COST AND PERFORMANCE
REPORT FOR LNAPL CHARACTERIZATION
AND REMEDIATION
Multi-Phase Extraction and Dual-Pump Recovery of LNAPL at the
BP Former Amoco Refinery, Sugar Creek, MO
March 2005
-------
Office of Solid Waste EPA 542-R-05-016
and Emergency Response March 2005
(5102G) www.epa.gov/tio
Cost and Performance Report for LNAPL Recovery
Multi-Phase Extraction and Dual-Pump Recovery of LNAPL
at the BP Former Amoco Refinery,
Sugar Creek, MO
-------
DISCLAIMER
This case study represents the views and opinions of the authors and those who participated in its
development. It has been subjected to U.S. Environmental Protection Agency (EPA) expert review,
however it does not necessarily reflect the views of the EPA or any other federal government
agency. The information is not intended, nor can it be relied upon, to create any rights enforceable
by any party in litigation with the United States or any other party. Reference herein to any specific
commercial product, process, or service by trade name, trademark, manufacturer, or otherwise
does not imply its endorsement or recommendation for use. The information provided maybe
revised periodically without public notice.
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
CONTENTS
Section Page
EXECUTIVE SUMMARY 5
1.0 INTRODUCTION 7
2.0 SITE INFORMATION 7
2.1 BACKGROUND 7
2.1.1 Lower Refinery Area 9
2.1.2 Crawford Area 12
3.0 MATRIX DESCRIPTION AND LNAPL CHARACTERISTICS 15
3.1 MATRIX CHARACTERISTICS AND OPERATING PARAMETERS AFFECTING
TECHNOLOGY COST OR PERFORMANCE 15
3.1.1 Dual-Pump Recovery 15
3.1.2 Multi-Phase Extraction 16
3.2 LNAPL CHARACTERIZATION 16
3.2.1 Dual-Pump Recovery Area 17
3.2.2 Multi-Phase Extraction Area 18
3.3 LNAPL SPECIFIC THICKNESS 18
3.4 LNAPL CHARACTERIZATION 21
3.5 LNAPL VOLUME ESTIMATES 21
4.0 REMEDIATION GOALS AND PERFORMANCE OBJECTIVES 22
4.1 LNAPL REMEDIATION GOAL 22
4.2 DUAL-PUMP RECOVERY AND MULTI-PHASE EXTRACTION PROTECTION
GOALS 22
4.3 DEFINING APPROPRIATE ENDPOINTS 23
4.3.1 API Distribution and Recovery Modeling 23
5.0 RECOVERY SYSTEMS DESCRIPTION, PERFORMANCE, AND COST 26
5.1 DUAL-PUMP RECOVERY SYSTEM 26
5.1.1 Dual-Pump Recovery System Description 26
5.1.2 Dual-Pump Recovery System Performance 26
5.1.3 Dual-Pump Recovery System Costs 31
5.1.4 Dual-Pump Recovery System Observations and Lessons Learned
32
5.2 MULTI-PHASE EXTRACTION SYSTEM 33
5.2.1 Multi-Phase Extraction System Performance 33
5.2.2 Multi-Phase Extraction System Costs 34
5.2.3 Multi-Phase Extraction System Observations and Lessons
Learned 35
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 1
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
5.3 COMPARISON OF DUAL-PUMP RECOVERY TO MULTI-PHASE EXTRACTION
35
5.4 TECHNOLOGY PERFORMANCE AND COST 36
5.4.1 Meeting Protection Goals and Endpoints 36
5.4.2 Cost per Gallon of LNAPL Removed 36
6.0 OBSERVATIONS AND LESSONS LEARNED 37
7.0 REFERENCES 41
8.0 SITE CONTACTS 43
9.0 ACKNOWLEDGEMENTS 44
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 2
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
FIGURES
1 Site Map: BP Products North America, Inc. Former Refinery, Sugar Creek, Missouri 7
2 Dual-Pump Recovery System Location Map 9
3 Dual-Pump Recovery Flow Diagram 9
4 Multi-Phase Extraction System Location Map 13
5 Multi-Phase Extraction Flow Diagram 13
6 LNAPL Schematic in a Typical Monitoring Well 18
7 LNAPL Recovery for Dual-Pump Recovery System: Modeled vs Actual 24
8 Predicted Yearly LNAPL Recovery from the Dual-Pump Recovery System 25
9 Dual-Pump Recovery System Annual LNAPL Recovery 28
10 Lower Refinery Recovery Well Network Cumulative LNAPL Recovery 28
11 Dual-Pump Recovery System Annual Groundwater Recovery 29
12 Well R-007 Fluid Level Measurements and Weekly LNAPL Recovery for 2003 31
13 Multi-Phase Extraction LNAPL Recovery Compared to API Modeled Results 34
TABLES
Table Page
1 Dual-Pump Recovery Well Status 10
2 Dual-Pump Recovery Well Construction Details 11
3 Multi-Phase Extraction Construction Details 12
4 Matrix Characteristics and Operating Parameters Affecting Technology Cost or
Performance of Dual-Pump Recovery 14
5 Matrix Characteristics and Operating Parameters Affecting Technology Cost or
Performance of the Multi-Phase Extraction System 15
6 LNAPL Saturations versus Depth forthe Dual Pump Recovery Area 16
7 LNAPL Saturations versus Depth forthe Multi-Phase Extraction Area 17
8 Comparison of Measured versus Specific LNAPL Thickness (D0) 19
9 Comparison of LNAPL Characteristics forthe Dual-Pump Recovery and Multi-Phase
Extraction Areas 20
10 Dual-Pump Recovery Annual Water and LNAPL Recovery 27
11 Measured LNAPL Thicknesses in Dual-Pump Recovery System Observation Wells 30
12 Dual-Pump Recovery System Estimated Capital Costs 31
13 Dual-Pump Recovery System Estimated O&M Costs 32
14 Multi-Phase Extract System Estimated Capital Costs 34
15 Multi-Phase Extraction System Estimated O&M Costs 35
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 3
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
amsl
API
ASTM
bgs
cm/sec
CMS
cp
former refinery
gpm
GC
LNAPL
mm
non-VOC
POTW
PTS
psi
psig
RCRA
RFI
ROC
ROI
SVOC
uses
VOC
ACRONYMS AND ABBREVIATIONS
micrometer
above mean sea level
American Petroleum Institute
American Society for Testing and Materials
below ground surface
centimeters per second
Corrective Measures Study
centipoise
BP Former Amoco Refinery
gallons per minute
gas chromatograph
light non-aqueous phase liquid
millimeters
non-volatile organic compound
publicly owned treatment works
PTS Laboratories, Inc.
pounds per square inch
pounds per square inch gauge
Resource Conservation and Recovery Act
RCRA Facility Investigation Report
radius of capture
radius of influence
semi-volatile organic compound
Unified Soil Classification System
Volatile Organic Compound
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
EXECUTIVE SUMMARY
This case study was prepared to summarize the recovery of light non-aqueous phase liquid
(LNAPL) at two locations at the BP Products of North America, Inc. Former Amoco Refinery
(former refinery) in Sugar Creek, Missouri. The purpose of this case study was to evaluate the cost
and performance of two remediation systems - one innovative (high-vacuum multi-phase
extraction) and one comprised of a more traditional approach (dual-pump LNAPL and groundwater
recovery). Two locations were selected for the case study based on differing soil lithology,
hydrogeologic and LNAPL characteristics, and remedial approach. This case study illustrates the
benefits of using site characterization and LNAPL recovery data to predict the effectiveness and
longevity of the technologies. In addition, a cost comparison is made between the two LNAPL
recovery technologies: dual-pump recovery applied in sand and multi-phase extraction applied in
silt.
Dual-Pump Recovery
Dual-pump recovery has been applied at the Lower Refinery Area to recover LNAPL beginning with
three wells in 1982, expanding to 15 wells in 1988. System optimization and asymptotic recovery
has resulted in a reduction in the number of operating wells to six by 2004. The geology of the
Lower Refinery Area consists of silty to fine sand with 11 to 69 percent fines, and hydraulic
conductivity ranging from 10~3 to 10~4 centimeters/second (cm/sec). Measured LNAPL saturations
in the dual-pump recovery area were as high as 36 percent of the pore volume and averaged
between 7 and 10-percent per soil core. The dual-pump recovery wells have collectively recovered
approximately 1.82 million gallons of LNAPL and 183.5 million gallons of groundwater from 1982
through 2003 and recovered another 79,500 gallons of LNAPL and 12 millions gallons of
groundwater during 2004. The total groundwater to LNAPL recovered ratio for the dual-pump
recovery system is 109:1.
Multi-Phase Extraction
Multi-phase extraction technology was applied in a portion of the former refinery known as the
Crawford Area. Multi-phase extraction involves application of a high vacuum to the subsurface to
recover LNAPL and groundwater and control migration of LNAPL from seeping into Sugar Creek, a
small stream that bisects the former refinery property. The Crawford Area site geology is silt loess,
with 92 to 98-percent fines, and is characterized by low hydraulic conductivity on the order of 10"6
cm/sec. Measured LNAPL thicknesses in monitoring wells in the Crawford Area were up to 16 feet,
but soil core analyses indicated the LNAPL was discontinuous and limited in volume with a
maximum LNAPL saturation of only 1.4 percent of the pore volume. The multi-phase extraction
system was operated from January 2001 until January 2003. Over two years of operation, it
operated at 26-inches of mercury vacuum and recovered only 151 gallons of LNAPL and 216,000
gallons of groundwater. The total groundwater to LNAPL recovered ratio for the multi-phase
extraction system is 1,430:1.
Cost and Performance
The dual-pump recovery system has recovered 1.899 million gallons of LNAPL and is predicted to
recover an additional 321,900 gallons of the estimated source volume of 1.1 million gallons over
the next six years. Overall, the dual pump recovery system is expected to recover a total of 2.25
million gallons of LNAPL, which is equal to 67-percent of the estimated recoverable LNAPL source
volume, based on the API Model-estimated LNAPL specific thickness over the plume area. The
percent recovery of the initial LNAPL spill volume is unknown, due to lack of spill data, and does
not take into account additional LNAPL in the unsaturated zone, smear zone or outside each
recovery well's radius of capture. Additional LNAPL likely exists outside of influence of the
pumping radius of capture. By comparison, the total LNAPL recovery for the multi-phase extraction
system amounted to less than 10-percent of the original estimated in-place LNAPL volume.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 5
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Each remedial system's performance and associated cost are very dependent on the different soil
matrices (silt versus sand). Although the total cost of the dual-pump recovery system ($3,554,349)
was much greater than the cost of the multi-phase extraction system ($183,053), the normalized
cost per gallon of LNAPL recovered was significantly less, with dual-pump recovery equal to only
$1.87 per gallon compared to multi-phase extraction at $1,212 per gallon. The dual-pump recovery
system continues to recover significant quantities of LNAPL, and recovered an additional 79,500
gallons in 2004. The multi-phase extraction system was shutdown in January 2003 due to no
LNAPL recovery over its last six months of operation.
Observations and Lessons Learned
Overall, the large difference in LNAPL recovery and performance of the two systems indicates that
LNAPL recovery is much more effective from higher-permeability sands than low-permeability silts
and clays, irrespective of the remediation technology. Although dual-pump recovery proves to be
continually effective at recovering LNAPL from sand, it is not expected to be an appropriate
technology for LNAPL recovery in silts and clays. Therefore, protection goals and LNAPL
endpoints should reflect the technical limitations of remediation technologies and soil type, with
appropriate performance expectations, operational timeframes, and shutdown criteria. This case
study suggests that LNAPL recovery for the purpose of source removal and migration control is a
viable remediation goal in sands whereas LNAPL source control and containment is more
attainable and an appropriate remediation goal in silts and clays.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 6
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
1.0 INTRODUCTION
Remediation of light non-aqueous phase liquid (LNAPL) in contaminated media is a particularly
challenging problem at large-scale sites. At former petroleum refineries, for example, LNAPL may consist
of volatile organic compounds (VOCs), semi-volatile organic compounds (SVOCs), non-volatile organic
compounds and trace metals. When released into the subsurface, they can release dissolved
contaminants to groundwater or VOCs into subsurface gas and potentially indoor air for an extended
period of time. In addition, if sites have low-permeability soils, characterization and remediation of LNAPL
is particularly difficult. No single technology has been identified as the best solution for all sites and all
soil types contaminated with LNAPL.
This case study summarizes the application of two LNAPL remedial technologies at two different
locations at a large former petroleum refinery, the BP Products North America, Inc. Former Amoco
Refinery in Sugar Creek, Missouri. This case study focuses on two LNAPL remediation systems at the
former refinery, a multi-phase extraction system and a dual-pump recovery system. The locations of the
two systems are in portions of the refinery known as the Crawford Area and the Lower Refinery Area,
which are shown on Figure 1. The remedial systems were initially selected as interim measures based
on their performance in two different soil matrices (i.e., silt/clay versus sand).
The purpose of this case study is to evaluate the cost and performance of the two LNAPL remediation
systems and highlight how soil type can have a considerable effect on LNAPL recovery, cost, and
performance. Remediation goals, LNAPL recovery, performance, lessons learned, data collection needs,
modeling approaches, and LNAPL endpoint strategies are also discussed herein.
2.0 SITE INFORMATION
2.1 BACKGROUND
From 1904 to 1982, the Standard Oil Company and then Amoco Oil Company operated a petroleum
refinery along the southern bank of the Missouri River east of Kansas City and north of Independence,
Missouri in the township of Sugar Creek, Missouri. The former refinery property occupies approximately
500 acres of the southern floodplain and bluffs along the Missouri River (Figure 1), and Sugar Creek is a
small urban stream that bisects the former refinery property and discharges to the Missouri River.
Refining operations ceased in 1982 and most tanks, process equipment, and buildings were dismantled
by 1989. Petroleum products refined at the Sugar Creek refinery included gasoline, jet fuel, kerosene,
furnace oil, liquified petroleum gases, petroleum coke, sulfur, and propylene polymers. Presently, BP
Products North America, Inc. (BP) owns the refinery property and the only active industrial operations
include a bulk storage and pipeline terminal for petroleum products, and an asphalt terminal. The
majority of the former refinery property is currently inactive. The only residential use is off-site in the
surrounding community of Sugar Creek.
Upon mutual agreement with United States Environmental Protection Agency (U.S. EPA) Region 7 and
the Missouri Department of Natural Resources (MDNR), the site was divided into ten parcels for RCRA
Corrective Action (Figure 1), based on prior refining use, proposed land use, geography, contaminants of
concern, and soil conditions. In each area, various interim measures technologies have been applied
including horizontal well total fluids extraction, vacuum truck enhanced fluid recovery, biosparging,
hydraulic control pumping, multi-phase extraction, dual-pump LNAPL recovery, and gravity-draining
interceptor trenches. In general, interim measures are designed and operated to abate imminent threats
to human health and the environment. These technologies have been applied with varying degrees of
success, and, based on their performance, are proposed for final corrective measures for the former
refinery. Two interim measures technologies discussed in this case study include the Lower Refinery
Recovery Well Network (i.e., dual-pump recovery) and Crawford Multi-Phase Extraction System (i.e.,
multi-phase extraction), are shown on Figure 1 and discussed in the following sections.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 7
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
The overall remediation goal of both systems is source reduction through LNAPL recovery to reduce
mobility and risk to the environment. Prior to implementation, LNAPL seeps and sheens were
periodically observed in Sugar Creek. Therefore, the goal of the interim measures systems was to
eliminate LNAPL seeps and sheens while reducing LNAPL sources to the practical limit of recovery.
Based on the historical performance of the systems, the dual-pump recovery system was proposed as a
final corrective measure for the Lower Refinery Area, while the multi-phase extraction system was
determined not to be effective as a long-term remedy for the Crawford Area. The multi-phase extraction
system was replaced with a hydraulic control pumping network in 2003, and presently both proposed final
corrective measures technologies are currently under review as final remedies by U.S. EPA Region 7
and MDNR.
Figure 1. Site Map: BP Products North America, Inc. Former Refinery, Sugar Creek, Missouri
>-:?v«t««i?!=Sf\
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
2.1.1 Lower Refinery Area
Initial operations at the refinery began in the Lower Refinery Area in 1904, and, over the course of its
operation, included numerous process units and storage tanks. Process units in the Lower Refinery
consisted of the coke, crude, and pressure stills, an acid treating plant, a clay plant, a batch agitator,
oil/water separator box, and storage tank area. Expansion added a naphtha sweetening plant, a paraffin
plant, cracking unit separator, pipe stills, and a heater oil treating plant. In addition, there were three
separate loading areas, a tetraethyl lead storage and blending area, and a total of 97 storage tanks in the
Lower Refinery Area. The tanks were used to store crude feedstocks, intermediates, heater oil, fuel oil,
diesel oil, gasoline oil, jet fuel, caustic, coalescer, prefractionator bottoms, slop oil, and process water.
Currently, the only remaining storage tanks include Tank 95R, which is used for recovered LNAPL
storage for the dual-pump recovery system, and two asphalt storage tanks.
Historically, LNAPL has been observed in 41 observation and monitoring wells in the Lower Refinery
Area (RETEC, 2004a). The LNAPL observed in the Lower Refinery Area is thought to be the result of
historical releases from process units and storage tanks present in the Lower Refinery Area dating back
to the start of the Amoco Refinery in 1904. Historic LNAPL releases at the surface may have migrated
vertically through the subsurface silty clay/clayey silt via macropores or preferential pathways (RETEC,
2004a). Upon reaching the deeper lithology sand and gravel, the LNAPL would have migrated laterally
due to its larger pore sizes and greater permeability. In general, groundwater flow is to the north towards
the Missouri River and to the west towards Sugar Creek. However, the sand and gravel are below the
water table for most of the year, and the LNAPL is therefore under pressure in a semi-confined condition
and is immobile under natural gradients. The LNAPL's initial lateral migration was limited by the pressure
head of LNAPL at the time of release, and the LNAPL has not migrated beyond the Lower Refinery Area
since its release. This semi-confined phenomenon is believed to be the reason LNAPL in the Lower
Refinery Area has not migrated downgradient to the Riverfront Area, Sugar Creek or the Missouri River
(Figure 1). The conceptual site model for the presence of LNAPL and the hydrogeology of the Lower
Refinery Area is discussed further in Section 3.
The presence of LNAPL in monitoring wells installed in the Lower Refinery Area in the early 1980s
prompted the installation of the first phase of the recovery well network (Woodward-Clyde, 1989). The
first three recovery wells (R-001, R-002, and R-003) were installed in 1982. Recovery well R-003 was
replaced by R-004 in 1984. Alternatives for expanding the recovery well system followed. Beginning in
1987, six additional recovery wells (R-005, R-006, R-007, R 008, R-009, and R-010) were added to the
area. Six other recovery wells (R-011 to R-016) were installed in 1988. The newer recovery wells were
placed in areas of significant LNAPL accumulation (i.e., greater than two feet thick) (Woodward-Clyde,
1987). The recovery wells (active and inactive) and their status are outlined in Table 1. Well construction
specifications are included in Table 2.
Today, only seven of the original 16 wells still operate. The other recovery wells were shut down due to
minimal LNAPL recovery, borehole collapse, or biological fouling. Six of the seven remaining operating
recovery wells are used to recover LNAPL from the subsurface and the seventh (R-015) is used solely for
hydraulic control of an LNAPL seep location along Sugar Creek (SCOP-08) (RETEC, 2004c). Thus,
recovery well R-015 is not discussed further in this case study.
In early 2002, it appeared that many of the recovery wells were not operating as originally designed,
primarily due to the age of the system, and the equipment was modified or replaced to increase the
efficiency of the active recovery wells. The remaining six recovery wells (R-001, R-002, R-006, R-007, R-
008, and R 009) constitute the dual-pump recovery system evaluated in this case study. The six
remaining operational recovery wells are shown in Figure 2, and a schematic of one of the recovery wells
and the dual-pump recovery flow diagram is provided in Figure 3. Groundwater is transferred to a
ClarifierTank and batch treated via air stripper and discharged to the local Publicly Owned Treatment
Works (POTW), while LNAPL is transferred to Tank 95R for eventual recycling at an off-site facility.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 9
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 2. Dual-Pump Recovery System Location Map
.
Figure 3. Dual-Pump Recovery Schematic
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
10
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Table 1. Dual-Pump Recovery Well Status
Recovery
Well
R-001
R-002
R-003
R-004
R-005
R-006
R-007
R-008
R-009
R-010
R-011
R-012
R-013
R-014
R-015
R-016
Date
Installed
1982
1982
1982
1984
1987
1987
1988
1988
1987
1987
1988
1988
1988
1988
1988
1988
Status
Modified and replaced equipment in
2003 to increase efficiency
Modified and replaced equipment in
2003 to increase efficiency
Limited operation from December
1982 to March 1984
Shutdown in 1996
Shutdown in 1992
Modified and replaced equipment in
2003 to increase efficiency
Modified and replaced equipment in
2002 to increase efficiency
Modified and replaced equipment in
2002 to increase efficiency
Modified and replaced equipment in
2002 to increase efficiency. Placed in
skimmer mode in April 2004 for
testing purposes
Temporarily shut down in 2002
Shutdown in 1992
Shutdown in 1995
Shutdown in July 2002
Shutdown in 1995
Hydraulic Barrier System
Shutdown in 1995
Comments
Operational
Operational
Borehole collapse and
suspected screen damage
prevented adequate recovery
and allowed silt to build up.
Replaced with R-004 in
1984.
Excessive biological growth
caused shut down; current
vacuum liquid recovery
location.
Shut down due to minimal
LNAPL recovery.
Operational
Operational
Operational
Operational
Temporarily shut down due
to minimal LNAPL recovery.
See Corrective Measures
Study for more information.
Shut down due to minimal
LNAPL recovery.
Shut down due to minimal
LNAPL recovery.
Shut down due to minimal
LNAPL recovery.
Shut down due to minimal
LNAPL recovery.
Acting as a hydraulic barrier
system; no longer part of the
recovery well network.
Shut down due to minimal
LNAPL recovery.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
11
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Table 2. Dual-Pump Recovery Well Construction Details
Recovery
Well
Number
R-001
R-002
R-003
R-004
R-005
R-006
R-007
R-008
R-009
R-010
R-011
R-012
R-013
R-014
R-015
R-016
Elev.
Top of
Casing
(ft amsl)
743.2
743.2
NA
742.8
743.1
743.9
743.3
744.3
743.4
743.5
739.2
741.6
738.8
744.5
745.8
748.7
Elev.
Ground
Surface
(ft amsl)
741.9
743.2
NA
740.8
740.6
741.4
740.9
741.3
741.3
741.7
736.7
736.2
739.0
741.7
743.5
745.6
Casing
Material
Steel
PVC SCH.80
PVC SCH.80
PVC SCH.80
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Screen
Material
Stain. Steel
PVC SCH.80
PVC SCH.80
Galv. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Stain. Steel
Approx.
Screened
Interval
(ft bgs)
11-46
15-50
N/A
17.5-52.5
20-55
11-46
19-54
22-57
25-60
19-54
16-51
14-49
14-49
16-51
5-40
5-40
Notes:
ft. amsl - feet above mean sea level
ft. bgs - feet below ground surface
All recovery wells are 12-inches in diameter and have a 20 slot screen size. Each recovery well is
equipped with an automated dual-pump system: one dedicated for groundwater and the other for LNAPL.
A cone of depression develops around the recovery well when the water pump extracts water from the
well and LNAPL within the cone of depression is drawn toward the well. Each well has a dedicated
LNAPL pump which removes the LNAPL automatically after it reaches a predetermined thickness.
Meters are attached to each water and LNAPL pump to record the quantity of liquids removed. The
system automatically adjusts for fluctuations in groundwater elevations, maintaining a constant
groundwater elevation in the well. All LNAPL is pumped to Tank 95R for storage and eventual recycling.
Field personnel record the volume of extracted LNAPL and groundwater and the fluid level
measurements weekly. The elevation of the Missouri River is also recorded on a weekly basis, due to its
effect on LNAPL and groundwater recovery. These monitoring requirements are outlined in the Interim
Measures Work Plan (Woodward-Clyde, 1989) and the data are provided in Quarterly Progress Reports
(BP, 1989 through 2004). Each recovery well has an associated observation well, typically installed
within 30 feet of the recovery wells, used to monitor LNAPL thicknesses. The depths to LNAPL and
groundwater in the observation wells are measured on a monthly basis.
2.1.2
Crawford Area
The Crawford Area is located in the south-central portion of the refinery along the western bank of Sugar
Creek, as shown in Figure 1. The Crawford Area was part of the refinery's major expansion efforts in the
1920s and 1930s on a tract of land west of Sugar Creek (Figure 1). Refining operations began in the
Crawford Area in 1921 and continued until the late 1950s. Currently there are no active petroleum
operations or tank storage in the Crawford Area.
LNAPL has been observed in monitoring wells and piezometers in the Crawford Area (RETEC, 2004b).
Groundwater flow is from the Crawford Area east towards Sugar Creek. Sugar Creek is a gaining stream
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
12
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
although average base flows are only 720 gallons per minute (gpm) (i.e., 1.6 cubic feet per second).
Based on the hydraulic conductivity and gradient, the estimated groundwater seepage rate from the
Crawford Area into Sugar Creek is only 0.14 gpm, which is approximately 0.02 percent of the average
base flow rate of 720 gpm for Sugar Creek (RETEC, 2004b).
Isolated and periodic LNAPL seeps were observed in Sugar Creek which are believed to be from
historical releases in the Crawford Area (RETEC, 2004b). To abate the seeps and due to the fine-
grained low-permeability soils in the Crawford Area, multi-phase extraction was selected as an interim
measure to extract LNAPL and groundwater under high vacuum. The system was pilot tested in 1998
and then full-scale operation on six extraction wells was started in January 2001. The extraction wells,
MW-078, SC-14, SC-15, SC-16, SC-24, and SC-25, are shown in Figure 4, and a process flow diagram
of the multi-phase extraction system is provided in Figure 5. Table 3 provides construction information
on the multi-phase extraction wells. The multi-phase extraction system was cycled for four months, and
then it was determined that more LNAPL recovery could be achieved through full-time extraction from
one well, MW-078, which was operated until January 2003. Total fluids were extracted via a 10
horsepower high-vacuum liquid ring vacuum pump, separated in a knockout tank and water was
discharged to an oil water separator and to the POTW. Vapors were treated with granular activated
carbon drums and discharged to the atmosphere. LNAPL was periodically transferred to the Tank 95R.
Table 3. Multi-Phase Extraction Well Details
Well ID
MW-078
SC-14
SC-15
SC-16
SC-24
SC-25
Date
Completed
4/12/1995
8/13/1999
8/13/1999
8/13/1999
8/29/2000
8/29/2000
Top Of
Casing
(ft. amsl)
764.87
764.44
763.92
764.27
763.81
763.91
Ground
elev.
(ft. amsl)
763.09
763.03
762.63
762.23
762.69
762.87
Well
Diameter
(inches)
2
0.75
0.75
0.75
1
1
Casing
Material
SCH 40 PVC
SCH 40 PVC
SCH 40 PVC
SCH 40 PVC
SCH 40 PVC
SCH 40 PVC
Total
Boring
Depth
(ft. bgs)
30.5
24
24
28
28
28
Screen
Interval
(ft. bgs)
13.5-28.5
9-24
9-24
8-28
8-28
8-28
Notes:
ft. amsl - feet above mean sea level
ft. bgs - feet below ground surface
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
13
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 4. Multi-Phase Extraction System Location Map
Figure 5. Multi-Phase Extraction Flow Diagram
MULTI-PHASE EXTRACTION FLOW
?,FLu£:',T "ROM
~RA::TIGn >A'EL_£
117 SAL.
CVCLOMC
KNOCKOUT
TANK
1
1
VAPOR
EFFLUENT
TO
ACTIVATED
3»HP
GSOvl iD'A'ATEK
EFFJJEUT
TO POTvV
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 14
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
3.0 MATRIX DESCRIPTION AND LNAPL CHARACTERISTICS
Matrix characteristics and operating parameters can affect the cost or performance of a treatment
technology. This section evaluates the key parameters that affect cost or performance of dual-pump
recovery and multi-phase extraction. The matrix characteristics documented in this section include soil
types, soil properties, hydrogeology, LNAPL characteristics, and LNAPL volume estimates. The operating
parameters include system parameters, such as pumping rates and applied vacuum.
3.1 MATRIX CHARACTERISTICS AND OPERATING PARAMETERS AFFECTING TECHNOLOGY
COST OR PERFORMANCE
3.1.1
Dual-Pump Recovery
Unconsolidated sediments in the Lower Refinery Area consist of colluvium overlying alluvial deposits
within the floodplains of both the Missouri River and Sugar Creek. The upper lithologic zone (Zone A)
exists from ground surface to approximately 25 to 35 feet bgs. Zone A is described on Lower Refinery
Area boring logs as silty clay or clayey silt with occasional fine sand (RETEC, 2004a). Zone B extends to
bedrock (approximately 60 feet bgs) and pinches out to the south. Zone B is described as a silty to
coarse sand which coarsens with depth in the Lower Refinery Area (RETEC, 2004a).
Table 4 lists the matrix characteristics of the soil in the Lower Refinery Area which affect the cost or
performance of the dual-pump recovery system.
Table 4. Matrix Characteristics and Operating Parameters Affecting Technology Cost or
Performance of Dual-Pump Recovery (RETEC, 2004a)
Parameter
Soil Classification
Clay Content and/or Particle Size
Distribution
Hydraulic conductivity
Air Permeability
Porosity
Depth of groundwater below ground surface
Total Organic Carbon
Groundwater Pumping Rate
Value
Zone A: silty clay or clayey silt with occasional fine sand.
Zone D: silty fine to coarse sand with a small percentage of
fine to coarse gravel.
Zone A: 69 to 99 percent fines
Zone D: 4 to 68 percent fines
Zone A: 1x10" to 4x10" cm/sec
Zone D: 7x10" to 6x10 3 cm/sec
Not measured
Zone A: 43 to 73%
Zone D: 32 to 06%
20 feet (average pre-pumping) - in Zone A
33 feet (average pumping) - at Zone A/D interface
Zone A: 1.7%
Zone D: 0.4%
R-001: 1.8 gpm (avg.1,4.6 gpm (maxJ
R-002: 3.2 gpm (avg.), 8.0 gpm (max.)
R-006: 3.6 gpm (avgJ, 7.9 gpm (max.)
R-007: 8.7 gpm (avg.), 12.0 gpm (max.)
R-008: 6.2 gpm (avg.), 9.1 gpm (max.)
R-009: 2.7 gpm (avg.), 7.7 gpm (max.)
Notes:
cm/sec - centimeters per second
gpm - gallons per minute
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
15
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
3.1.2
Multi-Phase Extraction
LNAPL in the Crawford Area is located in unconsolidated deposits consisting of fill; silt, clayey silt to silty
clay; and silt loess (RETEC, 2004b). Loess, consisting of windblown silt and clay, is the predominant soil
lithology in the Crawford Area. Lithologic data collected in the Crawford Area describe the loess as
poorly sorted silts and clayey silts with low to high plasticity, very stiff when unsaturated to soft when
saturated.
Table 5 lists the matrix characteristics and operating parameters of the soil in the Crawford Area which
affects the cost or performance of the multi-phase extraction system.
Table 5. Matrix Characteristics and Operating Parameters Affecting Technology Cost or
Performance of Multi-Phase Extraction at the Former Refinery (RETEC, 2004b)
Parameter
Soil Classification
Clay Content and/or Particle Size Distribution
Hydraulic conductivity
Air Permeability
Porosity
Depth of groundwater below ground surface
Total Organic Carbon
Operating Pressure/Vacuum
Air Flow Rate
Groundwater Pumping Rate
Value
Reworked loess and colluvium derived from the
upland bluffs, consisting of clayey silt to silty clay
sediments with occasional sands and gravels
Loess: 89 to 94% silt/clay content
2x10 cm/sec
K=0.03darcies
38to08%
9.7feet(pre-pumping)
17 feet (during MPE)
2% to 3%
Maximum: 27 inches of mercury
Average: 26 inches of mercury
Maximum: 98 acfm
Average: 13 acfm
Maximum: 3.8 gpm
Average: 0.02 gpm
Notes:
cm/sec - centimeters per second
acfm - actual feet per minute
gpm - gallons per minute
Overall, the soils in the multi-phase extraction area are predominantly cohesive silts and clays, and
classified as fine-grained soils. Overall they have less air permeability and lower hydraulic conductivity
than the soils in the dual-pump recovery area, which is reflected in the differences in groundwater
pumping rate. Even at a vacuum-enhanced 26-inches of mercury, the multi-phase extraction system only
averaged 0.52 gpm, while the dual-pump recovery wells average between 1.8 and 8.7 gpm per well
(RETEC, 2004d).
3.2 LNAPL CHARACTERIZATION
The following sections provide information on LNAPL characterization, LNAPL saturation, and distribution
at the dual-pump recovery and multi-phase extraction systems.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
16
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
3.2.1
Dual-Pump Recovery Area
Table 6 provides the LNAPL pore fluid saturations for the dual-pump recovery area. Sample
identification (i.e., "ID") is by soil boring ID and core interval in feet below ground surface (bgs). Sample
locations are shown in Figure 2.
Table 6. LNAPL Saturations versus Depth for the Dual Pump Recovery Area (RETEC, 2004d)
SAMPLE
ID
LRSB-4 A/20-22.51
LRSB-4 A/20-22.51
LRSB-4 A/22. 5-25'
LRSB-4 A/22. 5-25'
LRSB-4 A/25-27.51
LRSB-4 A/25-27.51
LRSB-4 A/27. 5-30'
LRSB-4 A/27. 5-30'
LRSB-4 A/30-321
LRSB-4 A/30-321
LRSB-4 A/32-32.51
LRSB-4 A/32. 5-35'
LSRB-5A/32.5-351
LSRB-5A/32.5-351
LSRB-5A/35-37
LSRB-5A/35-37
LRSB-5A/37-391
LRSB-5A/37-391
LSRB-5A/39-40.51
LSRB-5A/39-40.51
LRSB-5A/40.5-42.51
LRSB-5A/40.5-42.51
LRSB-6 A/30-32.51
LRSB-6 A/30-32.51
LRSB-6 A/32. 5-35'
LRSB-6 A/32. 5-35'
LRSB-6 A/35-37.51
LRSB-6 A/35-37.51
LRSB-6 A/37. 5-40'
LRSB-6 A/37. 5-40'
LRSB-6 A/40-421
LRSB-6 A/40-421
DEPTH,
ft.
21.0
22.0
23.0
24.1
26.0
27.0
28.0
29.0
30.5
31.5
32.4
33.5
33.0
34.1
35.5
36.6
37.5
38.5
39.25
40.25
41.0
42.0
31.0
32.1
33.0
34.1
35.5
36.9
38.5
39.5
40.5
41.5
MOISTURE
CONTENT
(% wt)
45.4
47.5
91.7
38.1
26.9
22.5
31.4
34.6
23.3
13.8
38.0
19.0
20.0
22.5
17.0
20.4
18.1
13.8
19.1
12.7
14.5
12.6
29.7
26.7
28.9
27.6
28.4
26.4
24.7
23.9
15.2
12.2
DENSITY
BULK
(g/cc)
1.09
1.07
0.71
1.26
1.39
1.42
1.31
1.26
1.45
1.50
1.17
1.52
1.48
1.41
1.62
1.52
1.54
1.66
1.52
1.79
1.75
1.81
1.30
1.31
1.27
1.29
1.23
1.36
1.40
1.43
1.63
1.75
GRAIN
(g/cc)
2.61
2.62
2.62
2.60
2.63
2.61
2.55
2.59
2.69
2.68
2.63
2.66
2.61
2.56
2.61
2.63
2.64
2.64
2.63
2.63
2.64
2.65
2.61
2.62
2.62
2.63
2.62
2.62
2.64
2.64
2.64
2.64
POROSITY, %
TOTAL
58.3
59.3
73.0
51.5
47.1
45.5
48.8
51.4
46.1
44.0
55.5
42.9
43.2
44.7
37.7
58.3
41.8
37.4
63.2
31.8
33.9
31.7
50.3
50.1
51.6
51.0
53.0
48.1
46.8
45.7
38.4
33.8
AIR
FILLED
7.4
8.6
8.0
3.2
8.8
11.6
7.8
7.3
11.8
22.0
26.3
12.6
13.0
11.5
8.1
26.8
13.8
14.2
33.7
9.0
8.2
8.8
11.5
13.3
13.9
14.0
16.4
10.4
11.5
11.2
13.3
12.2
Pore
Saturat
WATER
87.3
85.4
88.1
92.5
81.3
54.4
84.0
79.4
67.6
34.6
38.7
53.3
56.2
49.4
70.8
53.9
66.4
61.9
46.7
70.6
74.0
71.9
75.1
53.5
61.4
55.5
54.4
61.4
70.8
75.2
64.9
63.7
Fluid
ions %
NAPL
ND<0.1
0.1
0.9
1.2
ND<0.1
20.1
0.1
6.4
6.8
15.3
14.0
17.3
13.6
24.9
7.9
ND<0.1
0.6
ND<0.1
ND<0.1
1.0
1.8
0.3
2.2
19.9
11.7
17.1
14.7
17.0
4.5
0.3
0.4
0.2
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
17
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Table 6. LNAPL Saturations versus Depth for the Dual Pump Recovery Area, continued
(RETEC, 2004d)
SAMPLE
ID
SC-80B/5-71
SC-80B/7-8.51
SC-80B/10-11.51
SC-80B/11.5-131
SC-80B/15-161
SC-80B/16-171
SC-80B/20-221
SC-80B/25-271
SC-80B/30-31.51
SC-80B/30-31.51
SC-80B/31.5-331
SC-80B/35-36.51
SC-80B/35-36.51
SC-80B/36.5-371
SC-80B/40-41.51
SC-80B/41. 5-42.5'
DEPTH,
ft.
5.5
8.2
11.25
12.7
15.5
16.5
21.5
26.5
30.25
31.25
32.75
35.2
36.2
37.7
40.25
42.0
MOISTURE
CONTENT
(% wt)
26.7
21.4
21.8
21.8
19.7
24.9
21.5
20.5
25.5
24.9
14.2
7.4
11.9
14.8
12.9
20.3
DENSITY
BULK
(g/cc)
1.46
1.43
1.40
1.39
1.46
1.44
1.40
1.42
1.40
1.47
1.69
1.85
1.79
1.80
1.80
1.62
GRAIN
(g/cc)
2.54
2.57
2.58
2.63
2.60
2.60
2.61
2.62
2.61
2.60
2.64
2.63
2.62
2.63
2.64
2.65
POROSITY, %
TOTAL
42.4
44.4
45.7
47.2
44.0
44.7
46.3
45.9
46.6
43.5
36.0
29.6
31.7
31.6
31.8
38.9
AIR
FILLED
3.1
13.7
14.3
16.2
14.8
8.5
16.2
15.6
9.4
6.8
10.1
15.1
9.6
3.9
8.6
5.6
Pore
Saturat
WATER
90.8
68.9
56.5
55.1
61.1
77.5
63.3
53.0
60.0
84.2
36.1
32.2
57.4
70.3
72.8
84.4
Fluid
ions %
NAPL
2.0
0.2
12.3
10.7
5.4
3.5
1.8
13.0
19.9
0.1
35.9
16.8
12.3
17.4
ND<0.1
1.2
3.2.2
Multi-Phase Extraction Area
Table 7 provides the LNAPL pore fluid saturations for the multi-phase extraction system, and the soil
boring location is shown on Figure 4.
Table 7. LNAPL Saturations versus Depth for the Multi-Phase Extraction Area
(RETEC, 2004b)
Sample ID
SC-24/4.6'
SC-24/5.5'
SC-24/5.6'
SC-24/6.2'
SC-24/7.8'
SC-24/8.5'
SC-24/8.9'
SC-24/9.5'
SC-24/9.8'
SC-24/11.9'
Depth
(ft.)
4.6
5.5
5.6
6.2
7.8
8.5
8.9
9.5
9.8
11.9
Moisture
Content
%
24.3
24.1
22.1
20.4
23.3
22.3
20.2
25.8
26.3
24.4
Density
Bulk
(g/cc)
1.53
1.56
1.6
1.53
1.62
1.51
1.63
1.48
1.51
1.54
Grain
(g/cc)
2.63
2.64
2.63
2.48
2.65
2.63
2.65
2.63
2.61
2.64
Porosity
Total
41.8
40.9
39.1
38.1
38.9
42.6
38.4
43.8
42.2
41.7
Air
Filled
4.6
3.4
3.5
6.7
1.3
9
5.4
5.6
2.8
4.1
Pore Fluid
Saturation,
% Pore Volume
Water
89.8
90.6
92.1
83.6
99.9
82.0
88.5
87.8
96.7
91.5
LNAPL
0.1
1.4
<0.1
<0.1
<0.1
<0.1
<0.1
0.1
<0.1
<0.1
3.3 LNAPL SPECIFIC THICKNESS
Traditionally, the conceptual model for the occurrence of LNAPL in the subsurface pictured an LNAPL
layer, "pool", or "pancake" floating on a depressed representation of the capillary fringe or water table
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
18
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
(Ballestero et al. 1994). This was based on laboratory experience in highly uniform sand or glass beads
(idealized porous media), because those conditions are easiest to replicate in bench scale tests and can
be performed rapidly. The concept of measured LNAPL thickness in an observation well equalizing with
the LNAPL layer within the capillary fringe of the soil (even though the LNAPL saturation in the soil may
be low), made it possible to explain large (5 or more ft) accumulations of LNAPL in observations wells
while LNAPL recovery attempts in those conditions often resulted in recovery of little LNAPL.
Adamski, et al. (2004) explains the differences between LNAPL in homogenized sand versus fine-grained
soils like those in the multi-phase extraction area. Counter-intuitive behaviors include the fact that a very
small and discontinuous volume of LNAPL observed in the surrounding soil may result in several feet of
LNAPL in an observation well; very low LNAPL recovery volumes when large LNAPL accumulations are
present in neighboring wells; and apparent LNAPL migration below the water table. Because these
observations did not fit with the traditional understanding of LNAPL in ideal porous media, LNAPL
volumes were estimated based on a revised conceptual model for LNAPL in the subsurface.
To accurately estimate the quantity of LNAPL in the subsurface, the LNAPL specific thickness (i.e., "D0")
must be estimated from soil core or monitoring well data. D0 is the integral of the LNAPL saturation over
the depth of a soil column. It represents the total thickness of LNAPL that occurs as disseminated and
discontinuous pockets of LNAPL throughout the LNAPL-impacted porous medium (i.e., subsurface soil).
Therefore, D0 is defined as the specific thickness of LNAPL, which is representative of the amount of
LNAPL in a formation. For example, if you had a core of soil separated into its respective media (i.e., air,
water, LNAPL, and soil), D0 is a normalized volume of LNAPL (feet3/feet2) per unit surface area, but is
expressed as a thickness (in units of feet). At equilibrium, due to capillary forces in soil, the measured
LNAPL thickness in a monitoring well is always greater than D0.
The following schematic shows a conceptualization of a typical monitoring well in the subsurface with
groundwater and LNAPL:
Figure 6. LNAPL Schematic in a Typical Monitoring Well (Charbeneau et al., 1999)
J
bn
0
1
k
z
aw
r
^
Zao Water Table
_ _ _f
i
^^^
LNAPL
Layer
ow
The LNAPL thickness is located between the air-NAPL interface zao and the NAPL-water interface zow.
The total monitoring well LNAPL thickness is b0. The elevation of the water table, zaw, provides the
datum for fluid levels. While the water table is not measured in a monitoring well because of the LNAPL
layer, its elevation is determined from the elevations zao and zow, and the LNAPL specific gravity.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
19
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
The relationship between measured monitoring well LNAPL thickness, b0, and the specific LNAPL
volume, D0 may be calculated from the following equation:
^max
= ]nS0(z)dz
Where: Zmax = height of oil
Zow = height of the oil/water interface
S0 = saturation of oil
n = soil porosity
The function D0(b0) may be approximated piecewise by a linear function integration of soil core LNAPL
saturations with depth. LNAPL specific thickness. D0 is calculated as follows:
DO = LNAPL % * porosity * soil core interval (ft.)
Where:
LNAPL % = oil saturation (in % of pore volume)
porosity = site-specific total porosity (in %)
soil core interval = interval of LNAPL impacted core (in feet)
Tables 6 and 7 summarize soil core data for both the Lower Refinery and Crawford Area monitoring
wells. Table 8 provides the results of measured LNAPL thickness in wells compared to the integrated
specific thickness D0. The results demonstrate that the specific thickness of LNAPL in the soil
(expressed in units of feet) is a small fraction of the measured LNAPL thickness in monitoring wells.
Table 8. Comparison of Measured versus Specific LNAPL Thickness (D0) (RETEC, 2004b, 2004d)
Location
Crawford
Lower
Refinery
Lower
Refinery
Lower
Refinery
Lower
Refinery
Monitoring
Well
SC-24
MW-178
ROW-006A
MW-179
SC-80B
Soil Type
(USCS Code)
Loess (CL)
Silty Sand (SM)
Silty Sand (SM)
Silty Sand (SM)
Fine Sand (SW)
bo
Measured
LNAPL
Thickness
(ft.)
15
4.16
5.6
3.69
9.35
D0
(ft'm2)
0.09
0.51
0.24
0.5
1.55
Ratio of
bo/Do
164
8
23
7
6
Remediation
Technology
MPE
DPR
DPR
DPR
DPR
Notes:
USCS- Unified Soil Classification System (CL-Clay, SM - silty sand, SW-well graded sand)
D0 is defined as the integral of the LNAPL saturation over the depth of the soil column, also known as the specific thickness of
LNAPL over a given soil column area.
MPE - Multi-Phase Extraction
DPR - Dual Pump Recovery
For this case study, relative change in measured LNAPL thickness from observation wells within the
Lower Refinery Area is used as a measurable indicator of recovery well performance (i.e., LNAPL
recovery overtime). This comparative relationship is valid only because of the consistency in soil and
fluid properties within this area. Note that any measured LNAPL thickness presented throughout this
evaluation is not indicative of actual LNAPL thickness in the subsurface. In addition, measured LNAPL
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
20
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
thickness alone is not indicative of potential recoverability or the true volume of LNAPL in the subsurface.
Soil and fluid properties must also be known to make accurate estimates of these values.
3.4 LNAPL CHARACTERIZATION
Table 9 compares characteristics of LNAPL at the dual-pump recovery and multi-phase extraction areas.
Table 9. LNAPL Characterization for the Dual-Pump Recovery and Multi-Phase Extraction Areas
(RETEC, 2004b, 2004d)
Parameter
LNAPL Type
LNAPL Specific Thickness (Do)
LNAPL Pore Fluid Saturations
LNAPL Density
LNAPL Viscosity
Interfacial Tension (LNAPL/Water)
Surface Tension (Air/LNAPL)
Denzene Percentage of Total LNAPL
TEX Percentage of Total LNAPL
Dual -Pump Recovery Area
light crude oil, slightly to
moderately weathered
initially: unknown
Presently: 0.70 feet (avg.)
<0.1to30.9%
0.81 grams per milliliter (g/mL)
1.18 to 149 centipoise
14.3 to 20.0 dynes/cm
20.1 to 20.4 dynes/cm
2.82%
20.0%
M u It i -Phase Extraction Area
middle distillate range diesel or
fuel oil, heavily weathered
Initially: 0.09 feet
Presently: 0.04 feet
<0.1to1.4%
0.92 grams per milliliter (g/mL)
6.3 to 13.9 centipoise
17.3 to 20.2 dynes/cm
27.4 to 30.7 dynes/cm
0.36%
1.12%
Note: TEX - Toluene, Ethylbenzene, and Xylenes
The LNAPL in the multi-phase extraction system is less volatile, more viscous, and more degraded than
in the dual-pump recovery area, which limits the effectiveness of a high-vacuum extraction technology.
To overcome these factors, a 10 horsepower liquid ring vacuum pump was utilized to provide 26-inches
of mercury vacuum to the subsurface to enhance groundwater and LNAPL recovery rates.
3.5 LNAPL VOLUME ESTIMATES
Even though up to 15 feet of LNAPL were measured at multi-phase extraction wells in the Crawford Area,
the D0 specific thickness equals only 0.09 ft3/ft2. For an estimate of LNAPL volume, the estimated plume
extent equals approximately 50 ft. by 50 ft. is multiplied by D0 specific thickness. Under these
assumptions, the estimated volume of LNAPL in the Crawford Area was approximately 1,700 gallons.
In the dual-pump recovery area, LNAPL volume was estimated using an average D0 from all soil cores in
the Lower Refinery Area (Table 8, equal to 0.70 ft3/ft2) and a total plume size estimated at 5 acres. The
current in-place LNAPL volume estimate within each recovery well's radius of capture is 1.1 million
gallons, which is significantly higher than in the multi-phase extraction area. Additional LNAPL exists in
the unsaturated zone and outside of each well's radius of capture. The larger amount of LNAPL in the
dual-pump recovery area is due to a larger plume size, larger pore size, higher permeability soils and
higher LNAPL saturations (see Table 6 and Table 7).
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
21
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
4.0 REMEDIATION GOALS AND PERFORMANCE OBJECTIVES
Remediation goals and performance objectives for the former refinery are individual for each interim
measure, but are designed and operated to abate imminent threats to human health and the
environment. The overall remediation goal of both multi-phase extraction and dual-pump recovery
systems is source reduction through LNAPL recovery to reduce mobility and risk to the environment.
During the technology evaluation process in the CMS, particular technologies are evaluated as potential
final corrective measures technologies if they have a likelihood of meeting proposed LNAPL remediation
goals and endpoints in a reasonable timeframe. Remediation goals and performance objectives provide
a decision-making framework to guide a practicable and attainable approach for RCRA Corrective Action
and long-term management of contaminated media and LNAPL in the subsurface.
The Handbook of Groundwater Protection and Cleanup Policies for RCRA Corrective Action (U.S. EPA,
2004) provides guidance on defining short, intermediate and long-term protection goals, as well as a
timeframe for reaching endpoints at large RCRA corrective action sites. Overall protection goals are
broad objectives, while endpoints are specifically identified to measure the progress towards meeting the
goals. Short-term goals focus on immediate or imminent threats to human health and the environment,
which are characterized under Groundwater and Human Health Environmental Indicators (Els) (U.S.
EPA, 2001) and, at RCRA sites, are typically addressed under RCRA interim measures. The short-term
protection goal has been demonstrated through the issuance of the Human Health (U.S. EPA, 2002) and
Groundwater (U.S. EPA, 2004b) Els, which indicates that current human exposures and contaminated
groundwater are under control at the former refinery. Intermediate protection goals are broader and
longer-term than short-term goals and, for the former refinery, were defined for final corrective measures
which have reachable and measurable endpoints in a tangible time frame. Long-term protection goals
include final cleanup goals that ensure long-term protection of human health and the environment, control
the source of releases, and achieve media cleanup objectives.
4.1 LNAPL REMEDIATION GOAL
The overall remediation goal for LNAPL at the former refinery is to recover LNAPL to the maximum
extent reasonably, technically, and economically feasible and consistent with prudent engineering
practices (i.e., reduce LNAPL to its practicable limit of recoverability). For both the multi-phase extraction
and dual-pump recovery systems, the American Petroleum Institute's (API) LNAPL Distribution and
Recovery Model (Charbeneau, 2003) was used to define the practical limit of LNAPL recovery.
4.2 DUAL-PUMP RECOVERY AND MULTI-PHASE EXTRACTION PROTECTION GOALS
The former refinery is currently an interim status facility under RCRA, although final protection goals have
been proposed to U.S. EPA and MDNR and are currently under review. The proposed protection goals
for the LNAPL include the following:
Short-term protection goals:
Eliminating LNAPL seeps to Sugar Creek (source/boundary control)
Intermediate protection goals:
Eliminating LNAPL in the subsurface that may serve as a risk to surface water
Long-term protection goals:
Reducing interior sources of LNAPL such that dissolved-phase hydrocarbon
contaminants migrating from groundwater to surface water meet ECOTOX criteria in
surface water (U.S. EPA, 1996)
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 22
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
4.3 DEFINING APPROPRIATE ENDPOINTS
Endpoints are defined as site-specific, measurable criteria and milestones that demonstrate progress
towards meeting protection goals. The following subsection outlines the proposed endpoints for
shutdown of the dual-pump recovery and multi-phase extraction systems. The multi-phase extraction
system was shutdown in January 2003 after six-months of no LNAPL recovery.
4.3.1 API Distribution and Recovery Modeling
Dual-pump recovery is designed to meet the intermediate protection goal of LNAPL source reduction to
the practicable limit of LNAPL recovery. In order to define the practicable limit of LNAPL recovery, the
API LNAPL Distribution and Recovery Model (Charbeneau, 2003) was used to predict LNAPL recovery
from the dual-pump recovery system. The API modeling process includes two steps, distribution
modeling of LNAPL in the subsurface, and recovery modeling under dual-pumping conditions, as
summarized below. The API model results are then used to predict long-term LNAPL recovery to
propose shutdown criteria (endpoints) for the dual-pump recovery system.
The distribution portion of the model predicts LNAPL saturation and permeability based on measured
LNAPL thickness in a monitoring well and several site-specific soil, fluid (i.e., LNAPL and groundwater),
and soil-fluid interaction parameters. Using the results of the distribution model, the recovery portion of
the API model predicts LNAPL recovery from individual wells overtime. After predicting LNAPL recovery
for each recovery well, the model results were validated against one year of actual LNAPL recovery data
(Figure 7) (RETEC, 2004d).
Inputs necessary to run the API distribution model include the following:
Measured LNAPL thickness
Soil input parameters: porosity, the van Genuchten parameters "N" and "a," irreducible water
saturation, and residual LNAPL saturation in the vadose and saturated zones
Fluid input parameters: LNAPL density, air/water surface tension, air/LNAPL surface tension,
and LNAPL/water surface tension
These parameters were taken from field measurements and observations, the results of the lab analysis
of the soil plug samples and the fluid samples, or were based on professional judgment. The results from
the API distribution model for each recovery well are then used as the basis for the API recovery model.
For recovery modeling, the data required to predict LNAPL recovery overtime includes the radius of
capture (ROC) for the well, radius of influence (i.e., cone of drawdown), the LNAPL viscosity, and water
production rate. For a water-enhanced system, the effective depth of penetration of the well into the
aquifer must also be specified. The API recovery model gives estimates of the total volume of LNAPL
within the radius of capture of each recovery well, the amount of LNAPL that is recoverable at each well,
the rate of recovery, the total time in which that LNAPL can be recovered, the measured LNAPL
thickness overtime in the recovery well, and the LNAPL recovery rate overtime.
The API model was applied at the six remaining operational dual-pump recovery wells. Validation was
performed using 18 months of actual LNAPL recovery data (from June 2003 to December 2004). The
total LNAPL recovery of all six wells from June 2003 to December 2004 was equal to 133,533 gallons,
which is approximately 6-percent less than the API model prediction of 141,500 gallons. Modeled results
for the system, along with the first year validation are shown in Figure 7. Overall the API model results
show good calibration to actual LNAPL recovery results.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 23
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 7. LNAPL Recovery for Dual-Pump Recovery System: Modeled vs Actual (RETEC, 2004d)
350,000
= 300,000
_o
"55
7 250,000
> 200,000
£
o>
8 150,000
a>
100,000
v. 50,000
Actual LNAPL recovery
over 18 months =
133,533 gallons
Total Modeled NAPL Recovery (All Six Wells - Dual
Pumping)
- - Total Actual NAPL Recovery (All Six Wells -1.5 yrs)
3
Time [yr]
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 24
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 8. Predicted Yearly LNAPL Recovery from the Dual-Pump Recovery System
(RETEC, 2004d)
100,000
10 11
12 13 14 15
Years
Based on predictive LNAPL recovery modeling of the performance of the dual-pump recovery system
(Figure 8), the proposed endpoint will be when LNAPL reaches an asymptotic rate of recovery. Based
upon the API modeling results, the asymptote for LNAPL recovery is estimated to be reached at some
point after six to ten years of operation, although actual shut-down of the recovery wells will be based on
empirical recovery data. At that time, the recovery wells will be transitioned to skimming wells using the
former refinery's vacuum truck to continue LNAPL source reduction. Long-term management and
vacuum truck removal of residual LNAPL from monitoring and inactive recovery wells will be determined
in a forthcoming Long-Term Management Plan, which will be submitted to the agencies in 2005.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
25
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
5.0 RECOVERY SYSTEMS DESCRIPTION, PERFORMANCE, AND COST
The following section provides a summary of the performance and cost of the dual-pump recovery and
multi-phase extraction systems at the former refinery. Lessons learned and comparative performance
are also discussed herein.
5.1 DUAL-PUMP RECOVERY SYSTEM
The dual-pump recovery system has operated at the former refinery for more than 20 years. Over that
period of time, recovery well equipment has been replaced and modified following internal evaluations to
optimize the system and improve system performance. This section primarily concentrates on the
recovery well system performance since the system modifications were implemented, with pertinent
historical information provided as appropriate. This section also includes items that affect the system as
a whole; individual active recovery well performance is discussed in the next section.
5.1.1
Dual-Pump Recovery System Description
LNAPL recovered via the recovery well system is pumped directly from each well to LNAPL storage Tank
95R. Recovered LNAPL is stored in Tank 95R until it is sent off site for recycling. The Tank 95R gauge
is used to measure the quantity of water and LNAPL within the tank on a weekly basis. These
measurements are used to check the LNAPL mass balance.
The water treatment system includes a water totalizer, clarifier tank, and air stripper. Recovered
groundwater is piped from each active recovery well to the clarifier tank prior to batch-transfer to the air
stripper. Groundwater from the air stripper is discharged to the City of Independence publicly owned
treatment works (POTW) for final treatment and disposal.
5.1.2
Dual-Pump Recovery System Performance
Based on historical records presented in the Quarterly Progress Reports (BP, 2004), the dual-pump
recovery system has recovered 1.82 million gallons of LNAPL and over 200 million gallons of
groundwater between 1982 and 2003. The following table provides annual totals of numbers of operating
recovery wells, water recovery and LNAPL recovery.
Table 10. Dual-Pump Recovery System Annual Water and LNAPL Recovery (RETEC, 2004c)
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
No. of
Operating
Recovery
Wells
2
2
3
3
3
3
15
15
15
15
13
13
Water
Recovery
(gallons)
81,700
81,700
122,600
122,600
122,600
122,600
612,910
13,594,543
22,555,415
20,718,219
18,537,399
12,283,788
LNAPL
Recovery
(gallons)
12,000
12,000
18,000
15,000
15,000
15,000
92,365
179,799
180,000
180,000
180,000
150,000
Gallon
water /
gallon
LNAPL
6.8
6.8
6.8
8.2
8.2
8.2
6.6
76
125
115
103
82
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
26
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Year
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
No. Of
Operating
Recovery
Wells
13
10
9
9
9
9
9
9
7
6
6
Water
Recovery
(gallons)
3,823,121
13,810,350
15,687,000
10,745,430
13,531,570
13,609,270
11,573,710
8,421,900
2,607,390
12,100,000
12,462,984
LNAPL
Recovery
(gallons)
50,927
119,552
134,672
68,492
94,239
95,286
72,301
50,362
11,186
73,500
79,500
Gallon
water /
gallon
LNAPL
75
116
116
157
144
143
160
167
233
165
157
Total = 207,328,799 1,899,181
Total Water/LNAPL ratio = 109.2
The system recovered a total of 79,500 gallons of LNAPL in 2004. The increased LNAPL recovery from
2002 to 2004 is believed to be from operational enhancements detailed in the next section. The total
groundwater to LNAPL recovered ratio for the dual-pump recovery system is 109:1. Figures 9 and 11
show the respective annual LNAPL and groundwater recovery for the dual-pump recovery system from
1982 to 2004. Figure 10 shows the cumulative LNAPL recovery from 1982 through 2004, as well as the
superimposed API-model predicted recovery over the next ten years. In addition, the number of
operating recovery wells per year is plotted at the base of each figure. Model predictions were based on
LNAPL saturations collected in June 2003, so there is approximately 18 months of overlap for validation
purposes.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
27
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 9. Dual-Pump Recovery System Annual LNAPL Recovery (RETEC, 2004c)
180 000 -
170 000 -
160 000 -
"t/T 130 000 -
o
ion 000 -
O)
* ' no 000 -
o
EL 80 000 -
"Z. 70 000 -
30 000 -
20 000 -
j-jJLlLlLlL
CO CO CO CO CO CO C
C
D a
Ba
a
~) C
? g
D T-
3 8
r
3 §
-J O
§c
C
? i g
T CI
3 §
3 h
§a
a
a
§C
C
D a
? §
5 c
§c
c
^^
T^
3 8 8 £
3 8
2 2 3 3 3 3 15 15 15 15 13 13 13 10 9 9 9 9 9 9 7 6 6
Number of Wells In Operation
Figure 10. Lower Refinery Recovery Well Network Cumulative LNAPL Recovery
2,500,000
2,000,000
1,500,000
1,000,000
500,000
e a
COCOCOCOCOCQCQCQCQCQCQCQCQCQCQOOOOOOOOOO-*-*-*-*-*
aio^JoocDO-^roco^aio^JoocDO-^roco^aio^JoocDO-^roco^
Year
*Cumulative LNAPL Recovery (gal.) - B - Model-Predicted Future Recovery (gal.)
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
28
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 11. Dual-Pump Recovery System Annual Groundwater Recovery (RETEC, 2004c)
100,000,000
in
O
"ra
O)
0)
§
c
2
o
10,000,000
1,000,000
100,000
10,000
cococococococococncncncncncncncncncnooooo
0)0)0)0)0)0)0)0)0)0)0)0)0)0)0)0)0)0)OOOOO
2 2 3 3 3 3 15 15 15 15 13 13 13 10 9 9 9 9 9 97 6 6
Number of Wells in Operation
The effectiveness of LNAPL source recovery is indicated in reductions in measured LNAPL thickness
overtime. Table 11 shows measured LNAPL thickness reductions overtime in monitoring wells in the
dual-pump recovery network.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
29
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Table 11. Measured LNAPL Thicknesses Reductions in Dual-Pump Recovery System Observation
Wells over Time (RETEC, 2004d)
Observation
Well
A-014
A-015
A-015A
A-038
A-039
A-040
A-041
HB-005
MW-179
ROW-006
ROW-006A
ROW-007
ROW-008
ROW-009
ROW-010
Recovery
Well
NA
R-001
R-001
NA
R-002
R-005
NA
NA
NA
R-006
R-006
R-007
R-008
R-009
R-010
Measured LNAPL Thickness (feet)
1986/
1987
0.07
5.54
10.87
6.27
2.25
0.26
1.77
NM
NM
NM
NM
NM
NM
NM
NM
1989/
1990
0.8
NM
NM
0
NM
NM
NM
NM
NM
7.3
NM
8.7
7.8
7.6
10.3
1992/
1993
0
NM
NM
0.2
NM
NM
NM
NM
NM
0
NM
10.3
9.3
7.1
3.7
1996/
1997
0.24
2.1
4.51
2.74
1.62
0
0.72
0.84
NM
NM
NM
9.35
10.65
3.71
0
1999/
2000
0.02
NM
NM
0.27
NM
NM
NM
NM
NM
NM
NM
4.83
6.65
0.01
0.09
2003/
2004
0
NM
NM
0
0
NM
0.55
NM
7.12
NM
4.99
4.42
3.29
0
NM
% Decrease
in LNAPL
Thickness
1 00%
62%
59%
1 00%
100%
100%
69%
NA
NA
100%
NA
49%
58%
100%
99%
Notes:
NM - Not Measured
% Decrease in LNAPL thickness is calculated using 2003/2004 data, if it exists. Earlier numbers are used in some cases.
The data in Table 11 demonstrates that initial LNAPL thickness measurements averaged 7 feet before
source reduction and system optimization in 2002 (RETEC, 2002). Measured LNAPL thickness has
been significantly reduced overtime in each observation well located within the vicinity of the recovery
well network. The observation wells show a minimum 49 percent and average 83 percent decrease in
measured LNAPL in 2003/2004 when compared to initial measurements (Table 11), with many of the
wells showing no measurable LNAPL during the most recent gauging events. The decrease in measured
LNAPL thickness in the observation wells indicates that the dual-pump recovery system is successfully
removing LNAPL volume and reducing saturation in the subsurface.
The recovery well system optimization (RETEC, 2002) determined that groundwater elevation and
LNAPL thickness in the recovery wells should be maintained at approximately 705 feet above mean sea
level (amsl) to increase LNAPL recovery rates. This corresponds to the interface between lithologic
Zones A and B, and was considered to be the optimum groundwater level for LNAPL recovery in the
recovery wells (RETEC, 2002). As the Missouri River elevation increases above 705 feet amsl, the
gradient to the dual-pump recovery wells is increased and LNAPL recovery is increased. Figure 12
shows an example of LNAPL recovery at recovery well R-007 when the water level was maintained at
705 ft. amsl from January to July 2003 and lowered to 699 ft. amsl from July through December 2003.
The chart shows that LNAPL recovery increased with pump elevations lowered to 699 ft. amsl, which
corresponds with the lowest elevation the pumps can be placed in recovery well R-007.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
30
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 12. Well R-007 Fluid Level Measurements and LNAPL Recovery for 2003 (RETEC, 2004c)
R-007 Fluid Level Measurements
711 -r
r 1,400
695
Jan-03 Feb-03 Apr-03 May-03 Jul-03 Sep-03 Oct-03 Dec-03
Date
i LNAPL Recovered (gal/week) Water Elev. (ft. amsl) »LNAPL Elev. (ft. amsl)
Based on the LNAPL/water recovery ratio, well R-007 is one of the most effective recovery wells in the
system. It has the greatest daily LNAPL recovery rate at approximately 97 gallons, although it also has
the greatest daily water recovery rate at 9,900 gallons. In 2004, it recovered over 42,000 gallons of
LNAPL, which is greater than the estimated LNAPL recovery of all other recovery wells combined.
5.1.3
Dual-Pump Recovery System Costs
The following capital costs were estimated for the dual-pump recovery system, based on installation
dates of dual-pump recovery wells:
Table 12. Dual-Pump Recovery System Estimated Capital Costs
Year
1982
1984
1987
1988
No. of
Wells
Drilled
3
1
4
8
Drilling
Cost per
well
$ 10,000
$ 12,000
$ 14,000
$ 15,000
Pump cost
per well
$ 2,000
$ 2,400
$ 2,800
$ 3,000
Electrical
Controls
per well
$ 6,000
$ 8,000
$ 10,000
$ 11,000
Labor per
well
$ 4,000
$ 5,000
$ 7,000
$ 8,000
Design/
Oversight
(25%)
$ 5,500
$ 6,850
$ 8,450
$ 9,250
Subtotal
per well
$ 27,500
$ 34,250
$ 42,250
$ 46,250
Subtotal (per
year)
$ 82,500
$ 34,250
$ 169,000
$ 370,000
Total =
NPV =
Note: Net Present Value (NPV) estimated at a 3.5 percent discount rate
$ 655,750
$ 1,258,514
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
31
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
In addition, the following are estimated annual O&M Costs:
Table 13. Dual-Pump Recovery System Estimated O&M Costs
Year
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
No. Of
Operating
Recovery
Wells
2
2
3
3
3
3
15
15
15
15
13
13
13
10
9
9
9
9
9
9
7
6
6
Annual
O&M Costs
$ 20,000
$ 20,000
$ 30,000
$ 30,000
$ 30,000
$ 30,000
$ 150,000
$ 150,000
$ 150,000
$ 150,000
$ 130,000
$ 130,000
$ 130,000
$ 100,000
$ 90,000
$ 90,000
$ 90,000
$ 90,000
$ 90,000
$ 90,000
$ 90,000
$ 142,000
$ 100,000
POTW
Water
Disposal
Costs
$ 163
$ 163
$ 245
$ 245
$ 245
$ 245
$ 1 ,226
$ 27,189
$ 45,111
$ 41,436
$ 37,075
$ 24,568
$ 7,646
$ 27,621
$ 31,374
$ 21,491
$ 27,063
$ 27,219
$ 23,147
$ 16,844
$ 5,215
$ 24,200
$ 25,125
Oil Sale
($0.50/gallon)
$ (6,000)
$ (6,000)
$ (9,000)
$ (7,500)
$ (7,500)
$ (7,500)
$ (46,183)
$ (89,900)
$ (90,000)
$ (90,000)
$ (90,000)
$ (75,000)
$ (25,464)
$ (59,776)
$ (67,336)
$ (34,246)
$ (47,120)
$ (47,643)
$ (36,151)
$ (25,181)
$ (5,593)
$ (36,750)
$ (52,933)
Subtotal
O&M Costs
$ 14,163
$ 14,163
$ 21,245
$ 22,745
$ 22,745
$ 22,745
$ 105,043
$ 87,290
$ 105,111
$ 101,436
$ 77,075
$ 79,568
$ 112,183
$ 67,845
$ 54,038
$ 77,245
$ 69,944
$ 69,576
$ 76,997
$ 81,663
$ 89,622
$ 129,450
$ 72,192
Tote/ O
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
recovery was low (i.e., approximately 18 gallons a week on average). The Missouri River elevation,
however, affected the recovery because the river stage was lower than the elevation of the groundwater
pump and no groundwater gradient could be generated. For example, when the pump in recovery well
R-007 was set at an elevation of 699 ft amsl the average LNAPL recovery increased to over 1000 gallons
a week. Maintaining drawdown at an elevation of 699 ft. amsl is limited by the capacity of the
groundwater recovery well pump capacity, available drawdown (i.e., the elevation of the bottom of the
well), and the treatment system capacity. In June 2004, for example, higher-than-normal Missouri River
elevations and limited treatment system capacity forced raising the pumps to an elevation of 709 ft. amsl.
In September 2004, the pump in R-007 was lowered back to an elevation of 699 ft. amsl.
5.2 MULTI-PHASE EXTRACTION SYSTEM
Multi-phase extraction is a remediation process that applies a high vacuum (i.e., 26-inches of mercury) to
wells or extraction points to remove LNAPL, impacted groundwater, and vapor from subsurface soil.
From January 2001 to April 2001, the multi-phase extraction system was cycled on six extraction points
(SC-14, SC 15, SC-16, SC-24, SC-25, and MW-078). From April 2001 to January 2003, the multi-phase
extraction system operated full-time on well MW-078. In January 2003, due to the asymptotic and low
LNAPL recovery, the system was shut off. Initially, LNAPL was periodically recovered from six wells and
piezometers: MW-078, SC-14, SC-15, SC-16, SC-24 and SC-25. Multi-phase extraction operations
cycled for 15 minutes on three well pairs, and then the system shutdown for 15 minutes every hour.
Collected groundwater was treated and discharged to the city of Independence POTW. LNAPL was
transferred to Tank 95R for eventual recycling off site at an approved facility. Overall, application of a
high-vacuum multi-phase extraction system successfully recovered LNAPL to the practicable limit of its
technology.
5.2.1 Multi-Phase Extraction System Performance
During its two-year operation, the Crawford multi-phase extraction system recovered a total of 151
gallons of LNAPL and 215,000 gallons of groundwater. The total groundwater to LNAPL recovered ratio
for the multi-phase extraction system was 1,430:1. Performance monitoring data for the Crawford multi-
phase extraction system and LNAPL recovery overtime for2001 and 2002 is shown on Figure 13.
During the calendar year 2001, approximately 148 gallons of LNAPL were recovered from the multi-
phase extraction system. However, LNAPL recovery reached an asymptote after a few months of
operation, and only 2.5 gallons of additional LNAPL was recovered from the multi-phase extraction
system in 2002. Overall, the multi-phase extraction system was not effective at removing any addition
LNAPL from the area and the system was shutdown on January 3, 2003.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 33
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
Figure 13. Actual Multi-Phase Extraction LNAPL Recovery Compared to API Model Results
(RETEC, 2004b)
250
200 -
150 -
0- 100
50 -
.May 22, 2001 System Begins
Continuous Operation
13-Jan-01 13-Apr-01 12-Jul-01 10-Oct-01 8-Jan-02 8-Apr-02 7-Jul-02 5-Oct-02 3-Jan-03 3-Apr-03 2-Jul-03
-Model (15 ft case)
LNAPL Recovered (gal)
- System Down
5.2.2 Multi-Phase Extraction System Costs
The following capital costs were estimated for the multi-phase extraction system:
Table 14. Multi-Phase Extract System Estimated Capital Costs
Well Installation
Pilot Test/Design
MPE Equipment
Startup/Shakedown
Total Capital =
NPV =
Capital Costs
$ 10,000
$ 10,000
$ 53,000
$ 15,000
$ 88,000
$ 101,478
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
34
March 2005
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
In addition, the following O&M costs were estimated for the multi-phase extraction system:
Table 15. Multi-Phase Extraction System Estimated O&M Costs
2001 O&M
2002 O&M
Total O&M =
NPV =
O&M Costs
$
$
$
$
36,000
36,000
72,000
81,575
Notes:
Annual O&M costs include labor, electricity, replacement materials.
NPV = Net Present Value at a 3.5 percent discount rate
Accounting for inflation, the multi-phase extraction system had a capital cost ($101,478) and O&M cost
($81,575) equal to a total cost of $183,053. Assuming a total of 151 gallons of LNAPL have been
recovered to date, the cost per gallon to recover LNAPL using multi-phase extraction is estimated at
$1,212 per gallon.
5.2.3
Multi-Phase Extraction System Observations and Lessons Learned
Although the multi-phase extraction system reached its remediation goal (i.e., the practicable limit of
LNAPL recovery), the lack of significant LNAPL recovery confirmed that it is not effective at removing
LNAPL in areas of limited LNAPL saturation because of the low permeability of the silty loess soils, the
low percentage of LNAPL saturation (maximum of 1.4 percent), and discontinuous nature of LNAPL in
the subsurface. Even with high-vacuum enhancement (i.e., 26-inches of mercury vacuum), the system
was only able to recover 148 gallons of LNAPL in the first year, and only 2.5 gallons in the second year.
The total LNAPL recovery amounted to less than 10-percent of the original estimated in-place LNAPL
volume, with indicates that the majority of the LNAPL is unrecoverable by vacuum enhanced extraction
and not recoverable via in-situ technologies.
After LNAPL removal via multi-phase extraction reached an asymptote, the system was shutdown.
Within months after shutdown additional LNAPL accumulated in wells adjacent to Sugar Creek over time,
indicating a potential for LNAPL seeps in Sugar Creek. Therefore, an alternative remediation system
was installed to control and contain LNAPL seeps. In 2003, a series of hydraulic control pumps were
installed adjacent to Sugar Creek to reverse the hydraulic gradient and mitigate the seeps.
5.3 COMPARISON OF DUAL-PUMP RECOVERY TO MULTI-PHASE EXTRACTION
Dual-pump recovery has over 20 years of operational history and performance monitoring at the site,
recovering 1.9 million gallons of LNAPL over that time. The continuing success of dual-pump LNAPL
recovery, with 79,500 gallons of LNAPL recovered in 2004, underscores its longevity and effectiveness
as a source removal technology at the former refinery site. However, multi-phase extraction was less
successful, removing only 151 gallons of LNAPL over two years. The groundwater/LNAPL recovery ratio
of the dual-pump recovery system (109:1) is also much more effective than the multi-phase extraction
system (1,430:1).
It should be noted that dual-pump recovery is not expected to be more effective in fine-grained soils than
multi-phase extraction, based on lower permeability soils and lower groundwater recovery rates.
However, it was determined that a hydraulic control submersible pump could achieve the protection goal
of no LNAPL seeps to Sugar Creek at lower O&M and capital costs. Due to the fine-grained (i.e., silt and
clay) soils at the site, multi-phase extraction was determined to not be an effective LNAPL remediation
technology.
U.S. Environmental Protection Agency
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division
March 2005
35
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
5.4 TECHNOLOGY PERFORMANCE AND COST
5.4.1 Meeting Protection Goals and Endpoints
Overall the dual-pump recovery system has recovered significant quantities of LNAPL compared to multi-
phase extraction, although the initial source LNAPL volume estimate was orders of magnitude greater in
the dual-pump recovery area (1,700 gallons compared to 3.4 million gallons). The dual-pump recovery
system has recovered 1.899 million gallons of LNAPL to date and is model-predicted to recover an
additional 321,900 gallons of the remaining 1.1 million gallons over the next six years. The LNAPL
recovery has been corrected for small amounts of water recovered through the LNAPL skimming pumps
which is removed from Tank 95R. In addition, centrifuge tests performed on samples of the recovered
LNAPL from the recovery well skimmer pumps and Tank 95R do not show any entrained or emulsified
water in the recovered LNAPL. Therefore, BP has high confidence in the total recovered quantity of
LNAPL (1.9 million gallons).
Overall, the dual pump recovery system is expected to recover a total of 2.25 million gallons of LNAPL,
which is equal to 67-percent of the estimated LNAPL source volume within the recovery wells radius of
capture, based on the API Model-estimated LNAPL specific thickness over the plume area (3.34 million
gallons). The percent recovery of the initial LNAPL spill volume is unknown, due to lack of spill data, and
does not take into account additional LNAPL in the unsaturated zone, smear zone or outside each
recovery well's radius of capture. By comparison, the total LNAPL recovery for the multi-phase extraction
system accounted for less than 10-percent of the original estimated in-place LNAPL volume.
For the dual-pump recovery system, predicting remediation lifespan and time to reach endpoints was
done using the API Distribution and Recovery Model. Modeling results suggest that the remaining
LNAPL may be close to residual saturation and effectively immobile in the subsurface after 6 to 10 years
of additional dual-pump recovery. When recovery data indicates that the system reaches an asymptote
and a point of diminishing returns for dual-pump LNAPL recovery, the recovery wells will be transitioned
to LNAPL skimming wells and LNAPL recovery will take place using the site vacuum truck. The
frequency and duration of vacuum truck events will be determined at that time.
The multi-phase extraction system met its shutdown criteria (i.e., asymptotic LNAPL recovery) after two
years of operation, through asymptotic LNAPL recovery (e.g., the system was only able to recover 2.5
gallons of LNAPL over the calendar year 2002). To meet the additional protection goal of no LNAPL
seeps to Sugar Creek, hydraulic control pumps were installed in 4-inch diameter monitoring wells
adjacent to the creek to reverse the hydraulic gradient away from the creek. Overall the hydraulic control
pump requires less maintenance and oversight. The hydraulic control system has been successful at
reversing the gradient and the short term protection goals for the Crawford Area have been met (RETEC,
2004b).
5.4.2 Cost per Gallon of LNAPL Removed
Although the total cost of the dual-pump recovery system ($3,554,349) was much greater than the cost of
the Crawford multi-phase extraction system ($183,053), the normalized cost per gallon to recover LNAPL
from the Lower Refinery Area ($1.87 per gallon) was significantly less than the Crawford Area ($1,212).
The cost performance indicates that LNAPL is much more effectively recovered from higher-permeability
sand and gravel than low-permeability silt and clay.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 36
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
6.0 OBSERVATIONS AND LESSONS LEARNED
The purpose of this evaluation was to provide a comprehensive evaluation of the cost and performance
of the multi-phase extraction system and dual-pump recovery system, and show how soil type can have a
considerable effect of LNAPL recovery, cost and performance. Lessons learned are also discussed
herein.
Based on this case study, the following observations are made regarding the dual-pump recovery
system:
The dual-pump recovery system has collectively recovered approximately 1.899 million
gallons of LNAPL and over 200 million gallons of groundwaterfrom 1982 through 2004.
The measured LNAPL thickness has significantly reduced over time (i.e., average of 83
percent) in each observation well in the vicinity of the recovery well network. The
decreasing trend of measured LNAPL thickness in the observation wells suggests that
the system is successfully removing LNAPL at the former refinery.
If groundwater drawdown in a recovery well is maintained at a consistent level below the
Missouri River stage, LNAPL recovery is proportional to the river stage. This is due to
the increased groundwater gradient created by higher river stages. Since the river stage
changes seasonally, changes in LNAPL recovery are anticipated and the system
adjusted accordingly to increase LNAPL recovery rates.
Groundwater drawdown is necessary to optimize LNAPL recovery in a dual-pump
system. However, excessive groundwater drawdown will lead to increased groundwater
disposal costs. Thus, groundwater pumping rates at each recovery well were evaluated
to determine the optimal rate and level for each recovery well to ensure efficient
operation of the system.
Constant drawdown minimizes maintenance of the system and does not appear to
negatively impact LNAPL recovery. Constant drawdown also allows personnel to more
readily determine if a well and/or its equipment are not functioning properly during routine
data review. Recovery wells are operated to create the optimal groundwater drawdown
to maximize LNAPL recovery.
The API Distribution and Recovery Model predicted six to ten more years of dual-phase
pumping before asymptotic rates of LNAPL recovery are achieved, although actual
shutdown will be based on empirical LNAPL recovery data. At that time, recovery wells
will be transitioned to skimming wells using the site's vacuum truck.
The estimated total capital cost ($1,258,514) and O&M cost ($2,295,835) for the dual-
pump recovery system, accounting for inflation, is equal to $3,554,349. The cost to
recover 1.899 million gallons of LNAPL using dual-pump recovery is approximately $1.87
per gallon.
The dual-pump recovery system should continue to operate under current conditions
because it still meets the objective of LNAPL recovery and source removal. The system
recovered 79,500 gallons of LNAPL in 2004.
As determined from this evaluation, the dual-pump recovery system is operating effectively, recovering
significant quantities of LNAPL, and making progress toward achieving LNAPL remediation goals and
long-term protection goals and endpoints.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 37
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
The following conclusions were made regarding the multi-phase extraction system:
Based on the soil type in the Crawford Area (silt loess), a multi-phase extraction system
was required to remove LNAPL and groundwaterfrom the subsurface.
During its two-year operation, the Crawford multi-phase extraction system operated at
26-inches of mercury vacuum and only recovered 151 gallons of LNAPL and 215,000
gallons of groundwater. During the calendar year 2002, the system only recovered 2.5
gallons of LNAPL. The system reached an asymptote of impractical LNAPL recovery
and was shut down on January 3, 2003.
Cycling had little to no effect on increasing LNAPL recovery rates from multiple multi-
phase extraction wells, and the system was switched to full-time operation on one
extraction well.
Application of a high-vacuum multi-phase extraction system successfully recovered
LNAPL from low-permeability soils over a short period of time, and the system met its
LNAPL remediation goal of asymptotic recovery. However, the system was unable to
extract all recoverable LNAPL and eliminate potential seeps to Sugar Creek. At that
time, an alternative more cost-effective remediation system (i.e., hydraulic control
pumping) was implemented to control LNAPL seeps.
The multi-phase extraction system had a capital cost ($101,478) and O&M cost
($81,575) equal to a total cost of $183,053. The cost to recover total of 151 gallons of
LNAPL using multi-phase extraction is approximately $1,212 per gallon.
The comparison of the two systems provides the following key conclusions and recommendations:
Measured LNAPL thickness in a monitoring well does not necessarily correspond to the
actual amount of LNAPL in the subsurface. To determine accurate estimates of LNAPL
volume in the subsurface, soil cores and LNAPL saturations need to be quantitatively
analyzed by a specialized laboratory. Initial estimates of LNAPL volume and soil type
will greatly influence remediation technology selection, recovery performance,
remediation goals and endpoints.
Site soil characteristics should be considered before implementing remediation
technologies, and in the prioritization of remediation efforts at a site. Overall, the
permeability and grain size of soils can greatly influence the distribution and
recoverability of LNAPL at a site.
The groundwater/LNAPL recovery ratio of the dual-pump recovery system (109:1) is
more effective than the multi-phase extraction system (1,430:1), because LNAPL
recovery and source removal is significantly more technically feasible in coarser grained
materials, such as sands, irrespective of the remediation technology. Although dual-
pump recovery proves to be continually effective at recovering LNAPL from sand, it is not
expected to be an appropriate technology for LNAPL recovery in silts and clays.
The dual-pump recovery system recovered 1.899 million gallons of LNAPL to date and is
predicted to recover an additional 321,900 gallons of the remaining estimated in-place
volume of 1.1 million gallons. Overall, the dual pump recovery system is expected to
recover a total 2.25 million gallons of LNAPL, which is equal to approximately 67-percent
of the estimated LNAPL source volume, based on the API Model-estimated LNAPL
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 38
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
specific thickness over the plume area (3.34 million gallons). The percent recovery of
the initial LNAPL spill volume is unknown, due to lack of spill data, and does not take into
account additional LNAPL in the unsaturated zone, smear zone or outside each recovery
well's radius of capture. By comparison, the total LNAPL recovery for the multi-phase
extraction system accounted for less than 10-percent of the original estimated in-place
LNAPL volume.
At the Sugar Creek site, LNAPL recovery using dual-pump recovery in coarser grained
materials, had a cost per gallon differential of almost three orders of magnitude (i.e.,
$1,212 per gallon for multi-phase extraction, compared to only $1.87 per gallon for dual-
pump recovery). Although the capital and O&M cost of the dual-pump recovery system
is greater than the multi-phase extraction system, the cost effectiveness for LNAPL
source removal using dual-pump wells in sand is much greater than multi-phase
extraction in silt/clay.
Although multi-phase extraction is believed to be the most effective and demonstrated
technology to remediate LNAPL from silt and clay, it was not effective for Crawford Area
at the former refinery. The dual-pump recovery network was not a viable alternative in
silt and clay soils at the Crawford Area. Therefore, alternative remediation and
protection goals such as LNAPL source control and containment in silt and clay soils
may be more appropriate rather than LNAPL source reduction.
Overall, protection goals and LNAPL endpoints for remediation at large-scale RCRA sites must reflect the
technical limitations of remediation technologies applied in each soil type and the LNAPL distribution in
the subsurface, with appropriate performance expectations, remediation timeframes, and shutdown
criteria. The case study demonstrates that LNAPL source reduction and recovery is a viable remediation
goal in sands whereas LNAPL source control and containment may be more attainable in silts and clays.
However, site-specific factors must be considered in the design and implementation of any in-situ
remediation system for LNAPL source reduction.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 39
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
This page is intentionally left blank.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 40
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
7.0 REFERENCES
Adamski, M., V. Kremesec, R. Kolhatkar, C. Pearson, and B. Rowan, 2001. "LNAPL
Saturation, Distribution, and Recovery in Fine Grained Soils," Proceedings of the
Petroleum Hydrocarbons and Organic Chemicals in Ground Water Conference and
Exposition, pp.178-192. November 14-16, 2001.
Ballestero, T.P., Fiedler, F.R., and Kinner, N.E., 1994, An Investigation of the Relationship
Between Actual and Apparent Gasoline Thickness in a Uniform Sand Aquifer, Ground
Water, v. 32, no. 5, pp. 708-718
BP Products North America, Inc., (1989-2004). Quarterly Progress Reports, Amoco Oil
Company Former Refinery, Sugar Creek, Missouri, RCRA Docket No. VII-89-H-0028.
Submitted quarterly.
Charbeneau, R.J., R.T. Johns, L.W. Lake, and M.J. McAdams, 1999. "Free-Product
Recovery of Petroleum Hydrocarbon Liquid." American Petroleum Institute, Publication
No. 4682. Ground Water Monitoring & Remediation, 20(3), Summer, pp. 147-158. June
1999.
Charbeneau, R.J., 2003. "Models for Design of Free-Product Recovery Systems for
Petroleum Hydrocarbon Liquids." American Petroleum Institute, www.api.org/lnapl
RETEC, 2002. Lower Refinery Recovery Well System Evaluation, Amoco Former Refinery,
Sugar Creek, Missouri. The RETEC Group, Inc. Draft, October 4, 2002.
RETEC, 2004a. Volume 9 - Lower Refinery RCRA Facility Investigation Report, Revision 1,
Amoco Former Refinery, Sugar Creek, Missouri. The RETEC Group, Inc. January 30,
2004.
RETEC, 2004b. Corrective Measures Study for the Crawford and Sugar Creek Area,
Revision 2, Amoco Former Refinery, Sugar Creek, Missouri. The RETEC Group, Inc.
March 8, 2004.
RETEC, 2004c. Annual Interim Measures Performance Monitoring Summary, 2003, Amoco
Former Refinery, Sugar Creek, Missouri. Golden, CO: The RETEC Group, Inc. March
24,2004.
RETEC, 2004d. Corrective Measures Study for the Lower Refinery Area, Amoco Former
Refinery, Sugar Creek, Missouri. The RETEC Group, Inc. pending Novembers, 2004.
ThermoRetec, 2000. Interim Measures Work Plan, Sugar Creek. Amoco Oil Company,
Former Refinery, Sugar Creek Missouri. January 24, 2000.
U.S. EPA. 1996. OSWER Ecotox Thresholds, Eco Update 32, EPA 540/F 95/038. Tier II
Secondary Chronic Value from Suter and Tsao. United States Environmental Protection
Agency. 1996.
U.S. EPA, 2002. Human Health Environmental Indicator Determination, Interim Final -
RCRA Environmental Indicator (El) RCRIS code (CA725). Amoco Sugar Creek former
refinery. Facility ID #MOD007161425. October 3, 2002.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 41
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
U.S. EPA, 2004. Issuance of the Groundwater Environmental Indicator for BP Sugar Creek
- RCRA Environmental Indicator (El) RCRIS code (CA750). Amoco Sugar Creek former
refinery. Facility ID#MOD007161425. March 1, 2004.
U.S. EPA, 2004. Handbook of Groundwater Protection and Cleanup Policies for RCRA
Corrective Action, for Facilities Subject to Corrective Action Under Subtitle C of the
Resource Conservation Recovery Act. EPA/530/R-01/015. United States Environmental
Protection Agency, Office of Solid Waste and Emergency Response. April 2004.
Woodward-Clyde, (1987). Refinery-Wide Geology, Hydrology, and Groundwater
Investigation, Amoco Oil Company, Sugar Creek former refinery, Sugar Creek, Missouri.
May 29, 1987.
Woodward-Clyde, (1989). Interim Measures Work Plan, Amoco Oil Company, Former
Refinery, Sugar Creek, Missouri. Amended Novembers, 1989.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 42
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
8.0 SITE CONTACTS
BP Sugar Creek Site Contact:
Lloyd Dunlap
BP Products North America, Inc.
1000 N. Sterling Road
Sugar Creek, MO 64054
(816)836-6044
Email: dunlaple@bp.com
Technology System and Remediation Design Contact:
Christopher J. Pearson, P.E.
The RETEC Group, Inc.
1726 Cole Boulevard
Building 22, Suite 150
Golden, CO 80401
(303)271-2100
E-mail: cpearson@retec.com
EPA Regulatory Contact:
Robert E. Aston
Project Manager
U.S. EPA Region 7
(913)551-7392
E-mail: aston.robert@epa.gov
State Regulatory Contact:
Brian McCurren
Project Manager
Missouri Department of Natural Resources
(573)751-4130
Email: brian.mccurren@dnr.mo.gov
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 43
-------
BP Products North America, Inc., Former Refinery, Sugar Creek, MO
9.0 ACKNOWLEDGEMENTS
This report was prepared for the U.S. Environmental Protection Agency's Office of Solid Waste and
Emergency Response, Office of Superfund Remediation and Technology Innovation. Assistance was
provided by The RETEC Group, inc.
U.S. Environmental Protection Agency March 2005
Office of Solid Waste and Emergency Response
Technology Innovation and Field Services Division 44
------- |