United States
Environmental Protection
Agency
Solid Waste and
Emergency Response
(5102G)
EPA-542-R-00-009
August 2000
www.epa.gov/tio
cluin.org
&EPA
Innovations in Site
Characterization
Case Study: Site Cleanup of the
Wenatchee Tree Fruit Test Plot Site
Using a Dynamic Work Plan
-------
EPA-542-R-00-009
August 2000
Innovations in Site Characterization
Case Study: Site Cleanup of the Wenatchee Tree Fruit Test Plot
Site Using a Dynamic Work Plan
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation Office
Washington, DC 20460
August 2000
-------
Notice
This material has been funded wholly by the United States Environmental Protection Agency under
Contract Number 68-W6-0068. Mention of trade names or commercial products does not constitute
endorsement or recommendation for use.
Copies of this report are available free of charge from the National Service Center for Environmental
Publications (NSCEP), P.O. Box 42419, Cincinnati, OH 45242-2419; telephone (800) 490-9198 or (513)
489-8190 (voice) or (513) 489-8695 (facsimile). Refer to document EPA-542-R-00-009, Innovations in
Site Characterization Case Study: Site Cleanup of the Wenatchee Tree Fruit Test Plot Site Using a
Dynamic Work Plan. This document can also be obtained electronically through EPA's Clean Up
Information (CLU-IN) System on the World Wide Web at http://cluin.org or by modem at (301) 589-
8366. For assistance, call (301) 589-8368.
Comments or questions about this report may be directed to the United States Environmental Protection
Agency, Technology Innovation Office (5102G), 401 M Street, SW, Washington, DC 20460; telephone
(703) 603-9910.
August 2000
-------
Foreword
This case study is one in a series designed to provide cost and performance information for innovative
tools that support less costly and more representative site characterization. These case studies will
include reports on new technologies as well as novel applications of familiar tools or processes. They are
prepared to offer operational experience and to further disseminate information about ways to improve
the efficiency of data collection at hazardous waste sites. The ultimate goal is enhancing the cost-
effectiveness and defensibility of decisions regarding the disposition of hazardous waste sites.
Acknowledgments
This document was prepared by Science Applications International Corporation (SAIC) for the United
States Environmental Protection Agency's (EPA) Technology Innovation Office under EPA Contract No.
68-W6-0068. Special acknowledgment is given to the U.S. Army Corps of Engineers, Seattle District,
and Garry Struthers Associates, Inc. for their thoughtful suggestions and support in preparing this case
study.
iii August 2000
-------
Table of Contents
Notice ii
Foreword iii
Acknowledgments iii
CASE STUDY ABSTRACT vi
TECHNOLOGY QUICK REFERENCE SHEET vii
EnviroGard® DDT Immunoassay Test Kit vii
RaPID Assay® Cyclodienes Immunoassay Test Kit ix
EXECUTIVE SUMMARY 1
SITE INFORMATION 2
Identifying Information 2
Background 2
Site Logistics/Contacts 5
MEDIA AND CONTAMINANTS 6
Matrix Identification 6
Site Geology/Stratigraphy 6
Contaminant Characterization 6
Site Characteristics Affecting Characterization Cost or Performance 8
SITE CHARACTERIZATION AND REMEDIATION PROCESS 11
Systematic Planning and Sampling Work Plan 11
CHARACTERIZATION TECHNOLOGIES 21
Sampling Design and Methodology 21
Analytical Technologies and Method Modifications 24
Quality Assurance/Quality Control (QA/QC) Measures 27
PERFORMANCE EVALUATION 30
Performance Objectives 30
Strategy and Technologies Used to Attain the Performance Goals 30
COST COMPARISON 32
OBSERVATIONS AND LESSONS LEARNED 34
REFERENCES 35
iv August 2000
-------
List of Figures
Figure 1. Topographic map showing the location of the WTFREC relative to the town
of Wenatchee and the State of Washington 2
Figure 2. Site Plan for the WTFREC Test Plot 3
Figure 3. Disposal on the ground 3
Figure 4. Burial of concentrated pesticide products 3
Figure 5. Site plan showing the orientation of the rows and columns, sample locations,
and the two focused removal (RF) areas 10
Figure 6. Flow chart showing the integration of site characterization and remediation and
use of the dynamic work plan 12
List of Tables
Table 1. Established Contaminants of Concern for the WTFREC Test Plot Remediation 7
Table 2. Analytical Methods 17
Table 3. Sensitivities of Field and Fixed Laboratory Methods Relative to Cleanup Levels 18
Table 4. Example Removal Decision Matrix for Shallow Disposal 23
Table 5. Immunoassay Test Kit Performance Criteria 24
Table 6. Modifications to Reference Methods 26
Table 7. Summary of Field Duplicate and Equipment Blank QC Samples 29
Table 8. Time Frame for Activities 31
Table 9. Cost Comparison 33
August 2000
-------
CASE STUDY ABSTRACT
Wenatchee Tree Fruit Research and Extension Center (WTFREC) Test Plot
Wenatchee, Washington
Site Name and Location:
Wenatchee Tree Fruit Research and
Extension Center (WTFREC) Test
Plot
Wenatchee, Washington
Period of Operation:
1966-early 1980s
Operable Unit:
A 2,100-square foot test plot area
used for pesticide disposal testing
Sampling & Analytical Technologies:
1. Systematic planning process
2. Dynamic workplan
3. Direct push soil sampling
4. Field measurement immunoassay
analysis (IA) technologies combined
with limited fixed laboratory analyses
CERCLIS #:
None
Current Site Activities:
Washington State University test and
laboratory facilities; local residential
development.
Point of Contact:
Greg Gervais
Quality Assurance Representative
U.S. Army Corps of Engineers-
Seattle District
4735 East Marginal Way South
Seattle, WA 98134
Media and Contaminants:
Soil contaminated with organochlorine
pesticides, organophosphorus
pesticides, carbamate pesticides, and
paraquat
Technology Demonstrator:
Garry Struthers Associates, Inc.
3150 Richards Road, Suite 100
Bellevue, WA 98005-4446
(425)519-0300
Number of Samples Analyzed during Investigation:
A total of 271 samples were analyzed for the focused removal, characterization, final confirmation, waste profile, and
wastewater analysis phases of this project. Roughly two-thirds of analyses were performed in the field by IA kits. Field and
laboratory QC samples were also analyzed during this project.
Cost Savings:
The site characterization and cleanup approach used in this project resulted in savings of about 50% (over $500,000) over
traditional site characterization and remediation methods, which rely on fixed-base laboratory analysis with multiple rounds
of mobilization/demobilization to accomplish site cleanup.
Results:
Project was completed successfully and cost-effectively. The WTFREC test plot area was remediated, and shown to a high
degree of certainty that regulatory cleanup standards were achieved. The regulator, the client, and local stakeholders were
very satisfied with the project's outcome.
Description:
This case study describes an approach to site cleanup that includes the use of systematic planning, on-site measurement
technologies combined with limited fixed laboratory analyses, and rapid decision-making (using a dynamic work plan) to
facilitate quick cleanup. Site characterization information, obtained in the field through the use of IA kits, was used to guide
removal activities by means of an adaptive sampling strategy. This approach permitted a cost-effective cleanup of the
contaminated site.
VI
August 2000
-------
TECHNOLOGY QUICK REFERENCE SHEET
EnviroGard® DDT Immunoassay Test Kit
Case Study: Site Cleanup of the Wenatchee Tree Fruit Test Plot Site Using
a Dynamic Work Plan
Technology Name EnviroGard* DDT Immunoassay Test Kit
Summary of Case Study's Performance Information
Project Role:
Supporting in-field
decisions regarding further
characterization, removal,
waste segregation, and
disposal of soils
contaminated with DDT
and other pesticides.
Analytical Information Provided:
Semiquantitative concentration data for DDT and other organochlorine
pesticides in soil with sensitivity down to 0.2 mg/kg (ppm). The results are
reported as the concentration of DDT, but represent the sum of the responses
from the 2,4'- and 4,4'-isomers of DDT, ODD, and DDE. During the case study,
the test kit results were compared to fixed laboratory analyses for individual
pesticide compounds and site-specific action levels were developed for the
various decisions to be made (e.g., characterization, removal, waste segregation,
and disposal) using the test kit results.
Total Contract Cost: $13,036 for 230 samples
(includes project samples, PE samples, and blind field
duplicates)
Total Cost Per Sample: approx. $57 per sample
(includes QC costs)
Project Cost Breakdown
Spectrometer Cost:
$2000 for purchase, or
rentals available at
$175/day to $800/month
Consumables Cost:
$515fora20-testkit
Labor Cost:
approx. $20 per
sample (includes QC
costs)
Waste Disposal Cost:
Methanol extract waste:
$470 per lab pack (bulk)
disposal
Site-Specific Accuracy/Precision Achieved:
The test kit is intentionally biased 100% high by the manufacturer in order to
reduce the occurrence of false negative results. Based on a pilot study of the test
kits and fixed laboratory data for the individual organochlorine pesticides in soil
samples from the site, the project team determined that a DDT test kit result of 5
mg/kg (ppm) could indicate that the site-specific cleanup level for an individual
compound (e.g., DDT, DDE, or ODD) had been exceeded. An important aspect
of this project was that this initial determination was reviewed and revised as
needed during the latter phases of the project. For example, in the deeper soils
from the area of the site where bags of concentrated pesticides were buried, the
action level for DDT test kit results was raised to 10 mg/kg.
The precision achieved by the test kit was assessed by the analysis of a pair of
duplicate samples with each of 16 batches of field samples. The relative percent
difference of the duplicates ranged from 0% to 113% for these 16 batches, with a
mean RPD value of 38% and a median RPD of 28%.
Throughput Achieved:
A batch of 12 field samples
could be extracted and
analyzed in a half day by one
person.
vn
August 2000
-------
TECHNOLOGY QUICK REFERENCE SHEET
EnviroGard® DDT Immunoassay Test Kit (continued)
General Commercial Information (Information valid as of August 2000)
Vendor Contact:
Not available
Vendor Information:
Strategic Diagnostics, Inc.
Ill Pencader Drive
Newark, DE 19702
1-800-544-8881
www.sdix.com
Limitations on Performance:
This test kit is not specific for just DDT. It also
responds to the DDT daughter products DDE and
DDD, as well as some other organochlorine
pesticides.
Principle of Analytical Operation:
This test is based on a competitive enzyme-linked
immunosorbent assay (ELIS A) reaction between DDT
and related compounds extracted from the sample with
methanol and an antibody coated on a test tube
containing the extract.
The antibodies bound to the target analytes cannot bind
to an enzyme conjugate added to the tube. When a
color-developing reagent is added, the enzyme
conjugate forms a colored product. The color density is
read with a spectrometer and is proportional to the
amount of conjugate reagent present. Darker color
means less of the target analyte is present. The DDT
results are determined by comparison to 3-point
calibration.
Availability/Rates:
Test kits are commercially available as off-the-shelf
products. Associated test equipment, including hand-
held spectrometer, is available for purchase or rental
from manufacturer.
Power Requirements:
110 or 220 volt power is needed to charge the hand-
held spectrometer, which may then be used in the field
without additional power.
Instrument Weight and/or Footprint:
Approximately 5 square feet of space is required for
sample processing and analysis.
General Performance Information
Known or Potential Interferences: Other organochlorine pesticides can react with the antibodies to varying
degrees. The manufacturer provides cross-reactivity data with the test kit.
Applicable
Media/Matrices:
Soil and Water
Wastes Generated
Requiring Special
Disposal:
Small volumes of
methanol used for sample
extraction, plus the used
sample volume.
Analytes Measurable
with Expected
Detection Limits:
DDT 0.2mg/kg
DDD 0.05 mg/ kg
DDE 0.6mg/kg
Other General Accuracy/Precision Information:
See SW-846 Method 4042
Rate of Throughput:
Up to 17 samples can be assayed at one time, with
results available in 30 minutes.
Vlll
August 2000
-------
TECHNOLOGY QUICK REFERENCE SHEET
RaPID Assay® Cyclodienes Immunoassay Test Kit
Case Study: Site Cleanup of the Wenatchee Tree Fruit Test Plot Site Using
a Dynamic Work Plan
Technology Name RaPID Assay* Cyclodienes Immunoassay Test Kit
Summary of Case Study's Performance Information
Project Role:
Supporting in-field
decisions regarding further
characterization, removal,
waste segregation, and
disposal of soils
contaminated with
cyclodiene pesticides.
Analytical Information Provided:
Semiquantitative concentration data for cyclodiene pesticides in soil with
sensitivity down to 0.15 mg/kg (ppm). Greater sensitivity was achieved in this
project through method modifications. The results are reported as the
concentration of dieldrin, but other cyclodiene pesticides can be used to calibrate
the assay as well.
During the case study, the test kit results were compared to fixed laboratory
analyses for individual pesticide compounds and site-specific action levels were
developed for the various decisions to be made (e.g., characterization, removal,
waste segregation, and disposal) using the test kit results.
Total Contract Cost: $13,036 for 230 samples
(includes project samples, PE samples, and blind field
duplicates)
Total Cost Per Sample: approx. $57 per sample
(includes QC costs)
Project Cost Breakdown
Spectrometer Cost:
$2000 for purchase, or
rentals available at
$175/dayto$800/month
Consumables Cost:
$540 for a 20-test kit
Labor Cost:
approx. $20 per
sample (includes QC
costs)
Waste Disposal Cost:
Methanol extract waste:
$470 per lab pack (bulk)
disposal
Site-Specific Accuracy/Precision Achieved:
The test kit is intentionally biased 100% high by the manufacturer in order to
reduce the occurrence of false negative results. Based on a pilot study of the test
kits and fixed laboratory data for the individual organochlorine pesticides in soil
samples from the site, the project team determined that a Cyclodienes test kit result
of 0.086 mg/kg (ppm) could indicate that the site-specific cleanup level for an
individual compound (e.g.,dieldrin or endrin) had been exceeded. An important
aspect of this project was that this initial determination was reviewed and revised
as needed during the latter phases of the project.
The precision achieved by the test kit was assessed by the analysis of a pair of
duplicate samples with each of 14 batches of field samples. The relative percent
difference of the duplicates ranged from 0% to 110% for these 14 batches, with a
mean RPD value of 35% and a median RPD of 7%.
Throughput Achieved:
A batch of 12 field samples
could be extracted and
analyzed in a half day by one
person.
IX
August 2000
-------
TECHNOLOGY QUICK REFERENCE SHEET
RaPID Assay® Cyclodienes Immunoassay Test Kit (continued)
General Commercial Information (Information valid as of August 2000)
Vendor Contact:
Not available
Vendor Information:
Strategic Diagnostics, Inc.
Ill Pencader Drive
Newark, DE 19702
1-800-544-8881
www.sdix.com
Limitations on Performance:
This test kit is not specific for just a single
cyclodiene pesticide. It responds to:
dieldrin, aldrin, endrin, heptachlor, heptachlor
epoxide, chlordane, endosulfan (I and II), • -BHC,
• -BHC (lindane), • -BHC, and several other
organochlorine pesticides.
Principle of Analytical Operation:
This test is based on a competitive enzyme-linked
immunosorbent assay (ELISA) reaction between
cyclodiene compounds extracted from the sample with
methanol and an antibody bound to a magnetic particle
and added to a tube containing the extract.
The antibodies bound to the target analytes are
separated from the extract using by retaining the
magnetic particles with a magnetic field and decanting
off the extract. When a color-developing reagent is
added, the enzyme conjugate forms a colored product.
The color density is read with a spectrometer and is
proportional to the amount of conjugate reagent
present. Darker color means less of the target analyte is
present. The cyclodiene results are determined by
comparison to 3-point calibration.
Availability/Rates:
Test kits are commercially available as a special order
products. Associated test equipment, including hand-
held spectrometer, is available for purchase or rental
from manufacturer.
Power Requirements:
110 or 220 volt power is needed to charge the hand-
held spectrometer, which may then be used in the field
without additional power.
Instrument Weight and/or Footprint:
Approximately 5 square feet of space is required for
sample processing and analysis.
General Performance Information
Known or Potential Interferences: Other organochlorine pesticides can react with the antibodies to varying
degrees. The manufacturer provides cross-reactivity data with the test kit.
Applicable
Media/Matrices:
Soil and Water
Wastes Generated
Requiring Special
Disposal:
Small volumes of
methanol used for sample
extraction, plus the used
sample volume.
Analytes Measurable
with Expected
Detection Limits:
From manufacturer:
Cyclodienes, as dieldrin:
0.15 mg/kg in soil and
0.6 ug/kg in water
As employed for the
case study:
18 ug/kg (ppb) in soil.
Other General Accuracy/Precision Information:
See SW-846 Method 4041
Rate of Throughput:
Up to 50 samples can be assayed at one time, with
results available in 60 minutes.
August 2000
-------
Wenatchee Tree Fruit Test Plot
EXECUTIVE SUMMARY ^^^^^^^^^^^^^^^^^^^^^^^^^5
This case study describes an approach to site cleanup that includes systematic planning, on-site
measurement technologies combined with limited fixed laboratory analyses, and rapid decision-making
using a dynamic work plan to facilitate quick cleanup. The integration of site characterization, on-site
measurements, on-site remedial decision-making, and remedial action resulted in the expedited and cost-
effective cleanup of a site contaminated with pesticides.
The test plot area of the Wenatchee Tree Fruit Research and Extension Center (WTFREC) contained
soils contaminated with organochlorine pesticides, organophosphorus pesticides, and other pesticides due
to agriculture-related research activities conducted from 1966 until the mid-1980s. In 1997, the U.S.
Army Corps of Engineers (USAGE) implemented an integrated site characterization and remediation
project at the site. This approach permitted characterization, excavation, and segregation of soil based on
the results of rapid on-site analyses employing commercially-available immunoassay testing products.
Key to the project's success was a pilot test that assessed the suitability of the on-site analytical methods.
Site-specific contaminated soil was analyzed by both immunoassay (IA) methods and by traditional fixed
laboratory methods. The results of the pilot test demonstrated the applicability of the DDT and
cyclodiene pesticide IA methods and provided comparability data that the project team used to develop
site-specific action levels that would guide on-site decision-making using the IA results. The IA action
levels were refined during the course of project implementation as additional comparability data sets
(composed of matched IA and fixed laboratory results) became available.
A soil excavation profile was developed in the field using the analytical results according to a decision
matrix developed by the USAGE. Several phases of field activities were conducted under a dynamic
work plan framework using an adaptive sampling strategy. Characterization and cleanup were
accomplished within a single 4-month field mobilization, and the entire project cost was about half the
cost estimated according to a more traditional site characterization and remediation scenario relying on
multiple rounds of field mobilization, sampling, sample shipment, laboratory analysis, and data
assessment. The costs of waste disposal were significantly reduced by using field analyses to
characterize and segregate wastes that required costly incineration from other wastes that were suitable
for less expensive disposal methods. The "surgical" removal of contaminated materials ensured that
closure testing would demonstrate regulatory compliance to a high degree of certainty, while making
field activities such as sample collection, sample analysis, soil removal, soil segregation, and final
disposal of soil and wastewater highly efficient and effective.
The key features of the project that contributed to its success included:
Systematic planning accomplished by a team representing the USAGE, EPA, the site owners, and
state regulators with the appropriate mix of skills and decision-making authority.
A conceptual site model based on a review of historical records from the site.
A dynamic work plan that permitted the field team to make real-time decisions on the basis of
data generated in the field.
The pilot study that demonstrated the utility of the field analyses and provided data that were
used to establish site-specific action levels.
An adaptive sampling and remediation strategy that relied on the combination of the field
analyses and fixed laboratory data.
August 2000
-------
Wenatchee Tree Fruit Test Plot
SITE INFORMATION
Identifying Information
Site Name: Wenatchee Tree Fruit Research and Extension Center (WTFREC) Test Plot
Location: Wenatchee, Washington
Technology: Site Cleanup Using a Dynamic Work Plan and Immunoassay Field Kits
Operable Unit: None
CERCLIS #: None
ROD Date: None
Background
Physical Description: The Wenatchee Tree Fruit Research and Extension Center (WTFREC), an
agricultural research facility, is located in southeast Wenatchee, Washington (see Figure 1).
/, WenttAce
V •
Washington
--•V" -v» :• •
:$fcfe
..WSU Tree Fruit
Research C«mter
Scale 1:50,000
Figure 1. Topographic map showing the location of the WTFREC relative to the town of Wenatchee and
the State of Washington
August 2000
-------
SITE INFORMATION continued
Wenatchee Tree Fruit Test Plot
In the past, the U.S. Public Health Service (PHS), and
the U.S. Environmental Protection Agency (EPA)
used a 2,100 square-foot test plot area located in the
northeast corner as a pesticide disposal research area.
During the initial stage of the site remediation study,
the location and dimensions of that test plot were
determined based on the location of existing barbed
wire fencing. Based on the fence location, the
approximate dimensions of the test plot were 70 feet
by 30 feet, and the area was located approximately 23
feet south of the WTFREC facility's northern
property line. However, after evaluation of sampling
results from investigations conducted by Washington
State University (WSU) and EPA, the U.S. Army
Corps of Engineers (USAGE) concluded that lateral
contamination extended beyond the previously
identified edge of the test plot area. The new
dimensions of the contaminated area were then
determined to be 85 feet by 33 feet. The test plot is
adjacent to a graduate student mobile home, an
unpaved access road, and a nearby manufactured
home development (see Figure 2).
ELEW5ONS IN FEET
Figure 2. Site Plan for the WTFREC Test Plot
Site Use: The WTFREC was historically used as an agricultural research facility. The test plot area was
initially used by the PHS, and later by the EPA, as a test facility to determine the effectiveness of various
land disposal methods for pesticides.
Pesticide disposal testing reportedly began in 1966 and continued until the early 1980s. The disposal
experiments focused on organochlorine (OC) and organophosphorus (OP) pesticides, but could possibly
have included the testing of other pesticides. Pesticide burial was conducted at the site using the
following three methods:
(1) Pesticides were diluted with solvent and poured through the openings of
cinder blocks (see Figure 3); / ^H
(2) Pesticides were diluted with solvent and poured directly onto the ground
surface; and
3) Pesticides were mixed with lime, lye, or
Purex®, placed in paper bags and buried
two to three feet below the ground
surface (see Figure 4).
^j||||?l^fc'~- ~"
Jfc-
"
Figure 3. Disposal on
the ground.
Figure 4. Burial of
concentrated pesticide products
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE INFORMATION continued ^^^^^^^^^^^^^^^^^^^^^^5
In the mid-1980s, the property was transferred from EPA to the Washington State University (WSU).
WSU currently operates test and laboratory facilities at the WTFREC and uses the orchards shown in
Figure 2 as their primary research areas. Nearby residential development is changing the land use
pattern, increasing the concern that the test plot be remediated.
Release/Investigation History: Between 1985 and 1987, WSU performed limited sampling and
analysis of soil in and near the test plot in response to concerns about pesticide contamination. After this
initial sampling, WSU contacted EPA and asked for assistance in characterizing and remediating the test
plot site. EPA and its contractors performed site investigations, which included sampling and analysis, in
1990, 1991, and 1994. Sampling activities included the collection of four background samples from an
area approximately 1,200 feet west of the test plot.
EPA's Office of Research and Development (ORD) obtained assistance from the USAGE for the purpose
of remediating the test plot site. USAGE used sample results from the WSU and EPA sampling events to
determine the primary areas of OC and OP pesticide contamination at the site. Prior to writing
specifications for the test plot remediation, the USAGE reviewed records and publications from the
research facility and contacted several WTFREC researchers for additional information regarding
experiments at the site. Based on this research, the USAGE identified the three reported methods of
pesticide disposal used during pesticide research activities at the WTFREC.
Given the history of pesticide disposal at the site, there were significant concerns regarding the vertical
migration of pesticides in the test plot area. Research articles written by EPA researchers in the 1970s
indicated that no significant pesticide contamination was expected at depths greater than 8 inches below
any of the initial disposal depths in the test plot area. Sampling performed by WSU and EPA in the
1980s and 1990s at the test plot area confirmed this expectation. USAGE used the article findings and
sampling data from EPA's and WSU's investigations to develop initial plans for characterization and
excavation at the test plot area.
Regulatory Context: The Wenatchee Tree Fruit Test Plot cleanup was performed under the regulatory
oversight of the State of Washington Department of Ecology's Voluntary Cleanup Program.
August 2000
-------
^^m SITE INFORMATION continued
Site Logistics/Contacts
Wenatchee Tree Fruit Test Plot
"Customer" or Responsible Party:
Howard Wilson
U.S. Environmental Protection Agency (USEPA)
Office of Research and Development (ORD)
USEPA Headquarters/Ariel Rios Building
1200 Pennsylvania Avenue, NW
Washington, DC 20460
(202) 564-1646
Regulatory and Oversight Agency:
Washington State Department of Ecology
Thomas L. Mackie
Central Regional Office
15 West Yakima Ave -- Suite 200
Yakima,WA 98902-3401
(509) 454-7834
Project Manager:
Ralph Totorica
U.S. Army Corp of Engineers - Seattle District
4735 East Marginal Way South
Seattle, WA 98134
(206) 764-6837
Technical Site Contact/Quality Assurance
Contact:
Greg Gervais
Quality Assurance Representative
U.S. Army Corp of Engineers - Seattle District
4735 East Marginal Way South
Seattle, WA 98134
(206) 764-6837
Kira Lynch
Project Environmental Scientist/Chemist
U.S. Army Corp of Engineers - Seattle District
4735 East Marginal Way South
Seattle, WA 98134
(206) 764-6918
Technology Demonstrator:
Mike Webb
Garry Struthers Associates, Inc.
3150 Richards Road, Suite 100
Bellevue, WA 98005-4446
(425)519-0300 (x217)
August 2000
-------
Wenatchee Tree Fruit Test Plot
MEDIA AND CONTAMINANTS ^^^^^^^^^^^^^^^^^^^^^^5
Matrix Identification
Type of Matrix Sampled and Analyzed: Soil
Site Geology/Stratigraphy
The WTFREC is situated at approximately 800 feet above sea level and 194 feet above the normal
elevation of the Columbia River. The WTFREC is located approximately two miles east of the Columbia
River. The eastern foothills of the Cascade Mountains, which begin approximately one-half mile to the
west of WTFREC, rise to about 2,000 feet above sea level. The site lies on an alluvial fan deposited
along a steep drainage that flows eastward from the Cascade Mountains to the Columbia River. The
alluvial soils are composed of poorly sorted boulder gravel and gravelly sand with some clay layers. The
surface gradient in the area is approximately 200 feet per mile. The gradient portion becomes less steep
as the alluvial fan merges with the Columbia River flood plan.
Contaminant Characterization
Primary Contaminant Group: Table 1 contains a list of the established contaminants of concern and
action (cleanup) levels used for the WTFREC Test Plot remediation. The primary contaminant groups
include organochlorine pesticides, organophosphorus pesticides, carbamate pesticides, and paraquat. The
action levels in Table 1 were based on the specifications of the Washington State Model Toxics Control
Act (MTCA) and range over five orders of magnitude. See the "Site Characterization and Remediation
Process" section for more information on establishing cleanup levels during this study.
The on-site and fixed laboratory analyses performed for this project focused on two groups of
organochlorine pesticides: the cyclodienes and the DDT series. The cyclodiene group is characterized
by a six-membered ring with an endomethylene bridge structure (a double bond between two carbons at
one end of the ring). The specific cyclodienes of interest at the WTFREC site included: aldrin,
chlordane, dieldrin, endrin, endrin aldehyde, endrin ketone, endosulfan I and II, endosulfan sulfate,
heptachlor, heptachlor epoxide, and toxaphene.
The DDT series consists of the various isomers (2,4'- and 4,4'-) of DDT, as well as the isomers of the
related compounds DDE and ODD. The compounds of greatest toxicological concern are the 4,4'-
isomers, which are also typically the most prevalent compounds contained in commercial DDT
formulations. The toxicological data for the 2,4'-isomers are more limited, and 2,4'-DDT was generally
present in lesser amounts in commercial formulations than 4,4'-DDT (often a 20/80 percent mixture of
the 2,4'- and 4,4'- isomers), although the exact ratio varies with formulation and manufacturer. As a
result of the scarcity of toxicity data for the 2,4'-isomers alone and the desire to have protective action
levels, the action levels used for the WTFREC test plot remediation were based on the sum of both
isomers (2,4'- and 4,4'-) for all three compounds in the DDT series.
On-site analyses for DDT and cyclodienes were used to guide the decisions of the dynamic work plan.
Fixed laboratory analyses for the primary contaminant group in Table 1 were used to establish a closure
confirmation data set for regulatory compliance.
August 2000
-------
Wenatchee Tree Fruit Test Plot
MEDIA AND CONTAMINANTS continued
Table 1. Established Contaminants of Concern for the WTFREC Test Plot Remediation
Suspected Contaminant
MTCA
Method B*
Cleanup
Level (mg/kg)
Organochlorine Pesticides
Dieldrin
Endrin
Endrin aldehyde**
Endrin ketone**
Endosulfan I
Endosulfan II
Endosulfan sulfate**
DDT***
DDE***
ODD***
gamma-BHC (lindane)
Methoxychlor
Aldrin
alpha-BHC
beta-BHC
delta-BHC
Chlordane
Heptachlor
Heptachlor epoxide
Toxaphene
0.0625
24
24
24
480
480
480
2.94
2.94
4.17
0.769
40
0.0588
15.9
0.556
0.556
0.769
0.222
0.110
0.909
Suspected Contaminant
MTCA
Method B*
Cleanup
Level (mg/kg)
Organophosphorus Pesticides
Di-Syston (disulfoton)
Guthion (azinphosmethyl)**
Parathion
Methyl parathion
Aminomethyl parathion* *
Malathion
Ethion
DDVP (dichlorvos)
Diazinon
Dimethoate
Paraoxon-ethyl**
Paraoxon-methyl* *
3.20
3.20
480
20
20
1600
40
3.44
72
16
480
20
Carbamate Pesticides
Carbaryl
Furadan (carbofuran)
8000
400
Miscellaneous Pesticide
Paraquat
360
**
The Washington State Model Toxics Control Act (MTCA) specifies three methods for establishing
cleanup levels, Methods A, B, and C. Method B is the standard method for cleanup of soil and was
used at the WTFREC Test Plot remediation. See the "Site Characterization and Remediation
Process" section for more information on the use of MTCA Method B cleanup levels during this
study.
The action level is based on the parent compound's action level.
*** The action levels used for the site were based on the sum of the concentrations of the 2,4'-isomers
and the 4,4'-isomers of each compound (e.g., the sum of o,//-DDT and/?,/?-DDT).
August 2000
-------
Wenatchee Tree Fruit Test Plot
MEDIA AND CONTAMINANTS continued ^^^^^^^^^^^^^^^^^^5
Site Characteristics Affecting Characterization Cost or Performance
The design of the study and the implementation of field and laboratory activities were influenced by
several site-specific characteristics. These included:
Above-ground objects and vegetation that required removal prior to field sampling
The presence of concentrated pesticide products buried at the site
The need to segregate the excavated materials for cost-effective disposal
Removal of Above-Ground Objects and Vegetation: A number of objects that were in and
immediately adjacent to the test plot at the commencement of the work were removed and disposed of
according to the Remedial Action Management Plan (RAMP). These included the barbed wire fence and
fence posts, the chemical storage shed, and the trash cans. Additionally, all of the vegetation within the
boundaries of the test plot was cleared to a level of approximately two-inches above the ground surface
or less (GSA, Inc. 1998, p. 15).
Excavation and Removal of Concentrated Pesticide Products: Concentrated pesticide products had
been buried at two locations on the site. Prior to characterizing the entire site, these buried products were
removed during "focused removal" activities. These activities consisted of excavation of materials based
upon visual indicators, followed by closure confirmation sampling of the areas to ensure that all of the
contaminated materials had been removed.
Figure 5 is a site plan showing the orientation of the rows and columns established for the cleanup
activities as well as the locations of the various types of samples that were collected. The rows in Figure
5 were established based on historical data from the site regarding the pesticide disposal experiments that
were conducted there. As noted earlier, in addition to burying bags of concentrated pesticide products
mixed with lime, lye, or other chemicals on the site to monitor their breakdown, pesticides were diluted
with solvents and poured through concrete blocks on the site, and mixed with soil and placed directly
onto the surface. Each row includes areas used for similar disposal experiments. For example, during
the site characterization phase, samples collected from columns 1 and 9 were only analyzed for OC
pesticides, and samples collected from columns 2 through 8 were analyzed for both OP and OC
pesticides. The columns were drawn perpendicular to the rows to provide a grid spacing that was
statistically determined to allow detection of a hypothetical 5 foot by 10 foot elliptical hot spot.
The two focused removal areas were each approximately 10 feet wide (east-west direction) by
approximately 24 feet long. One area was identified as Focused Removal Area 2/3 (FR2/3) because it
spanned adjacent portions of columns 2 and 3 on the site; while the other area was identified as Focused
Removal Area 4/5 (FR 4/5), because it spanned portions of columns 4 and 5 (see Figure 5). Based upon
the USAGE review of the research records, the materials removed from FR2/3 were expected to contain
elevated levels of OP pesticides and the FR4/5 materials were expected to contain elevated levels of OC
pesticides.
Bags of concentrated pesticide materials were encountered within each of the two areas, at approximately
18" below ground surface (bgs). Excavation continued downwards until approximately 6" of soil was
removed below the last visually-observed bag remnant. Final excavation depths were approximately 27"
bgs for FR2/3 and approximately 33" bgs for FR4/5. Excavated materials were segregated according to
expected contaminant and concentration during excavation and placed directly into designated roll-off
bins. A total of 45.74 tons of material was excavated during the focused removal activity, 22.32 tons
from FR2/3 and 23.42 tons from FR4/5.
8 August 2000
-------
Wenatchee Tree Fruit Test Plot
MEDIA AND CONTAMINANTS continued ^^^^^^^^^^^^^^^^^^5
Segregation of Excavated Materials for Disposal: With over 45 tons of material excavated from the
focuses removal activities, the potential costs to dispose of those materials were significant. Of the
contaminants of concern shown in Table 1, endrin and lindane were significant disposal concerns
because of their presence on the list of constituents for the RCRA hazardous waste toxicity characteristic.
All wastes generated during the remediation activities were to be recycled, salvaged, incinerated, or
disposed of in a RCRA Subtitle C permitted landfill. The following three different "disposal"
classifications were anticipated, based on RCRA and the Washington State waste regulations:
• Dangerous waste
• Non-dangerous waste
• All other solid waste (including demolition debris, personal protective equipment, etc.)
The "dangerous waste" included soil containing pesticides and contaminated with endrin and lindane at
levels in excess of the RCRA toxicity characteristic limits. The "non-dangerous waste," a State of
Washington designation, consisted of soils that passed the toxicity characteristic, but contained
contaminants in excess of the State of Washington limits.
The IA testing product for the cyclodienes responds more strongly to endrin than to any other cyclodiene
other than chlordane. Therefore, after correlating the IA results with gas chromatographic analyses
conducted off-site during the pilot study, the on-site IA results for the cyclodienes were used to identify
those excavated materials that were high in endrin and therefore designated for the most costly disposal
option, incineration. The IA testing product for DDT responded to DDT, DDE, and ODD, and the on-
site results were similarly correlated with gas chromatographic analyses conducted off-site during the
pilot study.
The wastes in the roll-off bins were profiled in this fashion, based upon analytical data and generator
knowledge. In addition, TCLP leaching was conducted off-site, based on the IA results, and used for
final classification of the endrin-containing wastes.
August 2000
-------
II
z
7.
Row A
RowB
k
]
RowC
Legend
A
1 O
D
I
LJ
I
A
« «|s
53
S
» ^
+
I H
D
.
_-
O O
o_ o_
h- I-J
1-1
4
a
J
Q
"
1 [
D
I
d
n
D
LA
L '
H
L
1
D+
^
S)
§
4^
n
I
'
D
J !
A
D
I
D
1
A
n
A
*
n
I
D
n
A
|
i
n
D
i
n
o
n ^
'o.
A
1
D
1
v ^
O
z
' >
g
HH
Z
^
Z
H
05
O
o
a
rt-
5'
e
1 a
o o o o o o o
£.£.£.£.£. ° £.
w
% X-Y coordinate origin
+ Focused removal confirmation samples
4- (Jt ON
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS ^^^^^^^^S
Systematic Planning and Sampling Work Plan
Prior to implementing the remedial action at the WTFREC Test Plot, the USAGE and their contractor
(GSA, Inc.) planned the project by preparing narrative and quantitative acceptance and performance
criteria for data collection, a field sampling plan (FSP), and a quality assurance project plan (QAPP).
Project planning was based on the specifications set forth in the Remedial Action Management Plan
(RAMP). Current EPA guidance suggests that acceptance and performance criteria be developed for data
collection, evaluation, using the Data Quality Objectives (DQO) process. The DQO process is part of an
overall systematic data collection planning process and ensures that the right type, quality, and quantity
of data are collected to support overall project-level decision making (e.g., see Data Quality Objectives
for Superfund: Interim Final Guidance (USEPA 1993) and other guidances for the Data Quality
Objectives Process (USEPA 1994, 1999, and 2000). The use of systematic planning, and subsequently,
the use of a dynamic work plan, optimizes all site activities (not just data collection) and achieves the
most effective results.
Planning and Field Teams: Planning and field teams were created to include the appropriate mix of
skills and regulatory authorities needed to plan and implement cleanup of the WTFREC test plot. In
particular, the regulatory authority (Washington State Department of Ecology) was involved in the
planning process and approved the use of the dynamic work plan and the decision logic to be used during
the cleanup.
The Planning Team was comprised of representatives from EPA ORD (as the USACE's customer), the
regulator (Washington State Department of Ecology), stakeholders (Washington State University, as
property owner, represented by the Environmental Manager, the Facility Manager, and an Environmental
Scientist in charge of cleanup issues), the USAGE Project Manager/Team Leader, and the USAGE
Project Chemist/Scientist, Project Engineer, Health & Safety Industrial Hygienist, and a Construction
Engineer.
The Field Team was comprised of representatives from the USAGE (Project Manager/Team Leader,
Project Chemist/Scientist, Construction/Project Engineer, Field Quality Assurance Officer, and Health &
Safety); the prime contractor (Project Manager, Field Engineer, Project Chemist/QC Officer); and
subcontractors to perform excavation, IA, operate the Geoprobe, and manage soil disposal activities.
Conceptual Site Model: The initial conceptual site model (CSM) was developed by the USAGE after
review of records and publications available at the research facility and based on contacts with WTFREC
researchers. The information indicated that vertical migration of pesticides to a depth greater than eight
inches below the disposal point was not expected at the test plot area. In addition, the information
indicated that there would be negligible horizontal migration of pesticides at the site.
The initial remediation boundary of the investigation was established based on the location of an existing
barbed wire fence around the site. The approximate dimensions of the test plot were determined to be 70
feet by 30 feet. For additional information on delineation of the test plot area, see the discussion below
in DQO process Step 4, "Define the Boundaries."
Dynamic Work Plan: Based on a pilot study, the USAGE determined that site decisions could be made
in the field, aided by the use of semiquantitative data (i.e., data used to make a decision about whether
concentrations were above or below a certain action level) generated using on-site measurement
technologies. The use of data generated on-site would allow relatively quick decision-making regarding
subsequent steps. This approach would efficiently guide the characterization and removal efforts by
means of an adaptive dynamic sampling strategy. Using adaptive sampling and analysis strategies, field-
11 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
generated results were used to update the CSM and to better direct the analyses of the next batch of
samples (see Figure 6).
This study approach permitted rapid location and definition of "hot" areas, guided the removal of
contaminated soil, and quickly identified when enough information had been collected to address the
remedial decisions. With this approach, the project team minimized the collection and analysis of
uninformative samples, avoided unnecessary removal of soil, avoided multiple rounds of
mobilization/demobilization of equipment and personnel, and efficiently identified when the project was
"done," thus saving time and money.
Figure 6 shows the overall flow of work, including the systematic planning and the implementation of the
dynamic work plan. The use of the field analytical methods allowed for integration of the site
characterization with site remediation. In particular, site characterization information was used in the
field to make soil removal decisions. In Figure 6, the field sampling, field analysis, and decision-making
are shown in an iterative and dynamic "loop."
Systematic Planning
Implementation
Assessment
Assemble and Review |
Existing Data and
Develop/Revise Conceptual
Site Model (CSM) J
Implement Data Quality \
Objectives (DQO) Process
1. State the Problem
Identify/establish:
- planning & field team
- resources & deadlines
- communication strategy
2. Identify the Decision
3. Identify Inputs to the
Decision, including:
- Measurements to be
made
- Action levels
- Candidate analytical
methods
4. Define the Boundaries
5. Develop a Decision Rule
6. Specify Limits on Decision
Errors
7. Optimize the Design for
v Obtaining the Data J
Conduct Pilot Test Using IA Kits
and Set Test Kit Action Levels
Prepare Dynamic Work Plan
and QAPP
Mobilization
Focused Removal
Integrated Site Characterization and Remediation
Site Characterization
Field sampling
On-site analyses by IA and definitive
analysis at fixed laboratory.
Removal of Contaminated Soil [expand
boundaries if needed)
Characterization, Classification, and Disposal of
Contaminated Material
No
Confirmational Sampling and Analysis
Statistical Analysis To Confirm
Attainment of MTCA Cleanup
Standards
Site Restoration
Figure 6. Flow chart showing the integration of site characterization and remediation and use of the
dynamic work plan.
12
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
Application of the Data Quality Objectives Process: The initial planning steps, stated in terms of
EPA's DQO process, are described below:
Step 1: State the Problem - In this step of the DQO process, it is necessary to define the problem,
identify the planning team, and establish a budget and schedule. For the purpose of the remedial action,
the problem was to identify those soils and wastes which were contaminated.
The specific goals of the WTFREC Test Plot Remediation included:
• Focused removal of concentrated pesticide product
• Gross removal of pesticide-contaminated soil
• Restoration of the site to achieve the MTCA Method B Cleanup Levels
• Characterization, classification, and disposal of contaminated materials.
As described previously, planning and field teams were assembled with the appropriate mix of skills
needed to plan and implement the cleanup project. The planning team specified an expedited schedule
for completion of the remedial action.
Step 2: Identify the Decision - Three decisions were identified during this step of the DQO process. The
first decision was to determine whether the soil within each "exposure unit" (described below) was
contaminated above the action levels established under the MTCA for each contaminant of concern
(COC). Any soils contaminated above the action levels had to be removed. Any soil that was not
contaminated at or above those levels could remain in place.
After removal, a second decision was required to determine if the remaining soil attained the cleanup
standard.
Once they were removed from their original locations, soil and other wastes required appropriate
disposal, based upon RCRA and the Washington State Dangerous Waste Regulations (WAC 173-303).
Therefore, the third decision was to determine the appropriate classification of the remediation waste for
disposal purposes. Three different waste classifications were used: dangerous waste, non-dangerous
waste, and solid waste (including demolition debris, personal protective equipment, etc.). Each
classification involves different disposal methods, including incineration for the dangerous wastes, the
most costly approach. Therefore, it was critical that wastes from the site be segregated on the basis of
their waste classification in order to control disposal costs.
Step 3: Identify Inputs to the Decision - This step of the DQO process required a list of the information
inputs needed to resolve all parts of the decision statement. For example, to make remedial decisions
(i.e., to remove or not remove the soil), the necessary inputs included, at a minimum, a list of
contaminants of concern and action (cleanup) levels (see Table 1), the units of measure (e.g., mg/kg or
mg/L), target quantitation limits, candidate analytical methods capable of achieving the quantitation
limits, and measurement performance criteria.
A list of constituents of concern were identified based on previous investigations conducted by WSU and
the USEPA. The Washington State Model Toxics Control Act (MTCA) establishes three basic methods
for establishing cleanup levels: Methods A, B, and C. The MTCA Method B is the standard method for
determining cleanup levels for ground water, surface water, soil, and air. Cleanup levels are established
using applicable state and federal laws or by using the risk equations and criteria specified in the MTCA
regulations. The planning team determined that the Method B was an appropriate method for setting the
cleanup levels for those COCs with calculated MTCA Method B levels.
13 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
For COCs that do not have calculated MTCA Method B levels, the USAGE, EPA, Washington State
Department of Ecology, and WSU agreed to use the MTCA Method B cleanup levels for their parent
compounds (e.g., endrin ketone and endrin aldehyde had the action level of endrin and endosulfan sulfate
had the action level of endosulfan I).
Table 1 contains the list of the contaminants of concern and the MTCA Method B cleanup levels
established for this project. The quantitation limits for the field and fixed laboratory analyses were
established as described in Step 7.
It was determined that commercially-available immunoassay field test kits could measure two of the most
important classes of pesticides, DDT and two cyclodienes, dieldrin and endrin. The availability of the
test kits proved to be a critical element in optimizing the study design (see DQO Step 7), implementing a
dynamic work plan, and using real-time decision-making to streamline the cleanup process.
Step 4: Define the Boundaries - In this step, the planning team developed a detailed description of the
spatial and temporal boundaries of the cleanup problem.
Initially, the surface location and dimensions of the test plot area were established based upon the
location of the barbed wire fencing. The barbed wire fencing secured a rectangular area with
approximate dimensions of 69 feet-9 inches (from east to west) by 29 feet-9 inches (north to south).
From the previous investigations, however, the USAGE concluded the horizontal extent of
contamination, as defined by the MTCA Method B action levels, was not necessarily confined to the
fenced test plot. For the initial conceptual site model (GSM), the USAGE decided to extend the
boundary of the area of potential contamination as follows:
Another three feet beyond the northern edge of the test plot
An additional 5.5 feet beyond the eastern edge of the test plot
Another 10 feet beyond the western edge of the test plot.
Other locations within and near the test plot were identified by the USAGE as having minimal to no data
indicating the presence of contaminants. However, during the site characterization, as the CSM matured,
the boundaries were extended slightly beyond the original boundary established for the remedial action
(see Figure 5). Samples collected by EPA from the non-orchard area indicated that the background
pesticide levels in the area did not exceed the MTCA Method B cleanup levels (GSA, Inc. 1998).
The test plot was divided into nine columns (1 through 9) and three rows (A, B, and C), making 9
removal columns and 27 sampling grids. Each column was a separate "exposure unit" and was
established by the USAGE to correspond with a discrete potential removal location, based on historic
data on disposal locations, as well as past sampling and analysis actions. The final determination of
attainment of the cleanup standards was made based upon evaluation of the entire footprint of the test
plot site (i.e., all nine columns).
Depth of contamination was another spatial boundary of concern for site remediation. Within the site
boundary, two areas or were identified within which bags of concentrated pesticide product were buried.
Based on historical information, it was determined that pesticide product may have been buried to depths
up to 4 feet (48 inches) below ground surface (bgs). Historical data and research indicated that migration
of pesticide contamination beyond this depth was expected to be minimal (i.e., an additional 8 to 12
inches). These two areas were designated as FR2/3 and FR3/4 and were excavated as part of the focused
removal excavation (see previous discussion of "Excavation and Removal of Concentrated Pesticide
Product" on page 8) followed by closure confirmation sampling of the areas.
14 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
The temporal boundary (i.e., time frame for project completion) was established based on the desire to
complete on-site activities prior to the onset of winter. The winter climate at Wenatchee often includes
cold temperatures and snow. Therefore, completion of the site activities before winter was important to
ensure worker safety and to avoid weather-related delays of excavation and sampling. In addition, EPA
requested an expedited cleanup schedule in order to show good faith to the stakeholders.
Step 5: Develop a Decision Rule - In this step, the planning team specified the parameters of interest,
action levels, and developed a decision rule.
As noted previously in "Media and Contaminants" (see page 6), the DDT series consists of the various
isomers (2,4'- and 4,4'-) of DDT, as well as the isomers of the related compounds DDE and ODD. As a
result of the scarcity of toxicity data for the 2,4'-isomers alone and the desire to have protective action
levels, the USAGE, EPA, Washington State Department of Ecology, and WSU agreed that it was
appropriate to add up the soil concentrations of the 4,4'- and 2,4'-isomers of DDT and to compare this
value with an action level based on the sum of both isomers (2,4'- and 4,4'-) for all three compounds in
the DDT series.
A soil removal decision matrix was established for both the "shallow burial columns" and the "deep
burial columns" to guide the field sampling and establish a basis for removal and confirmation sampling,
or no further action. For example, if the immunoassay field kits found contamination in the interval 0 to
12" bgs at concentrations exceeding the action level established for the kit, then additional analyses were
performed on samples representing the interval 12" to 24" bgs. If no contamination was found above the
action level, then the 0 to 12" interval was removed and the removed soil was subjected to confirmation
sampling and analysis.
Based on the IA results and the decision matrix, more samples were actually collected than were
analyzed. This type of decision rule was applied to depths no greater than 72" bgs. Sampling was
limited to depths of 72 inches because the USAGE believe that all pesticide contamination would
effectively be found within that depth interval. This was based on the assumption that no pesticide
product was disposed below 4 feet (48 inches) bgs and that migration of pesticides would be minimal
(less that one foot) beyond that depth.
Finally, for the closure confirmation data to demonstrate attainment of the cleanup standards, the data
must pass three statistical tests. These tests are:
The analyte concentration for no more than 10 percent of the samples can exceed the cleanup
standard for that analyte;
No sample concentration can exceed a level more than two times the cleanup standard for
any particular analyte; and
The upper confidence limit (UCL) of the data for each analyte must be statistically shown to
be less than the cleanup criteria for that analyte.
The procedure to be used to calculate UCLs depends on the distributional assumptions that are made
about the data (e.g., normal, log normal, or other distribution) and the size of the sample population. For
the WTFREC test plot cleanup, UCLs were calculated using guidance published by the State of
Washington Department of Ecology (see Ecology 1992 and 1995). For most of the data sets, an
assumption of a log normal distribution was appropriate, and in these cases the UCL was calculated using
Land's method as described in the Washington State Department of Ecology guidance. For data sets that
15 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
contained a large percentage (>50%) of nondetects, the largest value in the data set was used as the UCL
in accordance with the Washington State Department of Ecology guidance.
Step 6: Specify Limits on Decision Errors - A decision error occurs when sampling data mislead the
decision maker into choosing a course of action that is different from or less desirable than the course of
action that would have been chosen with perfect information (i.e., with no constraints on sample size and
no measurement error). Data obtained from sampling and analysis are never perfectly representative and
accurate, and the costs of trying to achieve near-perfect results can outweigh the benefits. Uncertainty in
data must be tolerated to some degree. The DQO process controls the degree to which uncertainty in
data affects the outcomes of decisions that are based on those data. This step of the DQO process allows
the decision maker to set limits on the probabilities of making an incorrect decision.
When the data lead you to decide that the baseline condition (or "null hypothesis") is false when in fact it
is true, a "false rejection" decision error occurs (i.e., the null hypothesis is falsely rejected - also known
as a false positive decision error or Type I error). In the reverse case, a "false acceptance" decision
occurs when the data lead you to decide that the baseline condition is true when it is really false (i.e., the
null hypothesis is falsely accepted - also known as a false negative decision error or Type II error).
For the final calculation of upper confidence limits on the mean using the closure confirmation sampling
data, the Type I error rate (•) was set at 0.05 as specified by the requirements of the MTCA. Setting the
error rate at this level ensures there is only a 5% chance of falsely rejecting the null hypothesis. In other
words, when the MTCA standard has not truly been met, the chances are only 1 in 20 that the statistical
test will erroneously conclude it has been met.
Step 7: Optimize the Design for Obtaining the Data - The objective of this step is to use the outputs of
the first six steps of the DQO process to develop a sampling and analysis plan that obtains the requisite
information from the samples for the lowest cost and still satisfies the project objectives.
For this project, the overall DQOs were as follows:
• Provide field analytical results for DDT and cyclodienes (especially dieldrin and endrin)
with quantitation limits that are less than the field/operational action levels in order to guide
the removal of contaminated soil from each defined "column" of soil at the site such that
final cleanup goals will be met within a single field mobilization.
• Ensure that the turnaround time for the field-generated data supports the real-time decision-
making needs of the dynamic work plan.
• Collect sufficient soil data to confirm that the soil left in place meets the MTCA cleanup
standards such that:
no more than 10 percent of samples exceed the cleanup standard,
no sample can exceed two times the cleanup standard, and
the true mean concentration must be below the cleanup standard as measured by a 95%
upper confidence limit on the mean.
• Provide analytical results that can be used to segregate and classify excavated soil and other
remediation wastes for management as solid, hazardous, or dangerous waste according to
RCPvA and the Washington State Dangerous Waste Regulations.
16 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
Pilot Test
In an effort to develop the analytical plan and identify a cost-effective analytical strategy, a pilot test of
the IA methods was conducted using contaminated surface soil from the site. The pilot study was critical
to the success of this project in that it allowed the investigators to demonstrate the usefulness of the IA
methods for on-site analysis of soils for DDT and cyclodiene at their respective soil cleanup levels,
thereby providing an important tool for on-site decision making and implementation of the dynamic work
plan approach.
By their nature, the commercially-available IA testing products relevant to this study are not specific to a
single target compound. Rather, the antibodies used in the kits bind to a variety of structurally-similar
contaminants. Therefore, although the test kit may be calibrated using one specific pesticide, the
response generated during the test is due to all of the potential reactants present in the sample, each of
which elicits a response to a different degree. Since the cleanup levels for this and most other projects
are based on specific contaminants, the IA test results cannot be used to make cleanup decisions without
considering the site-specific nature of this limitation.
The pilot study was designed to evaluate the utility of the IA test kits by comparing their results to a more
traditional fixed-laboratory, contaminant-specific analytical approach. Samples of soil from the test plot
were collected and split into two portions, one for IA analysis and one for the traditional approach. The
results of both types of analyses were evaluated by the project team to determine the utility of the IA
results for site-specific decision making.
Analytical Method Selection
Analytical methods for the pilot study were selected that could achieve the method performance
requirements established by the project team and documented in the QAPP (GSA, Inc. 1997b). A list of
the analytical methods is presented in Table 2.
Table 2. Analytical Methods
Analyte
Cyclodiene IA field test
DDT IA field test
Organophosphorus pesticides
Organochlorine pesticides
Carbamates
Paraquat
Method
SW-8464041
SW-846 4042
SW-846 8 141, modified*
SW-846 8081
SW-846 8 141, modified*
RM-8-10**
* GC/MS was used in Method 8141 for the OP pesticides. The carbamate analyses used GC/NPD.
** This is a spectrophotometric method based on procedures developed by Chevron Oil
Modification of Methods under PBMS
As noted in Table 2, some of the reference methods were modified to accommodate the specific
contaminants of concern at the site. These modifications were designed by the project team that included
an analytical chemist and were conducted in accordance with the performance-based measurement
system (PBMS) approach adopted by EPA in recent years. The modifications are described in greater
17 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
detail in "Analytical Technologies and Method Modifications." In addition to the modifications to the
fixed laboratory reference methods, the IA methods were modified in order to allow a single sample
extract to be analyzed using both the DDT and cyclodienes test kits.
Establishing Site-Specific Action Levels for the Field Test Kits
The pilot study results confirmed that the IA test kits are intentionally biased 100% high by the
manufacturer in order to reduce the occurrence of false negative results. Combined with the fact that the
test kits respond to more than one of the contaminants of concern at the site, the project team determined
that a DDT test kit result of 5 mg/kg (ppm) could indicate that the site-specific cleanup level for an
individual compound (e.g., DDT, DDE, or ODD) had been exceeded. Similarly, they determined that a
cyclodienes test kit result of 0.086 mg/kg (ppm) could indicate that the site-specific cleanup level for an
individual compound (e.g., dieldrin or endrin) had been exceeded. These values (5 ppm and 0.086 ppm)
became the site-specific field action levels associated with the DDT IA test kit and the cyclodiene IA test
kit, respectively, at the start of field work.
Final Method Selection
The analytical methods used for cleanup phases of the project were based on the methods modified for
the pilot study (see Table 2). The sensitivities of the analytical methods selected for the field IA testing
and fixed laboratory confirmation analyses were evaluated relative to the MTCA Method B cleanup
levels established for this project. The goal was to employ a method that was sensitive enough to make
measurements at no more than one-half the MTCA Method B cleanup level. Table 3 illustrates the
sensitivities for the major contaminants of concern relative to the MTCA Method B cleanup levels.
Table 3. Sensitivities of Field and Fixed Laboratory Methods Relative to Cleanup Levels
Contaminant
Dieldrin
Endrin
4,4'-DDT
4,4'-DDE
4,4'-DDD
MTCA Method B
Cleanup Level (mg/kg)
0.0625
24
2.94
2.94
4.17
Field Method
Sensitivity* (mg/kg)
0.018
-
0.8
-
-
Fixed Laboratory Method
Sensitivity** (mg/kg)
0.00007
0.00012
0.0013
0.0036
0.00017
*The IA test kit sensitivities were established by the concentration of the lowest of the calibrator solutions analyzed
using the test kit. The cyclodiene kit used for dieldrin and endrin was calibrated using chlordane and the DDT test
kit was calibrated using DDT. Thus, the values above represent quantitation limits for the specific compounds used
for calibration.
**The fixed laboratory method sensitivities were based on the method detection limit (MDL) values reported by the
laboratory. Thus, the values above represent detection limits, and not quantitation limits, but they are specific to the
individual analytes listed. The MDL values were reported by the laboratory in units of ng/kg, and have been
converted to mg/kg in this table for ease on comparison with the cleanup levels.
Field Analytical Quality Control
Following the pilot test, the chemist and the project team designed a field analytical quality control (QC)
program that was used to monitor and ensure the quality of the field results. That program included the
18
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
use of such traditional QC operations such as calibrations and laboratory control samples, as well as
continuing to submit some split samples for fixed laboratory analyses in order to detect potential
interferences and to monitor the comparability of the field and fixed laboratory results over time and
across different areas of the site.
Monitoring and Refining the Action Levels
As a result of the continued generation of fixed laboratory results for a subset of all the samples collected
for field kit analyses, the field kit action levels were further refined after the characterization phase.
Comparison of the IA and fixed laboratory data sets generated during the characterization phase
determined that the 5 ppm field action level being used for the DDT IA kit was overly conservative.
With the approval of the regulator, the DDT IA field action level was raised to 10 ppm for the removal
phase of the project.
Site Cleanup Phases
Using information from previous site investigations and the results of the pilot study, the cleanup project
was designed to take place in seven phases.
Phase 1: Mobilization
Phase 2: Focused removal of pesticide product
This phase employed field test kit IA analyses with fixed laboratory confirmation of a
subset of those results.
Phase 3: Characterization of the remediation area
This phase employed field test kit analyses for DDT and cyclodienes, fixed laboratory
analyses for the organophosphorus and carbamates pesticides and Paraquat, as well as
fixed laboratory confirmation of a subset of the field test kit results, leading to the
revision of the action levels for the test kits in some areas of the site.
Phase 4: Gross removal of contaminated soil
This phase employed field test kit IA analyses
Phase 5: Final confirmation sampling for site closure
This phase employed fixed laboratory analyses.
Phase 6: Backfilling, grading, and restoration
Phase 7: Characterization and disposal of contaminated materials.
The final phase employed fixed laboratory analyses of soil samples as well as the
production and analysis of TCLP leachates to characterize RCRA-regulated wastes.
19 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S SITE CHARACTERIZATION AND REMEDIATION PROCESS continued ^^^^S
Optimizing the Sampling and Remediation Program
The optimization strategy focused on Phases 3, 4, 5, and 7 of the site cleanup. One of the key elements
of the optimization of the sampling and remediation program was the use of field methods to make
remedial decisions in the field (primarily during Phases 3 and 4).
In Phases 2 and 3, the sampling strategy for the site characterization was optimized by the use of a
"focused" sampling design in which sampling was conducted in areas where potential or suspected soil
contamination could reliably be expected to be found. Another example of the optimization was the use
of direct push soil sampling technology (i.e., Geoprobe) in lieu of traditional and more costly drill rig and
split-barrel samplers. Using homogenization and sample splitting techniques, the team was able to
provide sample volumes for IA analysis, fixed laboratory analysis (if needed), and archiving from a
single collection event (see additional discussion under "Sampling Design and Methodology" on page 21
of this report).
In addition, the team employed field analyses using IA and supported by limited fixed laboratory
analyses to increase the density of sample locations compared to that possible under traditional sampling
and analysis programs. This facilitated the "surgical" removal of contaminated materials and ensured
that closure confirmation testing would demonstrate compliance to a high degree of certainty. The
combined benefits of the optimized approach produced both time savings and significant reductions in
the overall project costs by making field activities such as sample collection, sample analysis, soil
removal, soil segregation, and final disposal of soil and wastewater highly efficient.
On-site activities in all phases were facilitated by the use of a mobile office trailer and a mobile
laboratory trailer. The cost of trailer rental was more than offset by savings realized from the on-site
analyses (see also "Cost Comparison" in this report).
Note that the advantages of using field methods include the ability to match the rate of sample processing
with the rate of sample collection providing efficient sample handling (e.g., minimal sample tracking,
transport, and storage) and rapid turnaround time of field results in relation to the desired on-site
decision-making abilities.
20 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S CHARACTERIZATION TECHNOLOGIES ^^^^^^^^^^^^^^^^^S
Sampling Design and Methodology
Sampling was performed at the site during various stages of the investigation including the following:
• After focused removal of pesticide products
During the site characterization (using a direct push sampling method combined with
IA analyses) prior to excavation
• After gross soil removal to evaluate attainment of the cleanup standards (closure
confirmation sampling) and to guide further soil removal activities, and
• Sampling of waste soil and decontamination water prior to waste characterization for
waste classification and disposal.
The text to follow discusses the sampling design and methodology for each of these sampling events.
Focused Removal Sampling Design: Focused Removal Area 2/3 (FR 2/3) and Focused Removal Area
4/5 (FR 4/5) (see Figure 5) were excavated until all visible evidence of pesticide disposal was removed.
Upon completion of excavation activities, confirmatory samples were collected. The sampling grids for
this effort were established by the row divisions of the test plot across the excavated areas. This resulted
in six sampling areas or grids. A single random sample was then taken from within each sampling grid,
except for one grid in which the sample location was biased towards a location with a piece of white
particulate matter. The particulate matter may have come from one of the bags of concentrated pesticide
products buried at the site.
Site Characterization Sampling Design: Site characterization sampling was initiated following
completion of the focused removal activities. The site characterization included collection of soil
samples throughout the test plot area. The samples were collected for the purpose of characterizing the
site so that an excavation plan and preliminary waste disposal plan could be developed. Samples were
collected using direct-push sampling equipment.
The sample collection approach was described as "focused sampling." Focused sampling is defined as
the selective sampling of areas where potential or suspected soil contamination can reliably be expected
to be found if present. One sample was collected from within each grid. The number and size of each
grid were determined in advance using a statistical analysis of the site and an estimate of potential hot
spot size. For sampling within each grid, biased locations were selected in the field based on visual
observations of surface conditions. If there was not sufficient information to select a biased location,
then a random sample was obtained instead.
At each sample location, a soil core was taken from the ground surface down to 72 inches. Samples were
taken from each core to represent each one-foot interval within the bore hole. Each sample representing
each one-foot interval was then homogenized and split into three subsamples - one for field analysis, one
for possible fixed laboratory analysis, and one to be archived for possible future analysis.
Gross soil removal was aided by the use of a decision matrix to guide the analysis of samples, develop a
removal profile, and select samples for fixed lab analysis. This approach was part of the adaptation of
the sampling design under the dynamic work plan. Table 4 is an example of the decision matrix used at
the WTFREC site for shallow soils. For example, if the field kits found contamination in the interval 0
to 12" bgs at concentrations exceeding the action level established for the kit, then the next interval (12"
to 24" bgs) was analyzed by the field kits. If no contamination was found above the action level, then the
0 to 12" interval would be slated for removal, and a split of the 12" to 24" interval was sent for fixed
laboratory analysis. (The fixed laboratory data helped ensure the accuracy of the removal profile, as well
21 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S CHARACTERIZATION TECHNOLOGIES continued ^^^^^^^^^^^^^5
as add to the data set establishing the comparability of the field results to fixed laboratory analyses with
respect to the action level.) This type of decision rule was applied to depths no greater than 72" bgs.
Sampling was limited to depths of 72" because the USAGE believed that all pesticide contamination
would effectively be found within that interval. This was based on the assumption that no pesticide
product was disposed below 4 feet (48 inches) bgs and that migration of pesticides would be minimal
(less that one foot) beyond that depth.
Confirmation Sampling Design: At the conclusion of the gross removal excavation, closure
confirmation sampling was conducted of the bottom and side walls of all 27 grids using IA analyses.
Each grid to be sampled was laid out into nine equal sub-grids, a random selection of the sub-grid to be
sampled was made, and the sampling point was marked with a wooden stake. Shallow soil samples were
collected from within a 12-inch diameter area around the sampling point, placed directly into the
sampling jar, and analyzed using the field IA method. Concentrations found above the IA action levels
resulted in further excavation. The modified action level of 10 ppm for the DDT test was used to direct
this excavation. The comparability data set had established that DDT IA results below 10 ppm correlated
well with the mix of individual DDT, DDE, and ODD concentrations that did not exceed their respective
MTCA standards.
When IA analyses indicated that no further excavation was needed, closure confirmation sampling for
fixed laboratory analysis was performed. This sampling consisted often samples, one for each column,
plus a sample for the second elevation in column 4. To ensure conservatism, the grid with the highest IA
result in a given column was the grid sampled for the fixed laboratory analysis. The ten final closure
confirmation samples for fixed laboratory analysis were discrete surface samples taken from the same
location as the previous IA sample (refer to Figure 5 on page 10 where the triangle symbol represents this
lA/fixed laboratory sampling location). The final closure confirmation samples submitted to the fixed
laboratory were analyzed for the OP and OC pesticides, paraquat, and carbamate pesticides listed in
Table 1.
Waste Characterization Sampling Design: Upon removal of the material from the ground, it becomes
a waste governed by the Washington State Dangerous Waste Regulations (WAC 173-303) and not by the
MTCA action levels. The waste was segregated into roll-off bins. See "Segregation of Excavated
Materials for Disposal" in the "Media and Contaminants" section of this report (page 9) for more
information on waste segregation. Waste stream characterization sampling was conducted at the
conclusion of the focused removal excavation and again as significant segments of the initial gross
removal excavation were completed.
During the focused removal, samples were collected from each of the segregated waste streams. Each
sample was collected as a composite sample from at least five different locations within either a single
roll-off bin or a grouping of roll-off bins. The proportion of sample collected from within any roll-off
bin was representative of the proportion of waste soil within the bin as compared to the collective
grouping of bins.
Some of the roll-off bins were not specifically sampled, particularly towards the end of the gross removal
activities. Based upon the information known about the contents of these bins, the judgement was made
that the relative contaminant concentrations within these bins were either at or lower than other bins,
which were already known to be in the non-Resource Conservation and Recovery Act (RCRA) regulated
waste category. All waste characterization samples were analyzed by fixed laboratory methods.
22 August 2000
-------
Table 4. Example Removal Decision Matrix for Shallow Disposal
(Contamination above MTCA Method B/Field Kit Action Level at depth)
Scenario# Otol2" 12 to 24" 24 to 36" 36 to 48" 48 to 60" 60 to 72" Action
No n/a n/a n/a n/a n/a Confirmation Sampling
Yes No n/a n/a n/a n/a Find contamination in 0-12 sample, field sample 12-24"
Find no contamination in 12-24" sample above MTCA:
0-12" of soil. Confirmation Sampling. No Further Action.
Yes Yes No n/a n/a n/a Find contamination in 0-12" sample, field sample 12-24"
Find contamination in 12-24" sample, field sample 24-36"
Find no contamination in 24-36" sample above MTCA:
0-24" of soil. Confirmation Sampling. No Further Action.
Yes Yes Yes No n/a n/a Find contamination in 0-12" sample, field sample 12-24"
Find contamination in 12-24" sample, field sample 24-36"
Find contamination in 24-36"sample, field sample 36-48"
Find no contamination in 36-48" sample above MTCA:
0-36" of soil. Confirmation Sampling. No Further Action.
Yes Yes Yes Yes No n/a Find contamination in 0-12" sample, field sample 12-24"
Find contamination in 12-24" sample, field sample 24-36"
Find contamination in 24-36" sample, field sample 36-48"
Find contamination in 36-48" above MTCA, field sample 48-
Find no contamination in 48-60" sample above MTCA:
0-48" of soil. Confirmation Sampling. No Further Action.
Yes Yes Yes Yes Yes No Find contamination in 0-12" sample, field sample 12-24"
Find contamination in 12-24" sample, field sample 24-36"
Find contamination in 24-36" sample, field sample 36-48"
Find contamination in 36-48 above MTCA, field sample 48-
Find contamination in 48-60 sample, field sample 60-72" soil
Find no contamination in 60-72" sample above MTCA:
0-60" of soil. Confirmation Sampling. No Further Action.
n/a = not applicable, i.e, the depth interval above the one specified was found to have no contamination above the MTCA Method B action level.
23 August 2000
-------
Wenatchee Tree Fruit Test Plot
CHARACTERIZATION TECHNOLOGIES continued
Analytical Technologies and Method Modifications
The project team used a selective mix of on-site analyses and fixed laboratory analyses to evaluate the
contaminants of concern. For the focused removal, site characterization, soil gross removal and final
confirmation sampling phases of this project, immunoassay field analysis (IA) kits were used at the site
for organochlorine pesticides, and results were supplemented by limited data from fixed laboratory
analyses. Waste characterization samples were analyzed for OP and OC pesticides, TCLP OC pesticides,
and TCLP metals at a fixed laboratory. The text to follow discusses the performance of these analyses
and related QC issues. The anticipation of such issues and related corrective actions was part of the
project planning process. Analytical chemists were involved in developing plans for using both IA and
fixed laboratory analyses.
Immunoassay Field Analysis: For on-site soil sampling and analysis during the focused removal and
site characterization phases, two on-site immunochemical analyses, one for DDT and one for
cyclodienes, were performed by GSA. The performance criteria for the immunoassay tests are outlined
in Table 5.
Table 5. Immunoassay Test Kit Performance Criteria
Compound
DDT - Method 4042
Cyclodienes - Method 4041
Matrix
Type
Soil
Soil
Correlation with
Definitive Analysis
(RPD and r2)
•50
>0.90
• 50
>0.90
Accuracy
(LCS
Recovery, %)
60-140*
60-140*
Precision
(Duplicate
% RPD)
•50
•50
* Verification of analytical accuracy was based on a mixed pesticide standard and a computed value based on the
sensitivities for the reactivity groups given above. If the mean LCS recovery was not near 100%, further evaluation
was performed to assess the accuracy.
The immunoassay tests were performed in batches of approximately 12 samples, at a rate of
approximately one batch per test kit per day. Each batch consisted of a set of project samples and quality
control (QC) samples; such as, calibration samples, field duplicates, lab duplicates and laboratory control
samples. Some of the calibration samples were conducted in duplicate. The calibration data were fit into
a straight line with linear regression and the resulting calibration line was used to compute the project
sample concentrations.
During the course of the field analysis, project chemists investigated quality control problems and
implemented corrective actions prior to releasing data for use. Most of the laboratory control sample
(LCS) results fell within an accuracy window from 100 to 300 percent, with a mean near 200 percent.
This was consistent with the known 100 percent calibration bias designed into the kits by the
manufacturer. However, for the first five sample batches, the concentration of the laboratory control
sample (LCS) was above the calibration range of the tests and the LCS recovery was high. This problem
was overcome by diluting the LCS solution starting with Batch 6. After dilution of the LCS into the
range of calibration, the mean LCS recovery was closer to the expected 200 percent. Other cases of LCS
recovery exceeding the accuracy goals were determined to be caused by dilution errors. These cases
were evaluated on a case-by-case basis and did not result in data rejection. The data in these instances
were still deemed usable for the intended purpose.
24
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S CHARACTERIZATION TECHNOLOGIES continued ^^^^^^^^^^^^^5
In some batches, other LCS non-conformances were identified that indicated calibration deficiencies. In
these cases, the LCS did not meet the acceptance criteria for LCS recovery. Such calibration deficiencies
resulted in these batches being rejected and rerun.
Despite the sample homogenization process used, the homogeneity of a sample was questionable in a few
cases. However, the overall conclusion was that sample inhomogeneity had not significantly affected the
site decisions.
Fixed Laboratory Analysis: A documented industry-developed method (Chevron, 1978) and SW-846
methods were used for all definitive confirmation sampling and waste characterization. Soxhlet
extraction (Method 3540 or 3541) and appropriate cleanup methods, as required by the interferences
encountered, were used for all soil samples to be analyzed for organochlorine pesticides and
organophosphorus pesticides. All pesticides listed on the quantitation limit tables for the IA kits were
reported by the laboratory. Modifications and equivalency of methods are described below.
Method Modifications: Some aspects of the fixed laboratory methods were modified for the purpose of
achieving the analytical performance required to support project goals. These modifications to reference
methods were evaluated and documented through the QC procedures, in order to provide data quality
indicators (e.g., precision and bias) appropriate to the intended data use. A list of the method
modifications applied to the EPA reference methods along with justification for these modifications is
presented in Table 6.
For the analysis of OP pesticides by Method 8141, gas chromatography/mass spectroscopy (GC/MS)
instrumentation was used instead of the gas chromatograph with nitrogen phosphorus detector (NPD)
specified in the method. As a result, improved selectivity and low quantitation limits were achieved. For
the analysis of OC pesticides by Method 8081, a GC with an electron capture detector (BCD) was used to
allow the analysis of both the primary compounds of interest and multi-component pesticides (technical
cyclodiene, reported as dieldrin and endrin, and toxaphene). The carbamates were analyzed by Method
8141 instead of Method 8321. The use of the less sensitive but more selective GC/NPD instead of the
high performance liquid chromatography (HPLC) technique usually recommended for these compounds
was possible due to the moderate project detection limit requirements and restricted analyte list. As a
result, improved performance was achieved due to reduction of interferences.
The IA tests were also modified slightly to make a single soil extraction serve for both the cyclodiene and
DDT field test kits. The immunoassay was calibrated to report the cyclodienes as dieldrin and endrin.
The overall goal of the method modifications was to improve sensitivity and selectivity for specific
analytes. Method modifications for the purpose of improving performance is consistent with the
performance-based measurement system (PBMS) approach being implemented by EPA. EPA defines
PBMS as a set of processes wherein the data quality needs, mandates or limitations of a program or
project are specified and are used as criteria for selecting methods that meet those needs in a cost-
effective manner. Under the PBMS approach, the regulated community has the option to select an
appropriate method other than those found, for example, in SW-846 or make method modifications that
are capable of measuring the analytes of concern, in the matrices of concern, at the regulatory levels of
concern, and at the confidence level of concern. The goal is to make compliance with EPA's regulations
easier and more cost effective by allowing more flexibility in method selection and use. For more
information on PBMS, go to http://www.cpa.gov/SW-846/pbms.htm.
In addition to the specific methods referenced, various sections of SW-846 contain specifications that
apply to the methods for this project. General gas chromatography method requirements are outlined in
25 August 2000
-------
Wenatchee Tree Fruit Test Plot
CHARACTERIZATION TECHNOLOGIES continued
Method 8000. Chapters Three and Four of SW-846 describe specific sample handling requirements for
metals and organics, respectively.
Table 6. Modifications to Reference Methods
Parameter
Method
Modification/Justification
Cyclodiene IA test
4041
Extraction fluids were pure methanol rather than
water/methanol mix. This made the test compatible with the
DDT test, allowing for a single sample extraction for both
tests. The extraction volume was doubled to 20 mL to better
bracket the action levels for these tests based on the pilot study
cross-sensitivity results.
DDT IA test
4042
The extraction volume was doubled to 20 mL to better bracket
the action levels for these tests based on the pilot study cross-
sensitivity results.
OP pesticides
8141
GC/MS rather than GC/NPD was used. The surrogates and
calibration requirements appropriate for this method were
utilized from the source method (8141). The modification
improved selectivity and maintained low enough quantitation
limits to meet the project DQOs.
Carbamates by GC
8141, modified
GC/NPD was used as directed in EPA Method 632, modified
for a soil matrix according to the SW-846 methods. The
moderate project detection limit requirements and restricted
analyte list allowed the less sensitive but more selective
GC/NPD technique to be used instead of HPLC (EPA Method
8321). The benefits were primarily in improved performance
due to reduction of interference. The surrogate selected was
bolstar. This pesticide was chosen as a surrogate since the
compound is rarely used for agricultural applications in this
geographical area.
Paraquat
RM-8-10
This spectrophotometric method accommodates paraquat in a
soil matrix according to procedures developed by Chevron Oil
(Chevron, 1978).
Correlation of Immunoassay Tests with Fixed Laboratory Results: During the pilot study and prior
to the development of the RAMP, the USAGE tested the IA kits against fixed laboratory results with
surface soils from the site. For the compound distributions found in these soils, it was apparent from the
pilot study that a DDT kit result of 5 ppm or a cyclodiene kit result of 0.1 ppm might indicate that a
clean-up standard for an individual compound was exceeded. The IA tests are most accurate at the
midpoint concentration level; therefore, the sample preparation procedures were customized to the
decision-making needs of the project by setting the calibration midpoint concentration at 5 ppm and
0.086 ppm for DDT and cyclodienes, respectively.
The particular test kits used for this project were intentionally biased high by the manufacturer by 100
percent in order to reduce the occurrence of false negative decision error. Thus, when quantitatively
comparing the IA results against the fixed-laboratory data and QC samples, the IA results are expected to
be twice as high (i.e. a 200 percent recovery on QC samples). DDT and dieldrin were thought to
26
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S CHARACTERIZATION TECHNOLOGIES continued ^^^^^^^^^^^^^5
respectively contribute the most to the response for the DDT and cyclodiene immunoassay kits.
However, because the project samples all contained a mixture of compounds, the immunoassay results
were expected to correlate better with the sum of the compounds (after taking into account their
respective reactivities toward the immunoassay test) than with any single component.
As expected, a plot of the correlation between the field and fixed laboratory results during the focused
removal and characterization phase of the remediation was not quantitatively consistent. A number of IA
results were higher than predicted by the regression line, particularly for the cyclodiene test. In some
cases, cross-reacting pesticides or other compounds were present to cause additional response. Most of
the samples were either well above or well below the IA action limit, so at few locations was the
proposed excavation profile uncertain based on the IA results alone.
For the most part, the proposed excavation profile based on IA results alone was confirmed to be correct
when compared to the excavation profile based on the fixed laboratory results. The excavation decisions
that were based on IA results below the action level (i.e., results indicating a "no further action required"
decision for that sampling location) were entirely confirmed by the fixed laboratory results. Therefore,
the IA tests produced no false negative decision errors with respect to the action level. Due to the
presence of cross-reacting compounds (i.e., interferences), a few cases of false positive decision errors
with respect to the action level were encountered. In particular, endosulfan compounds present in the
analyzed soils were found to respond strongly in the cyclodiene test, yet these compounds have a
relatively high clean-up standard. When endosulfans were present, even a high IA result (e.g., 2 ppm
cyclodienes, reported as dieldrin and endrin) did not necessarily indicate that a clean-up standard was
exceeded.
During the characterization phase (Phase 3), ongoing comparison between the IA results and fixed lab
results revealed that IA results below 10 ppm correlated well with the mix of individual DDT, DDE and
ODD concentrations that did not exceed their respective MTCA standards. As a result, the action level
for DDT was further refined to 10 ppm (i.e., raised from the 5 ppm field action level used at the start of
the project). The modified DDT action level was used during the gross soil removal phase (Phase 4) to
determine the need for further excavation.
Quality Assurance/Quality Control (QA/QC) Measures
A number of different QA/QC measures were implemented during sample collection and field and fixed
laboratory analyses. Table 7 provides a summary of field QC samples prepared and analyzed. The table
also provides the total number of field samples associated with the analyses. In addition, laboratory
control samples and blanks were analyzed for each parameter at a frequency of 1 per batch (up to 20
samples) for all analyses, both field and fixed laboratory analyses. Matrix spike and matrix spike
duplicates were also analyzed at a frequency of 1 per batch (up to 20 samples) for all parameters, with
the exception of cyclodienes, DDT and TSS. For those analyses, matrix spikes were not used and matrix
duplicates were analyzed at a frequency of 1 duplicate per batch. In addition, four performance
evaluation (PE) samples were analyzed by the fixed laboratory during the various sampling and analysis
phases of the project. The various QA/QC measures are described below.
27 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S CHARACTERIZATION TECHNOLOGIES continued ^^^^^^^^^^^^^5
Field Quality Control Samples: Field quality control samples were collected during field work to
monitor the performance of sample collection and measure the effects of sampling bias or variability.
Field QC samples included the following:
Equipment (rinsate) blank: An equipment blank is a rinse sample of the decontaminated
sampling equipment to evaluate the effectiveness of equipment decontamination or to detect
cross contamination. Equipment blanks were prepared during the focused removal, site
characterization, and final confirmation study phases. Equipment blanks were not prepared for
analysis by IA.
Field duplicate: Field duplicates are taken to evaluate the reproducibility of field sampling
procedures. Field duplicates were prepared during all phases of the cleanup project including
focused removal, site characterization, final confirmation, waste profiling, and wastewater
characterization. Field duplicates were collected for IA field analysis and fixed laboratory
analysis.
Field Analysis (IA) QA/QC Measures: Quality control checks employed during field analysis included
the following:
Calibration samples: High-purity materials provided by the kit manufacturer were used as
calibration samples to determine kit range, detection or quantitation limits, precision, and
instrument drift. For the IA tests, a set of three calibration standards were used. Calibration
verification was performed with each batch of 12 samples.
Negative control: An unspiked blank was used along with calibration samples during kit
calibration.
Matrix duplicates: An intralaboratory split sample was used to document the precision of the
method in a given sample matrix.
Laboratory control samples: A laboratory control sample was prepared from a solid matrix
performance evaluation (PE) sample containing known concentrations of target analytes.
Fixed Laboratory QA/QC Measures: In addition to periodic five-point calibrations, the following
laboratory internal analytical quality control measures were employed by the fixed laboratory to ensure
the quality of the analytical data:
Continuing calibration verification (CCV) compounds: CCV compounds were used daily to
verify calibration.
Internal standards: Internal standards were used for GC/MS analysis to monitor the
consistency of response factors, relative retention times, injection efficiency, instrument drift,
etc., for many organic analysis.
Surrogates: Surrogates are compounds which are similar to the target analytes in chemical
composition and behavior in the analytical process, but are not normally found in real-world
samples. They are added to each sample, blank and matrix spike prior to extraction or
processing. They were used to monitor the performance of the extraction, cleanup (when used),
and analytical system.
28 August 2000
-------
Wenatchee Tree Fruit Test Plot
CHARACTERIZATION TECHNOLOGIES continued
Method blank: A method blank is used to assess contamination levels in the laboratory. It is
prepared from clean reference matrix and carried through the complete sample preparation and
analytical procedure.
Matrix spike: A matrix spike is an aliquot of the sample spiked with known concentration of
target analytes. It is used to document the bias of the method.
Matrix spike duplicate (MSD): MSDs were used to document the precision and bias of the
method; the MSDs are intralaboratory split samples spiked with identical concentrations of target
analytes.
Laboratory control sample: Laboratory control samples were used by the fixed laboratory in
conjunction with the matrix spike results to differentiate matrix-related problems from laboratory
performance issues.
Performance evaluation (PE) samples: PE samples can be used to provide information on the
baseline performance of a laboratory. A total of four PE samples were submitted as blind QC
samples to the fixed laboratory during the various sampling and analysis phases of the project.
Table 7. Summary of Field Duplicate and Equipment Blank QC Samples
Analytical Parameter
Technique
Sample Type
No, Field
Samples
No. Field
Duplicates
No. Equip.
Blanks
Focused Removal
OC and OP Pesticides
GC/MS and GC
Soil
6
1
I/ day
Characterization
Cyclodienes and DDT
OC and OP Pesticides
IA
GC/MS and GC
Soil
Soil
162
36
16
4
0
11 day
Final Confirmation
Cyclodienes and DDT
OC and OP Pesticides,
Carbamate pesticides
Paraquat
IA
GC/MS and GC
Spectrometric
Soil
Soil
Soil
27
9
9
3
1
1
0
11 day
11 day
Waste Profile
Prelim OC, OP
Final OC, OP
Carbamate Pesticides
Paraquat
OC Pesticides
Metals
GC and GC/MS
GC and GC/MS
GC
Spectrometric
GC
3010/6010
Soil
Soil
Soil
Soil
TCLP extract
TCLP extract
6
3
1
1
3
5
1
0
1
1
1
1
0
0
0
0
0
0
Equipment Decontamination Rinse Water
OC and OP Pesticides
Metals
Total Suspended Solids (TSS)
GC/MS and GC
ICP/MS and GFAA
Gravimetric
Water
Water
Water
1
1
1
1
1
1
0
0
0
29
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S PERFORMANCE EVALUATION ^^^^^^^^^^^^^^^^^^^^^^5
Performance Objectives
The goal of the project was to identify, characterize, remove, and dispose of all pesticide-contaminated
soil and debris from the test plot area of the WTFREC. Action levels for soil removal on the project
were determined to be the MTCA Method B Cleanup Levels (see Table 1).
The final determination of whether the remedial action attained the cleanup standards was based on a
statistical analysis of the sample data representative of the final conditions at the entire footprint of the
site at the maximum extent of excavation. The statistical requirements to demonstrate cleanup were:
1. The analyte concentration for no more than 10 percent of the samples can exceed the cleanup
standard for that analyte;
2. No sample concentration can exceed a level more than two times the cleanup standard for
any particular analyte; and
3. The upper confidence limit of the data for each analyte must be statistically shown to be less
than the cleanup criteria for that analyte.
Approximately 230 soil samples were analyzed by IA to support focused removal, site characterization,
closure confirmation, waste characterization, and QA (including field and laboratory duplicates)
activities. Approximately 100 soil samples were analyzed in a fixed laboratory to support focused
removal, site characterization, closure confirmation, waste characterization (including wastewater
analysis, TCLP organics and inorganics, PCBs, total metals and total pesticides in preparation for waste
disposal) and QA (including equipment blanks and performance evaluation samples) activities.
Strategy and Technologies Used to Attain the Performance Goals
The strategy and technologies used to attained the project goals included:
Systematic planning
Use of an adaptive (dynamic) sampling plan
On-site analysis and "immediate" availability of results using immunoassay analysis (IA)
technologies combined with limited fixed laboratory analyses, and
Rapid on-site decision-making guided by a decision matrix (a dynamic work plan) that used
field analytical results to characterize, excavate, and segregate pesticide-contaminated soil.
Performance of the dynamic work plan approach was highly superior to a traditional scenario, had that
occurred at this site. Because of the ability to sample and test the sides of the excavated areas, it was
discovered that pesticide contamination exceeding the regulatory standard existed outside of the original
boundaries of the site (as determined from historical information). Since this was discovered
immediately, it was simple and convenient to continue excavating until compliant soil was reached. This
resulted in the removal of an additional 60 tons of soil by extending the sides of the original boundaries
(see Figure 5).
Under a traditional scenario, however, this discovery would not have been made until fixed laboratory
results for samples collected for cleanup attainment confirmation were received. Likely those sample
30 August 2000
-------
Wenatchee Tree Fruit Test Plot
PERFORMANCE EVALUATION continued
analysis results would not have been available until after the excavation team had left the site. Closure
would have been delayed and additional expenses would have been incurred to prepare a second work
plan and sampling and analysis plan, remobilize to the site to characterize the boundaries of the
remaining contamination, wait for the results to come back from the fixed lab, and then return to the site
to excavate yet again and perform additional closure testing. The use of on-site analyses and a dynamic
work plan avoided that unpleasant and inefficient chain-of-events.
The USACE's contractors completed the project work in conjunction with the USAGE, and the project
was successful. The Test Plot no longer contains soils exceeding the site action levels. The cleanup was
accomplished in a shorter time frame and at a lower cost than the traditional site characterization and
remediation approach in which multiple rounds of field mobilization, sampling, sample shipment,
laboratory analysis, and data assessment are required.
The time frame for various activities at the Wenatchee Tree Fruit Test Plot is presented in Table 8.
Once mobilization to the site occurred, all phases of site work were completed within 4 months.
Table 8. Time Frame for Activities
Date
1985-1987
1990, 1991, 1994
April 1996
June 1996
August 1997
Sept. 15-22, 1997
Sept. 22-24, 1997
Oct. 13, 23& 24, 1997
Oct23;Nov. 3, 4, 17 and
Dec. 10, 1997
Dec. 12, 1997
January 1998
Activity
WSU performs sampling and analysis at WTFREC
EPA performs 3 sampling and analysis events at WTFPvEC
USAGE begins project planning process to accommodate EPA ORD
request to remediate the WTFPvEC site
Pilot study performed with site-specific soils to assess IA and Geoprobe
performance
USAGE contracts with GSA to perform site work
Mobilization of construction support items to site
Focused Removal activities started/completed (45.74 tons excavated)
Gross Removal activities started/completed (271 tons excavated); initial
closure confirmation samples obtained and additional contamination
discovered
Additional excavation of sidewalls and floor performed; final closure
confirmation sampling completed (60 tons excavated)
Closure confirmation activities completed
463 tons of material used to backfill; site restoration completed
31
August 2000
-------
Wenatchee Tree Fruit Test Plot
COST COMPARISON ^^^^^^^^^^^^^^^^^^^^^^^^^^S
The approach to site cleanup employed in the WTFREC Test Plot resulted in considerable savings
compared to traditional site characterization and remediation approaches. The use of systematic
planning, a dynamic workplan, and on-site measurement technologies combined with limited fixed
laboratory analyses allowed for the cost-effective cleanup of the contaminated site with savings of
roughly 50% over traditional methods. Although it is extremely difficult to project a likely cost scenario
if a project were to be performed using a different work strategy, extrapolations are sometimes possible if
enough cost detail is available from the actual project. The USAGE made detailed unit and activity costs
available for preparing this case study. A cost comparison is projected based on the following
information and assumptions:
Assume that a more traditional approach would also use direct push sampling to produce a similar site
characterization profile in order to roughly delineate the boundaries of contaminated soil requiring
removal. Then a similar number of samples sent for traditional fixed laboratory analysis might be
assumed. Based on knowledge obtained during the actual cleanup, remediation of this area without the
use of a dynamic work plan could have possibly produced at least 391 tons of contaminated soil (see
Notes 4 and 7 of Table 9) requiring incineration, since segregation of less contaminated materials from
more contaminated materials during excavation would have been difficult without the immediate
feedback of real-time results. The excavation, transportation, and disposal cost alone for this volume of
contaminated soil would have exceeded $560,000 (see Table 9). The use of fixed laboratory methods
and/or more rapid turn-around times for fixed lab results would have resulted in a substantial increase in
analytical costs.
Furthermore, the dynamic work plan allowed the site team to discover immediately that unexpected
contamination existed outside of the original project boundaries and then to seamlessly extend sampling
and excavation until clean soil was reached. Under a traditional scenario, this discovery would likely not
have occurred until after the fixed lab results for anticipated closure confirmation had been returned,
examined, and reported to project decision-makers. In all likelihood, the discovery that the initial
removal did not attain regulatory cleanup standards would have incurred additional costs to prepare new
planning documents, remobilize to the size, and conduct yet another round of characterization sampling
and analysis, excavation, and closure confirmation sampling. In all, the estimated cost of cleanup
without the use of a dynamic work plan and field analytical methods may be projected as totaling nearly
$1.2 million. A simple analysis of cost repercussions also does not factor in the frustration of regulators,
clients, and stakeholders when "surprises" delay site closeout.
In contrast, the actual total cost for site characterization, remediation and closeout at WTFREC was
approximately $589,000. Of this total, $100,000 were expended by the USAGE for planning, design,
contracting and project management. (The cost for project oversight was assumed to be the same under a
traditional scenario.) A moderately detailed breakdown of actual and projected costs and assumptions is
shown in Table 9.
In addition, the USAGE had prepared a different cost comparison estimate for remediating the site that
assumed excavating and incinerating the entire 70-foot long by 30-foot wide by 7-foot deep original plot
(estimated as 708 tons of soil) without performing any site characterization. The estimate for this was
$1,122,049. Although this estimate included closure testing, it did not include the cost of remobilization
to extend the excavation after sidewall contamination was discovered. It is notable that the cost of
traditional site characterization could have been approximately equivalent to the cost of the most
conservative treatment option for this site.
32 August 2000
-------
Wenatchee Tree Fruit Test Plot
COST COMPARISON continued
Table 9. Cost Comparison
Item
Design
Procurement
Oversight/Contract Management
Technical Review
General, Mobilization, Construction,
Data Analysis, Demobilization
Contaminated Material Excavation
Soil Analysis
Backfilling, Grading, and
Revegetation of Test Plot
Waste Transport and Disposal
Environmental Planning and
Reporting
Additional Characterization
(including revised planning
documents and remobilization)
Additional Sample Analysis
Additional Soil Excavation
Additional Backfilling of Test Plot
Additional Waste Transport and
Disposal
Data Validation
TOTAL PROJECT COST
Estimated Cost Without Use of Dynamic
Work Plan and Field Analysis (i. e., a
"traditional" approach)
$36,000
$9,000
$45,000
$10,000
$128,846
(See Note 1)
$35,959
(see Note 2)
$235,942
(see Note 4)
$11,486
$353,358
(see Note 5)
$15,304
$29,563
$101,356
(see Note 6)
$9,773
(see Note 7)
$3,046
$168,193
$4,053
$1,196,880
Actual Cost Using Dynamic
Work Plan and Field
Analysis
$36,000
$9,000
$45,000
$10,000
$129,446
$46,052
(see Note 3)
$79,412
$11,486
$112,622
$15,304
Not applicable.
$28,364
$10,615
$2,031
$49,627
$4,053
$589,012
Notes:
1. Mobilization would not require rental of a trailer for the field laboratory, therefore, mobilization costs are
slightly less than that required for the dynamic work plan with field laboratory.
2. Cost estimates assumes 271 tons of soil excavated with no on-site temporary storage.
3. Cost includes on-site temporary storage.
4. Cost assumes 230 field and QC samples analyzed by fixed lab for OC pesticides, OP pesticides, carbamates,
and paraquat to delineate the 271 tons of soil to be removed.
5. Cost assumes that all excavated soil would be managed as dangerous waste (i.e., incinerated).
6. Cost assumes 80 samples analyzed at fixed lab for OC pesticides, OP pesticides, carbamates, and paraquat.
7. Cost estimates for additional soil excavation, backfilling, transport, and disposal assume that 120 additional
tons of soil would be removed to avoid another remobilization. Note that the actual quantity of additional soil
removed was approximately 60 tons.
33
August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S OBSERVATIONS AND LESSONS LEARNED ^^^^^^^^^^^^^^^^^5
The involvement of the regulator and stakeholders during project planning allowed the team to develop a
decision-making strategy that all parties would follow during the removal action. This reduced the
amount of risk and cost associated with clean closure disagreements that can cause schedule delays,
especially during contractor mobilization on site. However, it relied on a planning team with the
appropriate mix of both skills and regulatory authorities.
The conceptual model of the site was based on a thorough review of historical records of site activities.
However, the project team still encountered contaminants in areas that were not originally anticipated.
Without the ability to generate analytical data on site and in near real time, the costs to remediate the test
plot and the time required would have increased greatly.
Substantial cost-savings were realized through the use of IA and an adaptive sampling plan. Cost savings
were realized through reduced analytical costs (compared to traditional fixed based laboratory analysis)
and reduced mobilization/demobilization costs that would be incurred if multiple mobilizations were
required.
The on-site analysis was designed to support in-field decisions regarding further characterization,
removal, waste segregation, and waste disposal. By conducting the pilot study and using additional
fixed-laboratory results to correlate with the immunoassay results, the action levels for the field analyses
were continually updated and adapted to changing site conditions. This approach reserved resources
(both time and dollars) that could then be applied to the relatively expensive fixed-laboratory analyses, or
used to increase the number of samples that were collected and analyzed by immunoassay.
The ability to increase the number and density of samples that were collected also helped to minimize the
amount of soil that was removed, as well as reducing the amount of soil sent for incineration, the most
expensive possible disposal option.
The length of the project from mobilization to site restoration of the site was relatively quick compared
to traditional methods.
The adaptive sampling strategy allowed several different sampling strategies to be employed throughout
the cleanup, based on the intended use of the data and the need to optimize the overall design. For
example, during the focused removal phase, random sampling was conducted within grid blocks, except
where there was a need to bias a sample location towards an observed stain in the soil. During site
characterization, soil cores were purposefully located near visual indicators of contamination within grid
blocks. In the absence of visual indicators of contamination, sample locations were randomly selected.
Finally, samples collected for confirmation of cleanup were discrete samples randomly located within
grid blocks. The assumptions of random samples is required for application of the statistical tests to
determine attainment of the cleanup standards.
The combined benefits of this optimized approach facilitated the "surgical" removal of contaminated
materials and ensured that closure confirmation testing would demonstrate compliance to a high degree
of certainty. Significant time and cost savings over the life of the project were possible by making field
activities such as sample collection, sample analysis, soil removal, soil segregation, and final disposal of
soil and wastewater as efficient and effective as possible.
34 August 2000
-------
^^^ Wenatchee Tree Fruit Test Plot
^S REFERENCES ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^S
1. GSA, Inc. (Gary Struthers Associates, Inc.). 1998. Final Report, Wenatchee TFREC Test Plot
Remediation, Wenatchee, Washington. DACA67-95-G-0001-38. Prepared for USAGE Seattle
District, Seattle, WA. August.
2. GSA, Inc. 1997a. Sampling and Analysis Plan Field Sampling Plan, Part B-Wenatchee TFREC
Test Plot Remediation. August.
3. GSA, Inc. 1997b. Quality Assurance Project Plan Sampling and Analysis Plan, Part C-
Wenatchee TFREC Test Plot Remediation. August.
4. USAGE (U.S. Army Corp of Engineers). 1997. Wenatchee Tree Fruit Research Center (TFRC)
Test Plot Removal Action. Basic Ordering Agreement 1013-95-510711/DACA67-95-G-0001.
August.
5. USAGE. 2000. Cost and Performance Report - Expedited Characterization and Soil
Remediation at the Test Plot Area, Wenatchee Tree Fruit Research Center, Wenatchee,
Washington. May. httpi//wvTOLfirtLgQy/CQS^dJ(Wenatchee^df
6. USEPA. 1993b. Data Quality Objectives Process for Superfund. Interim Final Guidance.
EPA/540.G-93/071. Office of Solid Waste and Emergency Response. Washington, DC.
7. USEPA. 1994. Guidance for the Data Quality Objectives Process, EPA QA/G-4.
http://es.epa.gov/ncerqa/qa/qa_docs.html#G-4. EPA/600/R-96/055. Office of Research and
Development. Washington, DC.
8. USEPA 1999. Guidance for the Data Quality Objectives Process, EPA QA/G-4, Peer Review
Draft, http://www.epa.gov/qualit>-1 /qs-docs/g4-prd.pdf. Office of Environmental Information.
Washington, DC.
9. Washington State Department of Ecology (Ecology). 1992. Statistical Guidance for Ecology Site
Managers. August 1992.
10. Washington State Department of Ecology (Ecology). 1995. Guidance on Sampling and Data
Analysis Methods. Publication No. 94-49. January 1995.
11. Chevron Chemical Company. April 1978. Method RM-8-10: Analysis of Paraquat Residues.
Ortho Research & Development Department. Richmond, CA. (Paraquat analyses performed by
North Coast Laboratories, Arcata, CA.)
12. USEPA. 2000. Data Quality Objectives Process for Hazardous Waste Site Investigations, EPA
QA/G-4HW, Final, http://www.epa.gov/qualitv 1 /qs-docs/g4hw-final.pdf. EPA/600/R-00/007.
Office of Environmental Information. Washington, DC.
35 August 2000
------- |