United States
Environmental Protection
Vgency
Site Characterization and Monitoring
Technical Support Center
FY08 Report
October 2007 - September 2008
Office of Research and Development
Office of Science Policy www.epa.gov/ord/osp
Superfund and Technology Liaisons
-------
Site Characterization and Monitoring
Technical Support Center
FY08 Report
October 2007 - September 2008
Prepared by:
U.S. Environmental Protection Agency
Office of Research and Development
Office of Science Policy
1200 Pennsylvania Avenue (8104R)
Washington, DC 20460
-------
CONTENTS
BACKGROUND 2
FY08 TECHNICAL SUPPORT 2
GENERAL SUPPORT
Scout- Online Modeling Software 2
ProUCL 4.00.02 Upgrades 3
Parallel Axes Software 3
Short-Term Implementation Requests 3
REGION/PROGRAM-SPECIFIC SUPPORT
Site-Specific Support by Region 4
- Central Chemical Site - Region 1 4
- NewportNAS Site- Region 1 4
- Centredale Manor Site - Region 1 5
- Natick Site - Region 1 5
- Massachusetts Military Reservation - Region 1 5
- Fort Devens - Region 1 6
- Passaic River- Region 2 6
- Cornell-Dubilier Electronics Site - Region 2 6
- OnondagaLake Site -Region2 6
- Patuxent River NAS Site-Region 3 7
- Hanlin-Allied-Olin Site - Region 3 7
- Delaware Sand & Gravel Site - Region 3 8
- Drake Chemical Site - Region 3 8
- FormerNansemond Ordnance Depot- Region 3 8
- Central Chemical Site - Region 3 8
- Olin Superfund Site - Region 4 9
- B.F. Goodrich Site - Region 4 9
- Barite Mine Cap Monitoring Design- Region 4 9
- USAF-Lockheed, Hanger 6- Region 4 9
- Remote Mine Monitoring System (Lutrell) - Region 8 10
- Rimmini Remote Mine Monitoring System - Region 8 10
- Goldstrike Mine - Region 9 10
- Navajo Nation Radioactive Structures - Region 9 10
- Jackson Park Site - Region 10 11
- Lower Duwamish Waterway Site - Region 10 11
- Eastern Michaud Flats - Region 10 11
- Bunker Hill MMC, Sequential Extraction Test- Region 10 11
- Sediment Background Statistical Workshop - Region 10 12
- Asbestos Sampler Training - Region 10 and OSWER 12
RARE Project (Asbestos) - Region 10 12
• Explosives Contaminated Site Issue Paper
- Federal Facilities Forum... ..12
-------
BACKGROUND
As of October 1, 2007, the Site Characterization and
Monitoring Technical Support Center (SCMTSC)
changed operation from the National Exposure
Research Laboratory (NERL) in Las Vegas to the
Superfund and Technology Liaison (STL) Program
of ORD's Office of Science Policy (OSP).
Felicia Barnett, the STL in Region 4, is the new
SCMTSC Director. The SCMTSC Director receives
requests for technical support from individual STLs
on behalf of their regional waste program staff. Each
STL works with his or her regional staff to determine
if and how ORD can best handle their technical
support needs. STL contact information is provided
at http://www.epa.gov/OSP/hstl.htm.
Each ORD Technical Support Center (TSC) provides
support on a different focus area. The SCMTSC
provides support on sampling and monitoring-related
issues.
Technical Focus of the SCMTSC
Soil-Gas Measurements
Special Analytical Services
GIS and Data Interpretation
Site Characterization Technologies
(e.g., field portable X-ray fluorescence)
Fingerprinting of Wastes
• Geophysics
Quality Assurance
• Geostatistics Statistical Design/Analysis
FY08 TECHNICAL SUPPORT
OCTOBER 1,2007-
SEPTEMBER 30, 2008
The following is a listing of all projects and requests
handled by the SCMTSC during FY08.
GENERAL SUPPORT
Scout - Online Modeling Software
The Scout software April 1 deliverable was completed
and delivered to John Nocerino, the EPA technical
representative, on March 23, 2008. Deliverables included
Scout software executable files and the Users' Guide.
During this reporting period, Scout software capabilities
were expanded to include univariate and multivariate
statistical analyses. The following capabilities were
developed and tested:
The Parallax software program was included as a
module within Scout. This program utilizes parallel
axes algorithms for data analyses.
Enhanced three-dimensional graphics capability.
Statistical capability modules [Orthoganolized
Ketterling Gnanadesikan (OKG) and Minimum
Covariance Determinant (MCD)].
Three upper-tailed upper confidence limit (UCL) menus
were added to the Classical Intervals module.
Least median squares (LMS) regression module.
Upper-tailed upper prediction limit/upper tolerance
limit (UPL/UTL) menu was added to the Classical
Intervals module [full and with non-detects (NDs) for
normal, gamma, lognormal, nonparametric, and all].
• Robust Intervals module (calculations on Confidence,
Prediction, and Simultaneous Intervals).
Tolerance limit calculations and graphics for all
modules.
MCD routines used in outliers, regression, and
discriminant analyses.
Cross-validation in discriminant analysis.
Iterative Regression module that contains separate
Huber, Proposed Method (PROP), Multivariate
Trimming (MVT), and biweight menus.
Module to generate ranked sets of correlation of
bivariate coefficients.
-------
After the deliverable was completed, enhancements to the
Scout software were requested by John Nocerino. The
following enhancements are being developed and once
completed, will be turned over from the SCMTSC to the
Las Vegas Lab for updating and management:
• A "grouping" option will be incorporated into the
intervals module. This will be useful to compare
confidence, prediction, and simultaneous intervals
for various populations (site area, monitoring wells)
using classical and robust statistical methods.
Add a prediction and confidence band option within the
linear-fitted regression module. Develop robust linear
modules.
• Add a quality assurance/quality control (QA/QC)
module.
Include options for selecting (user-selected) second-
and third-order interaction terms.
• Add a chapter describing the QA/QC module and
related updates to the Scout Users' Guide.
PROUCL 4.00.02 UPGRADES
ProUCL is a statistical environmental applications
software package designed for calculating upper
confidence limits (UCLs) for data sets with and without
ND observations. Provided online by the SCMTSC, it is
intended for use by government site and risk managers,
states, and environmental contractors. The software
package includes statistical methods that can be used
to estimate exposure point concentration (EPC) terms,
not-to-exceed, and background threshold values (BTVs).
ProUCL addresses various statistical issues arising in
exposure and risk assessment studies, background
evaluations, and background versus site comparison
applications. Some of the statistical methods (e.g., two-
sample hypotheses, upper prediction, and tolerance limits),
as incorporated in ProUCL 4.0, can be used in groundwater
(GW) monitoring applications (EPA, 1992). ProUCL can
be accessed at http://www.epa.gOv/esd//tsc/software.htm.
The ProUCL software is being upgraded with the
following capabilities:
1. Prediction and tolerance limits based upon gamma
distribution. These limits are often used to estimate
background-level concentrations and will be added
separately for data sets with and without ND
observations.
2. Include lower percentiles (upper percentiles are already
in ProUCL 4.0).
3. Expand Shapiro-Wilk (S-W) normality test up to
sample sizes of 2,000. At present, ProUCL 4.00.02 tests
normality using S-W for sample sizes up to 50.
4. Formatting upgrades. For data sets with very small
values, the outlier module of ProUCL 4.00.02 does not
provide the appropriate number of significant digits.
ND menus will be excluded when there are no NDs in a
data set.
PARALLEL AXES SOFTWARE
Commercially available software utilizing parallel axes for
analysis of data was evaluated for inclusion into the Scout
software. The parallel axes software chosen for inclusion in
Scout required unlimited distribution rights. The Evaluation
Report was delivered to John Nocerino on November 5,
2007. A commercial product, Parallax, was compared with
several freeware products. The Parallax program had more
capabilities compared to the freeware products, but the
Parallax Users' Manual required improvements. The
Parallax source code was requested from the developer and
was determined to be compatible with the Scout software.
The Parallax software was paid for and incorporated into
the Scout software as of August 2008.
SHORT-TERM IMPLEMENTATION
REQUESTS
ProUCL 4.00.02 was officially released on the EPA public
Web site in January 2008. The SCMTSC provided ProUCL
usage assistance to 68 users that requested support.
Assistance included administrative and technical guidance.
Administrative guidance included assistance with installing
the software, help with error messages, importing data, and
constructing data matrices. Technical guidance included the
use of statistical options within the software, identifying the
minimum number of samples required for a valid statistical
analysis, incorporating sample NDs into the statistical
analysis, identification of outliers, and identification of
multiple sample populations.
There were 36 requests for experts in Las Vegas or within
the SCMTSC Lockheed contract to respond to questions
concerning sampling issues, such as using bioassays for
water quality criteria or establishing minimum
requirements for a statistical analysis.
-------
REGION/PROGRAM -
SPECIFIC SUPPORT
In FY08, the SCMTSC provided support for 27 sites in
seven regions - 1, 2, 3, 4, 8, 9, and 10 (see Chart 1)
performing 52 separate tasks, including site-specific work,
requested training, a Regional Applied Research Effort
(RARE) project, and work for the Federal Facilities Forum
(see Chart 2).
CHART 1
FY08 Technical Support by
Number of Sites per Region
Region 1
Region 4
4 Sites
27 Total Sites
CHART 2
FY08 Technical Support by
Number of Tasks per Region
Region 4
7 Tasks 52 Total Tasks
SITE-SPECIFIC SUPPORT
BY REGION
CENTRAL CHEMICAL SITE - Region 1
Statistical methods (normal probability plots or Q-Q plots)
used to extract a background data set from a potentially
mixed sample were evaluated for the Central Chemical
Site. The following primary comments were delivered on
December 6, 2007:
1. The normal probability plot, or Q-Q plot-based method.
was considered an adequate method for extracting a
background data set from a potentially mixture data set.
provided it was used correctly.
2. A Q-Q plot should be generated using raw data (not
transformed data), because the objective is to extract the
background data (determined by the lowest segment of
a Q-Q plot) from a potentially mixture data set.
3. It was advised to avoid the use of a lognormal
distribution in the construction of a Q-Q plot or a
normal probability plot, because the use of a lognormal
distribution tends to hide contamination and
accommodate mixture populations as one single
population.
NEWPORT NAS SITE - Region 1
The statistical approaches used in the draft Background
Study Report for the Newport Naval Air Station (NAS) Site
were evaluated, and comments were submitted to the Task
Order Project Officer (TOPO) on November 26, 2007. The
following comments were provided:
1. Outliers should be properly identified in surface as well
as subsurface soil data sets. It was suggested not to use
outliers in the computation of background statistics and
control charts.
2. Inclusion and/or exclusion of outliers in the background
data set should be decided by the Project Team.
Decisions based upon distorted statistics (by outliers)
can be incorrect, misleading, and expensive. The Project
Team must define the objectives of the background
evaluations before deciding about the disposition of
identified outliers.
-------
The use of appropriate statistical methods is important
to derive correct and defensible conclusions. For an
example, a Mann-Whitney test used on data sets with
potential outliers may lead to incorrect conclusions.
In order to derive correct results and decisions, it was
recommended to use appropriate hypotheses testing
methods for data sets with and without NDs. It was
suggested to exclude outliers from the computations
of test statistics, which become distorted by potential
outliers.
Avoid the use of transformations (log-transformations
or power transformations), because cleanup and
remediation decisions are made in the original scale.
Therefore, statistics obtained in the transformed space
need to be back-transformed in the original scale. The
back-transformed statistics thus obtained often suffer
from significant and unknown amount of transformation
bias.
CENTREDALE MANOR SITE - Region 1
The technique used for fingerprinting the 2,3,7,8
tetrachlorodibenzo-p-dioxin (TCDD) contamination at the
Centredale Manor Restoration Project (CMRP) Superfund
site was evaluated, and the Summary Report was delivered
on January 21, 2008. The technique was provided by the
New England Container Company (NECC) Customer
Group Proposal (prepared by Exponent and Limno-Tech).
The described technique was vague and lacked the
necessary detail to reach meaningful conclusions.
Additional information was requested if the region wishes
to determine the validity of the fingerprinting results.
NATICK SITE - Region 1
Task 1 - An evaluation of the U.S. Army's response
to EPA comments on the Draft Feasibility Study Report
for the Natick Site was developed during January and
February 2008, and the report was delivered to the
Remedial Project Manager (RPM) on February 21, 2008.
Suggested improvements included consulting the ProUCL
4.00.02 Users' Guide to obtain the minimum sample size
requirements needed to perform two sample hypotheses
tests, including a test for proportions (exact and not
approximate), a t-test, a Mann-Whitney test, a Q test,
and a Gehan test. Appropriate goodness of fit tests should
be performed properly on data sets with ND observations.
The standard deviation of the site data should be compared
with the standard deviation of reference data before
computing a pooled standard deviation.
Task 2 - The report titled "Revision of Statistical Testing
Approach for U.S. Army SSC Sediment Feasibility Study"
was received on March 18, 2008. The report addressed the
February 21 evaluation comments of the original feasibility
study. Evaluation comments for the March 18 report were
delivered to the RPM on April 8, 2008.
Task 3 - The Draft Fall 2007 Sediment and Fish
Sampling Memo and appendices were received on April
30, 2008, with a request to evaluate and provide comments.
The statistical analysis of the April 30 materials was
delivered to the RPM on June 9, 2008.
Task 4 - A conference call was held on June 16, 2008,
to discuss the statistical analysis delivered on June 9.
Two additional outputs addressing polychlorinated
biphenyl (PCB) compound concentrations were
delivered to the RPM on June 16, 2008. Comments
from the June 16 conference call were incorporated
into the statistical analysis of the Draft Fall 2007
Sediment and Fish Sampling Memo and appendices,
and an updated Statistical Analysis Report was delivered
to the RPM on June 20, 2008. '
MASSACHUSETTS MILITARY
RESERVATION - Region 1
The SCMTSC was asked to review, refine, and further
develop Maximum Allowable Soil Concentrations and
provide oversight of sorption/desorption studies. As apart
of this effort, Idaho National Laboratory (INL) staff
conducted reviews of several documents and provided
comments on identification of contaminants of concern in
soils at the site. Specifically, the region was concerned that
if the contaminants became mobilized in the water, the
water would exceed the Maximum Contaminant Level
(MCL) or other critical standard. INL also provided
expertise on sorption and desorption experiments.
5
-------
FORT DEVENS - Region 1
Technical support was requested for the purposes of
conducting several radiation surveys using the INL portable
backpack-mounted system (BaSIS system) with real-time
measurement and integrated GIS capabilities. Use of the
BaSIS system supports an ongoing bedrock mapping
project at a granitic bedrock upland area. The purpose of
the surveys is to provide input to the site conceptual model
for defining locations of bedrock fractures based on the
measurement of naturally occurring radioactive materials.
primarily radon decay products. The challenges facing the
area in which the site resides center around the incidence of
fracture zones that may provide a pathway for arsenic
flowing into ground and surface water systems that supply
surrounding communities.
A detailed survey of this area using the BaSIS system
was scheduled for November 2008. The Conceptual Site
Model (CSM) will directly dictate drilling locations for an
EPA Region 1 bedrock drilling program to be undertaken
in late 2008/early 2009.
PASSAIC RIVER - Region 2
The Passaic River Creel Survey for the Diamond Alkali
Superfund Site was evaluated in October of 2007. The
report noted that the fish consumption rate analyses
appeared to have been based on insufficient data to
support the conclusions made from the survey.
CORNELL-DUBILIER ELECTRONICS SITE -
Region 2
Two data sets of PCB analytical results were compared
from the creek sediments at the Cornell-Dubilier
Electronics Site. The data sets were generated 10 years
apart, in 1997 and 2007. The evaluation was conducted
to determine if there had been a statistically significant
increase in the mean concentrations of PCBs over the
time period. The data sets contained PCB concentrations
obtained from soil samples (surface and subsurface) of
the site and from the sediment samples (shallow and deep)
from the nearby stream, Bound Brook. The study involved
the use of box plots, index plots, Q-Q plots, and hypothesis
tests. Gehan's two-sample hypothesis tests were conducted
because of the presence of NDs with multiple detection
limits. The hypothesis tests were reported to the RPM
on September 9, 2008, and showed that the soil PCB
concentrations from 2007 were higher than the
concentrations from 1997. This finding was also
supported by the box plots.
ONONDAGA LAKE SITE - Region 2
Statistical procedures used in the interim Human Health
Risk Assessment (HHRA) Report submitted to EPA on
September 8, 2008, were reviewed to determine if the
statistical analyses were conducted properly. The report
describes chromium speciation into trivalent Cr3+ (total
chromium, Cr) and hexavalent (Cr6*") forms. The chromium
ratio of hexavalent to trivalent is important because the
hexavalent form is more toxic. Speciated total Cr and Cr64"
data from all areas (including the biosolids area) of the
Wastebeds 1-8 subsite were used, and the species ratio was
estimated by the 97.5% upper bound of the slope of the
regression line (regressing Cr6*" on total Cr).
The statistical review submitted on September 16 had the
following conclusions:
The chromium concentrations (both Cr^ and total
chromium) were significantly higher in the biosolids
area than in the rest of the areas in the Wastebeds 1-8
subsite.
• As computed in the report, two-sided 95% confidence
intervals were computed for ratios: Cr^/Cr, with and
without using biosolids area data. The 97.5% upper
bounds of the regression slopes estimating the ratios are
summarized as follows.
Area
97.5 % Limit
of the Ratio
All areas in the site (from HRAA)
All areas in the site
All areas excluding biosolids area
All areas excluding biosolids area
and one outlier (WB18-SB-134)
Biosolids area
4%
4.41%
0.97%
0.877%
11.24%
The biosolids area yielded the highest estimate of the
Cr^/total Cr ratio. Chromium concentrations from the
biosolids area may not have been included in estimation
of Cr^/totalCr ratios.
The chromium concentrations of surface soil were
higher than the concentrations in subsurface soil in
the New York State Fair parking area.
-------
PATUXENT RIVER MAS SITE - Region 3
Task 1 - A report was developed summarizing the results
of a statistical analysis (including an outlier analysis)
performed on background soil data collected from the
Patuxent River Naval Air Station (NAS) Site. The report
was delivered on February 8, 2008.
Task 2 - An evaluation was conducted of a Background
Outlier Analysis Report that was developed by a consultant
for the site's potentially responsible party (PRP). On April
15, 2008, the following recommendation was provided:
In background evaluation studies, upper tolerance limits
(UTLs) at the 95% or 90% levels should be used to
estimate the background level concentrations.
Inclusion of outliers will typically yield elevated
background estimates (e.g., UTLs). In background
evaluations, it is a common and recommended practice
not to include outliers in the computation of UTLs. This is
especially true when background data are collected from
various onsite locations of Federal facilities (as is the case
for the Patuxent site).
Task 3 - The SCMTSC responded to e-mail questions
from the RPM and PRP in May and June concerning the
previous review.
HANLIN-ALLIED-OLIN SITE - Region 3
Task 1 - The Bayesian Sampling Design proposed for the
Hanlin-Allied-Olin site was evaluated, and the following
comments were developed and were delivered to the RPM
on December 17, 2007. Comments included:
1. The proposed Bayesian approach is an acceptable
approach. However, successful and practical application
of the proposed approach needs to be demonstrated. The
successful application of a Bayesian approach depends
upon the accuracy and adequacy of the prior
information. The uncertainties associated with the
estimates of the sample sizes (based upon regression
equations on simulated and estimated statistics)
remained unknown and intractable. Appropriate
references and related success stories should be
provided, or perform simulation experiments to
demonstrate the appropriateness of the proposed
Bayesian approach based upon (to determine sample
size) regression models on estimated statistics.
2. Alternative approaches, such as kriging, should be
evaluated. If kriging methodology cannot be used,
then it was suggested to use straightforward standard
deviation based upon sample size determination
separately for each of the three areas of the site. The
use of this sample size determination approach should
not be a problem because enough data were already
available to compute the standard deviations for the
three areas of concern.
Task 2 - In January and February 2008, Brian
Schumacher and Dale Werkema of the Las Vegas NERL
provided technical expertise and reviewed and commented
on a sampling plan for attempting to identify a possible
source of elemental mercury beneath concrete slab at
the site.
Task 3 - A conference call was held on February 20,
2008, to discuss the regression results used to compute
sample sizes that will be collected to statistically
characterize the analysis area of the Hanlin-Allied-Olin
site. The Bayesian sampling approach was used to
statistically characterize the analysis area of the site. The
sampling approach proposed by the remediation contractor
was evaluated to ensure the proper statistical approach.
Additional materials addressing the sampling design were
received on February 20 for discussion and evaluation. An
Evaluation Report of the sampling design materials and the
conference call was delivered to the RPM on March 26.
2008.
Task 4 - Another conference call was held on April 15 to
discuss comments on the sampling design approach. A
Statistical Report (Bayesian Statistical Analysis for Olin
Allied Site), developed by Geosyntec Inc., was received on
May 9 with a request to evaluate the report and provide
comments. The evaluation of the Statistical Report and
comments from the April 15 conference call were delivered
to the RPM on July 1, 2008.
Task 5 - A third conference call was held on August 27,
2008, to address the July 1 Evaluation Report and to
discuss the consumer risk and posterior probability
statistical approaches recommended for the site.
-------
DELAWARE SAND & GRAVEL SITE -
Region 3
Anita Singh participated in a conference call on March 3,
2008, with the EPA regional representative to discuss the
sampling plan for the Delaware Sand & Gravel site.
Statistical approaches were discussed, including several
sampling options. Recommendations were provided to
the regional representative to improve the sampling plan.
DRAKE CHEMICAL SITE - Region 3
SCMTSC completed the evaluation for the Drake
Chemical Site and delivered the report on November 18,
2007. The statistical approaches used for detecting trends in
concentrations of the contaminants of potential concern in
the groundwater monitoring wells were evaluated to verify
that appropriate methods were used.
FORMER NANSEMOND ORDNANCE DEPOT-
Region 3
An evaluation of the multiple increment sampling (MIS)
approach is being conducted to determine the applicability
and performance of the approach in risk assessment
applications (estimating the exposure point concentration
term). A preliminary simulation computer program was
developed and a field scenario was generated. Initial
simulation results did not support the claim that average
estimate of a decision unit based upon the MIS approach
would meet the data quality objectives (DQOs) for the site.
It appeared that MIS might dilute contamination "hot
spots," which could lead evaluators away from polluted
sites. MIS also might yield higher mean concentrations for
cleaner areas, leading to cleanup of sites with lower
contamination levels. A draft evaluation of the simulation
was submitted to the Region 3 risk assessor on September
18,2008.
CENTRAL CHEMICAL SITE - Region 3
The Probability Plot Analysis methodology for
determining background concentrations of arsenic in
soils proposed by the PRP in the Feasibility Study was
reviewed and evaluated. Comments and recommendations
on the methodology were provided to the RPM on
December 3, 2007.
-------
OLIN SUPERFUND SITE - Region 4
SCMTSC initiated support to Beth Walden, EPA Region 4,
to analyze a contamination plume using the FALCON
(Fingerprint Analysis of Leachate CQNtaminants)
fingerprinting technique. Releases of mercury and organic
chemicals have contaminated the shallow GW beneath the
Olin industrial site. Discharges of wastewater containing
mercury have contaminated wetlands adjacent to the
Tombigbee River. The FALCON technique will be used to
determine if the source of the mercury can be identified.
Data received to date is not adequate to identify a source
of the mercury. Sample analyses received in the fall of
2007 have variable mercury concentration along the flow
path of the waterway and do not indicate a pattern of
increasing mercury concentration near an industrial site.
When additional data are provided, a further evaluation
will be performed.
B.F. GOODRICH SITE - Region 4
An active Resource Conservation and Recovery Act
(RCRA) component, a former landfill, and a burn pit area
at the B.F. Goodrich Site are now being addressed under
Superfund. The PRPs were considering soil flushing to
address remediation, but Region 4 asked for help with the
dense non-aqueous phase liquid (DNAPL) assessment so
they could make a more informed decision about the site.
The SCMTSC is providing technical support sampling
expertise through a subcontract with Northwind via the
STREAMS contract.
Task 1 - Request for assistance during a Remedial
Investigation/Feasibility Study (RI/FS) Technical Meeting
with the PRPs on May 1, 2008. Bob Starr from Northwind
attended.
Task 2 - On June 7, 2008, reviewed and provided
comments to the RPM on the first draft of the "RI/FS Work
Plan for the B.F. Goodrich/Airco Superfund Site, Calvert
City, KY." The report was submitted on June 3, 2008.
Task 3 - On July 8, 2008, reviewed and provided
comments to the RPM on the second draft of the
"RI/FS Work Plan for the B.F. Goodrich/Airco
Superfund Site, Calvert City, KY." The report
was submitted on June 27, 2008.
BARITEMINE
BARITE MINE CAP MONITORING DESIGN -
Region 4
Region 4 requested TSC aid in the development of an
Integrated Cap Monitoring and Pit Monitoring Design for
the innovative mine pit lake neutralization using soluble
carbon sources and an evaporative waste rock dump cap.
The design will provide continuous monitoring of the site
while the natural cap is being established. The continuous
monitoring data will be used to evaluate the innovative
technology to identify and prevent any problems early. Gail
Heath of DOE's INL met with the RPM during the week of
August 1, 2008, and visited the site later in the week. A
draft design was expected in October 2008.
USAF-LOCKHEED, HANGER 6 - Region 4
Task 1 - On May 7, 2008, the RCRA Project Manager,
working with the Georgia Environmental Protection
Division (EPD), asked for expertise to answer questions
about studies used to analyze geochemical background data
as a means for determining whether metals in the soil are
releases or naturally occurring. Evan Englund in Las Vegas
responded the same day with pros and cons of different
approaches and then spoke with the RPM the next day.
Task 2 - On May 12, 2008, the RPM requested that Evan
review the report and provide comments. On June 12,
2008, Evan submitted comments to the RPM through the
SCMTSC, noting that the study was following standard
practices for site investigations, but recommended ways for
the study to focus on identifying barriers to making a final
decision for the site.
9
-------
REMOTE MINE MONITORING SYSTEM
(LUTRELL) - Region 8
The goal of this project is to design, build, install, and
evaluate an autonomous monitoring system that can
provide near real-time information on environmental
conditions and aqueous chemistry at remote mine sites.
During November and December 2007, the system
in Helena was operable and provided imagery and
temperature data from the site. Activities have
focused on the following tasks at this site.
Task 1 -Web site enhancement.
Task 2 - Deployment of the chemistry system to a
test site.
Task 3 - July fact-finding trip to Luttrell. The trip
included:
1. Planning and brainstorming meetings with the EPA
RPM and David Reisman at Helena.
2. A site walkdown with David Reisman and a
documentation effort of the existing system and issues.
3. A small report to INL management covering existing
issues and planned remediation.
4. Initiating a new scope of work (SOW).
Task 3 was completed the week of August 4, 2008, and a
new SOW is being negotiated with David Reisman/
Cincinnati.
RIMMINI REMOTE MINE MONITORING
SYSTEM - Region 8
Task 1 - Provide a remote monitoring system for a
treatment process for acid mine discharge from the Susie
mine. Funding was provided by TSC to install the final
upgrades to the control software, complete the final QA/QC
tests, and attend a walkdown by EPA and project review
meeting. Tests were run on the system in September 2007
to look at stability overtime. In December 2007, INL made
an overnight trip to the site, retrieved the system, and
brought it back to get it ready to install in May 2008.
Task 2 - A site trip to discuss issues with Dave Reisman
occurred the week of August 4, 2008.
GOLDSTRIKE MINE - Region 9
On March 19, 2008, Dale Hoff of the National Health and
Environmental Effects Research Laboratory (NHEERL) in
Duluth provided technical expertise to the Project Manager
on the issue of use of bioassays to determine the national
recommended Water Quality Criteria, particularly for
cadmium (Cd). He provided expertise on how reasonable it
was to decide an acceptable risk to fish in warm lake water
based on fathead minnow bioassays.
NAVAJO NATION RADIOACTIVE
STRUCTURES - Region 9
Task 1 - A statistical evaluation was performed on data
collected at locations on the Navajo Nation Reservation and
the results were delivered on May 21, 2008. Counts per
minute (cpm) data from four regions [Tee Nos Pec (TNP),
Cane Valley (CV), Red Valley (RV), and Tuba City (TC)] of
the Navajo site were analyzed separately and combined.
The primary objective of the analysis was to establish
background reference value(s) for the Navajo site. The
background values were computed for the four regions
(TNP, CV, RV, and TC) separately and also combined.
Region TC has only two structures, and both structures
exhibited elevated cpm values, suggesting that home sites
in Region TC may be potentially contaminated by the
Navajo site activities. Therefore, no separate background
statistics were computed for Region TC.
Task 2 - An additional request was received to evaluate a
report entitled "Structures & Homesite Yard Trigger Level
Determination Paper." Statistical comments were delivered
on June 11,2008.
10
-------
JACKSON PARK SITE - Region 10
Anita Singh attended a meeting on September 10, 2008,
at the Jackson Park Site (Bremerton, Washington) to
review and comment on the recommended sampling
approaches for evaluating the potential for buried
unexploded ordnance (UXO). Harry Craig (EPA
Region 10 technical representative) facilitated the
meeting to discuss the sampling plan for Operating
Unit 3. Methods recommended include the Visual
Sampling Plan (VSP). The potential use of kriging
was also discussed. VSP has a geostatistical module
to perform kriging (both indicator and ordinary) using
data collected along linear parallel transects. One item
recommended being included in the UXO module in VSP
5.1 software was the kriging standard deviation (KSD).
LOWER DUWAMISH WATERWAY SITE -
Region 10
Task 1 - Guidance was provided on August 5, 2008,
regarding the sample size determination for estimation
of the population mean. The number of samples required
depends upon the confidence coefficient and other
performance objectives, such as the specified departure
(difference) from the unknown true population mean.
Several options were presented, including the use of
the Chebyshev inequality method. Depending upon the
confidence coefficient (and therefore, a), the sample
size based upon Chebyshev inequality can be 2.5 or
more times larger than the variations available to compute
the sample size needed to estimate the population mean
with a specified confidence coefficient and desirable
difference/error.
Task 2 - A conference call was held on August 11, 2008,
to discuss the site background evaluations.
LOWER Dl SH WATERWAY
EASTERN MICHAUD FLATS - Region 10
On April 16, 2008, the RPM requested a review of the
"Geophysical Investigation for a Superfund Site in
Pocatello Idaho, Eastern Michuad Flats Superfund Site."
The document was a work plan for conducting a resistivity
geophysical investigation at the site. On April 23, 2008,
Dale Werkema of the Las Vegas Laboratory provided
review comments on the work plan, which included
suggestions about QA/QC, concerns about the inverse
modeling of the results, and how to make it as accurate
as possible.
BUNKER HILL MMC, SEQUENTIAL
EXTRACTION TEST - Region 10
The region requested assistance in performing sequential
extraction experiments on soil cores collected from the site.
There was a quick turnaround need for this project since
crews working for EPA were in the field and collecting
samples. The SCMTSC provided for INL to perform the
first phase of the work, sample collection. The sequential
extraction tests would be performed by INL for the region
under a separate interagency agreement (IAG).
Task 1 - On July 21, 2008, INL transmitted a draft
Quality Assurance Project Plan (QAPP) responding
to the SOW for the project.
Task 2 - Collection of the samples was completed by
August 7, 2008. The samples were stored in liquid nitrogen
pending approval of the second phase.
Due to costs of the project, the region then entered into
a regional IAG with INL for the second and third phases
of the work, including the sequential extraction and
batch or column studies. The IAG was signed on
September 18, 2008.
11
-------
SEDIMENT BACKGROUND STATISTICAL
WORKSHOP-Region 10
Preparation was initiated for Anita Singh to attend the
Sediment Background Workshop, hosted by EPA Region
10, in Seattle on October 7, 2008.
ASBESTOS SAMPLER TRAINING - Region 10
and OSWER
Task 1 - Provided hands-on training to regional and
headquarters personnel on June 24, 2008.
Task 2 - Provided a Standard Operating Procedure on
July 21, 2008, for the asbestos segregator developed with
RARE and OSWER program funds.
RARE PROJECT (ASBESTOS) - Region 10
The current phase of work consisted of 3 parts:
1. Design fluidized bed/cyclone separator for
separation of 5-20 [im particles.
2. Build the proposed separator designed in Task 1.
3. Demonstrate that the separator built in Task 2 is capable
of separating the desired size fraction and recovering
asbestos fibers from spiked oil samples without
significantly altering the aspect ratio of the fibers.
On October 2007, Jed Januch, Region 10, sent final
comments from the Technical Review Workgroup (TRW)
for Metals and Asbestos on the INL Fluidized Bed Report.
This project was completed in December 2007.
EXPLOSIVES CONTAMINATED
SITE ISSUE PAPER
Federal Facilities Forum
The SCMTSC provided technical information to assist
RPMs in developing CSMs and sampling plans to
investigate munitions residues at military training,
testing, and disposal ranges.
Task 1 - On November 4, 2007, Harry Craig sent
Joel Hubbard a report titled "Protocols for the
Collection of Surface Soil at Military Training and
Testing Ranges for the Characterization of Energetic
Munitions Constituents" with the comment that he
found the report to have good general format flow and
content that he thought would be a useful model for
the Issue Paper.
Task 2 - Joel Hubbard of INL provided an overview
of this Issue Paper to the entire Forum at the
Technical Support meeting in Las Vegas, Nevada, held
November 5-8, 2007.
Task 3 - INL then provided a draft report in March
2008.
Upon review, it was determined that the Federal
Facilities Forum wanted a more extensive report and
wished to have the Corps of Engineers Cold Regions
Research and Engineering Laboratory (CRREL) take
over using the draft as a starting place. The SCMTSC
is working with the Federal Facilities Forum and the
Technical Support Project (TSP) to support
completion of the paper if additional funding can be
provided by the TSP, because the project will be far
more expensive than the original SCMTSC budget.
12
-------
-------
&EPA
United States
Environmental
Protection Agency
Office of Research and Development
Office of Science Policy (8104R)
Washington, DC 20460
www.epa.gov/ord/osp
Recycled/Recyclable. Printed with vegetable oil-based inks on 100% postconsumer, process chlorine-free recycled paper.
------- |