\J J^lf^
United States
Environmental Protection
Agency
Dynamic Field Activity Case Study:
Soil and Groundwater Characterization,
Marine Corps Air Station Tustin
-^/\// l "^A /""x1
/\// \ /7?<4^>V • Release
/s. / / ''ft sSJsL ^ //Areal
L^y /¥/Cw'^/^K /1
-------
-------
Office of Solid Waste and
Emergency Response (5201G)
EPA/540/R-02/005
OSWER No. 9200.1-43
November 2002
Dynamic Field Activity Case Study:
Soil and Groundwater Characterization,
Marine Corps Air Station Tustin
-------
-------
Notice
This document has been funded by the United States Environmental Protection Agency (EPA)
under Contract 68-W-02-033. The document was subjected to the Agency's administrative and
expert review and was approved for publication as an EPA document. Mention of trade names
or commercial products does not constitute endorsement or recommendation for use.
-------
-------
Acknowledgments
The Office of Emergency and Remedial Response would like to acknowledge and thank the
individuals who reviewed and provided comments on draft documents. The reviewers include
EPA headquarters and regional offices, state environmental programs, United States Department
of Defense, United States Department of Energy, and representatives from the private sector.
-------
-------
Contents
Notice iii
Acknowledgments v
Exhibits viii
Abbreviations ix
Abstract 1
Background 1
Using a Systematic Planning Process 3
Reviewing Existing Site Information 4
Selecting Key Personnel 4
Identifying the Project Objectives 6
Developing a Conceptual Site Model 6
Preparing Sampling and Measurement Strategies 10
Selecting Appropriate Analytical Methods, Equipment, and Contractors 12
Writing a Dynamic Work Plan 14
Conducting the Dynamic Field Activity 14
Selection of Sampling Locations 15
Refinement of Conceptual Site Model 15
Completion of Field Program 17
Writing a Final Report 20
Nature and Extent of Contamination 20
Contaminant Fate and Transport 21
Human Health Risk Assessment 21
Data Quality Assessment 22
Estimated Cost and Time Savings 23
Lessons Learned 24
Off-Site Analytical Program Oversight 25
Unexpected Chemicals of Concern 25
Flexible Work Planning 26
Performance Evaluation Samples 26
Geophysics 26
References 27
VII
-------
Exhibits
Number Title Page
1 Marine Corps Air Station Tustin Map Showing IRP Sites 2
2 Distribution of Contaminants Above Screening Levels Discovered During
Expanded Site Inspection 5
3 Initial Conceptual Site Model Pictorial for IRP-12 7
4 Initial Conceptual Site Model in Plan View 8
5 Initial Exposure Conceptual Site Model 9
6 Initial Sampling Locations at IRP-12 11
7 Interim Conceptual Site Model of IRP-12 TCE Contamination 16
8 Conceptual Model of Shallow Ground Water Plumes at IRP-12 18
9 Conceptual Model of Larger Plume in Cross Section 19
10 Dynamic Field Activity vs. Staged Investigation
Summary of Cost and Time Comparisons 24
VIM
-------
Abbreviations
ARAR applicable and relevant or appropriate requirement
bgs below ground surface
BRAC Base Realignment & Closure
CERCLA Comprehensive Environmental Response, Compensation, and Liability Act
DNAPL dense non-aqueous phase liquids
DP direct push
DQO data quality objective
EE/CA engineering evaluation/cost analysis
EPA U.S. Environmental Protection Agency
ESI expanded site inspection
FAM field-based analytical method
FID flame ionization detector
ft foot
GC gas chromatograph
gpd gallons per day
gpm gallons per minute
IRP Installation Restoration Program
MCL maximum contaminant level
MCLG maximum contaminant level goal
Mg/L micrograms per liter
mg/kg milligrams per kilogram
mg/L milligrams per liter
NPL National Priorities List
PAH polyaromatic hydrocarbon
PARCC precision, accuracy, representativeness, completeness,
and comparability
PCB polychlorinated biphenyl
PCOC potential chemicals of concern
PID photoionization detector
QAPP quality assurance project plan
RI/FS remedial investigation/feasibility study
SWMU solid waste management unit
TCE trichloroethene
TDS total dissolved solids
TRPH total recoverable petroleum hydrocarbons
UST underground storage tank
VOC volatile organic compound
IX
-------
-------
Dynamic Field Activity Case Study:
Soil and Groundwater Characterization,
Marine Corps Air Station Tustin
Abstract
The Navy used a dynamic field activity (i.e., a project that combines on-site data
generation with on-site decision making) for a CERCLA remedial investigation (RI) at the
Marine Corps Air Station (MCAS) Tustin between 1995 and 1996. Field-based analytical
methods (FAMs) provided defensible data that met project objectives for delineating the nature
and extent of contamination in the base-wide soil, surface water, and groundwater investigations.
The FAMs were also used to choose monitoring well locations and select a subset of risk
assessment samples for off-site analysis. At one location on the base, the dynamic field activity
saved the Navy over 15 percent of the total site cost of the investigation and helped to compress
the investigation schedule by an estimated 60 percent.
Background
The Navy planned, implemented, and completed a dynamic field activity at MCAS
Tustin, in southern California, between July 1995 and June 1996. The 1,600-acre base was part
of the Department of Defense's Base Realignment and Closure (BRAC) program, which
designated the land for redevelopment and integration into the surrounding community of Tustin,
located just north of Irvine in Orange County. The dynamic field activity at the site demonstrated
cost savings of $90,000, a temporal savings of several years, and regulator satisfaction with the
on-site decision-making process. A map of the entire site is presented in Exhibit 1.
The Navy conducted a preliminary assessment and an expanded site inspection (ESI)
between 1991 and 1993 at all areas with documented releases. Although the site had not yet been
included on the National Priorities List (NPL) at the beginning of the CERCLA remedial
investigation/feasibility study (RI/FS), regulators had already collected all of the required data. If
EPA had determined that Superfund remedial funds were necessary, the site could have been
listed quickly on the NPL. The regulatory authorities involved in the dynamic field activity were
the California Department of Toxic Substances Control, the California Regional Water Board,
and EPA Region 9. Representatives of these groups, along with the Navy and their contractor,
made up the BRAC Cleanup Team. Cooperation among these stakeholders was vital for the
success of the investigation.
Based on background information, the BRAC Cleanup Team documented 15 separate
areas with hazardous substance releases. They designated these areas for further investigation
and named them Installation Restoration Program (IRP) sites. Because they believed seven of
these IRP sites had substantial releases, they placed them in the remedial program. The BRAC
-------
yoday uoi)e6i)S9AU| {eipauiau ujjsnj. SVDW :tuojj pajdepv
V * - " 4k r *• Si - i^ •^/^"
X^ _^fc' ^ ^ ^ t* ^1?
!S didl
de|/\| uj
J!V
auuB|/\|
-------
Cleanup Team planned Engineering Evaluation/Cost Analyses (EE/CAs) at the remaining eight
IRPs, with the stipulation that if the contamination was worse than anticipated, they would
transfer the sites to the remedial program. In addition to these 15 IRP sites, the Navy was
responsible for investigating:
• Approximately 70 solid waste management units (SWMUs) that had operated under
RCRA authority;
• Several underground storage tanks (USTs), including heating oil tanks and on-base
gasoline facilities;
• Several underground jet-fuel lines;
• Agricultural fields to determine if there was any lasting impact due to pesticide
application; and
• Base residential areas in order to confirm and formally declare that no further response
was required.
During preliminary discussions among BRAC Cleanup Team members, the Navy
contractor suggested that the investigation could be carried out faster and cheaper if a dynamic
approach with FAMs was used. Once this approach was accepted by the BRAC Cleanup Team,
the investigators developed a plan to complete the field work in a single mobilization. This plan
would let them integrate the work at all of the contaminated sites so that they could benefit from
the economies of scale by moving between locations as equipment and personnel were needed.
For the purposes of providing a succinct case study of how dynamic field activities can be
conducted for a CERCLA RI/FS, this discussion focuses exclusively on the investigation of only
one site: IRP-12, Drum Storage Area No. 2. The field team required a total of only six weeks at
IRP-12; however, they worked on the site intermittently over the ten months of the entire base
investigation.
Using a Systematic Planning Process
Investigators used the EPA's seven-step data quality objective (DQO) process (U.S. EPA,
1994a) to guide the planning, and they based their decisions on information available from earlier
studies. The systematic planning process included the following information:
• Reviewing existing site information;
• Selecting key personnel;
• Identifying the proj ect obj ectives;
• Developing a conceptual site model;
• Preparing sampling and measurement strategies; and
• Selecting appropriate analytical methods, equipment, and contractors.
-------
By following the steps laid out in EPA's guidance, they developed DQOs that were reviewed and
approved by EPA Region 9. These DQOs can be reviewed on EPA's web site at:
http://www.epa.gov/superfund/programs/dfa/casestudies.
Reviewing Existing Site Information
Investigators reviewed existing data and found that the 2.5 acre IRP-12 site had been used
as a drum storage area from the mid-1960s until July 1975. Records indicated that these drums
contained new and spent solvents, used motor oil, and hydraulic fluids. There were three
reported releases in three separate drum storage areas, ranging in quantity from 600 to 1,000
gallons. In addition, previous investigations indicated that trichloroethene (TCE) had
contaminated groundwater and polyaromatic hydrocarbons (PAHs) had been released in a
drainage ditch that ran through the site. Exhibit 2 provides the location and concentration of
contamination discovered during the ESI. This field work also found that the upper 15 to 20 feet
of soil was a silty clay. The next 5 to 10 feet was a silty sand.
Selecting Key Personnel
As is necessary for the successful execution of dynamic field activities, the contractor's
planning team members were very qualified and experienced in their areas of responsibility.
Team members and their backgrounds included:
• Project manager: A civil engineer with over 20 years of engineering and management
experience. His primary functions were client relations, program office interactions,
management oversight of the activities, and ensuring that adequate resources were
available to the investigation teams.
• Technical team leader: The technical team leader was cross-trained in hydrogeology,
chemistry, and chemical fate and transport. He had more than 15 years of experience in
site investigations and generally was in the field full-time. He had authority over the field
team members' activities, and he was responsible for providing recommendations to the
Navy and regulators.
• Project chemist: Two project chemists took part in the initial planning of the
investigations and were available throughout the project for consultations. Each had
more than 20 years of experience in analytical chemistry and QA/QC.
• Project hydrogeologist: The project hydrogeologist was involved in the planning and
implementation of the investigation. Although he worked on this project full-time, he
was based at the home office. Geologic information was faxed to him on a regular basis
so that he could analyze the data and provide advice to the technical team leader. He also
was available for meetings on an "as needed" basis and was on site at times when the
-------
Exhibit 2
Distribution of Contaminants Above Screening Levels
Discovered During Expanded Site Inspection
fin o-3» « H FT) /
"' ^ ""• /
\
-------
technical team leader determined his presence was required. He had PhDs in
geotechnical engineering and geology as well as more than 10 years of experience.
• Risk assessor: A risk assessor was involved in the initial planning and was responsible for
producing the baseline risk assessment document. He held a PhD in toxicology, had
more than 20 years experience in risk assessment, and was available to the project on an
"as needed" basis.
• Data management: A data manager worked full time on the project. She had more than
10 years experience in environmental work and held a PhD in geostatistics.
• Community involvement: A community involvement specialist with more than 10 years
experience was involved in the planning stages and was responsible for all aspects of
community involvement. She worked part-time for the project and was available on an
"as needed" basis.
At full strength, the base field team for the entire site also consisted of seven geologists of
varying experience (1-7 years), associated field technicians also of varying experience (0- 3
years), two chemists for the mobile laboratory (each with more than 10 years experience), a site
supervisor, and a part-time health and safety officer.
Identifying the Project Objectives
The DQO process produced five principal study questions for IRP-12 that the
investigation sought to answer:
• Do the analytical results from the ESI indicate widespread PAH contamination of the
drainage ditch, and is there any vertical migration?
• Are there any direct release areas containing high concentrations of contaminants (i.e., hot
spots) in the soils of the three storage areas?
• Has there been any vertical migration of contaminants from the three storage areas?
• What is the lateral and vertical extent of the TCE contamination in the groundwater, and
what is its source?
• Have any of the four SWMUs found on IRP-12 released any hazardous materials to the
soil?
Developing a Conceptual Site Model
Based on the existing information, the project planners developed a preliminary
conceptual site model with the following scenario: waste oils (containing PAHs and metals) and
solvents (primarily TCE laden with metals from paint stripping) were released from the drum
-------
storage area 20 to 30 years earlier, either through decay of the drums or by spillage during filling
and handling. Because the soils on the site are alkaline with low permeability, the contaminants
migrated slowly into the soil to form a secondary release source. The groundwater had been
impacted in at least one area of the site, but the source was unknown. In addition, depending on
the quantity of releases, groundwater in other areas of the site may have been impacted.
Additional aspects of the initial conceptual site model included the following points:
• The water table at the site would be 7 feet below ground surface (bgs) (based on
information from a nearby UST site and another nearby investigation);
• Movement of contaminants in the groundwater would be very slow in the silty clay and
somewhat more rapid in the silty sand;
• The 1,000 |ig/L TCE concentration in a monitoring well near Building 90, with no known
source, required an up gradient investigation; and
• The site was sufficiently close to the fuel farm to assume that groundwater at IRP-12 also
moved in a south-southwesterly direction.
Exhibits 3 and 4 provide the initial conceptual site model of IRP-12, and Exhibit 5 provides the
initial exposure model.
Exhibit 3
Initial Conceptual Site Model Pictorial for IRP-12
Soil Surface
\
Silty Clay
Sand/Clay
Silt) Clay
1M5!eelbgs(l991! —
y Water
TCE/Melals
15-20 feet bgs
Silly Sand
Silly Sand
Adapted From: MCASTustfn R@m@d[aMnvestigatfon Report
Legend
tfp - below gm LI rtd surlace
TCE - t
-------
00
|eip9Uiau ujjsni SVOIAI :UIOJJ pajdepv
-------
Exhibit 5
Initial Exposure Conceptual Site Model
PRIMARY
SOURCE
DRUM
STQPACE
(IRP- US
PRIMARY SECONDARY SECONDARY PATHWAY
RELEASE SOURCE RELEASE
M E C H A MIS M M E C H A NISM
»•{ SPILLS
VDtATiLtZATfOM/
WIND EROSION
DIRECT CONTACT
mu-WATER I JsifRfACE WAIER/1
RullOrf I *1 SEDIMENTS |
RECEPTOR
EXPOSURE
ROUTE
HUMAN
«««
**£SD'Ems
WO**KlfiS/
VISITORS
BIOTA
FISflfSfftsAU * QUA IK
iNCCSfiOll (jfe
~4—~— -
COHIACI 1 vfi)
LL2LHD
• CURRENT POTENTIAL RECEPTOR
O FUTURE POTENTIAL RECEPIOR
Adapted From: MCAS Tustin Remedial Investigation Report
-------
Preparing Sampling and Measurement Strategies
The BRAC Cleanup Team designed a shallow soil sampling strategy for IRP-12 to detect
hot spots based on statistical probabilities. First, they randomly placed a grid over the entire site
based on 60-foot centers. Second, in the two areas where contamination was identified in the
ESI (north of Building 90 and south of Building 20B), the grid was subdivided into 20-foot
centers. Finally, in the area where contamination was suspected but not found during the ESI,
project planners used a grid based on 30-foot centers. In addition, they decided to take two
judgmental samples in the drainage ditch to better define the level of contamination detected in
the ESI. Exhibit 6 presents a site map with the proposed sampling locations.
This sampling design satisfied two limits of uncertainty for risk assessment:
• The probability of declaring that the site is not posing any risk to human health, when in
fact it is, will be 5 percent or lower; and
• The probability of characterizing the site as posing a threat to human health, when it does
not, will be 20 percent or lower.
If on-site analysis indicated the initial sample taken between 1 and 2 feet bgs was
contaminated, investigators would hand-auger the location to 5 feet bgs because of a base
requirement to avoid damaging utilities. They then would use direct push (DP) equipment to
take continuous cores to the first permeable zone (15 to 25 feet bgs). Unless the flame ionization
detector (FID) or visual inspection by the geologist suggested a better sampling interval, samples
would be taken at seven feet (estimated top of the water table), 12 feet bgs, and in the first
permeable zone. Whenever soil samples from the first permeable zone demonstrated high
volatile organic compound (VOC) levels, investigators would take groundwater samples with a
driven sampling point (i.e., equipment for taking a one-time groundwater sample that is pushed
to the sampling point, such as the HydroPunch™) for screening analysis.
If any groundwater samples were found to be contaminated, investigators would search
for the source area and horizontal extent by sampling groundwater with DP equipment in
upgradient and downgradient directions on 20-foot centers. After finding the source area, they
would delineate the lateral extent of the plume by step-out sampling on an axis perpendicular to
the groundwater flow direction.
Since one of the potential chemicals of concern (PCOCs) was capable of forming dense
non-aqueous phase liquids (DNAPLs), the BRAC Cleanup Team recognized the need for
sampling multiple permeable zones. Therefore, they sought sampling equipment that would
provide the hydrogeologist with the flexibility of reaching the required sampling zones without
posing a risk of dragging contamination into previously uncontaminated areas. At the same time
the equipment would have to be capable of obtaining a quality groundwater sample in each
permeable zone it encountered.
10
-------
Exhibit 6
Initial Sampling Locations at IRP-12
ARMORY PARKING
502
Adapted From: MC AS Tustin Remedial Investigation Work Plan
-------
Only after the extent of the plume had been fully defined would the Navy propose
monitoring well locations and screening intervals to the regulatory agencies for their approval.
This procedure would help ensure that each monitoring well provides useful information for
long-term study of the contaminant plume(s). If the data indicated that a pump-and-treat system
was a viable remedial option, then a hydrogeologist would complete an aquifer pumping test in
the same mobilization as part of the feasibility study.
Selecting Appropriate Analytical Methods, Equipment, and Contractors
The first step the BRAC Cleanup Team used in selecting appropriate analytical methods
for IRP-12 was identifying the PCOCs, from the historical information, and their associated risk
concentrations, from chemical ARARs and "To Be Considered" guidance documents. Using
these documents, the BRAC Cleanup Team determined that their FAMs needed to provide a
detection limit of a least 5 |ig/L for TCE, based on the federal MCL, and a total recoverable
petroleum hydrocarbon (TRPH) detection limit of 10 mg/kg, based on the preliminary
remediation goals for PAHs found in waste oils.
The review of historic site data revealed several items that helped streamline sample
analysis. First, the site had obvious VOC contamination (i.e., fuels and halogenated solvents).
Second, the risk drivers for cleanup in many cases were PAHs related to engine waste oils that
had been disposed improperly. Third, metal concentrations that appeared to be above naturally
occurring levels were always related to either VOCs (paint strippers) or waste oils.
Consequently, any FAMs that could be used for VOCs and waste oils (PAHs) could also ensure
that metal contamination would not be missed.
Based on this information the project team selected the following FAMs:
• A hand-held FID for VOC screening of soil samples;
• A gas chromatograph/photoionization detector (GC/PID) method (U.S. EPA, 1994b) with
detection limits of 5 |ig/L TCE in water and 25 to 50 |ig/kg TCE in soil (as well as other
VOCs); and
• A project-modified EPA Method 418.1 (infrared spectroscopy) with a method detection
limit of 10 ppm for analysis of TRPH, which would be used as a surrogate for the
presence of PAHs and metals in oils.
A throughput of 60 to 70 samples per day was expected with the GC/PID. EPA provided
performance evaluation (PE) standards for both the on-site and off-site laboratories. All other
VOCs that were detected by the instrument would require additional analysis for definitive
identification.
The quality assurance project plan (QAPP) proposed confirmatory analyses on 10 to 15
percent of the field analytical data at an off-site laboratory. Two-thirds of these samples would
12
-------
be the most contaminated according to the on-site analysis, and one third would be non-detects.
The type of analyses performed by the off-site laboratory also would vary depending upon what
caused the sample to be sent. If a VOC was identified on-site, then VOCs and metals would be
tested. If TRPH was identified on-site, then PAHs, PCBs, and metals would be requested. The
one-third of confirmatory samples that were non-detects would be tested for all constituents. To
accommodate the on-site instrumentation, a laboratory trailer with hood, sinks, and counter space
was contracted for $2,900 per month.
To collect soil samples, a dual-tube DP rig was chosen (i.e., an outer casing is driven at
the same time as an inner sampling tube). This procedure prevents the walls of a hole from
sloughing as a sample is removed. The equipment provides continuous samples in 3-inch by
1.65-inch sleeves, and its production rate is six to seven 20-foot holes per 8-hour day.
For shallow groundwater sampling a drive-point sampler (i.e., HydroPunch™) was
selected so that samples could be collected quickly without the expense of installing a complete
monitoring well. Hollow stem auger equipment was selected for setting shallow to medium
depth monitoring wells and well points, and a dual-tube air percussion rotary rig was selected for
investigating the deep regional drinking water aquifer. The air rotary rig is faster than hollow
stem augers and provides a continuous core while driving an outer casing as part of its normal
drilling operation. The outer casing prevents cross contamination of different aquifer zones. The
firm chosen to supply this equipment was large enough to be able to add and subtract drilling rigs
with little advance notice. This feature provided added flexibility because it allowed for as little
as one piece of equipment to be on site or as many as three hollow stem augers, one well
development rig, and one dual tube air percussion rig, depending on project needs.
In sampling groundwater, the planning team proposed measuring pH, temperature,
specific conductance, dissolved oxygen, redox, and turbidity with a flow-through cell. The cell
provided continuous real-time measurement of these parameters that could be used for evaluating
water quality as well as providing information on whether formation water was flowing through
the cell. The real-time turbidity measurements were especially helpful in taking water samples
for metals analysis.
Selection of additional contracts and arrangements included:
• A geophysics firm to perform utility clearance and to locate any subsurface anomalies
that required further investigation;
• A survey team to provide reference points and well point location data;
• A fixed laboratory to perform off-site confirmatory analysis using methods published by
EPA's Contract Laboratory Program and SW-846; and
• A data validation firm to validate off-site laboratory results.
The prime contractor provided oversight of the work with experienced personnel as it occurred.
13
-------
Writing a Dynamic Work Plan
The contractor submitted a dynamic work plan containing a number of elements that
made it different from a staged approach. These elements included the use of:
On-site analysis of groundwater samples to locate monitoring wells;
• On-site field instruments for determining the extent of contamination;
One type of organic detection as a surrogate for other organic and inorganic
contaminants;
Sampling decision trees that would be implemented if contamination was found during
the initial preset sampling phase; and
Flexible sampling and analysis planning that included contingencies for alternative
methods if problems occurred.
During the review of the dynamic work plan, the regulatory agencies objected to the first
four of these five elements. Items two and three were resolved when the contractor prepared an
issue paper for the regulatory agencies showing that the field instruments could detect both
VOCs and oil contaminants at levels of concern, and that the detection limits for these surrogates
also would capture any metals or PAHs that were in the oil, or VOCs at concentrations of
concern.
With items one and four, the state and EPA regulators were concerned that the on-site
decision making would result in insufficient regulatory oversight. As a result, the work plan was
amended to ensure full participation of the regulatory agencies in the decision-making process
through the submission of weekly summaries for teleconference discussion followed by weekly
progress meetings. In addition, if any serious problems arose during the week, a conference call
would be scheduled. This arrangement satisfied the agencies' concerns and the plan was
approved.
Conducting the Dynamic Field Activity
The field team was able to adjust the dynamic work plan before beginning the actual field
work because IRP-12 was not the first site to be investigated. The major change resulting from
their experience was to use the dual-tube DP rig to collect both groundwater and soil samples at
the initial sampling locations, rather than the drive point sampler. The dual-tube DP rig had been
shown to produce acceptable VOC sample results at other locations on MCAS Tustin. The drive
point sampler was used only when groundwater was the only medium to be sampled or when a
metals analysis was to be performed on the groundwater sample because of regulator concerns
about potential turbidity levels.
14
-------
Selection of Sampling Locations
The IRP-12 field work began with a utility clearance of the sampling locations through a
geophysical survey of the site. Subsequently, two hand auger crews consisting of a geologist and
technician began the shallow soil sampling at the statistically selected locations. The two hand-
auger crews sampled up to 16 locations a day to depths of one to two feet bgs.
Refinement of Conceptual Site Model
The initial sampling and analysis revealed one major surprise: significant TCE
contamination outside the boundaries of all three storage areas, near Building 20B. In addition,
the investigation confirmed the ESI findings of widespread—but low level—TRPH
contamination in the shallow soil at Storage Area 1 adjacent to Building 90 (see Exhibit 2), and it
found TCE contamination in the groundwater and shallow soil in the same area. In contrast to
the findings of the ESI, however, the initial sampling did not find any TRPH in Storage Area 2
adjacent to Building 20B, while it did find widespread but low-level TRPH contamination in
shallow soil at the Storage Area 3.
The investigators then initiated a second round of sampling with the dual-tube DP rig at
sampling points that had shown signs of contamination following the decision tree outlined in the
work plan. At these locations, continuous cores were collected from 5 feet bgs to the silty sand
layer at about 20 feet bgs. In addition, groundwater samples were collected in the silty sand
layer. This round of sampling delineated the source area of the contamination next to Building
20B. However, the groundwater underneath the clean, upgradient soil samples was
contaminated, indicating the presence of an additional source area. Further sampling delineated a
second source area about 50 feet outside of the initial sampling zone. Upgradient groundwater
samples were clean, confirming that there was no additional source for the plume.
The investigation of the source area next to Building 90 followed a similar scenario.
Shallow soil samples collected upgradient of the initially identified source area were "clean,"
however, the groundwater at these points had significant TCE concentrations. The subsequent
soil sampling identified another source upgradient. Attempts to bound groundwater
contamination upgradient failed a second time, and an additional source area was sought. By
continuing to sample soil and groundwater in an upgradient direction, the fifth and final source
area in IRP-12 was identified and delineated.
The completed source investigation showed shallow low-level (less than 200 ppm) TRPH
contamination in Storage Areas 1 and 3 but insignificant contamination in Storage Area 2 (refer
to Exhibit 2 for locations). It also found five TCE source areas that produced two shallow
groundwater plumes, both continuing beyond the boundaries of the original IRP site. The
resulting interim conceptual site model is presented in Exhibit 7. With the source areas
identified, the dual-tube DP rig was used to continue the shallow plume delineation. The field
15
-------
o>
uoi)EDi|S9AU| icipaiuay uitsni SVOW :iuojj paidepv
|en)daouoQ iuua)U|
-------
team was able to define the shallow plumes within eight days with a total of 21 samples. This
was accomplished by using only on-site laboratory results from water samples that were available
within the time it took to decontaminate the equipment and move to a different location. Exhibit
8 presents the results of this phase of the investigation.
The next step was to determine how deep the TCE had migrated. This was accomplished
by taking water samples in the outer casing of the dual-tube air percussion rig. This investigation
demonstrated that the plume originating near Building 90 was confined to the first permeable
zone, but that the larger plume, originating near Building 20B, had penetrated to a second
permeable zone, within the shallow aquifer, between 42 to 53 feet bgs. This information was
then integrated into the conceptual site model presented in Exhibit 9.
Completion of Field Program
In October 1995, the BRAC Cleanup Team examined the data delineating the two plumes
and decided how to place the necessary number of monitoring wells efficiently. In all, only 14
monitoring wells were installed and developed, with the work being completed in November and
December 1995. Since pump-and-treat was a viable option for aquifer remediation, two aquifer
pumping tests were designed and conducted to aid in the feasibility study evaluation and future
remedial design.
In December 1995 the Navy submitted a short summary of findings to the regulatory
agencies. The summary presented all available data along with a discussion that indicated the
nature and extent of contamination at the site had been characterized. In response, the regulators
suggested that one more deep boring be placed in the central area of the larger plume to better
define how far the plume in the second permeable zone had traveled. This was completed in
January 1996 along with the first round of permanent groundwater monitoring well sampling.
Following monitoring well sampling, the field team performed a final quality assurance
check on the existing data before equipment was demobilized. This review revealed a few data
discrepancies, and the technical team leader ordered the field team to collect three additional soil
samples. Furthermore, there appeared to be a gap in the soil characterization to the east of
Building 533. To ensure that another source area did not exist at this location, the field team
collected three shallow judgmental soil samples. All of these additional samples were collected
in March 1996.
Over the life of the investigation, the effort was expanded from an initial 68 sampling
points to 147. Of these, 60 were between 1 to 2 feet bgs, and 54 were taken in the first
permeable zone (i.e., 15 to 25 feet bgs). Approximately 390 soil samples were screened by the
on-site laboratory, and approximately 70 were sent off-site for some form of confirmatory
analysis (i.e., VOCs and metals; or metals, PAHs, and PCBs). In general, the two sets of
17
-------
00
sauinid jajBMpunojQ MOHBLJS jo |apo|/\| |enjdaouoo
8 iiq
-------
Northwest
Exhibit 9
Conceptual Model of Larger Plume in Cross Section
Southeast
IRP-12
Explanation
GRAVEL ? SILT (SU-ULI
SANOV SILT (MLI
SIL' MLI.'O- 10
m*diuw plasticity
SU.T IMHI. high tfoslie
ORCAMC SILI (OL1
| SANDY C'.AT rcil
CLAr ta>. to. lo
| m^diu-r1 prnsticily
CLAV (CM). Wjn projl
ORGANIC CLAT IQH:
WOOD, PEAT |P|I
FILL
| ASPHALT
NO RECOVERV
CALICHE
-------
results were in agreement. However, the on-site laboratory reported consistently higher
concentrations for the soil samples than the off-site laboratory, probably due to the loss of VOCs
during sample handling.
Approximately 110 groundwater samples were screened by the on-site laboratory of
which 79 were split with the off-site laboratory for confirmation. The high number of off-site
analyses was due primarily to the fact that a number of these groundwater samples were collected
from the permanent monitoring wells that were installed after the completion of the dynamic
field activity portion of the RI/FS. In addition, another group of samples could not be analyzed
on-site because the on-site laboratory was not set up to perform analysis on metals in
groundwater. The on-site analysis of groundwater samples was used primarily for identifying the
presence of TCE for plume delineation. In this group, there were 69 samples that were analyzed
on-site, 14 of which were sent off-site for confirmation. There was very good agreement
between the results of these samples, with the off-site laboratory generally reporting slightly
higher concentrations, perhaps due to more efficient purging capabilities of the off-site
laboratory.
Writing a Final Report
The three major sections of the final report that are of interest to this case study included:
• Nature and extent of the contamination;
• Contaminant fate and transport;
• Human health risk assessment; and
• Data quality assessment.
Nature and Extent of Contamination
The soil and groundwater investigation discovered five TCE source areas that contributed
to two groundwater plumes. The plume originating near Building 90 had three source areas and
was confined to the first sandy zone in the shallow aquifer. It was approximately 125 feet wide
and 400 feet long. The plume originating near Building 20B had two source areas and penetrated
into a deeper sandy gravel layer within the shallow aquifer but did not reach the regional drinking
water aquifer. It was approximately 150 to 300 feet wide and 1,500 feet long.
In addition, to the surprise of the BRAC Cleanup Team, Freon 113™ was detected in
confirmatory laboratory results that arrived during the delineation of the plume originating at
Building 20B. The results showed that the Freon 113™ plume was slightly smaller than, and
within, the TCE plume, so they concluded that characterization of the TCE plume would also
encompass the less hazardous Freon plume. No background information, nor any previous
sampling, provided any indication that Freon 113™ was a potential contaminant of concern, and
20
-------
the selected instrumentation (GC/PID) was unable to detect it. However, in choosing the
instrumentation, the investigation team had considered the possibility of not detecting potential
chemicals of concern in the planning process. For this reason, a third of the samples sent to the
off-site laboratory were on-site laboratory non-detects. A more complete discussion of this issue
is provided in the "Lessons Learned" section that follows.
Contaminant Fate and Transport
With the help of MODFLOW (McDonald and Harbaugh, 1988) and MT3D (Zheng,
1990) modeling software, the BRAC Cleanup Team predicted that the two plumes in the first
permeable zone would slowly migrate to the south and co-mingle through dispersion as they
dilute. After 100 years, the model indicated that the plume may migrate off-base at a TCE
concentration in the 10 to 20 |ig/L range. The low-concentration TCE plume in the second
permeable zone also was expected to migrate south and dilute. Because of the slightly upward
groundwater gradient, none of the plumes were expected to reach the regional drinking water
aquifer (approximately 100-120 feet bgs).
Human Health Risk Assessment
The potential risk drivers for the site at the beginning of the investigation were believed
to be heavy metals (primarily arsenic, cadmium, chromium, and lead), PAHs found in waste oils,
and TCE. The soil investigation found that the risk from metal and petroleum hydrocarbon
contamination was negligible. TCE soil contamination was found in five general areas at
concentrations below the state preliminary remediation goal for soil. However, given the depth
to groundwater, these areas were considered secondary groundwater contamination sources and
were candidates for remediation to prevent further degradation of the aquifer.
In addition, although TCE concentrations in groundwater were above MCLs, human
exposure to the contaminants through drinking water was not considered a risk factor because the
contaminated groundwater was not potable. The BRAC Cleanup Team came to this conclusion
by noting that the concentration of total dissolved solids (TDS) was as high as 15,000 mg/L with
manganese concentrations up to 2 mg/L in the shallow aquifer, and secondary drinking water
standards for TDS are 500 mg/L and 0.05 mg/L for manganese. In addition, the aquifer was not
appropriate for irrigation, not only because of the high TDS, but also because of the low
transmissivity (84 gpd/ft). Furthermore, an aquifer pumping test using an agricultural supply
well (3,000 gpm) screened in the underlying regional aquifer indicated that there was no
hydraulic connection with the shallow aquifer.
In spite of these issues, the BRAC Cleanup Team decided remediation of the aquifer was
appropriate for two reasons. First, California law requires groundwater remediation for all its
aquifers unless treatment is proven to be technically infeasible. Second, the model of the plume
21
-------
indicated that it would eventually discharge to a nearby surface water and potentially threaten a
neighboring wildlife refuge. Consequently, the RI recommended that remedial measures be
evaluated for the groundwater at IRP-12.
Data Quality Assessment
In order to demonstrate that data collected at the site were adequate for decision making,
the final report included a section on data quality assessment. This section examined whether
investigators met the site performance criteria with regard to precision, accuracy,
representativeness, completeness, and comparability (PARCC). Although the on-site GC and oil
analyzer were used only for determining the extent of contamination, a comparison of the data
generated was made with appropriate off-site results. Based on an agreement between the Navy
and regulators, the following types of off-site laboratory data received full validation in
accordance with EPA's Contract Laboratory Program National Functional Guidelines:
• All well-point samples;
All PAH data;
• All hexavalent chromium data;
Deep soil and water samples;
10 percent of VOCs;
10 percent of metals; and
• 10 percent of PCBs/pesticides.
All other off-site data was reviewed to ensure that it was reported adequately and consistently.
The results of the PARCC analysis stated:
• Precision: Precision for both laboratory and field duplicates was determined to be of
acceptable quality. Outliers did not affect the overall conclusions of any of the
investigations.
• Accuracy: Attainment of accuracy performance criteria was judged to be acceptable, with
the exception of PAH analyses by the fixed laboratory with Method 8310. This finding
was discovered early in the investigation and resulted in a formal project audit of the
laboratory when complaints did not rectify the problem. The audit resulted in the
dismissal of the chemist running the method.
• Representativeness: With the exception of only a few small anomalies the data were
deemed to be representative. An analysis of these problems indicated they would not
change any risk or remedial recommendations being made for the site.
22
-------
Completeness: Completeness generally is judged by the number of planned samples
versus the number of actual samples taken that are useable. Since this program was
dynamic, the number of planned samples was not specified at the beginning of the
project. However, completeness can also be measured by the number of samples shipped
for analysis (i.e., samples for which data are required) and number of useable results
obtained. Under this modified definition, a completeness factor of 99 percent was
achieved at all but one of the IRP sites (IRP 16 at 97 percent). IRP-12 was 99.6 percent
complete.
Comparison of On-Site and Off-Site Data: Because soil samples analyzed by the two
laboratories were duplicates rather than splits, they could not be compared directly.
Consequently, there was some disparity between the two laboratories. On the other hand,
because groundwater samples tend to be more homogeneous, the data evaluated for the
two laboratory groundwater data sets were largely comparable.
Estimated Cost and Time Savings
Although the field work for IRP-12 began in August 1995 and was completed in March
1996, the actual time spent on the site only totaled six weeks. The long period of time for the
field work was not the result of staging the field activities, but rather from the integration of all
dynamic field activities so that the problems at the base could be treated as a whole. This
approach allowed for the free movement of equipment and personnel between locations. When
the technical team leader reached a point where his evaluation of the on-site data indicated that
subsequent activities should be discussed with the BRAC Cleanup Team, the field team would
move to a new site. It should be noted here that had IRP-12 been the only site under
investigation the review process would have been altered to allow for uninterrupted activities.
The investigation often continued at the new location until another decision point was reached, at
which time they moved back to the original site or on to a different location. Since the goal of
the investigation was to characterize the contamination at the entire military base as quickly as
possible, not just one location, equipment and personnel were not bound to a specific spot.
Although it is difficult to distinguish the resources required for only one location at this
site, we have estimated that the cost of IRP-12 was $496,000, and that it would have taken a total
of 10 months (including project planning and the writing of a final report). Although it is even
more difficult to estimate the resources that would have been spent at the site had dynamic field
activities not been conducted, we estimate that a staged approach could have cost $587,488 and
would take approximately two years. This is based on the assumption that an accurate
characterization would have required four mobilizations: the first for initial sampling, a second to
install the first round of monitoring wells, a third to find the source areas outside the initial
sampling area, and a fourth to delineate the contaminant plumes.
23
-------
Each of these mobilizations would have resulted in extra time and cost associated with
the preparation and review of iterative work plans as well as interim reports. This estimate does
not include the administrative expenses EPA and the State of California would have incurred
through regulatory review of multiple report and work plan iterations, nor does it include the
additional time and money that is often required during the CERCLA remedial design and
remedial action to refine the conceptual site model (the dynamic approach was able to completely
characterize the site so there would have been no additional expense for this activity). Finally,
incalculable but significant costs would have been incurred with the staged approach through the
long-term expense of sampling and monitoring large numbers of wells that are ineffectively
located. A summary of the cost and time comparison is presented in Exhibit 10, a detailed
discussion of how they were calculated is available at
http://www.epa.gov/superfund/programs/dfa/casestudies.
Exhibit 10
Dynamic Field Activity vs. Staged Investigation
Summary of Cost and Time Comparisons
Category
Work Plan Development
Field
Activities
On-site Analysis
Off-site Analysis
Sampling Equipment
Costs
Prime Contractor's
Field Support
Misc. Field Costs
Report Writing
Grand Total
Dynamic
Cost
$64,920
$40,140
$107,920
$93,800
$60,620
$20,000
$109,200
$496.600
Time
(weeks)
14
6
24
44
Staged
Totals For All Phases
Cost
$105,295
$180,285
$62,288
$44,310
$20,000
$169,260
$587.488
Time
(weeks)
35
16-24
6
50
110-118
Lessons Learned
Because this RI/FS work was organized and proceeded differently than any investigation
either the contractors, regulators, or Navy had previously been involved in, there were a number
of lessons learned. Although there were no major problems caused by the new process used at
this site, with experience, the investigation could have proceeded more smoothly and more
resources could have been saved.
24
-------
Off-Site Analytical Program Oversight
There were several problems that resulted from poor communication or insufficient
oversight of the off-site analytical program. The first resulted from a failure to warn the off-site
laboratory when a sample was highly contaminated. Although it was program policy to do so,
there was no good system in place to ensure that the warning occurred. This communication
problem resulted in analytical delays and associated QC difficulties, especially with highly
contaminated samples analyzed by SW-846 method 8310.
Second, although the sampling and analysis plan limited the type of analysis that the off-
site laboratory would conduct based on contaminants detected by the on-site laboratory, samples
going to the off-site laboratory were generally analyzed for all constituents. This additional
analysis was caused by a poorly trained shipping technician and a distracted site supervisor. The
result, though not detrimental to the investigation, raised the cost at IRP-12 by about $10,000.
Finally, the number of non-detect samples sent to the off-site laboratory was much higher
than the analytical plan had prescribed. Non-detects were supposed to represent a third of the
samples shipped, but they actually represented approximately 45 percent. The cause for these
excess analyses was that the QC program was ineffective in tracking the ratio. Although spot
checks were performed by the technical team leader, these were not done on a sufficiently
frequent basis to correct problems.
Unexpected Chemicals of Concern
The information in the PA and ESI provided no indication that Freon 113™ had ever
been used, stored, or detected at MCAS Tustin. In addition, the field analytical equipment used
during the investigation was not able to detect it. Investigators had considered the possibility of
an unexpected chemical of concern being missed by the on-site instrumentation, so the work plan
included a provision for sending one-third of non-detect samples to an off-site laboratory for
confirmatory analysis. Although this provision was sufficient for delineating Freon 113™ in this
instance, if an undetected chemical had formed a separate plume or been at the leading edge of a
plume, there may have been a need to revisit the site because the confirmatory analysis did not
arrive for four weeks. Consequently, this case study emphasizes the need for analytical plans to
contain provisions that will enable detection of unsuspected contaminants. Provisions could
include using quick turnaround confirmatory analysis at the beginning of an investigation phase
(e.g., delineation of source areas, groundwater sampling) or having analytical equipment on site
that can detect broad ranges of contaminants (e.g., GC with two detectors).
25
-------
Flexible Work Planning
Flexible work planning was an important aspect of the investigation because it allowed
efficient use of resources and enabled changes to be made in how equipment was used as new
information was obtained. An example of this was in the groundwater sampling program. The
initial investigation plan proposed that soil samples be taken by a dual-tube DP rig, and that
groundwater samples be taken exclusively by driven sampling point equipment (i.e.,
HydroPunch™). However, investigators learned soon after the start of the sampling program that
a bailer could be lowered through the outer casing of the DP rod to collect a groundwater sample.
Although the sample was turbid and therefore not useful for metals analysis because of the
potential to bias the results high, it was still useful for on-site screening of VOCs. By being able
to make this sampling plan adjustment in the field, investigators were able to save considerable
resources, and they were able to supply the on-site laboratory with many more groundwater
samples than had been projected, thereby obtaining a higher density of site data that improved the
certainty of their decisions.
Performance Evaluation Samples
Because the PE samples that were used by the on-site laboratory were developed for the
Superfund Contact Laboratory Program, they contained a number of analytes that the on-site
laboratory was not set up to identify. Although regulators still found the results acceptable, more
useful data could have been obtained if the PE samples were specifically designed for
contaminants and concentrations that were most important to the project. The Contract
Laboratory Program, operated by the EPA Office of Emergency and Remedial Response, can
now provide site specific PE samples.
Geophysics
Apart from the dynamic field activities at this site, the Navy hired a third party to conduct
an extensive geophysical survey. However, the study was not helpful because the upper layer of
soil contained large concentrations of clay, effectively eliminating the usefulness of some
geophysical methods, and there were no clear breaks in the site stratigraphy, effectively
eliminating the usefulness of many other geophysical methods. Since this basic stratigraphic
information was available before implementing the geophysical survey, significant resources
could have been saved if the investigators had reviewed the data before concluding that a
complete geophysical investigation was an appropriate activity for this site. Consequently, the
usefulness of geophysical methods should be viewed on a site-specific basis.
26
-------
References
Argonne National Laboratory. 1995. Expedited Site Characterization at the Marine Corps Air
Station, Tustin, California: Phase I Report and Recommendations for Phase II. Prepared for
U.S. Department of Defense, Department of Navy, Southwest Divisions, Naval Facilities
Engineering Command, by Argonne National Laboratory, Argonne, Illinois.
McDonald, M.G. and A.W. Harbaugh. 1988. A modular three-dimensional finite difference
ground-water flow model. U.S. Geological Survey Techniques of Water Resources
Investigations, Book 6, Chapter Al.
USEPA. 1989. Methods for Evaluating the Attainment of Cleanup Standards, Volume 1: Soils
and Solid Media, EPA/230/02-89/042.
USEPA. 1994a. QA/G-4 Guidance for the Data Quality Objectives Process, EPA/600/R-
96/055.
USEPA. 1994b. SOP #2109: Photovac GC analysis for air, soil gas, water, and soil.
Compendium of ERT Field Analytical Procedures, OSWER Directive 9360.4-04.
USEPA. 1997a. Draft Final Remedial Investigation Report for Operable Units 1 and 2 Marine
Corps Air Facility Tustin, California, CTO-0049/1165.
USEPA. 1997b. Test Methods for Evaluating Solid Waste, SW-846. Office of Solid Waste,
Washington, DC. http://www.epa.gov/epaoswer/hazwaste/test/main.htm
U.S. Naval Facilities Engineering, Southwest Division. 1995a. Appendix A Data Quality
Objectives. Draft Final Remedial Investigation Work Plan for OU-1 andOU-2, Marine Corps
Air Station, Tustin, California. U.S. Naval Facilities Engineering, Southwest Division.
http://www.epa.gov/superfund/programs/dfa/casestudies.
U.S. Naval Facilities Engineering, Southwest Division. 1995b. Draft Final Remedial
Investigation Work Plan for OU-1 andOU-2, Marine Corps Air Station, Tustin, California. U.S.
Naval Facilities Engineering, Southwest Division.
U.S. Naval Facilities Engineering, Southwest Division. 1995c. Marine Corps Air Station Tustin
Data Quality Objectives IRP-12 (Drum Storage Area No. 2). Draft Final Remedial Investigation
Work Plan for OU-1 andOU-2, Marine Corps Air Station, Tustin, California. U.S. Naval
Facilities Engineering, Southwest Division.
http://www.epa.gov/superfund/programs/dfa/casestudies
27
-------
U.S. Naval Facilities Engineering, Southwest Division. 1995c. Draft Final Sampling and
Analysis Plan Marine Corps Air Station, Tustin, California. U.S. Naval Facilities Engineering,
Southwest Division.
U.S. Naval Facilities Engineering, Southwest Division. 1997. Draft Final Remedial
Investigation Report for Operable Units 1 and 2, Marine Corps Air Facility, Tustin, California.
U.S. Naval Facilities Engineering, Southwest Division.
Zheng, C. 1990. MT3D: A Modular Three-Dimensional Transport Model for Simulation of
Advection, Dispersion, and Chemical Reactions of Contaminants in Groundwater Systems. S.S.
Papadopulos & Associates, Inc. Software and user guide prepared for U.S. EPA.
http://www.epa.gov/ada/csmos/models.html
28
------- |