of the mainstem Tualatin suggests that the most important sources of
oxygen demand include (in order of importance) SOD, CBOD, algal respiration, zooplankton respiration
and nitrogenous biochemical oxygen demand (NBOD). Important sources of the factors listed above
include runoff and effluents from WWTPs.
Phosphorus
Important sources of phosphorus are identified as groundwater, which is higher than previously attributed
in the original TMDL, WWTPs, only two of which discharge during the summer TMDL period, and runoff
from urban, agricultural and forested lands. Other smaller but potential sources include failing septic
systems, instream and riparian phosphorus loading under anoxic conditions, tile drains and other
permitted point sources.
DRAFT A-55
-------
Handbook for Developing Watershed TMDLs December 2008
Roles and Responsibilities
Oregon DEQ was responsible for developing the TMDL. In addition to state funding, a lawsuit against
USA in the 1990's regarding compliance issues over their treatment facilities resulted in additional
financial resources being made available to support a DEQ Tualatin Coordinator.
Stakeholder Involvement and Implementation
The original TMDL set an expectation for heavy stakeholder involvement. Both Citizen and Technical
Advisory Committees were established in the first TMDL. This basic framework carried forward and
resulted in recommendations for the revised TMDL.
Technical Approach
As has been mentioned, the analyses performed to develop the Tualatin TMDLs for DO, Phosphorus and
Temperature were linked, either through concurrent modeling analysis or through the application of
similar assumptions and targets. The bacteria TMDL utilized an event based, unit loading model which
uses storm volumes, runoff concentrations for land uses, and bacteria die off rates to predict
concentrations in streams. Input data for the bacteria modeling was derived from CIS databases with
information on soils, land use, precipitation patterns, watersheds and distance from streams.
In comparison to the bacteria analysis, more complex modeling was performed for the temperature and
DO modeling, which incorporated the phosphorus analysis. For the temperature TMDL, a continuous
simulation model called Heat Source was developed to identify loading capacities and allocate to
sources. DO modeling was conducted using a modified version of the U.S. Army Corps of Engineers
model CE-QUAL-W2. The model provides a good fit to the measured data for streamflow, water
temperature, and water quality constituents such as chloride, ammonia, nitrate, total phosphorus,
orthophosphate, phytoplankton, and dissolved oxygen. The CE-QUAL-W2 model utilized certain
temperature related inputs from the Heat Source temperature model. For additional details regarding
each of the modeling analyses, the reader is refered to the TMDL report and its supporting appendices.
The DO modeling is futher detailed in a USGS report, Modeling Water Quality in the Tualatin River,
Oregon, 1991-1997 (Rounds and Wood 2001).
Allocations
Allocations for each of the TMDLs, Temperature, Bacteria, DO and Phosphorus are made for both point
and nonpoint sources. Important factors are noted below.
Temperature
While the point source load is not the predominant current source of thermal load to the system
(representing approximately 7%), under TMDL conditions, it represents a significant source and thus
receives a significant reduction relative to currently permitted levels, approximately 95%. Fifteen facilities
received WLAs. Background nonpoint sources were given a load allocation equal to the system potential,
which is defined as the median concentration of total phosphorus during non-runoff periods.
Anthropogenic nonpoint sources were given a load of zero.
The loading capacity for heat energy as expressed in the TMDL (kcal / day) is of limited value for
management activities, thus the TMDL also allocates "other appropriate measures" as provided under
EPA regulations [40 CFR 130.2(i)]. As a result, percent effective shade (percent reduction in potential
solar radiation load delivered to the water surface) is used to translate the reduction targets into
measurable practices (restoration, riparian planting, etc.) that can be expected to meet those targets.
Allocations derived for the temperature TMDL were also used in the analysis to derive the DO TMDL.
Bacteria
Allocations during the non-runoff periods are based on the single sample maximum of 406 counts /100
mL. Combined with reductions in septic system loads, illegal dumping, and CAFO loads, which received
A-56 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
allocations of zero, this level of allocation is expected to achieve both the 406 single sample maximum
and the 126 counts/100 ml criterion.
Allocations during the runoff periods are set to a geometric mean of 126 counts /100 ml as measured at
the mouth of each fifth-field watershed. Instream samples taken during storm events provide sufficient
data (5 samples or more per storm event) to use this criterion. This criterion is used because the load
allocations are derived using event mean concentrations (EMCs) for storm events. The achievement of
this loading capacity is also expected to achieve the single sample criterion of 406 /100 ml. LA and
WLAs are specified for the entire year; however specific allocations are identified for the summer (May1 -
Oct. 31) and the Winter (Nov.1 - Apr. 30) to reflect seasonal patterns and to coincide with seasonal
periods for other TMDLs.
DO
LAs and WLAs for parameters related to DO were derived from model results. Allocations for ammonia
and SOD are presented in the DO TMDL, while allocations for temperature and phosphorus are
presented in those sections. LA from the previous ammonia TMDL were deemed to be still appropriate;
WLAs were updated. SOD reductions were addressed through allocations for settleable volatile solids
and total phosphorus. Due to a lack of data in the watershed on levels of settleable volatile solids, DEQ
expected to base initial allocations on a similar parameter for which data exist, such as total suspended
solids (TSS).
Phosphorus
LAs and WLAs for phosphorus are related to issues addressed in the DO TMDL and will help ensure that
large algal blooms on the mainstem are controlled as much as possible, minimizing impacts on SOD from
algal detritus. Seasonal LAs and WLAs were assigned to meet the background level loading capacities
identified in the TMDL analysis.
Implementation
In Oregon, water quality management plans (WQMP) are designed to reduce pollutant loads to meet
TMDLs. A WQMP was developed with strategies for how the Tualatin River TMDLs will be implemented.
Portions of the WQMP were developed by the various DMAs who share responsibilities for managing
areas within the Tualatin watershed and thus, for assisting with implementation efforts. Critical elements
of the WQMP include programs developed in response to the earlier ammonia and phosphorus TMDLs.
In addition, structured programs on which DEQ intends to rely for implementation include the NPDES and
Water Pollution Control Facility permit programs, local ordinances, and programs within the state's
Departments of Forestry and Agriculture.
A particularly novel and interesting element of the implementation effort has been the state's effort to
authorize water quality trading to achieve portions of the load reductions required in the temperature and
DO TMDLs. With the help of grant funding from EPA, Oregon DEQ began developing a prototype trading
program with Clean Water Services, which operates four WWTPs and implements an MS4 permit in the
watershed. In 2005, DEQ issued CWS a revised NPDES watershed-based permit that combined its
formerly separate WWTP and MS4 permits into a single permit.
Watershed Based Permit
The new single permit issued to CWS includes provisions for trading to implement the required WLAs for
the DO and temperature TMDLs. The permit establishes the authority for trading and spells out the
specifics of DO related trades in an appendix. In addition, it lays out the basics of temperature trading,
which were to be augmented in a detailed Temperature Management Plan (TMP) to be developed by the
permittee.
DO Trading
Permit provisions allow for CWS to shift loads of oxygen demanding substances between two facilities
(the Rock Creek and Durham treatment plants). The permit contains no numeric limits for BOD and
ammonia but rather model-derived formulas that define allowable daily and weekly mass loads of
ammonia, NBOD and CBOD for the plants. Modeling was performed to determine what levels and
DRAFT A-57
-------
Handbook for Developing Watershed TMDLs
December 2008
combinations of BOD and ammonia can be released from the plants without violating criteria at Oswego
Dam, the location where oxygen levels have consistently been lowest. Formulas in the permit cover
varying stream conditions, accounting for instream flows, DO levels, months and water temperature.
Figure 9. Locations of WWTPs
Temperature
The permit provisions that cover temperature trading include the following options:
• Increase riparian shade with plantings
• Augmenting flows with cooler waters
• Effluent reuse
• Other mechanisms as identified by CWS
In its detailed IMP, the CWS discusses various options available for controlling temperature, some
involve trading while others do not. Non-trading options include use of cooling equipment, effluent reuse
to irrigate non-food crops, and source control. Trading options include flow augmentation (CWS owns
water rights at Scoggins and Barney reservoirs) and riparian shade creation.
With regard to flow augmentation, CWS would receive credits toward its required reductions by releasing
reservoir waters into the relatively warm mainstem. The TMP includes a methodology for calculating the
impacts of flow augmentation for purposes of compliance.
Credits for shading are generated through programs that provide incentives to landowners to plant shade-
producing vegetation along streams. The TMP provides a calculated estimate of the stream miles to be
planted each year. Once shade is established, CWS's efforts to demonstrate compliance shift from
showing adequate number of surviving plantings to showing how much shade they are producing. Annual
report submission will detail the credit trading activities for the previous year. To compensate for the fact
that the heat offset will take years to establish through shading, DEQ determined that at the end of 20
years, the load offset must actually be two times the actual excess thermal load.
A-58
DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
Based on calculations by CWS, after accounting for heat reductions resulting from flow augmentation,
approximately 35 miles of stream must be shaded in the 5 year permit period to reach the offset goal.
CWS is implementing the riparian shading program by enhancing US Department of Agriculture's existing
Conservation Restoration and Enhancement Program (CREP) which provides financial incentives to
landowners to implement restoration projects such as riparian shade plantings. Enhanced CREP cost
breakdowns are shown in Table 20.
Table 20. Estimated 5- Year Costs for Enhanced CREP
Clean Water Services
Partners a' D
Five-year cost total
Costs
$820,000
$1.38 million
$2.2 million
%
37
63
100
USDA still provides funding for the basic CREP program
b Partners include USDA, Oregon Department of Forestry, Oregon Water Trust, Tualatin Soil and Water Conservation District
The total cost of $2.2 million is in stark contrast to estimates for mechanical effluent refrigeration of more
than $50 million, plus yearly operations and maintenance costs of $2 million. (Source: Charles Logue,
CWS, 2nd National Water Quality Trading Conference, May 2006)
Lessons Learned
The Tualatin provides a good example of how TMDLs can work to improve water quality in a watershed
as there have been measurable improving trends in the parameters that have been addressed. It is also
a good example of the need for adaptive management and the need to revisit the TMDL and
implementation structure, on a periodic basis, in the case of Oregon DEQ, this is about every ten years.
Adaptive Management
Oregon DEQ's experience with the Tualatin TMDLs and implementing a statewide program is that TMDLs
are something that need to be revisited about every 10 years. Based on comments from staff involved in
the development, implementation and revision of the Tualatin TMDLs, DEQ's watershed TMDL history
has seen TMDLs (for the Tualatin) initially developed in 1988, modified in 2001 and which are now being
proposed for modification in 2009. DEQ has been trying to work this into its Watershed Approach where
they revisit TMDLs in the year prior to permit renewals. DEQ now has the Tualatin Subbasin on a
schedule for permit renewal in 2010 and will modify the TMDL in 2009. According to staff, while it would
be nice to revisit each TMDL on the 5-year cycle that permits are scheduled to be renewed, they have
found that there aren't the resources to do that and a 10-year cycle seems to appropriate to allow time for
implementation and development of new information/tools, etc.
Updating the existing TMDL
The process for updating the existing ammonia and phosphorus TMDLs was relatively straightforward
and easy. The various DMAs were involved in the collection of the data, were able to see the results and
had experience with the implementation. Similarly, environmental groups had a chance to see the data
and the results of implementation as well. While there was some controversy around the revision, it was
minor compared with the controversy around the new TMDLs - temperature, bacteria and DO in the
tributaries.
Between 1988 and 2001, CWS supported ongoing work by USGS to refine the Tualatin mainstem model.
Similarly, they continued to invest in expanding and improving the mainstem and tributary models for
TMDL updates in 2009. It should be noted that the models were also enhanced for other reasons as well
- mainly to help in CWS' efforts to understand and better manage the river through flow augmentation,
treatment, etc. Again, this all feeds into the first lesson above, that a 10 year TMDL revisit (adaptive
management) appears appropriate as there is new information, new or refined analytical techniques,
policy changes, and lessons learned in watershed management that minor or major modifications to the
TMDL are often needed.
DRAFT A-59
-------
Handbook for Developing Watershed TMDLs December 2008
Particular Difficulties
One specific difficulty related to development of the Tualatin TMDLs, according to staff, is that the
Tualatin is often on the leading edge of TMDL development in the state. It often is one of the first
subbasins receiving a type of TMDL (e.g., TP, NH3 in 1988, temperature, bacteria, and settleable volatile
solids in 2001.) Being one of the first, it becomes the "test case" for working out targets, allocations and
working them into permits (including MS4 permits) and NPS implementation mechanisms, which is
always challenging. In addition, the presence and continued involvement of a very active environmental
watchdog group that has litigated in the past, makes things very challenging. Finally, working WLAs into
MS4 permits continues to be a challenge.
Implementation
DEQ's report, Water Quality Credit Trading in Oregon: A Case Study Report (Oregon DEQ, date
unspecified) identifies several lessons learned since the trading agreement with CWS was finalized.
Trade efforts should seek to encourage restoration of contiguous areas to produce a higher value benefit.
For the trading program, DEQ offers that providing additional incentives for higher value riparian shading
projects might encourage this. Implementation efforts, whether related to trading or not, will tend to be
enhanced when high priority areas can be identified and addressed early.
Because the trading program relies on planting trees to provide shade, it is important that those trees be
allowed to grow and remain at the site in order to provide the needed shading services. Interestingly,
while conservation easement options are provided in the Enhanced CREP program to encourage farmers
not to remove trees, landowners are reluctant to relinquish longterm control of their property. Only one
farmer participated in a conservation easement, committing only one acre of land.
DEQ also notes that water quality standards and their interpretation in the permit process impacts the
achievability of potential trading schemes. The temperature standard under which the trading agreement
was negotiated required that a source develop a temperature management plan within the five-year
permit cycle. The current stipulation is that the source must meet wasteload allocations within a single
permit cycle. As trees cannot be expected to provide the needed amount of shade in 5 years, the current
situation would preclude riparian restoration as an option.
Finally, trading can be a way to focus limited resources in areas where they will do the most good. The
Tualatin watershed has many miles of deteriorated streams; point sources represent a relatively small
portion of the problem. Requiring point sources simply to reduce their loads at the discharge point misses
the real problem, which is that riparian areas all over the watershed are deteriorated.
Implementation work related to bacteria has been tied to continuing to sewer unsewered areas, correct
problems with aging infrastructure, and implementing MS4 storm water permits. Additional work has
been done to do DMA testing of bacteria to determine the source - avian sources were found to be a high
contributor in some areas. Additionally, further work has been done by CWS on the influence of flow on
the low DO concentration found in tributaries with some work investigating ways to get additional flow in
the tributaries and resultant changes in the DO. DEQ will be exploring ways to make the tributary DO
TMDL more flow based.
References
Oregon DEQ, 2001. Tualatin River Subbasin Total Maximum Daily Load. Available online at:
http://www.deq.state.or.us/wq/tmdls/willamette.htmtft
Rounds, Stewart A, and Tamara M. Wood 2001. Modeling Water Quality in the Tualatin River, Oregon,
1991-1997. Water-Resources Investigations Report 01-4041. Prepared in cooperation with the Unified
Sewerage Agency of Washington County, Oregon Portland, Oregon: 2001
Oregon DEQ, unspecified. Water Quality Credit Trading in Oregon: A Case Study Report. Available
online at: http://www.deq.state.or.us/wq/trading/docs/wqtradingcasestudv.pdf
A-60 DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
Case Study 8: Virgin River TMDLs, Utah
Watershed TMDL at a Glance
Jurisdictions:
Utah's portion of the watershed drains portions of
3 counties
# Impaired Segments Addressed:
2002 303(d) List included 8 impaired streams
Technical Approach:
Streams - Load Duration Curve
Reservoirs - BATHTUB
Urban Runoff- Simple Method
Source Types:
PS and NPS
Waterbody:
Virgin River
Drainage Area:
2,800 square miles
Parameters:
total dissolved solids, temperature, total
phosphorus, and dissolved oxygen
Development Status:
Approved by EPA on September 20, 2004
Developed by:
Utah Department of Environmental Quality
Washington County Water Conservancy District
Background
The Virgin River watershed, part of the larger Lower Colorado River-Lake Mead watershed, occupies
approximately 2,800 square miles within Utah's borders (Figure 10). The majority of the watershed
(approximately 76 percent) is in Washington County, while 19 percent is in Kane County and 5 percent is
in Iron County. The principal drainage for the watershed is provided by the Virgin River and its tributaries:
the East Fork Virgin River, North Fork Virgin River, North Creek, La Verkin Creek, Ash Creek, Fort Pearce
Wash, Santa Clara River, and Beaver Dam Wash. Although permanent flow is observed year round in
the Virgin River and its major tributaries, the majority of the watershed is drained by intermittent streams
that are dry most of the year. The watershed contains artificial canals and ditches, as well as diversions
and reservoirs, for agricultural irrigation and drinking water supplies. The major land uses in the
watershed include shrubland and forest and the Bureau of Land Management is responsible for
managing 43 percent of the land in the Virgin River watershed. Only 23 percent of the land in the
watershed is privately owned, with agricultural and residential land uses each approximately one percent
of the total watershed land uses. However, accessibility to larger metropolitan areas in surrounding states
and favorable climate conditions are causing the watershed to experience rapid population growth.
DRAFT
-------
Handbook for Developing Watershed TMDLs
December 2008
Virgin River Watershed HUCs
i |15010003
15010008
15010009
15010010
15010013
Unincorporated Places
' Incorporated Urban Areas
Utah State Parks
Raule Indian Reservation
| Zion National Park
Dixie National Forest
/\. State Boundaries
County boundaries
/\/ Inters! ates
U.S. Highways
Sale Routes
/\/ Major Streams
10
20 Miles
Figure JO. Virgin River Watershed, Utah
In Utah, the development of TMDLs is integrated within a larger watershed management framework that
emphasizes a common-sense approach aimed at protecting and restoring water quality. Key elements of
this approach include:
• Water quality monitoring and assessment
• Local stakeholder leadership
• Problem targeting and prioritization
• Integrated solutions that coordinate multiple agencies and interest groups.
Stakeholder involvement for the TMDL development process tapped into an existing watershed
stakeholder group referred to as the Virgin River Watershed Advisory Committee (VRWAC). The
VRWAC initiated watershed management planning activities in 1998, including conducting public
outreach and involvement activities to educate watershed stakeholders on issues related to water quality,
water quantity, land use, and groundwater. In 2002, the VRWAC decided to take a more comprehensive
approach to watershed management planning and hired a consultant to facilitate the development of the
locally-led watershed management plan, a drinking water source protection plan for the three surface
water intakes in the watershed, as well as TMDLs for impaired segments. As a result, stakeholder
involvement in the Virgin River watershed TMDL development process was also integrated with other
watershed management activities.
Impairment Listing Information
Various segments of the Virgin River are listed on Utah's 2002 Section 303(d) list of impaired waters for
total dissolved solids, dissolved oxygen, temperature, and total phosphorus. The beneficial uses that are
listed as impaired include cold water aquatic life (3A), other aquatic life (3C), and agriculture (4). Several
of the listings are due to naturally high concentrations of TDS and therefore the adoption of site-specific
criteria are being recommended as part of the Total Maximum Daily Load (TMDL) development process.
Other listings (i.e., for high temperatures) were made in error and are being corrected.
Beaver Dam Wash, from Motoqua to the Utah/Nevada state line, is listed on Utah's 2002 303(d) list as
impaired for temperature for the beneficial use of cold water game fish and other cold water aquatic life
A-62
DRAFT
-------
December 2008
Handbook for Developing Watershed TMDLs
(3A). However, DWQ believes that high water temperatures are naturally occurring and the impairment is
due to an incorrect beneficial use designation, rather than an actual water quality impairment.
Applicable Standards and Water Quality Targets
Designated
Use
3A
3B
3C
4
Description
Cold water
aquatic life
Warm water
aquatic life
Other aquatic
life
Agricultural
use
IDS
1200 mg/L
(max)
Temperature
Max.:20°C
Max. change: 2
°C
Max.:27°C
Max. change: 4
°C
Max.:27°C
Max. change: 4
°C
Dissolved
Oxygen'1 '
30 day avg: 6.5
7 day avg:
9.5/5.0
1 day avg:
8.0/4.0
30 day avg: 5.5
7 day avg:
6.0/4.0
1 day avg:
5.0/3.0
30-day avg:
5.0
1 day avg: 3.0
Selenium
4 day avg:
4.86(ug/L)
1 hour avg: 20
(Mg/L)
4 day avg: 4.86
(Mg/L)
1 hour avg: 20
(Mg/L)
4 day avg: 4.86
(Mg/L)
1 hour avg: 20
(Mg/L)
0.05 mg/L
(Max)
TP Pollution
Indicator
0.025 mg/L
(max) for lakes
0.025 mg/L
(max) for lakes
—
Pollutant Sources
Pollutant sources contributing to impairments in the Virgin River watershed include both point and
nonpoint sources. Field assessments supplemented with an aerial reconnaissance and photo analysis
facilitated a better understanding of pollutant sources in the watershed. Sources identified through the
field assessments included animal feeding operations (AFOs), wastewater lagoons, industrial sources,
areas of disturbance, streambank erosion, agricultural practices, agricultural return flows, sand and gravel
operations, and natural sources. Other sources contributing to impairments in the watershed include
septic systems, urban wet weather and dry weather flows, and geothermal activities. Different methods
were applied for developing loading estimations for pollutants associated with each source category.
Total Dissolved Solids
The TMDL identifies a variety of factors affecting total dissolved solids include natural erosion, geology,
geothermal, stormwater runoff, AFOs, irrigation return flow, reservoirs, streambank erosion, exotic
vegetation, and sand and gravel mining. According to the TMDL, TDS loads in North Creek, a segment
of the Santa Clara River (from Gunlock Reservoir to the confluence of the Virgin River), and the lower
Virgin River are primarily due to natural erosion, geology, and geothermal sources. Anthropogenic
nonpoint and point sources are relatively small percentages of the overall load.
Temperature
According to the TMDL, factors affecting heat loading include natural conditions and reservoir
management. The aerial and ground-based field assessment revealed a healthy riparian stream corridor
and natural channel geomorphology for the Santa Clara River segment listed as impaired due to
temperature, largest source of heat loading is from nonpoint sources, with anthropogenic nonpoint
source heat loading as the dominant pollutant source. NPDES point source loading is relatively small.
Nonpoint source thermal loading is attributed to lack of riparian shading (e.g, near-stream vegetation
disturbance and removal). For the Beaver Dam Wash subwatershed, DWQ conducted a specific
vegetative cover and temperature data analysis in to determine if the appropriate beneficial use was
assigned to this waterbody.
DRAFT
A-63
-------
Handbook for Developing Watershed TMDLs December 2008
Selenium
According to the TMDL, the largest source of selenium in the listed segment of the Santa Clara River is
from streambank erosion, with additional significant contributions from irrigation return flows. NPDES
point source loading from permitted stormwater sources is relatively small.
Total Phosphorus and Dissolved Oxygen
According to the TMDL, streambank erosion, livestock, and failing septic systems are the predominant
sources of total phosphorus for the Baker Dam and Gunlock Reservoirs. Internal loading is also a
significant source for Gunlock Reservoir. Streambank erosion is due to upstream sand and gravel mining
operations. The TMDL does not attribute any of the phosphorus load to point sources.
TMDL Development Approach
The technical approaches to develop the Virgin River watershed TMDLs varied based on type of impaired
waterbody. For streams, a statistical approach to linking sources and water quality was chosen rather
than a watershed model due to modeling challenges associated with diversions, canals, and irrigation
pathways that have altered the natural flow of the Virgin River, as well as seasonal snowmelt events. The
statistical approach used to calculate the TMDLs for impaired stream segments in the Virgin River
watershed included the following steps:
1. A flow duration curve for each segment was developed using the available flow data. This was done
by generating a flow frequency table that consisted of ranking all of the observed flows from the least
observed flow to the greatest observed flow and plotting those points.
2. The flow curve was translated into a load duration (or TMDL) curve by multiplying each flow by the
water quality standard and a conversion factor and plotting the resulting points.
3. Each water quality sample was converted to a daily load by multiplying the sample concentration by
the corresponding average daily flow on the day the sample was taken. The load was then plotted on
the TMDL graph.
4. Points plotting above the curve represent deviations from the water quality standard and unallowable
loads. Those plotting below the curve represent compliance with standards and represent allowable
daily loads.
5. The area beneath the TMDL curve is the loading capacity of the stream. The difference between this
area and the area representing current loading conditions is the load that must be reduced to meet
water quality standards.
6. Average annual loads were calculated by using a weighted-average approach based on the total
number of days associated with each flow percentile.
Data used for the load duration curve approach includes daily flow records for USGS gages and water
quality data collected by DWQ. Where the period of records for the USGS flow gage and the DWQ water
quality station do not coincide with each other, DWQ instantaneous flow values supplemented the USGS
daily flow records for use in the load duration analysis.
The load duration curve approach is not appropriate for calculating loading capacities for reservoirs.
Therefore, the BATHTUB model was selected to develop total phosphorus TMDLs for Gunlock and Baker
Dam reservoirs. For Gunlock Reservoir, the BATHTUB model facilitated determining the internal total
phosphorus loading. The phosphorus sedimentation term in BATHTUB is net sedimentation-that is, it
represents the rate of phosphorus settling minus the rate of resuspension/regeneration from the
sediment. The TMDL analysis interpreted internal loading as the difference between an estimate of
phosphorus deposition based on a phosphorus budget model (Chapra, 1997) and the BATHTUB net
sedimentation rate.
Stakeholder coordination in the TMDL development process occurred through VRWAC activities planned
and implemented to support comprehensive watershed management activities, including the development
of a locally-led watershed management plan and drinking water source protection plans for three surface
water intakes. Stakeholders contributed to the development of the Virgin River watershed TMDLs
through their participation in several meetings at key junctures in the project:
A-64 DRAFT
-------
December 2008 Handbook for Developing Watershed TMDLs
• Project Kickoff Meeting on June 11, 2002
• Field Survey Meeting on October 10, 2002
• Key Issues Public Meetings on July 9th and 10th, 2003
• TMDL Overview and Status Report Meeting on October 27, 2003
Allocations
The TMDL report for the Virgin River watershed TMDLs presents allocations for each of the impaired
segments. The source assessment and data analysis for several of the listed segments in the Virgin
River watershed revealed that exceedances of water quality standards for several parameters (i.e.,
temperature and total dissolved solids) are due to natural causes. As a result, the analysis conducted
through this watershed TMDL approach resulted in DWQ either delisting or not recommending a TMDL
for specific segments. For some segments, DWQ recommended site-specific standards. A summary of
the recommendations made in the final TMDL report is provided below for specific segments.
Beaver Dam Wash Temperature Listing
The TMDL report includes the results of DWQ's vegetative cover and temperature data analysis.
Through the analysis, DWQ determined that the high water temperatures in Beaver Dam Wash are
naturally occurring and the listing was due to an incorrect beneficial use designation, rather than an actual
water quality impairment. As a result, DWQ removed this segment from the 303(d) list.
North Creek Total Dissolved Solids Listing
Review of the available data for this subwatershed revealed that most sources of TDS are naturally
occurring and very few are anthropogenic. Through the analysis, DWQ recommended the development
of a site-specific standard of 2,035 mg/L (the 90th percentile of data for North Creek). Although the
recommendation is to remove North Creek from 303(d) list, the TMDL report does recommend some
implementation activities to address the small percentage of anthropogenic-related load.
Santa Clara River Total Dissolved Solids and Temperature Listings
The available temperature data for the Santa Clara River subwatershed revealed that a temperature
TMDL was not warranted; only four percent of the recent samples exceeded the standard of 27 °C, which
is within the allowable (ten percent) frequency of exceedances for aquatic life beneficial use support
classes. Therefore, DWQ recommended removing this waterbody from the 303(d) list for temperature.
Allocations for TDS and selenium were made for both point and nonpoint sources, although nonpoint
sources (streambank erosion, irrigation return flows) and upstream sources were found to have the most
significant contributions. A 24 percent reduction in annual TDS loads and a 9 percent reduction in
selenium loads is necessary to achieve water quality standards. Urban stormwater and dry weather flows
received the wasteload allocation for both TDS and selenium.
Baker Dam and Gunlock Reservoir Total Phosphorus and Dissolved Oxygen Listings
The TMDL analysis using the BATHTUB model revealed that approximately a 70 percent reduction in
total phosphorus loads to Baker Dam Reservoir is necessary to meet the 0.025 mg/L target, which is
predicted to result in meeting the dissolved oxygen standards. For Gunlock Reservoir, the BATHTUB
model determined that a 32 percent reduction in total phosphorus loads is necessary. The primary
sources of total phosphorus to both Baker Dam and Gunlock Reservoirs are livestock and failing septic
systems, indicated in part by the dissolved form of total phosphorus. The wasteload allocation is zero for
both reservoirs because there are no point sources.
Virgin River Total Dissolved Solids Listing
The recommendations for a portion of the Virgin River are similar to those for North Creek. Review of the
available data for the Virgin River below Pah Tempe Springs, a natural source of salts and other minerals,
indicate that the statewide TDS water quality standard is not achievable and that a site-specific standard
would be more appropriate. A TDS concentration of 2,360 mg/L, rather than 1,200 mg/L, is considered to
represent natural background conditions and is proposed as the site-specific criterion for the Virgin River
DRAFT A-65
-------
Handbook for Developing Watershed TMDLs December 2008
below Pah Tempe Springs to the Utah/Arizona border. The TMDL analysis used the recommended site-
specific criterion to determine the need for load reductions in the lower Virgin River. Based on the
analysis with the site-specific criterion, the TMDL recommends a 5 percent reduction in annual TDS
loads. The St. George wastewater treatment plant received the wasteload allocation. Because a majority
of the load comes from nonpoint sources (streambank erosion, irrigation return flow), nonpoint sources
received a larger load allocation.
Implementation
Because the Virgin River watershed TMDLs were developed as part of a comprehensive watershed
management approach, implementation recommendations to achieve the pollutant load reductions
appear in both the TMDL report and the locally-led Virgin River Watershed Management Plan.
Section 6 of the final TMDL report provides a detailed analysis of potential BMPs that stakeholders could
implement to achieve the load reductions for impaired segments in the Virgin River watershed. Figure 6-
1 in the TMDL report provides a graphical overview of the Virgin River watershed and the load reductions
necessary to achieve water quality standards. Appendix C of the TMDL report contains detailed fact
sheets best management practices (BMPs), including a description of the practice and the expected
pollutant load reductions. Tables in Section 6 of the TMDL report use the information from Appendix C to
present recommended sets of BMPs that could be implemented together to achieve the necessary
pollutant load reductions.
Lessons Learned
Utah's TMDL program is watershed-based in that TMDL analyses consider all contributing sources and
the process attempts to involve all affected stakeholders. The TMDLs for the impaired segments in the
Virgin River watershed were approached on a watershed-basis because there was a critical mass of
listed waters that were inter-related due to similar land-use, geology, and hydrology, as well as affected
stakeholders. Addressing these impairments on a watershed-basis provided significant savings in staff
and stakeholder time and resources.
Using a watershed approach allowed DWQ to characterize the watershed once and provided a more
holistic perspective on water quality issues. For example, because TDS was so prevalent in the listed
streams throughout the watershed, finding it in several places provided additional support that TDS loads
were due to natural sources and not site-specific anthropogenic sources.
Although DWQ ensures coordination among programs on all TMDLs, this project was unique in that it had
a high level of involvement from watershed stakeholders, particularly the Washington County Water
Conservancy District (WCWCD). The District provided funding and leadership for many of the
comprehensive watershed management activities, including development of the locally-led watershed
management plan and the drinking water source protection plans, that generated a large amount of
stakeholder participation. Through the VRWAC, stakeholders raised issues for consideration that aren't
typically considered, including water resource developments and detailed information on anticipated land
use changes.
References
DWQ (Utah Department of Environmental Quality, Division of Water Quality). 2004. TMDL Water Quality
Study of the Virgin River Watershed. September 2004. Salt Lake City, Utah.
VRWAC (Virgin River Watershed Advisory Committee). 2006. Virgin River Watershed Management
Plan. February 2006. St. George, Utah.
A-66 DRAFT
------- |