Office of Solid Waste and EPA-540-R-11 -018
Emergency Response March 2011
(5102G) www.epa.gov/tio
www.clu-in.org/optimization
Remediation System Evaluation (RSE)
Moss-American Superfund Site
Milwaukee, Wisconsin
-------
REMEDIATION SYSTEM EVALUATION
Moss-AMERICAN SUPERFUND SITE
MILWAUKEE, WISCONSIN
Final Report
March 2011
Prepared by:
US Army Corps of Engineers
Environmental and Munitions Center of Expertise
and
Seattle District
Prepared for:
US Environmental Protection Agency
Region 5
US Army US Environmental
Corps of Engineers Protection Agency
-------
Table of Contents
EXECUTIVE SUMMARY ii
1.0 INTRODUCTION 1
1.1 Purpose 1
1.2 Team Composition 1
1.3 Documents Reviewed 1
1.4 Site Location, History, and Characteristics 2
2.0 PERFORMANCE OBJECTIVES 5
3.0 SYSTEM DESCRIPTION 6
3.1 Groundwater Treatment System 6
3.2 Monitoring Program 7
4.0 SYSTEM PERFORMANCE 8
4.1 Groundwater Flow 8
4.2 Groundwater Chemical Concentrations 8
4.3 Treatment Gates 9
5.0 REMEDY OPTIMIZATION OPTIONS 11
5.1 Recommendations to Improve Effectiveness 11
5.2 Recommendations to Improve Site Closeout 12
6.0 SUMMARY 16
List of Tables
1. Groundwater Cleanup Goals
2. Monitoring Program
3. Cost Assumptions
4. Remedy Optimization Options Evaluation Summary
List of Figures
1. Site Location
2. Proposed Additional Monitoring Locations for Chemical Analysis
3. Potential NAPL Investigation Program
List of Appendices
A. Groundwater Modeling Documentation
B. Trend Analyses
C. Effective Solubility of Naphthalene
-------
EXECUTIVE SUMMARY
This document presents the results of a Remediation System Evaluation (RSE) conducted for the Moss-
American Superfund Site in Milwaukee, Wisconsin. The RSE process is designed to help site operators
and managers improve effectiveness, reduce operation costs, improve technical operation, and gain site
closeout. The observations and recommendations given within this RSE report are not intended to imply a
deficiency in the work of either the designers or operators, but are offered as constructive suggestions to
fill data gaps and optimize remedy performance.
This RSE report focuses primarily on optimizing system performance, in particular addressing the
stagnant groundwater zone that is limiting flow through the treatment gates and elevated COC
concentrations in the vicinity of MW-34S. Recommendations include:
• Monitoring program modifications to further delineate source and dissolved-phase contaminant
extent. These modifications would result in additional costs of approximately $22,500. Benefits
include ensuring that contaminants are not migrating through or around the sheet pile wall, as
well as providing necessary information for implementing treatment enhancements, which would
ultimately lead to earlier site closeout.
• Additional NAPL investigation. This investigation would cost approximately $72,000.
Identification of source areas would allow targeted removal, thereby diminishing long-term
contributions to the dissolved-phase plume and shortening time to achievement of cleanup
objectives.
• Depending on results of characterization efforts, it is recommended that one of the following
treatment modifications be implemented:
1) NAPL-impacted soil excavation and enhanced dissolved-phase treatment. This option
would cost roughly $381,000 for the stagnant zone near MW-34S; costs for similar work
near TG1-1 have not been developed but could be readily scaled from the estimate for the
MW-34S area based on results from field investigations. Aggressive removal of identified
source material (NAPL) and subsurface amendments of ORC Advanced® would greatly
shorten time until achievement of cleanup objectives.
2) Limited NAPL-impacted soil removal and installation of additional gate in NW corner.
Costs for this option are estimated to be roughly $979,000. This option adheres closely to
the original design, which included a gate in the northern portion of the sheet pile wall.
Installation of a gate in the wall should improve flow and eliminate the stagnant zone,
thereby resulting in more effective treatment of the dissolved-phase plume. Risk
management and design considerations would determine whether the gate is installed near
MW-34SorMW-7S.
-------
1.0 INTRODUCTION
1.1 Purpose
The Remediation System Evaluation (RSE) as identified in the U.S. Army Corps of Engineers
(USAGE) Guidance is intended to achieve a number of goals, including:
• Assuring there is a clear system objective (an end to the project),
• Reducing costs and optimizing the system performance considering current conditions and new
technologies,
• Evaluating the protectiveness of the system in accordance with the National Contingency Plan
(the NCP and CERCLA requires reviews at least every five years), and
• Assuring adequate maintenance of government-owned equipment by operators, [not directly
applicable to this RP-run system]
The Third Five-Year Review Report (EPA, 2010) concluded that the site is currently protective, but
recommended that an optimization study be performed "to develop a solution to remediate the elevated"
contaminant of concern (COC) levels found in areas within the funnel and gate system. Due to
development of stagnation in groundwater flow and resulting reduction in flow through the treatment
gates, these elevated COC levels persist, with consequences for long-term operations and overall costs.
Because a site visit was not included in the scope for this study, the focus of this RSE was directed at
optimizing system performance, with the intent of ensuring cleanup objectives can be reached within a
reasonable timeframe, thereby reducing long-term costs. This report provides a brief background on the
site, current operations, and recommendations for changes and additional actions. The cost impacts of the
recommendations are also discussed.
1.2 Team Composition
This team conducting the RSE consisted of Mike Bailey (hydrogeologist, USAGE Environmental &
Munitions Center of Expertise), Mandy Michalsen (engineer, USAGE Seattle District), and Sharon
Gelinas (hydrogeologist, USAGE Seattle District).
1.3 Documents Reviewed
Remedial Investigation Report, Moss-American Site, January 9, 1990
Superfund Record of Decision (ROD), Moss-American Co., Inc, USEPA, September 27, 1990
Explanation of Significant Differences (ESD), Moss-American Co., Inc, USEPA, April, 29, 1997
Superfund ROD Amendment, Moss-American Co., Inc, USEPA, September 30, 1998
ESD, Moss-American Co., Inc, USEPA, November 2007
Third Five-Year Review Report for Moss-American Superfund Site, USEPA, April 2010
Groundwater Monitoring Reports for the Moss-American Site from 1998-2008, Roy F. Weston, Inc
(Weston)
Groundwater Remedial System Drawings, Weston , Kerr-McGee Corporation, March 1998
1
-------
Response to Comments on Focused Remedial Alternatives Evaluation for Soil and Sediment, Moss-
American Site, Weston, January 12, 1996
Integrated Review Comments of Soil and Groundwater Remedy, Moss-American Site, Weston, January
20,1997
Response to Comments on Intermediate (60%) Groundwater Design, Moss-American Site, Weston,
February 3, 1997
Comments on Prefmal Design - Groundwater, Moss-American Site, USEPA, October 30, 1997
Supplemental GeoProbe Soil Investigation Report, Moss-American Site, Weston, May 2, 2001
1.4 Site Location, History, and Characteristics
1.4.1 Location
The Moss-American site is located in the northwestern section of the City of Milwaukee (Figure 1). The
88-acre site is comprised of a former wood treating facility plus several miles of the Little Menomonee
River and its adjacent floodplain soils. The wood treating, using creosote, was conducted on land bounded
roughly by the intersection of Brown Deer and Granville Roads on the west, and Brown Deer and 91st
Street on the east.
With the cessation of wood treating operations, 23 acres of site land are now owned by the Union Pacific
Railroad (railroad), which, until very recently, used this land as an automobile/light truck loading and
storage area. Recent business conditions curtailed most of the vehicle storage/transfer function. Industrial
site zoning and usage of this portion of the site remain intact. Milwaukee County (the county) owns the
remainder of the land comprising the former wood treating facility, approximately 65 acres.
The Little Menomonee River flows approximately 5 miles to its confluence with the Menomonee River.
Land along the floodplain corridor is owned primarily by the City of Milwaukee, the County, and to a
much lesser extent, private owners.
1.4.2 History
Wood treating operations using creosote were conducted from approximately 1921 to 1976. Past site
aerial photos show that land usage patterns have changed considerably with the passage of time. Photos
from the 1930s to the 1950s show the wood treating plant operating in a relatively sparsely populated
setting, where several farms surrounded the manufacturing operation. From the 1960s to the present,
residential and commercial use of nearby property has increased considerably, and agricultural and
farming operations have been phased out almost completely. Industrial parks and multi-lane highways
also traverse the site setting. County owned land along the river corridor now features recreational hiking
and bicycle trails. These features have had a direct bearing on site soil cleanup standards and sediment
management at the site.
In 1921, the T. J. Moss Tie Company established a wood preserving facility west of the Little
Menomonee River. The plant preserved railroad ties, poles, and fence posts with creosote, a mixture of
numerous chemical compounds derived from coal tar. Creosote plant operations often contain storage
facilities for creosote and fuels, a boiler for making steam, heating the creosote and applying the creosote
-------
to the wood, areas for unloading and storing incoming timbers, rail cars for transporting the creosote, and
a drying area for subsequent storage. Creosote is the major source of a class of contaminants called
poly cyclic aromatic hydrocarbons (PAHs) which are the main driver of risk at this site. Potential for
release of PAHs existed throughout the storage, application, and drying processes.
From 1921 to 1971, the facility discharged wastes to settling ponds that ultimately discharged to the Little
Menomonee River. These discharges ceased when the plant diverted its process water discharge to the
Milwaukee sanitary sewerage system. Production at the facility ceased in 1976.
Kerr-McGee purchased the facility in 1963 and changed the facility's name to Moss-American. The name
was changed again in 1974 to Kerr-McGee Chemical Corporation - Forest Products Division. In 1998, the
name of this company changed to Kerr-McGee Chemical LLC (KMC). Tronox assumed ownership of the
site in 2006 when it was spun off from Kerr-McGee. In January 2009, Tronox filed for Chapter 11
bankruptcy.
1.4.3 Hydrogeology Setting
The site overlies a surficial water-bearing unit and confining bed. The water-bearing unit consists of a
thin mantle of fill, alluvium, and weathered till. This thin layer of material would not yield sufficient
water to wells to be classified as a true aquifer. The confining bed is the unweathered till of the Oak
Creek Formation.
The surficial unit comprises everything above the confining bed. It includes extensive fill deposits,
alluvial deposits along the river, and the weathered upper few feet of the Oak Creek Formation. The fill is
highly variable and has been added to the site at different times for different reasons. Alluvial deposits are
associated with the Little Menomonee River. They consist of sand and gravel channel deposits and silt
and clay flood deposits. The till is part of the Oak Creek Formation, which consists of glacial till,
lacustrine clay, silt and sand, and some glaciofluvial sand and gravel. The till is fine grained, commonly
containing 80 to 90 percent silt and clay. The till was generally weathered to a depth of 2 to 10 feet.
The unweathered part of the Oak Creek Formation consists of a confining bed between the surficial
water-bearing unit and underlying regional aquifers. The formation is a dense, silty clay till with
interbedded lacustrine units. Below the site, the glacial deposits are approximately 150 feet thick and
underlain by the dolomite aquifer. The minimum thickness of the confining bed below the site is at least
40 feet. Slug tests conducted during the RI on the most permeable parts of the Oak Creek Formation
indicate average hydraulic conductivities of 10"5 to 10"6 cm/s [0.03 to 0.003 feet per day (ft/day)]. The
overall hydraulic conductivity of the entire unit is probably less than the values reported.
Prior to implementation of the remedy, groundwater flowed toward the low-lying areas adjacent to the
river. Groundwater discharged to these areas either migrates downriver through alluvial sands, or is lost to
the atmosphere by evapotranspiration. Groundwater and surface water elevation data suggest that
discharge to the river may vary seasonally. During dry periods, the Little Menomonee River is probably a
losing stream (the river discharges to groundwater). Conversely, during wetter conditions, it is likely a
gaining stream.
Constrained and channeled by the funnel and gate system, the groundwater within the shallow
groundwater-bearing zone generally flows northeastward toward the Little Menomonee River. A review
-------
of data presented in the quarterly and annual groundwater monitoring reports by Weston indicate that in
the topographically higher (western) portion of the site, the horizontal hydraulic gradient is relatively
steep, at approximately 0.032 feet per foot (ft/ft) to the northeast. The topography of the site levels out
near the river, as does the potentiometric surface with a northerly hydraulic gradient of approximately
0.013 ft/ft. The estimated hydraulic gradients within the treatment gates ranged from 0.0007 to 0.0043
ft/ft. The hydraulic gradient is relatively flat within the treatment gate area with an overall hydraulic
gradient from TGI to TG5 of approximately 0.0026 ft/ft in an easterly direction. Lowest hydraulic
gradients are found in the area encompassing monitoring wells MW-7S, MW-33S, MW-34S, and MW-
38S.
The hydraulic conductivity of the deposits located on the topographically higher, western portion of the
site is in the range of 10"5 to 10"6 cm/s. In contrast, the hydraulic conductivity of material used to backfill
areas within the funnel and gate remedial system is approximately 10"3 cm/s (3 ft/day). Using a hydraulic
gradient of 0.032 ft/ft, an assumed effective porosity of 0.3, and a hydraulic conductivity of 0.03 ft/day,
the groundwater flow velocity in the western portion of the site is calculated to be approximately 0.0032
ft/day. Near the river, using a hydraulic gradient of 0.013 ft/ft, a porosity of 0.3, and a hydraulic
conductivity of 3 ft/day, the velocity of groundwater flow is calculated to be approximately 0.13 ft/day.
The groundwater flow velocities within the treatment gates are estimated to range from 0.0066 to 0.1049
ft/day.
1.4.4 Description of Groun dwater Plume
Historically, non-aqueous phase liquid (NAPL) has been identified in monitoring wells MW-34S, MW-
7S and TG1-1. Recent NAPL occurrences in these wells have been limited to observations of sheen. The
current dissolved-phase plume boundary is primarily in an area encompassing monitoring wells MW-7S,
MW-33S, MW-34S, and MW-38S (Figure 2), which coincides in large part with the groundwater
stagnation zone. There are also exceedances of State groundwater standards at MW-35S and treatment
gate wells TG1-1, TG2-3 and TG4-1. In general, PAH concentrations measured in groundwater samples
collected from the rest of the site were at relatively low levels with only sporadic detections.
Monitoring well MW-34S exceeds cleanup standards for numerous contaminants of concern including
anthracene, benzene, benzo(a)pyrene, benzo(b)fluoranthene, chrysene, fluoranthene, fluorene,
naphthalene, and pyrene. Monitoring well MW-7S exceeds standards for benzene and naphthalene,
although trends for both contaminants are decreasing. In addition, increasing concentrations are
identified for several COCs at these, and other, wells. Statistical analysis by EPA Region 5 indicates that
multiple PAH contaminant concentrations are increasing, with current concentrations higher than the
period just after construction of the funnel and gate system. Monitoring well MW-33S continues to
exceed standards for naphthalene. Current contaminant concentrations from well MW-33S are also higher
for anthracene and fluorene than they were shortly after implementation of the remedy.
-------
2.0 PERFORMANCE OBJECTIVES
The focus of this RSE was on the groundwater remedy; the soil and sediment remedies were not
evaluated. Groundwater remediation goals were to prevent migration of contaminated site groundwater
into the Little Menomonee River and to attain concentrations in NR 140 of the Wisconsin Administration
Code for COCs at the site. Groundwater contaminants of concern and their associated State preventative
action levels (PAL) are listed in Table 1.
The remedial action objective (RAO) for groundwater as stated in the ROD was to: Prevent release of
contaminants through the surficlal groundwater aquifer to the Little Menomonee River surface water or
sediment and remove contaminants from groundwater such that concentrations don't exceed applicable
State groundwater standards.
-------
3.0 SYSTEM DESCRIPTION
The groundwater remedy consisted of a funnel and gate system to capture and treat contaminated
groundwater prior to discharge to the Little Menomonee River. The following section provides a
description of the groundwater treatment system and associated monitoring program.
3.1 Groundwater Treatment System
A funnel and gate system was selected as the preferred alternative in the 1997 BSD. Pre-design results
indicated that the relatively fine-grained site sediments would be well suited for this type of system.
Groundwater flow was relatively uniform toward the Little Menomonee River with discontinuous zones
of increased permeability (i.e. gravel fill and silty sand) acting to guide the direction of the contaminant
plume. In the BSD, groundwater was predicted to move slowly through the treatment gates, which would
provide adequate residence time for contaminant treatment.
The funnel and gate system is constructed of Waterloo sheet piling, which has an internal cavity scalable
joint. This type of joint reduces the potential for leakage of contaminants through the joints. Early
designs (60%) of the funnel and gate system showed two sets of funnel and gates: two gates on an upper
funnel and three gates on a lower funnel located adjacent/parallel to the river. Installation was proposed
in a phased approach. The upper funnel and gates would be installed and tested for performance. The
lower funnel and gates, which had a higher potential to negatively impact the river, would then be
installed following verification of the upper funnel and gate performance. This phased approach was not
approved by the regulators because contaminants adjacent to the river would continue to be discharged
during the test performance period.
The final design of the funnel and gate system changed the lower funnel and gates to a sheet pile
containment wall with two sets of funnel/treatment gates to the east. Using this design, the entire system
could be installed at one time and the potential for untreated contaminants reaching the river would be
reduced. In considering the design change for the final funnel and gate system, it is uncertain if this
system was thought to be capable of mobilizing contaminants located in the northwest corner of the sheet
pile area toward the eastern gates for treatment. A groundwater model was reportedly developed for the
60% design, but was not available for review during this RSE.
The treatment gates consist of an area backfilled with a mixture of clean sand/soil and line of injection
wells. The injection wells were installed at the up-gradient edge of the gate area and were designed to
distribute air or other nutrients, as necessary. NAPL collection sumps were installed up-gradient of the
gates to prevent potential plugging and/or treatment performance problems.
Treatment at the gates consists of air injection to enhance biodegradation of COCs. Dissolved oxygen
concentrations in the gate area have been measured at less than 1 to over 4 mg/L. Well packers were
installed at Gate 5 in June 2000 to help direct the air injection; however, no discernable changes in
dissolved oxygen levels were observed until 2003. Packers were also proposed at Gates 1 and 2, but
could not be properly installed. Nutrients were added at Gate 1 from June 2001 through October 2002
using a solution containing potassium nitrate (KNO3) and potassium phosphate (KHPO4). Nutrient
augmentation was discontinued due to inconclusive evidence that it was enhancing biodegradation. Air
injection has been the only treatment since that time.
-------
3.2 Monitoring Program
Performance monitoring for the funnel and gate system consists of an evaluation of groundwater
hydraulics and groundwater chemical analyses. The groundwater monitoring program has been revised
several times, most recently in 2006/2007. During this last revision, twenty-two monitoring wells and
piezometers across the site that were no longer sampled were abandoned. In addition, two monitoring
wells were installed within the northwest area of the sheet pile for the funnel and gate system.
Monitoring wells currently sampled as part of the monitoring program are shown in Table 2. All of the
wells and piezometers are screened in the shallow groundwater-bearing zone underlying the site (surficial
aquifer).
Water level measurements are collected on an annual basis at all monitoring wells and piezometers at the
site to evaluate groundwater hydraulics. Chemical analyses are collected annually except at monitoring
wells MW-7S, MW-34S, MW-38S, and MW-39S, where samples are collected semi-annually.
Piezometers installed in 2002 and the middle performance monitoring well at each gate are not included
in the chemical monitoring program. In addition to the on-site monitoring wells listed in Table 2, 11
shallow groundwater monitoring wells (MW-A through MW-K) located along the Little Menomonee
River are sampled to monitor groundwater chemical conditions between the old and new river channels.
Analytical parameters collected at each well include benzene, toluene, ethylbenzene, and xylene (BTEX),
poly cyclic aromatic hydrocarbons (PAHs), and field parameters: pH, oxidation-reduction potential,
dissolved oxygen, specific conductance, temperature, and turbidity. Samples collected at the treatment
performance monitoring wells at each gate also are analyzed for microbial enumeration, nitrate-nitrogen
(NO3-N), nitrite-nitrogen (NO2-N), total Kjeldahl nitrogen (TKN), ammonia-nitrogen (NH3-N),
phosphate-phosphorous (PO4-P), orthophosphate (ORP), biological oxygen demand (BOD), chemical
oxygen demand (COD), and total organic carbon (TOC).
-------
4.0 SYSTEM PERFORMANCE
4.1 Groundwater Flow
Groundwater elevation data collected since the funnel and gate system was completed in 2000 were
reviewed to evaluate flow through the system. Groundwater at the site generally flows from south to
north toward the Little Menomonee River. Due to the presence of the sheet pile wall along the north and
west portion of the system, groundwater is directed toward the eastern treatment gates.
The groundwater flow evaluation indicates that there are several areas of concern where groundwater may
not be hydraulically contained or treated by the gates:
• Groundwater flow maps consistently indicate the presence of a stagnation zone in the northwest
corner of the sheet pile area near MW-34S and MW-7S. Groundwater elevation data show that
there is only a very slight gradient between these two wells. The boring log for MW-7S indicates
the surficial aquifer in this area is composed of low permeable materials (very fine sand and silt),
which, coupled with the low gradient, would result in a very low groundwater velocity. The
borelog for MW-34S was not available for review.
• Groundwater elevation data at MW-33S and PZ-02 indicate that groundwater may be flowing
around the end of the sheet pile wall. A head difference of about 0.5 feet is typically measured
between MW-33S and PZ-02. Borelogs for these two wells were not available for review.
• Groundwater elevation data from performance wells at gates 1,3, and 4 frequently show the
gradient is reversed (flowing from down-gradient of the gate toward the up-gradient side). It
should be noted that the magnitude of the calculated gradient is very low, so the possibility of
measurement error (i.e water levels, top of casing survey) should also be considered.
Two monitoring wells, MW-38S and MW-39S, located near the groundwater stagnation zone, were
installed in 2006 to help delineate the remaining dissolved-phase plume in the northwestern portion of the
system. These wells were never surveyed and have never been used in the preparation of groundwater
flow maps. These wells could be surveyed and used in future construction of groundwater flow maps to
help evaluate groundwater flow across the site.
4.2 Groundwater Chemical Concentrations
Contaminants in groundwater are consistently detected above cleanup goals in two areas: 1) in the
northwest section of the sheet pile area in the groundwater stagnation zone at monitoring wells MW-7S,
MW-33S, MW-34S, and MW-38S, and 2) up-gradient of Gate 1 in TG1-1.
4.2.1 Contaminant Concentrations in Northwest Corner of Site
Trend analyses for the most prevalent contaminants (benzene, naphthalene, fluorene, and benzo(a)pyrene)
show that there are decreasing trends or no trends for wells in the northwest corner (Appendix B). Trend
testing results confirmed decreasing naphthalene concentrations in MW-7S and MW-38S and decreasing
benzene concentrations in MW-7S, indicating that natural attenuation is occurring in these areas.
However, these trends cannot be used in a predictive sense, because overall trends indicate that PALs
-------
should have been achieved within the past year or two. Instead, recent sampling results suggest that
trends may be asymptotically "bottoming-out."
Measurable NAPL has historically been detected at MW-34S. In 2008, 3.24 inches of NAPL was
measured. Since that time measurements have decreased to trace detections, although dissolved-phase
concentrations of naphthalene continue to exceed 10,000 (ig/L (September 2009 data). Given high
dissolved-phase PAH concentrations and typical inaccuracies with NAPL measurements, it is assumed
that some NAPL remains in the vicinity of MW-34S and could be a continued source to the dissolved-
phase plume. It should also be noted that the soil excavation completed during the installation of the
funnel and gate system only occurred to the southeast of MW-34S and did not extend into the current
dissolved-phase plume area (see Groundwater Remedial System drawings, March 1998). Presence of
NAPL and the development of a stagnation zone in the funnel and gate system have the potential to
greatly extend time to restoration.
Besides the extended time to restoration, there are several potential issues with the remaining dissolved-
phase plume. As suggested in the 2010 Five-Year Review, the pattern of water levels near MW-7S/MW-
34S could indicate that the sheet pile barrier to the north does not form a sufficiently competent barrier to
groundwater flow. Thus, contaminated groundwater could be flowing through joints in the sheet pile wall
near MW-34S and discharging to the river. In addition, the flow evaluation indicated that groundwater
has been moving around the end of the sheet pile wall near MW-33S. Since there are no chemical
samples collected north of the sheet pile wall, contamination migration along this pathway cannot be
ruled out.
4.2.2 Contaminant Concentrations Up-gradient of Gate 1
Concentrations of benzene and PAHs in groundwater are typically measured above PALs at up-gradient
performance monitoring well TG1-1. Trend tests show concentrations of naphthalene, fluorene, and
benzo(a)pyrene have been increasing, indicating a continued source of contamination in this area
(Appendix B). NAPL was historically detected in TG1-1 up to 11 inches thick; however, only trace or
sheen thickness has been observed since 2003. As with MW-34S, naphthalene concentrations in TG1-1
currently exceed 10,000 (ig/L (September 2009 data), which suggests that a NAPL source persists in the
area. Since the extent and magnitude of the remaining contamination in soil and groundwater near Gate 1
is uncertain and contaminant concentrations continue to rise, time to restoration cannot currently be
estimated. Most of the monitoring wells used to define the historical extent of the groundwater
contamination near Gate 1 have been abandoned. However, there are several piezometers used only for
hydraulic monitoring near Gate 1 that could be sampled to help delineate the remaining dissolved-phase
plume.
4.3 Treatment Gates
With the exception of Gate 1, contaminant concentrations up-gradient and down-gradient of the treatment
gates indicate that much of the historical groundwater contamination has been removed. Several PAHs
(benzo(a)pyrene, benzo(f)fluorene, and chrysene) are sporadically detected above PALs in monitoring
wells near Gates 3 and 4, however, concentrations are low, just above the cleanup goal of 0.02 (ig/L.
Even with the potential gradient reversal at Gates 3 and 4, the treatment gates appear to be functioning
adequately.
-------
The only gate area with significant remaining contamination is Gate 1. Even though groundwater
concentrations are elevated at TG1-1, there are typically no detections of PAHs in the down-gradient
performance monitoring well, TGI-3. Oxygen levels measured in Gate 1 are also low, signifying that the
injected oxygen is being consumed, and the gate is functioning adequately.
10
-------
5.0 REMEDY OPTIMIZATION OPTIONS
Previous assessments in annual reports and Five-Year Reviews determined that the existing runnel and
gate remedy was having limited success in the northwest corner of the site due to development of a
stagnant zone in groundwater. Investigations recommended to ensure effectiveness of the remedy and to
inform decisions about ways to improve effectiveness and shorten time to site closeout are discussed
below (Section 5.1). Section 5.2 evaluates three options to hasten site closeout through source removal
and/or groundwater gradient enhancements.
5.1 Recommendations to Improve Effectiveness
5.1.1 Monitoring Program Modification
The primary areas of concern for the monitoring program are the lack of chemical data outside the sheet
pile wall near MW-7S and MW-34S, where there is a possibility that contaminants could be passing
through the joints or migrating around the end of the wall, and the extent of remaining contamination near
TG1-1. A secondary area of concern is the extent of the dissolved-phase plume in the interior of the
funnel and gate system. The following enhancements to the monitoring program are recommended (see
Figure 2 for we 11 locations):
• Install two monitoring wells outside the sheet pile wall to the north of MW-34S and to the west of
MW-7S to determine if contaminants are migrating through the sheet pile wall.
• Develop and sample piezometer PZ-02 to determine if contaminants are migrating around the end
of the sheet pile wall.
• Develop and sample piezometers PZ-07, -09, and -10 to determine the up-gradient extent of
remaining contamination near TG1-1.
• Develop and sample piezometer PZ-03 to confirm the extent of the dissolved-phase plume in the
interior of the funnel and gate system.
• Survey MW-38S and MW-39S and include water levels from these wells in groundwater flow
maps.
Costs for modifying the monitoring program include $13,100 for the installation and development of two
monitoring wells (includes oversight and reporting) and $5,000 for development of five existing
piezometers. Prior to development of the piezometers, their construction should be verified (i.e. depth,
well screen interval). Additional costs of about $5,900 for labor and laboratory analysis would also be
accrued during each sampling event. Costing assumptions are described in Table 3. If contaminants are
not detected in new monitoring locations after four sampling events, the wells/piezometers could be
dropped from the program.
5.7.2 NAPL Investigation
Removal of residual NAPL in areas near MW-34S and TG1-1 would eliminate this continued
contaminant source to the dissolved-phase plume and shorten time to site closeout. A localized direct
push soil and groundwater investigation could be implemented to spatially delineate residual NAPL
contamination in these areas. NAPL is likely not uniformly distributed in site soil, which means absence
of NAPL in a particular soil boring would not necessarily preclude NAPL presence in nearby soil. In
order to improve NAPL delineation during the investigation, grab groundwater samples could be
collected by the direct push rig during completion of soil borings. Groundwater samples with
11
-------
naphthalene concentrations approaching 9,100 (ig/L1 would indicate NAPL presence in the vicinity of the
soil boring. A schematic of a potential NAPL investigation program is provided on Figure 3. Locations
where NAPL presence, soil concentrations or groundwater naphthalene concentrations greater than 9,100
(ig/L were detected would be considered for inclusion in an excavation footprint. This investigation for
each area would cost an estimated $36,000 based on assumptions described in Table 3.
5.2 Recommendations to Improve Site Closeout
Remedy optimization options were developed primarily to address the elevated COC concentrations in
the vicinity of MW-34S and the stagnant groundwater zone that is limiting flow through the treatment
gates. Because treatment at Gate 1 is currently effective and the remedy is functioning as intended, future
work to shorten time to site closeout in that area is discretionary and of secondary importance to work in
the MW-34S area. Consequently, costs for enhancements to the remedy near Gate 1 have not been
developed but should be readily scalable from those for the MW-34S area. Implementation of these
options would be influenced by the results of investigations discussed in Section 5.1.
Options were evaluated for effectiveness using a simplified numerical groundwater model and by
considering implementability, and if applicable, cost (Table 4). It should be noted that a more robust
numerical model would likely be needed if the selected remedy optimization includes significant
modifications to the groundwater flow system, such as with the installation of a new gate or extraction
wells. For those options which were deemed technically ineffective or for which there was insufficient
site information, costs have not been developed and are not presented herein.
The groundwater model was designed to simulate groundwater flow only in the vicinity of the funnel and
gate system and was calibrated to water level data collected during the 3rd quarter of 2009. Details on the
model setup, calibration, and results are presented in Appendix A. The following simplifying
assumptions were utilized:
• The flow system is steady state,
• The surficial unit (shallow aquifer zone) is uniformly 15-feet thick,
• The topographically higher, western portion of the site has a lower hydraulic conductivity than
the topographically lower portion within the funnel and gate system, and
• The sheet pile barrier has a bulk hydraulic conductivity of 1 x 10"7 cm/s.
5.2.1 NAPL-Impacted Soil Excavation and Enhanced Dissolved-Phase Treatment
Locations identified during the NAPL investigation where NAPL presence, soil concentrations or
groundwater naphthalene concentrations representing a significant percentage of the solubility level were
detected could be considered for inclusion in an excavation footprint. We have assumed that an area
centered around MW-34S extending 50 ft from the wall and 75 ft along the wall would be included in the
excavation footprint (Figure 3). Excavation costs near TG1-1 are not included but could be scaled from
MW-34S, depending on the results of field investigations. Based on current data, it is believed that
excavation near TG1-1 would be less extensive than near MW-34S and costs proportionally lower.
1 Estimated effective naphthalene groundwater water solubility in presence of NAPL calculated assuming a typical creosote
composition; calculations are included in Appendix C for reference.
12
-------
Available boring logs2 for nearby wells MW-7S and MW-39S indicate that depth to the confining clay
layer is 10-12 feet bgs. An average depth of 15 feet has been assumed for the thickness of the surficial
unit in the numerical groundwater model, so this excavation depth was assumed as well. A lined staging
and dewatering area for excavated soil could be prepared near the excavation pit and could be sloped to
allow dewatering water to collect in the excavation pit. A sump could be included to capture any product
seeping from the dewatering water. Groundwater could be allowed to accumulate in the excavation pit,
the bottom of which could be sloped to function as a sump as well. Any accumulated product in the
excavation could be removed by pumping. Excavation, materials, handling and associated activities
would cost an estimated $202,000 based on assumptions described in Table 3.
Although the final depth of sheet pile wall installation into the clay layer is not known, preliminary design
documents indicate a target final depth of 3 ft below the clay layer surface, i.e. a final sheet pile wall
depth of- 18 ft bgs. Because the sheet pile wall will function as a retaining wall during excavation, and
the engineering rule for minimum wall depth is 2x the excavation height, the wall section adjacent to the
excavation area will need to be improved to safely meet depth requirements. Assuming a 15 ft
excavation, the required improved sheet pile wall depth in this area would be 50 ft bgs. Materials and
installation for the improved 50 ft x 75 ft section of sheet pile wall would cost an estimated $94,000 based
on assumptions described in Table 3.
Oxygen Releasing Compound Advanced (ORC Advanced®) could be incorporated into the excavation
backfill to enhance biodegradation of dissolved-phase contaminants in both the excavation and
groundwater. Because molecular oxygen would subsequently diffuse into groundwater surrounding the
ORC Advanced® amended backfilled area, biodegradation of dissolved-phase contaminants would be
enhanced in surrounding groundwater as well. The groundwater model also showed that there would be
some localized groundwater flow into the ORC backfilled area (Figure A-4).
ORC Advanced® is a proprietary formulation of food-grade, calcium oxy-hydroxide that produces a
controlled release of molecular oxygen for a period of up to 12 months upon hydration by groundwater3
and has been demonstrated to enhance treatment of PAHs4 and benzene5 in groundwater. The
recommended application rate for ORC Advanced® is 0.1-0.3 percent by weight of excavated soil.
Approximately 5.2 tons of ORC Advanced® would be required for an excavated soil mass of 2,600 tons6,
which would cost an estimated $86,000 based on assumptions described in Table 3.
Total cost for this option, assuming excavation only in the MW-34S area, would be approximately
$381,000. In addition, limited design work not included in this estimate may be necessary for sheet pile
shoring and excavation.
2 The MW-34S boring log was not available during our analysis.
3 Information for ORC Advanced is available online: http://www.regenesis.com/contaminated-site-remediation-
products/enhanced-aerobic-bioremediation/orc-advanced/
4 Koenigsberg, S. and Sandefur C. The Use of Oxygen Release Compound for the Accelerated Bioremediation of
Aerobically Degradable Contaminants: The Advent of Time-Release Electron Acceptors. (1999, Winter)
Remediation. 6(4), 3-29.
5 Bianchi-Mosquera, G. C., Allen-King, R. M, Mackay, D. M. Enhanced Degradation of Dissolved Benzene and
Toluene Using a Solid Oxygen-Releasing Compound. (1994, Winter). GWMR X(X), 120-128.
6 Assumes excavation volume of 2083 cy and bulk density of 1.26 ton/cy.
13
-------
Despite evidence for decreasing trends in some wells, groundwater in the vicinity of NAPL-impacted
wells MW-34S and TGI will likely not attenuate within a reasonable timeframe. Targeted NAPL
removal in these areas followed by addition of ORC Advanced® would enhance dissolved-phase
attenuation in the TGI and MW-34S areas and decrease restoration timeframes in nearby wells MW-7S
andMW-38Saswell.
5.2.2 Limited NAPL-impacted Soil Removal and Installation of Additional Gate in NW Corner
The installation of a new treatment gate with air injection system in the northwest corner of the sheet pile,
similar to the original design concept, could also be adopted. A new gate would increase the hydraulic
gradient in the NW corner and eliminate the stagnation zone and the potential for groundwater to flow
around the end of the sheet pile, as well as provide long-term treatment for any remaining dissolved-phase
contaminants. Excavation of NAPL-containing soils near MW-34S could be conducted in conjunction
with the installation of the gate system, thereby potentially eliminating the need for structural sheet pile
during excavation as discussed in Section 5.2.1.
Two gate scenarios were evaluated: one installed to the north of MW-34S and one installed to the west of
MW-7S. Both scenarios include limited excavation of NAPL-containing soil near MW-34S that is easily
accessible without requiring reinforcement of the sheet pile wall. The groundwater model shows that if a
new gate is installed to the north of MW-34S, the majority of groundwater flow from the upper treatment
gates (Gate 1 and 2) would be directed toward the new gate (Figure A-8), eliminating the stagnation zone.
Potential issues with installation of this gate include the proximity to the river, slope stability issues and a
limited buffer zone between the treatment gate and the river. Concern about contaminant discharge to the
river from the treatment gate should be alleviated by performance data from existing gates. Engineering
complications associated with proximity of the river would have to be resolved during design.
A new gate to the west of MW-7S could also induce groundwater flow in the area of the stagnant
dissolved-phase plume. The groundwater model shows that groundwater from Gates 1 and 2 would
continue to flow toward the eastern treatment gates and groundwater within the dissolved-phase plume
would flow toward the new gate near MW-7S. Costs for either gate scenario would total approximately
$979,000. These costs do not include additional modeling or design work that may be necessary,
especially if proximity to the river requires special design considerations.
It should be noted that a gate near NW-34S is preferred over one near MW-7S for hydraulic reasons,
because it does a better job of improving flow through the stagnant zone. However, risk management and
design considerations may make a gate near MW-7S preferable.
5.2.5 Groundwater Flow Modification to Enhance Treatment of Existing Funnel & Gate System
Groundwater flow modifications using the existing funnel and gate configuration could be implemented
to induce a hydraulic gradient across the site and eliminate the zone of stagnation in the northwest corner.
Excavation of NAPL-containing soils around MW-34S could also be conducted in conjunction with the
flow modifications as described in Section 5.2.1.
Two model scenarios were evaluated: 1) installation of extraction wells down-gradient of Gates 5 and 6
and 2) installation of a large scale re-circulation cell that includes an injection well near MW-7S and an
extraction well down-gradient of Gate 5. The groundwater model shows that even with extraction wells,
the groundwater stagnation area may still exist (Figure A-6). The extraction wells induce a slight gradient
14
-------
across the site as there is a reduction in flow around the end of the sheet pile near MW-33S. Due to the
low permeability soils, groundwater extraction rates were predicted to be less than 1 gpm. Since the
gradient across the site would still be very low, it could take over 30 years for contaminated groundwater
near the stagnation zone to reach the eastern treatment gates.
The groundwater model shows that with a large scale re-circulation cell groundwater within the
stagnation zone would flow toward the eastern treatment gates; however, there could be increased flow
around the end of the sheet pile near MW-33S due to mounding effects (Figure A-7). Again, the low
permeability materials would limit the extraction/injection rates. When compared to the extraction well
scenario, the gradient across the site is increased, but it could still take over 20 years for contaminated
groundwater near the stagnation zone to reach the eastern treatment gates. In addition, such flow
modification would encourage contaminated groundwater flow into areas that currently contain low-level
contamination, thereby potentially increasing the volume of groundwater contaminated above cleanup
levels at the site.
Planting poplar trees by the final gate pairs has also been proposed in lieu of extraction wells to induce a
gradient across the site. In addition to the low gradient issues stated above, poplar trees would only have
a seasonal influence on the water levels at the site. Also rejected as ineffective was extension of the sheet
pile wall near MW-33S. Preliminary modeling showed no improvements to flow in the stagnant zone.
Due to problems associated with persistence of the stagnation zone, sheet pile wall bypassing due to
groundwater mounding, and excessive transport times to reach treatment gates, manipulations to
hydraulic gradients (in the context of the existing funnel & gate system) are of questionable effectiveness.
Costs were not developed for these scenarios due to perceived ineffectiveness at achieving desired results.
15
-------
6.0 SUMMARY
The observations and recommendations contained in this report are not intended to imply a deficiency in
the work of either the designers or operators, but are offered as constructive suggestions to fill data gaps
and optimize remedy performance. These recommendations obviously have the benefit of operational
data unavailable to the original designers. The RSE process is designed to help site operators and
managers improve effectiveness, reduce operation cost, improve technical operation, and expedite site
closeout.
Improvements to site characterization and the groundwater monitoring program were recommended in
order to evaluate effectiveness and protectiveness of the system as installed and better understand
subsurface conditions in advance of remedy alterations. At a minimum it is recommended that the
limited monitoring program adjustments and subsurface characterization activities discussed in Sections
5.1.1 and 5.1.2 be seriously considered. These recommendations include:
• Installation of two monitoring wells outside the sheet pile wall to determine if contaminants are
migrating through the wall [addresses effectiveness of the wall and evaluates protectiveness for
receptors in the river]
• Conversion of PZ-02 (by developing and sampling) to a monitoring well to determine if
contaminants are migrating around the end of the wall [addresses effectiveness of the wall and
evaluates protectiveness for receptors in the river]
• Conversion of several piezometers (PZ-03, -07, -09, and -10) to monitoring wells to better
understand residual source and dissolved-phase contaminant extent [feeds into design for system
modifications leading to quicker site closeout]
• Direct push soil and groundwater investigation in the stagnant zone to delineate persistent source
area [feeds into design for system modifications leading to quicker site closeout]
In addition, the following options were evaluated with the goal of improving system performance and
shortening time to achievement of cleanup objectives:
• NAPL-impacted soil excavation and enhanced dissolved-phase treatment
• Limited NAPL-impacted soil removal and installation of additional gate in NW corner
• Groundwater flow modification to enhance treatment of existing funnel & gate system
Of these, the first two have the greatest potential to improve treatment efficiency and shorten time to
achievement of cleanup objectives. However, the second option, which is most similar to the original
design, has the potential to discharge contaminants above PALs to the Little Menomonee River. This
potential is considered unlikely given a considerable record of successful treatment in the existing gates
at the site. The third option was found to be ineffective or of limited benefit because of the difficulty
associated with enhancing the hydraulic gradient in the low permeability soils and protracted times to site
closeout.
Results from field investigations could determine the most cost-effective option for improving system
performance. If minimal amounts of NAPL are encountered, the assumed need for sheet pile wall
improvement and volume of soil excavation and ORC Advanced® quantities required may be reduced
thereby resulting in a lower estimated cost. Likewise, institution of the original design concept of a
16
-------
treatment gate in the NW corner may be sufficient to flush and treat remaining dissolved-phase
contaminants. If significant quantities of NAPL are found, more aggressive excavation, followed by
amending the backfilled area with ORC Advanced®, may be more suitable to achieving site cleanup
goals in a reasonable timeframe. A determination may have to be made whether the latter option requires
an additional decision document.
17
-------
TABLES AND FIGURES
18
-------
Table 1. Groundwater Cleanup Goals
Constituent
Anthracene
Benzo(a)pyrene
Benzo(b)fluoranthene
Chrysene
Fluoranthene
Fluorene
Naphthalene
Pyrene
Benzene
Toluene
Ethylbenzene
Xylene
PAL (jig/L)
600
0.02
0.02
0.02
80
80
8
50
0.5
68.6
140
124
Notes:
PAL - Wisconsin Department of Natural Resources (WDNR)
Preventative Action Level, Ch. NR 140, Wis. Adm. Code
(ig/L - microgram per liter
19
-------
Table 2. Monitoring Program
Well ID
MW-7S
MW-34S
MW-38S
MW-39S
MW-5S
MW-9S
MW-27S
MW-30S
MW-31S
MW-32S
MW-33S
MW-34S
MW-37S
MW-38S
MW-39S
TG1-1
TGI -2
TGI -3
TG2-1
TG2-2
TG2-3
TG3-1
TG3-2
TG3-3
TG4-1
TG4-2
TG4-3
TG5-1
TG5-2
TG5-3
TG6-1
TG6-2
TG6-3
PZ-01
PZ-02
PZ-03
PZ-04
PZ-05
PZ-06
PZ-07
PZ-09
PZ-10
Monitoring
Purpose
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Containment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Treatment
Piezometer
Piezometer
Piezometer
Piezometer
Piezometer
Piezometer
Piezometer
Piezometer
Piezometer
Screened Interval
(feet bgs)
10-15
*
10-15
10-15
12-17
8-13
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Analytical
Sampling
Semi-Annual
Semi-Annual
Semi-Annual
Semi-Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
~
Annual
Annual
—
Annual
Annual
~
Annual
Annual
~
Annual
Annual
~
Annual
Annual
~
Annual
~
~
~
~
—
~
~
~
~
Water Level
Measurements
Semi-Annual
Semi-Annual
Semi-Annual
Semi-Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
Annual
20
-------
Table 2 Notes:
Piezometer - Additional water level measurements locations to verify hydraulic containment
Containment - Shallow and Containment Performance Monitoring Wells
Treatment - Treatment Performance Monitoring Wells
Annual - Sampled during 3rd Quarter (September)
Semi-Annual - Sampled during 1st and 3rd Quarter (March and September)
~ Not sampled
* Well construction details not available, proposed construction included a 5-foot screen interval and
total depth of 10-12 feet bgs.
21
-------
Table 3. Cost Assumptions
REPORT SECTION,
TASK
ITEM
QUANTITY
UNIT
UNIT
PRICE
AMOUNT
ASSUMPTIONS
5.1.1, Monitoring Program Modifications1
Install and develop
new monitoring
wells
Develop existing
piezometers
Sampling and
analysis for new
monitoring locations
Mob/Demob
Drill and Install MW
Development
Survey
Workplan/Oversight/Reporting
1
30
8
1
40
each
foot
hr
each
hr
$2,000
$120
$250
$1,500
$100
Subtotal
Development
20
hr
$250
Subtotal
Sampling
GW PAH Analysis, EPA
625/8270
GW BTEX Analysis, EPA
624/8260
16
8
8
hr
each
each
$200
$160
$173
Subtotal
Total Cost for Monitoring Program Modification
$2,000
$3,600
$2,000
$1,500
$4,000
$13,100
$5,000
$5,000
$3,200
$1,280
$1,384
$5,864
$23,964
15-foot wells, 5-foot screen
4 hours/well
2 new wells, plus MW-38S and -39S
4 hours/well
2 days, 2 people
Includes 1 duplicate
Includes 1 duplicate
22
-------
Table 3 Cont. Cost Assumptions
REPORT SECTION,
TASK
ITEM
QUANTITY
UNIT
UNIT
PRICE
AMOUNT
ASSUMPTIONS
5.1.2, NAPL Investigation1
MW-34S GeoProbe
Investigation
TG1-1 GeoProbe
Investigation
Drilling (Geoprobe)
Workplan/Oversight/Reporting
Soil BTEX Analysis EPA
624/8260
Soil PAH Analysis EPA 625/8270
GW BTEX Analysis, EPA
624/8260
GW PAH Analysis, EPA
625/8270
3
150
30
30
15
15
day
hr
each
each
each
each
$2,000
$100
$173
$160
$173
$160
subtotal
Drilling (Geoprobe)
Workplan/Oversight/Reporting
Soil BTEX Ana lysis EPA
624/8260
Soil PAH Analysis EPA 625/8270
GW BTEX Analysis, EPA
624/8260
GW PAH Analysis, EPA
625/8270
3
150
30
30
15
15
day
hr
each
each
each
each
$2,000
$100
$173
$160
$173
$160
subtotal
Total Cost for NAPL Investigation
$6,000
$15,000
$5,190
$4,800
$2,595
$2,400
$35,985
$6,000
$15,000
$5,190
$4,800
$2,595
$2,400
$35,985
$71,970
5 borings/day
2 soil samples/well
2 soil samples/well
5 borings/day
2 soil samples/well
2 soil samples/well
23
-------
Table 3 Cont. Cost Assumptions
REPORT SECTION,
TASK
ITEM
QUANTITY
UNIT
UNIT
PRICE
AMOUNT
ASSUMPTIONS
5.2.1, NAPL-lmpacted Soil Excavation and Enhanced Dissolved-Phase Treatment2
Excavation
Workplan/Oversight/Reporting
Project Manager
Project Scientist
QA/QC Officer
Field Technician
Clerical
CADD
Excavate and load, bank
measure, medium material, 2
CY bucket, hydraulic excavator1
12 CY Dum Truck Haul/Hour
Backfill with crushed stone1
Unclassified fill, 6" lifts, off-site,
includes delivery, spreading and
compaction1
Disposable materials per
sample1
Soil testing for soil disposal
EPA 625/8270 (SVOCs)
EPA 624/8260 (BTEX)
Haul & Dispose Debris, 16.5 CY
Truck, 10 mi Haul Distance,
Non-hazardous Landfill1
40
40
16
100
8
8
2084
130
70
2708
12
5
5
2084
hr
hr
hr
hr
hr
hr
CY
BCY
CY
CY
each
each
each
CY
$200
$203
$201
$120
$74
$82
$2.5
$174
$54
$15
$15
$160
$173
$45
subtotal
$8,000
$8,136
$3,216
$12,000
$592
$656
$5,175
$
22,742
$3,768
$41,129
$175
$800
$865
$94,269
$201,523
assumed 10, 10 hr days in field
Labor unit cost 1.48; equipment unit cost 1.01
Labor unit cost 107.4; equipment unit cost
67.54
Labor unit cost 1.91; equipment unit cost 1.1;
material unit costs 1.10
Labor unit cost 1.52; equipment unit cost 1.32;
material unit costs 12.32; subbid unit cost 0.02
Labor unit cost 2.62; equipment unit cost 2.01;
subbid unit cost 40.6
24
-------
Table 3 Cont. Cost Assumptions
REPORT SECTION,
TASK
Sheet Pile Wall
Improvement
Oxygen Release
Compound (ORC)
Advanced®
Enhancement
ITEM
50ft x 75ft sheet pile wall
installed
QUANTITY
3750
UNIT
sqft
UNIT
PRICE
$25
subtotal
ORC Advanced® Product
10535
Ib
$8
subtotal
Total Cost for NAPL Excavation
AMOUNT
$93,750
$93,750
$85,856
$85,856
$381,129
ASSUMPTIONS
Ref: 2007 RS Means Building Construction
Cost Data 65th Ed. Included 2x markup on
unit cost.
Does not include freight for product
shipment.
25
-------
Table 3 Cont. Cost Assumptions
REPORT SECTION,
TASK
ITEM
QUANTITY
UNIT
UNIT
PRICE
AMOUNT
ASSUMPTIONS
5.2.2, Limited NAPL-lmpacted Soil Removal and Installation of Additional Gate in NW Corner
Treatment Gate
Installation3
Soil Disposal
Mobilization/Demobilization
Equipment Setup/Teardown
Steel Sheeting Removal
5-30' Long Self Hardening
2'0"wide Perimeter and
Internal cantilever support
walls 24' deep
l-60'x30'x!5'deep Biopolymer
Treatment Gate Pit and Piping
accessories (4- 15'x30' pits)
BP Breakdown and Pumping
Operation
1
1
1080
3,600
1
1
LS
LS
VF
SF
LS
LS
$93,729
$18,492
$35.00
$38.57
$609,444
$32,645
subtotal
Gate plus limited additional soil
disposal
0.5
LS
$96,109
subtotal
Total Cost for Limited NAPL-lmpacted Soil Removal and Installation of Gate
$93,729
$18,492
$37,800
$138,852
$609,444
$32,645
$930,962
$48,054.51
$48,055
$979,017
Scaled (0.5x) from disposal costs in
5.2.1
Notes:
2 RACER 10.3 used to prepare costs for NAPL-impacted soil excavation and enhanced dissolved-phase treatment. Costs are only for excavation
and treatment near MW-34S; similar work in the TGI area could cost as much, although it is likely investigation results would limit the scope of
excavation and treatment required.
3 Industry estimate
26
-------
Table 4. Remedy Optimization Options Evaluation Summary
Recommendation
Effectiveness
Implementability
Cost
5.1.1 Monitoring program modification
Evaluates effectiveness of remedy to
gain site closure.
Easily implemented by installing two
new wells and using existing
piezometers.
$22K
5.1.2 NAPL investigation
Evaluates the extent of residual NAPL.
Reduces uncertainty in the required
excavation extent to gain site closeout.
Easily implemented using direct-
push technology.
$72K
5.2.1 NAPL-impacted soil excavation
and enhanced dissolved-phase treatment
(MW-34S area only)
Removal of residual NAPL would
eliminate the continued source to the
dissolved-phase plume and shorten the
time to site closeout. ORC will
enhance bioremediation in the vicinity
of the excavation.
Moderate effort to improve sheet pile
wall near MW-34S prior to
excavation. ORC Advanced can
easily be incorporated into
excavation backfill.
$381K
5.2.2a Limited NAPL-impacted soil
removal and installation of additional
gate in NW corner
Limited removal of residual NAPL
would eliminate a continued source to
the dissolved-phase plume and shorten
the time to site closeout. The treatment
gate near the excavation would
eliminate the groundwater zone of
stagnation and provide long-term
treatment of any remaining dissolved-
phase contaminants. More
hydraulically effective than a gate near
MW-7S.
Moderate effort to remove sheet pile
wall, excavate residual NAPL, install
gate near MW-34S and install air
injection system. State no longer has
concerns with a treatment gate close
to the river. Proximity to river may
make this more complicated than a
gate near MW-7S.
$979K
5.2.2b Limited NAPL-impacted soil
removal and installation of additional
gate west of MW-7S
Limited removal of easily accessible
residual NAPL would eliminate a
continued source to the dissolved-
phase plume and shorten time to site
closeout. A treatment gate to the west
of MW-7S would eliminate the
groundwater zone of stagnation and
provide long-term treatment of any
remaining dissolved-phase
contaminants. Less hydraulically
effective than gate near MW-34S.
Moderate effort to remove sheet pile
wall, excavate residual NAPL, install
new gate near MW-7S and install air
injection system. The State no
longer has concerns with a treatment
gate close to the river. Possibly
easier to implement than a gate near
MW-34S.
$979K
27
-------
Recommendation
Effectiveness
Implementability
Cost
5.2.3a Groundwater flow modification
to enhance treatment of existing funnel
& gate system - install extraction wells
Installation of extraction wells down-
gradient of Gates 5 & 6 would only
induce a slight hydraulic gradient
across the site; thus it would take years
for contaminants to reach the treatment
gates. Deemed ineffective.
Moderate effort to install extraction
wells and treat groundwater prior to
discharge. Long-term treatment of
remaining dissolved-phase
contaminants may not be necessary if
source removed.
Not costed,
ineffective
5.2.3b Groundwater flow modification
to enhance treatment of existing funnel
& gate system - large scale re-
circulation cell
The re-circulation cell would induce
flow in the groundwater zone of
stagnation, however, there could be
increased flow around the end of the
sheet pile. Flow modification would
encourage contaminated groundwater
to migrate into areas that currently
contain low-level contamination.
Deemed ineffective.
Moderate effort to install
extraction/injection wells and piping.
Long-term treatment of remaining
dissolved-phase contaminants may
not be necessary if source removed.
Not costed,
ineffective.
28
-------
Former Moss-American Production Area
2) Milwaukee County
Figure 1. Site Location
29
-------
CABLE FENCE
CATCH BASIN
HYDRANT
SIGN
FREE PRODUCT COLLECTION SUMP
UTILITY POLE
SAMPLING MANHOLE
MONITORING WELL
INJECTION WELL
STAFF GAUGE
PIEZOMETER
CURRENT RIVER CHANNEL
FORMER RWER CHANNEL
DIRECTION OF GROUNDWATER FLOW
GROUNDWATER ELEVATION CONTOUR
DASHED WHERE INFERRED
ESTIMATED BOUNDARY OF
CONTAMINANT PLUME
GROUNDWATER ELEVATION NOT
MEASURED
Legend
•^Monitoring Locations
Type
• Existing
• New Monitoring Well
• Piezometer to be Sampled
FGURE 2-1
RT-80,570
E:1177.070
ELEV = 737,17
CP-6
GROUNDWATER ELEVATION CONTOUR MAP
3RD QUARTER 2009
TRONQX, LLC
MOSS-AMERICAN SITE
Milwaukee. Wisconsin
750 E. Bunker Ct.
Suite 500
Vernon Hills, Illinois
60061
Figure 2. Proposed Additional Monitoring Locations for Chemical Analysis
30
-------
N:678,
E: 1134.580
ELEV = 720.
CP-4
Legend
• Proposed Boring
1 Proposed Excavation Area
~\~ * f. iX\\V\W.VW*A-V—-
Figure 3. Potential NAPL Investigation Program
31
-------
Appendix A
Groundwater Modeling Documentation
-------
1. Computer Code
MODFLOW-2000 (Harbaugh et al., 2000) was utilized for the groundwater flow model. The Department
of Defense Groundwater Modeling System (GMS) version 7.1 (EMRL, 2005) was used as the software
platform and graphical-user interface for the groundwater flow model.
MODFLOW has a modular structure that allows it to be easily modified to simulate different aspects of
the project. The model must use one flow and one solver package available. Those utilized for the Moss
American model are:
• Layer Property Flow Package - This package defines how hydraulic properties of the model
layers are defined, read, and utilized during the simulation. It differs from other flow packages in
that all input data that define hydraulic properties are independent of model cell dimensions.
• Pre-conditioned Conjugate Gradient Solver Package - This package contains the information that
defines the simultaneous equations that must be solved at each cell. Convergence information is
output with this package if the solver fails to meet closure criteria.
Boundary condition packages are optional packages used to simulate various site-specific features of the
project. The boundary condition packages utilized for the Moss American model are:
• Horizontal Flow Barrier (HFB) - This package is used to simulate the effects of the sheet pile
walls, slurry trenches, or other objects which act as a barrier (or partial barrier) to horizontal
flow.
• Well - This package is used to simulate injection wells or extraction wells.
2. Groundwater Model Design
Due to the limited site information, a simplified model was developed to screen groundwater flow
modification alternatives at the Moss American site.
2.1. Domain and Grid
The model domain includes the area surrounding the funnel and gate system from just up-gradient of the
southern-most gate system to the river. The simplified model consists of one layer with a uniform cell
size of 10 feet horizontal and 15 feet thick and is shown in Figure A-l. The top elevation of each cell was
interpolated from survey data of existing wells. It was assumed that the model lower boundary (top of the
confining till unit) was uniformly 15 feet below ground surface (bgs).
2.2. Boundaries
Numerical models require boundary conditions, such that the hydraulic head or groundwater flux must be
specified along all the outer edges of the system and any internal cells to which conditional head values
must be determined (i.e., extraction well cells, drain cells). The boundary conditions used for the Moss
American model include:
• A specified head boundary was used to represent the river elevation at the north-eastern
boundary.
-------
• A specified head boundary was used to simulate groundwater flow from upgradient of the model
domain. Due to the limited site information, recharge was accounted for in the upgradient
specified head instead of using the recharge package.
• Groundwater flows from the south to the north toward the river; therefore the north-western and
south-eastern boundaries were specified as no flow.
2.3. Material Properties
Hydrologic properties were assigned to individual grid cells based on average properties referenced in the
quarterly/annual groundwater monitoring reports. Based on slug tests completed during the remedial
investigation (RI), the hydraulic conductivity of material location on the topographically higher, western
portion of the site ranged from 0.03 to 0.003 ft/d. Based on the laboratory-performed hydraulic
conductivity analyses conducted on material used to backfill areas of the site located along the river, the
hydraulic conductivity of the material on the topographically lower portion of the site within the funnel
and gate system is approximately 3 ft/d.
According to design documents, the funnel and gate system was constructed using internal cavity scalable
joint sheet piles. Bulk hydraulic conductivity values for Waterloo Barriers, which have a scalable joint,
have been reported at less than 1 x 10"8 cm/s. A conservative estimate for the hydraulic conductivity of 1
x 10"7 cm/s (0.00028 ft/d) was used to represent the sheet pile at the Moss American site.
2.4. Calibration
The purpose of model calibration is to establish that the model can reproduce field-measured hydraulic
heads and flows. During the calibration process, model input parameters are adjusted so that field-
measured heads and flows are reasonably correlated and are considered to provide a good representation
of actual site conditions.
The Moss American groundwater model was calibrated to water levels collected during the 3rd quarter of
2009. Hydraulic conductivity values were varied until modeled water levels provided a reasonable match
to the observed values and the residuals of the modeled versus observed heads were minimized. All water
level values were weighted equally. Table A-l presents the residual calibration statistics and Figure A-2
shows the graphical representation.
Table A-l. Residual Calibration Statistics
Mean Residual (Head)
Mean Absolute Residual (Head)
Root Mean Squared Residual (Head)
Mean Weighted Residual (Head+Flow)
Mean Absolute Weighted Residual (Head+Flow)
Root Mean Squared Weighted Residual
(Head+Flow)
Sum of Squared Weighted Residual (Head+Flow)
-0.076
0.611
0.715
-0.149
1.20
1.40
62.8
-------
The final hydraulic conductivity values used for the model are shown on Figure A-l and were:
• South/Western area - 0.2 and 0.5 ft/d
• Funnel and gate area -3.0 ft/d
3. Predictive Simulations
The calibrated model was used to evaluate modifications to the funnel and gate system that could improve
groundwater flow in the north-west section near monitoring wells MW-7S and MW-34S. MODPATH
was used to depict the flow paths of fictitious contaminant particles for each scenario, which are shown in
green on the Figures A-3 through A-9. Arrows along the flow paths were placed every 10-years to
represent the relative time-frame for contaminant migration. It should be noted that since the model was
run at steady state, particles are shown to eventually pass through the sheet pile walls if the groundwater
does not flow toward the treatment gates.
3.1. Current Conditions
Figure A-3 shows the groundwater elevation contours for the current funnel and gate configuration. The
model shows that there is a stagnation point area near MW-7S and MW-34S as indicated by the slow
particles moving through the sheet pile wall and that groundwater near MW-33S may be moving around
the end of the sheet pile wall. Particles generated at Gate 1 are shown to migrate toward the eastern gates
indicating that this part of the flow system is functioning as intended.
3.2. Excavation at MW-34S
Figure A-4 shows the groundwater elevation contours for the Excavation at MW-34S scenario. This
scenario includes excavation of NAPL containing soils around MW-34S (shown in red on Figure A-4)
and backfill with sand and ORC. The model shows that there will still be a stagnation area near MW-7S
and MW-34S, however, the presence of the higher permeability backfill material may induce localized
flow toward the treated excavation area. This scenario does not impact the potential groundwater moving
around the end of the sheet pile near MW-33S.
3.3. Small Scale Re-Circulation Cell, Excavation at MW-34S
Figure A-5 shows the groundwater elevation contours for the small scale re-circulation cell and
excavation at MW-34S. This scenario includes excavation of NAPL containing soils around MW-34S
(shown in red on Figure A-5) and backfill with sand and ORC. In addition, a small re-circulation cell
would be installed in the north east portion of the system to help distribute ORC to the dissolved phase
plume. An extraction well would be installed near MW-34S and an injection well would be installed near
MW-38S. Due to the low permeability soils near this area, pumping/injection would be very low (0.5
gpm). The model shows that this type of circulation cell could adequately distribute ORC throughout the
remaining dissolved phase plume, however, there will likely be some groundwater mounding near MW-
33S that could increase the amount of flow around the end of the sheetpile wall. Additional costs may
include treatment of contaminated groundwater prior to re-injection.
-------
3.4. Groundwater Extraction near Gate 5 and 6, Excavation at MW-34S
Figure A-6 shows the groundwater elevation contours for groundwater extraction near Gates 5 and 6 and
excavation at MW-34S. This scenario includes excavation of NAPL containing soils around MW-34S
(shown in red on Figure A-6) and backfill with sand and ORC. Two groundwater extraction wells would
be installed east of Gates 5 and 6. Due to the low permeability materials, groundwater extraction rates
would only be about 0.75 gpm near Gate 5 and 0.25 near Gate 6. The model shows that the groundwater
stagnation area near MW-7S and MW-34S still exists, however, flow no longer goes around the end of
the sheet pile near MW-33S and groundwater near MW-38S will eventually reach the eastern treatment
gates. Since the gradient is very low, it may still take over 30 years for the contaminated groundwater to
reach the eastern treatment gates.
3.5. Large Scale Re-Circulation Cell, Excavation at MW-34S
Figure A-7 shows the groundwater elevation contours for the large scale re-circulation cell and excavation
at MW-34S. This scenario includes excavation of NAPL containing soils around MW-34S (shown in red
on Figure A-7) and backfill with sand and ORC. One extraction well would be installed near Gate 5 and
one injection well would be installed near MW-7S to induce flow across the system. Due to the low
permeability materials, groundwater extraction/injection rates would be very low (0.25 gpm). The model
shows that groundwater near MW-7S and MW-34S would flow toward the eastern treatment gates.
Groundwater mounding near MW-33S could increase the amount of flow around the end of the sheet pile
wall.
3.6. New Gate North of MW-34S, Excavation at MW-34S
Figure A-8 shows the groundwater elevation contours for a new gate north of MW-34S and excavation at
MW-34S. This scenario includes excavation of NAPL containing soils around MW-34S (shown in red on
Figure A-8) and backfill with sand and ORC. A new gate with air injection treatment would be installed
to the north of MW-34S. The model shows that flow is induced toward the gate from the up-gradient
treatment gates, near the area of stagnation at MW-7S, and near MW-33S where groundwater is
potentially migrating around the end of the sheet pile.
3.7. New Gate West of MW-7S, Excavation at MW-34S
Figure A-9 shows the groundwater elevation contours for a new gate west of MW-7S and excavation at
MW-34S. This scenario includes excavation of NAPL containing soils around MW-34S (shown in red on
Figure A-9) and backfill with sand and ORC. A new gate with air injection treatment would be installed
to the west of MW-7S. The model shows that flow is induced toward the gate from the area of stagnation
and near MW-33S where groundwater is potentially migration around the end of the sheet pile. This new
gate configuration shows that groundwater flow from the up-gradient Gates 1 and 2 still flows toward the
eastern gates.
4. References
Environmental Modeling Research Laboratory (EMRL), 2005. Groundwater Modeling System (GMS)
version 6.5. Brigham Young University, Provo, UT. 2005.
-------
Harbaugh, A.W., Banta, E.R., Hill, M.C., and McDonald, M.G., 2000. MODFLOW-2000, the US
Geological Survey modular ground-water model - User guide to modularization concepts and the ground-
water flow process; USGS Open File Report 00-92, 121 p. 2000.
-------
Sheet pile (HFB Package)
River (Specified Head Cells)
Up-gradient
(Specified Head Cells)
Figure A-l. Model grid and hydraulic conductivity zones.
-------
Modeled vs. Observed Values
Head
722--
712--
712 713 714 715 716
717 718
Observed (ft)
719 720
722
Figure A-2. Modeled versus observed heads.
-------
\ \
Figure
pfX _ . __
A-3. Current Conditions
-------
Figure A-4. Excavation at MW-34S.
-------
Figure A-5. Small Scale Re-Circulation Cell, Excavation at MW-34S
-------
Figure A-6. Groundwater Extraction near Gate 5 and 6, Excavation at MW-34S
-------
Figure A-7. Large Scale Re-Circulation Cell, Excavation at MW-34S
-------
Figure A-8. New Gate North of MW-34S, Excavation at MW-34S
-------
New Gate West of MW-7S, Excavation at MW-34S
-------
Appendix B
-------
Trend Testing Methods.
Trend presence was determined at the 5% significance level using the censored Mann-Kendall
trend test, which is a non-parametric procedure that accommodates datasets with non-detects.
The censored Mann-Kendall test looks for trends in rankings of the data, rather than in absolute
values of the data. If the Mann-Kendall test indicated a significant trend, the Theil-Sen slope
was computed to quantify the rate of change of concentrations in each well. Both the censored
Mann-Kendall and Theil-Sen computations were performed using the MiniTab statistical
software program using MiniTab scripts from Helsel 2005a (available from PracticalStats.com).
Trend testing was completed for wells and contaminants that had sufficient number of non-detect
values over time.
Regression plots for wells where significant trends were detected are presented in this Appendix.
Increasing trends were detected for naphthalene, fluorene and benzo(a)pyrene in TG1-1.
Decreasing trends were detected for naphthalene and benzene in MW-7S and naphthalene in
MW-38S and corresponding regression equations were used to estimate timeframes to achieve
PAL levels in these wells. Caution should be applied when interpreting these predicted
restoration timeframes because (a) trend testing results are based on current site conditions and
conditions could change in the future resulting in a different restoration timeframes and (b)
uncertainties inherent in trend testing translates into uncertainties in predicted timeframes.
-------
Trend Testing Results.
8-
7-
Qj
N ,-
C 5
Qj
m
Jj 4-1
I 3
2-
1-
0-
Akritas-Theil-Sen line for censored data
MW7S-Benzene = 2,50038 -0,39972*x 7
3
x 7
Detect
HD
Predicted Time to PALs: Benzene in MW-7S
y = 2.500-0.3997 x
[Benzene PAL concentration, |ig/L] = 2.500 - 0.3997 * [Predicted Time to PAL, years]
[0.5 |ig/L] = 2.500 - 0.3997 * [Predicted Time to PAL, years]
[Predicted Time to PAL, years] = {[0.5 |ig/L] - 2.500} + {-0.3997}
[Predicted Time to PAL, years] = 5 years
-------
Qj
u
to
3000-
2500-
2000-
1500-
cn
; 1000-
500-
0-
Akritas-Theil-Sen line for censored data
MW7S-Napthalene = 2424,77 -57Q,QQO*x
* *
* *
MW7S-Napthalene.
» Detect
- HD
Predicted Time to PAL: Naphthalene in MW7S
y = 2425 - 570 x
[Naphthalene PAL concentration, |ig/L] = 2425 - 570 * [Predicted Time to PAL, years]
[8 |ig/L] = 2425 - 570 * [Predicted Time to PAL, years]
[Predicted Time to PAL, years] = {[8 |ig/L] - 2425} + {-570}
[Predicted Time to PAL, years] = 4.2 years
-------
OJ
1000
800
Ł 600
CL
I
§ 400
200-
Akritas-Theil-Sen line for censored data
MW38S-NapNhalene = 1921,94 -348.178*x_9
Detect
3,0 3.5
4.0
4.5
X 9
5.0
5,5
6.0
Predicted Time to PALs: Naphthalene in MW-38S
y = 1922-348.2 x
[Benzene PAL concentration, |ig/L] = 1922 - 348.2 * [Predicted Time to PAL, years]
[8 |ig/L] = 1922 - 348.2 * [Predicted Time to PAL, years]
[Predicted Time to PAL, years] = {[8 |ig/L] - 1922 } + {-348.2}
[Predicted Time to PAL, years] = 5.5 years
-------
7000-
6000-
c 5000 H
.0)
(0
n. 4000-
(0
3000-
2000-
1000
Akritas-Theil-Sen line for censored data
TGl-Napthalene = 1514.17 + 749,487*x_2
3
x_2
TGl-Hapth^lene.
Detect
-------
4000 f
Akritas-Theil-Sen line for censored data
TGl-Fluorene = 78.2743 + 340,792*x 4
TGl-Fluorene.
Detect
3000-
Qj
Qj
3 2000
c
rH
a
1000-
o-L
3
x 4
-------
3000-T
2500-
Akritas-Theil-Sen line for censored data
TGl-B(a)P = 3.45964 + 29,4596*x_6
TGl-B(a)P.
Detect
Q.
t^+
03
2000
1500-
1000-
500-
o-
3
x 6
-------
Appendix C
-------
constituent
naphthalene
phenanthrene
acenaphthene
fluoranthene
2-methylnaphthal(
fluorene
dibenzofuran
pyrene
anthracene
benzo(a)anthracer
checksum
weight percent NAPL
25.1
22.4
9.2
8.2
7.5
6.7
6.1
4.8
2.9
1.8
95
molecular weight,
g/mol
128.17
178.23
154.21
202.25
142.2
166.22
168.19
202.25
178.23
228.29
mole fraction
0.29
0.19
0.089
0.061
0.079
0.060
0.054
0.036
0.024
0.012
0.90
single compound solubility in effective solubility
water, ug/L assuming y = 1
31000 9094
equivalent MWT creosote
149.80401
Estimated effectivewatersolubiliyty of naphthalene in ground water assuming typical creosote weight fraction, where NAPL
constituents less than 2 percent were not included (Pacific Sound Resources RI/FS, 1998). A ground water activity correction
factor (gam ma) of 1 was used for this estimate but the actual value is less less than 1, which means thae actual effective
solublity estimate for naphthalene would be less than 9094 u.g/L
------- |