Technology News & Trends | a quarterly newsletter
Winter 2015 I Issue 68
A newsletter about soil, sediment, and groundwater character!
Technology
News & Trends
remediation technologies
This issue of Technology News & Trends highlights approaches for improving and streamlining site cleanup
through a broad strategy that begins with upfront planning with an eye toward project completion, as
described in the U.S. Environmental Protection Agency's (EPA's) fiscal year 2014 Superfund Remedial
Program Review Action Plan. EPA's plan describes short- as well as long-term measures and activities the
Agency is undertaking to maintain an effective remedial cleanup program under Superfund program budget
constraints. An important component of the plan is the use of an adaptive management process—an iterative
approach to site investigation and remedy implementation that facilitates responding to new information and
conditions throughout the lifecycle of a site. The plan also focuses on assessment, study, design and
construction phases of the remedial process and outlines modified priorities for related resource management
to be combined with additional increases in efficiencies.
One action underway as part of the plan is development of an Agency directive providing examples of tools
and approaches that leverage the adaptive management process, such as:
• Life-cycle conceptual site models (CSMs).
• Dynamic work plan strategies.
• Groundwater remedy completion strategies.
• Early source treatment response actions followed by effectiveness monitoring.
• Phased or iterative risk assessment approaches incorporating multiple lines of evidence and
ecological monitoring.
A related action involves identifying best management practices and technical resources for remedial
investigation and CSM development with a focus on streamlining data collection through smart scoping,
strategic sampling that uses high-resolution and real-time analytical techniques, and improved data
management. Better planning and scoping of a remedial investigation/feasibility study (RI/FS) up front can
help reduce RI/FS costs (including data collection costs), which have significantly increased in the Superfund
remedial program over the past decade.
The projects featured in this issue illustrate ways to more effectively compile information as part of optimizing
the design, implementation and monitoring of remedies and to strategically schedule key activities
accordingly. Each of the featured projects involves site characterization and remediation at a National
Priorities List (NPL) site.
-------
Site Characterization: Strategic Sampling and Adaptive Remedy Implementation for Improved
Cleanup Performance at Commencement Bay-South Tacoma Channel
Contributed by Kira Lynch, U.S. EPA/Region 10, and Steve Dyment, U.S. EPA/Office of Superfund Remediation and
Technology Innovation
The U.S. EPA Region 10 office uses an adaptive site management approach that relies on high-resolution site
characterization (HRSC) techniques for strategic sampling at the Well 12A project area. This approximate one-
square-mile project area is one of three at the 2.5-square-mile Commencement Bay-South Tacoma Channel
Superfund site in Tacoma, Washington. The vicinity of Well 12A, one of 13 wells used by the City of Tacoma to
meet peak summer and emergency water demands, contains high concentrations of chlorinated volatile organic
compounds (VOCs) including light and dense non-aqueous phase liquids (NAPL). Use of HRSC at Well 12A is
critical to evaluating the mass discharge (flux) of contaminants associated with implementation of several
treatment technologies and to determining the point at which active treatment may transition to monitored natural
attenuation (MNA). The Well 12A record of decision (ROD) as amended in 2009 is the first EPA ROD specifying a
mass discharge reduction as an interim goal fora remediation performance metric.
Well12A contamination is associated with past waste oil recycling operations, which resulted in release of solvent-
related NAPL from drum and tank storage areas and disposal of filter cake containing NAPL. Trichloroethene
(TCE) and its degradation products are the primary contaminants of concern (COCs). The well was taken out of
operation when it was found to be contaminated but has operated since 1983 through use of five air-stripping
systems. After the source area was identified, approximately 1,200 cubic yards of filter cake mixed with
contaminated soil were excavated along a rail line and an additional 5,000 cubic yards of filter cake were removed
during construction of a soil vapor extraction (SVE) system. Between 1993 and 1997, the SVE system removed
an estimated 54,100 pounds of VOCs. A carbon treatment-based pump and treat (P&T) system began operating
in 1988 and was expanded in 1993 with additional extraction wells. By 2011, the P&T systems had extracted and
treated 860 million gallons of water and removed approximately 18,625 pounds of VOCs; however, progress
towards aquifer restoration was slow, and the capture and treatment of all site-related chemicals remained a
challenge.
Remedies to address this challenge recently involved excavation and offsite disposal of remaining filter cake and
contaminated shallow surface soil, in situ thermal remediation (ISTR) of deeper vadose zone and saturated zone
soil and groundwater, and in situ enhanced anaerobic bioremediation (EAB) of groundwater. Continued operation
of the P&T system is needed to prevent migration of contaminants until their mass is significantly reduced through
excavation, ISTR and EAB. The adaptive management strategy for implementing ISTR and EAB technologies
involves an overlapping operating schedule to maximize use of the ISTR applied and residual heat for advancing
EAB.
The highest-priority remedial action objectives (RAOs) for Well 12A are to reduce risk from contaminated surface
soil and achieve at least a 90% reduction in contaminant mass discharge from the source area (below and around
the former recycling building known as the Time Oil Building) to the dissolved-phase contaminant plume. Other
priorities are to achieve chemical-specific applicable or relevant and appropriate requirements (ARARs) measured
at alternate points of compliance and to determine if MNA can be used to achieve ARAR requirements throughout
the plume.
The conceptual site model (CSM) has been updated throughout the remedy design and remediation process. A
major CSM refinement involved determining how to best measure and assess the mass discharge reduction goal
while considering the significantly varying hydraulic conductivities, which range from 0.3 to 3,550 feet/day (ft/d)
over relatively small vertical distances within the heterogeneous soil. Three-dimensional (3-D) imaging software is
used to analyze the data and effectively portray the site's complex hydrogeology and the VOC spatial distribution,
masses and volumes, and mass discharge transects (Figure 1).
-------
I Time Oil Building I
Elev. (ft msl)
300 +
-^''
Conductivity (
Surface Fill
Qvs Gravel
Total VOC
MD % of
(ko/vrt Total MD
Transect 1
Qva
Qpfc1/Qpf
,Qpfc2
.Qpfc3
iQpogc
Total
0.1 1%
2.9 T 31%
5.996%H 64%
0.06 L 1%
0.3 4%
9.3
Qpfc
-North
Transect 2
Qva
Qpfd/Qpf
Qpfc2
QpfcS
Qpogc
Total
0.01
0.2
1.7
0.4%
7%
57%
0.1 3%
1.0 33%
3.0
Figure 1. 3-D image of subsurface transects, and corresponding estimates of mass discharge (MD) at Well 12A
Surface
Area (ft2)
VOC Mass I Discharge to
(kilograms) I P&T Systems
224 g/day
(53%)
The refined CSM reflects two mass flux transects suggesting that most of the mass discharge occurs in three
distinct hydrostatrographic units. Close to the source area in transect 1, 95% of the mass discharge occurs in two
units (Figure 1) designated as Qpfc1/Qpf (accounting for 31% of the discharge) and Qpfc2 (64% of the
discharge). Both units consist of primarily coarse-grained
soil with a hydraulic conductivity ranging from 35 to 782
ft/d. However, the Qpf sub-unit consists of fine-grained
material (conductivity below 1 ft/d) that stores
contaminant mass and acts as a secondary diffusional
source feeding units above and below it. Farther
downgradient in transect 2, 90% of the mass discharge
occurs in the Qpfc2 unit (57% of the discharge) plus a
unit designated as Qpogc (33% of the discharge).This
portion of the Qpfc2 unit is coarse grained (conductivity
of 35-782 ft/d) while the Qpogc unit represents a mass
storage area in a relatively thin finer-grained unit
(hydraulic conductivity of approximately 1 ft/d) that
transitions to the aquitard. The hydraulic conductivity
ranges and mass discharge estimates of these
hydrostratigraphic units are key elements of the lifecycle
CSM as it is used throughout the adaptive remedy
implementation process.
Other tools for analyzing and facilitating decision-making
on the VOC mass discharge include analytical mapping
software that highlights the estimated discharges to the
P&T system from the thermal treatment zone and the
bioremediation zone (Figure 2). Based on this
information, the P&T operations were slightly modified
and will be used as the method for RAO compliance.
3-D imagery and other HRSC techniques also were
used to develop vertical profiles to help visualize
subsurface contaminant distribution, calculate
contaminant mass, and delineate treatment zones for
excavation, ISTR and EAB (Figure 3). One significant
Figure 2. Analytical mapping imagery depicting
contaminant mass distribution across the Well 12A
site, which was used to define zones of excavation,
thermal treatment, and in situ bioremediation
Legend
Monitoring Well
Injection Well
P&T Extraction Well
In Situ Thermal Treatment Area
Total VOCs in Soil >5,000 ug/kg
Union of TCE and cis-DCE > 300 ug/L above Qpf silt
v ) Amendment Injection Location
Pilot Study Injection Well
-------
benefit of using vertical profiling in this adaptive management approach is the capability to prioritize treatment
based on the mass expected to be discharging to more transmissive zones. For EAB, this capability also helped
reduce the target vertical interval from approximately 60 ft to 15 ft, which substantially reduced the EAB treatment
volume, saved approximately $1 million in EAB amendment purchasing costs, and significantly reduced the
environmental footprint of EAB implementation.
Total VOC
^•l 1,000 mg/kg
9 300 mg/kg
Figure 3. Vertical profile of the Well 12A VOC extent in soil of the excavation, ISTR, and EAB zones.
Excavation was completed in 2012, and a portion of the Time Oil Building was demolished in 2013. In early 2014,
remediation focused on installing and operating the ISTR system and performing two rounds of EAB injections.
The first round, which was conducted as a pilot test, involved injecting waste vegetable oil in two wells and a
commercial, emulsified vegetable oil in one additional well. These oils were used to promote microbial activity and
improve vertical distribution of other amendments to be injected a few months later.
The second injection round, completed in November 2014, entailed injecting over a million gallons of amendment
including bioaugmentation culture in 47 wells at select locations to treat an area covering 3.7 acres. The injection
strategy was adaptive, using different design specifications to adjust for variability in the soil permeabilities,
observed contaminant mass levels, geochemical conditions, and presence of important contaminant-degrading
microbes throughout the treatment zone.
Active ISTR heating of the subsurface began in April
2014 and continued for 117 days at an average
temperature of 96.8°C, using a total of approximately
4,026,000 kWh of electricity. The operation involved
use of three isolation transformers inside a power
control unit to generate electrical resistance heating,
70 electrodes, 35 independent recovery wells, 299
temperature monitoring points and 23 pressure
monitoring points (Figure 4).
Based on the evaluated data, it is estimated that a
minimum of 22,300 pounds of contaminant mass
was removed from the target treatment zone during
ISTR. This total mass estimate includes
approximately 9,600 pounds of VOC mass removed
in the vapor phase, 7 pounds of VOC mass removed
in water and 12,700 pounds of NAPL. This estimate
does not include aliphatic hydrocarbons removed Figure 4. Well 12A ISTR electrodes at 15- to 17- foot spacing
-------
from the treatment volume via vapor recovery due to the difficulty in quantifying aliphatic carbons. However,
based on the carbon consumption rates and other available data, it is assumed that the total mass of aliphatic
hydrocarbons removed in the vapor phase is equivalent to or greater than the total mass of VOCs (9,600 pounds)
removed in the vapor phase. Accounting for this additional recovery, it is estimated that a minimum of 31,900
pounds of total contaminant mass was recovered from the treatment zone.
By implementing ISTR and EAB in parallel (Figure 5), excess or residual heat from active heating have increased
temperatures in the EAB zone, fostering additional activity of anaerobic microorganisms capable of degrading the
contaminants. Performance monitoring of the EAB injections indicates that high levels of carbon may persist in
the target zone for at least one year, with an overall longevity of approximately two years.
EPA is currently monitoring
remedial performance and progress
towards reducing mass discharge.
In addition, EPA is considering
strategies to address previously
unidentified hotspots of dense non-
aqueous phase liquids that were
encountered during installation of
the EAB wells.
ISTR Construct ISTR Operation Post-ISTR Monitoring Period
EAB Injections: Round 1 EAB Initial Monitoring Period EAB Injections: Round 2
Combined Technology Optimization Period
Figure 5. ISTR and EAB implementation overlap leading to availability of
residual heat that can improve EAB performance at Well 12A
-------
Streamlining Remedy Design through Early Optimization Evaluations
Contributed by Vince Malott and Camilla Hueni. U.S. EPA/Region 6
The U.S. EPA Region 6 office recently consolidated efforts to finalize remedy designs for three Superfund sites in
Texas with similar conditions: the Jones Road site, the Sandy Beach Groundwater Plume site and the East 67th
Street Groundwater Plume site. Efforts focused on optimization of the preliminary designs by prioritizing each
site's issues in light of budget restrictions, sequencing remediation activities in light of those priorities, and
streamlining the final designs in ways expected to reduce the cleanup costs and achieve cleanup levels more
efficiently. This adaptive approach to site management relied on the use of site-specific environmental data to
interpret a single conceptual site model (CSM) reflecting similarities among the three sites, including the existence
of chlorinated solvent plumes in relatively deep aquifers used for drinking water. Based on remedial investigations
and preliminary designs, cleanup at each site also was expected to involve the same technologies: soil vapor
extraction (SVE) or in situ chemical oxidation (ISCO) to address source area soil and shallow groundwater, in situ
bioremediation (ISB) to address deeper groundwater, and groundwater extraction and treatment to control plume
migration. Additionally, affected private groundwater wells will be abandoned at each site and the properties
connected to pubic water supplies. Technical experts from EPA's Region 6 and national Superfund program and
a third-party optimization review team collaborated to finalize design recommendations over a two-week period in
early 2013.
The Jones Road Superfund site is located in a residential, commercial and light industrial area just outside the city
limits of northwest Houston. Due to past releases from an onsite dry cleaning facility, the primary contaminants
are tetrachloroethene (PCE) and its breakdown products trichloroethene (TCE), 1,2-dicloroethene (DCE) and
vinyl chloride (VC). In 2008, EPA connected 144 affected properties to public water lines under a time-critical
removal action. Components of the site's CSM (Figure 1) are based on information collected during development
of the site's 2010 record of decision (ROD), 2011 environmental assessments and 2012 preliminary remedy
designs.
Direct contact
(Shop
Cet
ping
ter j ParhngLot
Containment — ^^^i /^x^l Volatilization Indoor Air
Mass AP(1N/m*)AC(2-)
SHALLOW CLAY J DlBchargeto GW
SHALLOW WBZ ' Groundwater Transport » }
• ;
w Discharge to Unsaturated Chicot \
T \
UNSATURATED CHICOT 4 \
-+- VAPOR MIGRATION -^ I Ap™^°C,r,
• /SAND
A Condensation of Vapors
y Discharge to Lower Chicot A
1 VolatilizAion to
SATURATED LOWER CHICOT ' Unsatura\edZone
|
' ^
—
^
APPROXIMATE
THICKNESS
25'
10'
25'
50'
_^ Drinking
water supply
Figure 1. Contaminant migration and potential exposure pathways at the Jones Road site
The 2010 ROD selected in situ enhancements (via ISCO) and pump and treat (P&T) for the shallow water-bearing
zone (WBZ) (25-35 feet below ground surface [bgs]) and the Lower Chicot WBZ (110-400 feet bgs) to contain and
treat impacted ground water. Additional tests during the remedial design (RD) confirmed that natural degradation
in the shallow WBZ was occurring after occasional enhancements. RD tests also confirmed high concentrations of
contaminants in the vapor phase at depths of 35 to 110 feet bgs, which were likely sourcing contaminants to the
underlying Lower Chicot Aquifer.
-------
The 2014 optimization review prioritized mitigation of the two predominant vapor sources and implementation of
the shallow WBZ treatment design. Although the ROD identified P&T with hydraulic containment as a remedy for
the Lower Chicot WBZ, the review recommended a limited P&T system be considered only after mitigating the
source and if deemed necessary based on groundwater monitoring trends. If source reduction is effective at
reducing impacts to the Lower Chicot, a P&T system may not be needed or may be needed at a much smaller
scale than originally anticipated. Other optimization recommendations focused on identifying data gaps for all
target subsurface zones; reducing costs associated with using SVE for contaminant source reduction; further
assessing possible indoor air exposure in the area of initial contaminant release; and implementing a data
management system to integrate historical and future monitoring data that support this adaptive site management
approach.
Based on results of the review, the ROD will be amended in 2015 to prioritize mitigation of the two vapor-phase
sources. Prioritization will involve evaluating SVE and other technologies capable of more aggressively
addressing volatile organic compounds in the shallow soil (0-25 feet bgs) and high PCE vapor concentrations in
the upper unsaturated zone (60-110 feet bgs) of the underlying Chicot Aquifer that were previously unknown.
Work to resolve existing data gaps will continue in 2015.
At the Sandy Beach Road Groundwater Plume Superfund site, 2010 groundwater analysis showed a TCE
groundwater plume approximately one-half mile wide and one mile long (Figure 2 [a]), including zones beneath
some residences in the City of Pelican Bay, the adjacent City of Azle and unincorporated areas of Tarrant County
in north central Texas. Many private and some public water supply wells have drawn or continue to draw water
from affected depths of the two aquifers (the Paluxy and Twin Mountains Aquifers). EPA has completed
residential water line connections to the City of Azle's water supply system for those residences with
contaminated private water supply wells.
Early remedial investigations, including a soil gas survey, indicated the source area is located at a former dump
that operated in the 1960s (Figure 2 [b]). Due to associated TCE contamination, two public water supply wells in
|b. iww™™* ^^^H^^^H
GW-34 o Private water supply with TCE below reporting limit
Pelican Bay 10 A Public water supply with TCE below reporting limit
SVE-1 $ SVE well with TCE below reporting limit
I TCE above 5 ng/L
TCE above 50 |ag/L
TCE above 100 ng/L
Figure 2. [a] Sandy Beach site plume boundaries identified in 2010 groundwater sampling, and [b] recent aerial photo overlaid
with TCE vapor boundaries identified in 2007 soil gas survey
-------
Pelican Bay were shut down and filtration units or public water connections were installed to replace nine
residential drinking water wells. The selected remedy includes a groundwater P&T system to contain the TCE
plume and restore two underlying aquifers; an SVE system to remove TCE vapors in the source area above the
water table; and replacement water supply wells for the residents with the remaining well filtration systems.
The optimization review identified several data gaps in the CSM. One gap concerns the materials remaining in a
ravine formerly used for TCE disposal and now considered the primary source of contamination. The presence
and impacts of any chemical or waste drums potentially buried in or near the ravine, however, is not fully defined.
Another data gap concerns the precise distribution of TCE in shallow and saturated soils, including soils of
varying porosity. Other gaps involve the magnitude and persistence of the dissolved phase plume and the
potential water quality impact from proposed ISB treatments.
Through use of an adaptive site management approach that reflects the optimization review, the SVE installation
activities are now prioritized in ways that simultaneously address the data gaps and enable source treatment to
begin. The potential for additional sources of TCE (such as buried drums) will be evaluated during SVE system
installation. Instead of the typical sequence of installing extraction wells followed by trenches and piping, the
trenches and piping will be installed first in order to observe the nature of buried debris. If evidence of additional
sources is encountered, field activities will shift to excavation of the additional source material. During installation
of the SVE wells, saturated soil in the source area will be characterized further using sonic drilling techniques to
collect soil samples from various intervals. Also, an ISB pilot test will be conducted during or following SVE
operations (depending on the results of additional site characterization) to optimize treatment efficacy and identify
potential impacts on water quality. Plans for the groundwater control and treatment system, now prioritized to
follow source control, will include fewer extraction wells than originally anticipated but an up-scaled treatment
system (up to 150 gallons per minute) in case the extraction system is expanded in the future. The plans also
involve a phased approach, beginning with use of horizontal extraction/reinjection wells to contain the plume.
After approximately five years of monitoring aquifer restoration, more informed decisions can be made about
efficient scaling and implementation timing of the designed groundwater treatment system.
At the East 67th Street site, which is
located immediately north of the
City of Odessa, a groundwater
contaminant plume exists due to an
intentional release of over 15,000
gallons of chemicals (including 635
gallons of PCE) from a former
chemical manufacturing facility
(Figure 3). The underlying Trinity
Aquifer is the only source of
drinking water in this area. The
Texas Commission on
Environmental Quality installed and
is maintaining filtration systems on
14 private wells within the plume.
Currently, the site is being
evaluated as a groundwater plume
of PCE and TCE with no identified
source. Although the outer
boundary of the plume is not yet
defined, private well samples
indicate a plume size of at least 0.3
by 0.3 miles.
The selected remedy includes
installation of a water supply line
from the city to three new service
areas within the plume
boundaries. The second remedy
component involves injection of nutritional amendments to enhance ISB of contaminants in the plume interior, and
groundwater P&T for hydraulic containment of the plume front. The third component entails SVE to remove
Monitoring well screened in Lower Sand
Private water supply
MW-21
2.8
<0,5
PCE above the 5 ug/L MCL
Well designation
PCE concentration (ug/L)
PCE below reporting limit
Figure 3 East 6f» street site map ///USfraf/ng the PCE plume (exceeding the 5
drinking water limit) based on June 2013 sampling data
-------
residual contamination beneath and surrounding the former chemical facility, ISB of contaminants in groundwater
above the drinking water aquifer, and abandonment and replacement of selected private supply wells that act as a
conduit for vertical migration of contaminants to the aquifer.
CSM data gaps identified through the optimization review include: the contaminant mass remaining in the vadose
zone that is capable of causing long-term groundwater contamination and its response to SVE; the extent of
dissolved groundwater contamination in a target layer of upper sand in the site's stratigraphy; the potential effect
of ISB on secondary water quality issues such as mobilization of arsenic and metals; potential mechanisms for
vertical migration of contaminants; and the extent of contaminant migration and timeframe for aquifer restoration
after relevant supply wells are plugged and abandoned.
Based on optimization review recommendations that were incorporated into an adaptive site management
approach, the finalized remedy design now reflects the following prioritized activities to improve project
efficiencies, minimize cleanup costs and reduce the project's environmental footprint: (1) replace specific private
water supply wells that may function as contaminant conduits; (2) install new groundwater monitoring wells; (3)
implement ISB for the plume interior before implementing the proposed P&T system; (4) use extracted
groundwater rather than fresh water for ISB substrate blending and delivery; (5) conduct a small-scale SVE pilot
test near the former chemical facility to better characterize contaminant mass remaining in the vadose zone; and
(6) evaluate the need for other remediation activities, such as additional ISB, after plugging supply wells that
potentially act as a contaminant conduit. Work will commence accordingly in 2015.
-------
Performance Evaluation: The Role of Monitoring During Remedy Transition Planning in a Complex
Hydrogeologic Setting
Contributed by William O'Steen and Ralph Howard, U.S. EPA/Region 4
Evaluation of remediation performance at the Medley Farm Drum Dump Superfund site, a former waste solvent
disposal area in northern South Carolina, illustrates the need for comprehensive and structured groundwater
monitoring when planning a transition to a new groundwater remedy. In this case, enhanced reductive
dechlorination (ERD) will eventually transition to monitored natural attenuation (MNA) as cleanup progresses.
ERD injections of lactate over the past 10 years have varied with respect to injection volumes, locations and
timing. Associated groundwater monitoring and data analysis were structured to evaluate responses to each
injection event rather than the injection series, which significantly limits data interpretation and predictive analysis
due to its variable and complex groundwater flow patterns. As a result, evaluation of MNA as a potential final
remedial action will require changing the current monitoring program and data evaluation paradigm. An adaptive
site management approach is facilitating the needed monitoring changes and helping to identify optimal
timeframes for any future injections.
Prior remediation work involved installation and startup of a groundwater pump and treat (P&T) system and a soil
vapor extraction (SVE) system in 1985. Both systems were optimized from the late 1990s until 2004. The SVE
system was shut down when performance objectives for soil remediation were met in 2004. Concurrently,
operation of the P&T system was suspended to pilot test injections of a sodium lactate and oxygen scavenger
solution to enhance conditions for reductive dechlorination of the remaining groundwater contaminants of concern
(primarily chlorinated ethenes). Favorable ERD pilot results led to the amendment of the site's record of decision
in 2012, specifying ERD for groundwater remediation with MNA as a contingency remedy.
The hydrogeologic setting of the site is complex. Exploratory trenches, geophysical investigation, and detailed
geologic mapping identified two distinct near-surface lithologies separated by a northeastward-striking, reverse
fault that dips steeply to the southeast (Figure 1). The former waste solvent disposal areas are upslope from the
fault. The depth to the top of the schist and phyllite bedrock differs by about 20 to 70 feet across the fault. Two
dominant orientations of rock fractures were identified at the site, one set striking parallel to the fault and the
second set striking at near right angles. Highly weathered bedrock (the "transition zone") of varying thickness
overlies the bedrock, and a 50- to 70-foot thick saprolite (primarily silt) overlies weathered bedrock in the disposal
areas.
approximate scale, feet
0 200 400
TN
"\ drum/waste
} disposal area
fault; location
approximate
Q monitoring well
Q extraction well
f~l dual-phase
extraction well
(installed later)
Figure 1. Location of extraction wells at the Medley Farm site in relation to onsite waste disposal areas and the reverse fault
-------
The saprolite, transition zone and bedrock are hydraulically connected. Hydraulic head measurements at most
locations indicated upward hydraulic gradients from the shallow bedrock into the transition zone and saprolite
across the site. The water table may be as deep as 70 feet below ground surface (bgs) in upslope areas,
including the area of contaminant waste disposal, and about 10 feet bgs in downslope areas. The hydraulic
conductivity (K) is estimated at 10~5 to 10~3 centimeters per second (cm/s) in the saprolite and shallow fractured
bedrock and 10"4 to 10"3 cm/s in the transition zone. Limited data from deeper bedrock wells indicate a K of
approximately 10~7 cm/s.
During remedy design, site investigation revealed that the bedrock groundwater flow is strongly controlled by
geologic structures. The fault bisecting the site, for example, is a zone of preferential flow; the fault separates an
upslope area of significant groundwater contamination from a downslope area of inconsequential groundwater
contamination. Other fault-parallel fractures considerably upslope from the fault zone are potential conduits for
northeastward groundwater flow, although some groundwater flow in this part of the site is also oriented
downslope towards a southeastward trending tributary stream that appears to be aligned along a bedrock
structure.
Baseline groundwater monitoring was conducted in 2004 before the first lactate injection. At that time,
tetrachloroethene (PCE), trichloroethene (TCE) and their daughter products were present at a maximum summed
concentration of approximately 0.4 milligrams per liter (mg/L) in the groundwater. Saprolite in the area of the most
highly contaminated groundwater had been dewatered by groundwater extraction; thus, baseline data were
unavailable for saprolite monitoring wells in this area. Baseline monitoring of key geochemical indicators indicated
that no wells in the area of significant groundwater contamination were in an optimal environment for reductive
dechlorination.
Seven lactate injections were performed beginning
in late 2004 and ending in spring 2012 (Figures 2
and 3). Initially, lactate was injected in 14 wells that
were used previously for groundwater or dual-
phase recovery. Overtime, wells were added to the
injection program and some of the initial injection
points were omitted from some injection events.
Both the lactate injection and post-injection
monitoring timing and location and the magnitude
of lactate injections were adjusted to fit observed
groundwater quality responses to previous
injections.
In September 2007, after more than a year
following the fourth injection, groundwater
monitoring showed that geochemical conditions
were substantially improved relative to the
baseline. Over half of the bedrock or
saprolite/transition zone/bedrock monitoring wells
and injection points in the area of significant contamination showed dissolved oxygen levels ranging from non-
detect to less than 0.5 mg/L, and a groundwater pH that was much more uniform (clustered around 6 to 7) than at
baseline. Monitoring showed that PCE and TCE were no longer the principal chlorinated ethenes at key
monitoring and injection wells; instead, c/s-1,2-dichloroethene (DCE) and vinyl chloride predominated at many
monitoring points (Figure 4).
Figure 2. Lactate injection at the Medley Farm site
-------
approximate scale, feet Q monitoring point />\ drum/waste
O lactate solution injection point \$ disposal area
well-specific lactate injection
volume per event, gallons
total lactate injection
volume per injection
event (all wells), gallons
400000
200000
llil
S*o 10
po
s
i
Figure 3. Lactate injection history at the Medley Farm site
September 2004 (baseline data)
cisl,2-DCE
D vinyl chloride
data obtained
after four
lactate
injections
Figure 4. Chlorinated ethene concentrations in September 2004 versus September 2007, after lour
lactate injections at Medley Farm
-------
Early 2008 review of the data indicated that different areas of groundwater were responding differently to the
lactate injections. In particular, the re-saturated saprolite in the zone of past P&T dewatering was not favorably
responding to the lactate injections. The differences suggested uncertainties regarding hydrogeologic aspects of
the conceptual site model.
In 2010, priorities shifted to address concerns about the lack of groundwater monitoring points between two
principal areas of residual contamination. Other uncertainties concerned the inability to discern any natural
attenuation component to groundwater cleanup and the insufficient understanding of whether MNA could be an
effective remedial strategy for any portion of the site. As a result, three new lactate injection wells were installed
from the lower saprolite through the upper bedrock in the site's more upgradient area of groundwater
contamination and upslope of the area without monitoring wells. Baseline monitoring of the new wells revealed
groundwater quality similar to that observed elsewhere before the initial lactate injection, with minimal c/s-1,2-
DCE and vinyl chloride and substantive concentrations of PCE and TCE.
In the next (March 2012) lactate injection, a bromide tracer was added to the lactate solution injected into two of
the new wells and three of the original injection wells to evaluate the injectate retention time and movement.
Groundwater samples collected approximately nine months later from the three new wells indicated that the
relative amounts of chlorinated ethene parents and daughters had reversed from baseline conditions. Two of
these samples contained chlorinated ethenes at concentrations below groundwater performance standards.
Dissolved oxygen was non-detect in all three samples.
Bromide tracer tests yielded a varied bromide retention pattern at the tracer injection wells and variable rates and
directions of bromide movement. For example, although multi-level sampling in one of the new injection wells
identified inconsequential flushing of bromide out of bedrock, sampling at a new well without bromide injection
found that bromide injected approximately 200 feet upgradient was present three to four months after the
injection. Some of the tracer remained in samples from the new injection wells almost a year after the bromide
injection.
An adaptive site management approach is facilitating necessary monitoring changes and helping identify optimal
timeframes for any future injections.
Monitoring results to date indicate that the current understanding of ERD progress is insufficient to enable
transition to MNA as a groundwater remedy. Primary concerns are the irregular lactate injection and post-injection
monitoring events, complex and incompletely understood groundwater flow paths, variable lactate migration rates,
and absence of monitoring points in key locations. Conditions are either unfavorable or uncertain for continued
reductive dechlorination without further lactate injection at 21 of the 24 evaluated wells completed in the transition
zone and/or bedrock. Saprolite monitoring wells in the core area of residual groundwater contamination have
shown limited response to the injections, and an alternate ERD treatment strategy is needed to address this
contamination.
Based on these findings, EPA is using an adaptive management approach to prioritize future steps and modify
the site's monitoring program. Currently, a sufficient monitoring period without further ERD injections is needed to
avoid confounding interpretation of data collected for MNA assessment. Future injections and associated tracer
tests are anticipated after that period to better target chlorinated ethenes in saprolite. Also, more frequent and
structured groundwater monitoring is anticipated for periods following future injections, in order to more effectively
assess contaminant and geochemical indicator rebound and to let any residual effects of prior lactate injections
dissipate. As the new information becomes available, it will be integrated into the conceptual site model to
facilitate transition from ERD to MNA.
To learn more about recent findings at this site, download the full paper. "Challenges in Planning for Groundwater Remedy
Transition at a Complex Site"
------- |