-------

-------
                    METAL BANK SUPERFUND SITE
                       RECORD OF DECISION

                        TABLE OF CONTENTS
PART I - DECLARATION 	1

I.    SITE NAME AND LOCATION	1

H.   STATEMENT OF BASIS AND PURPOSE	1

m.   ASSESSMENT OF THE SITE	1

IV.   DESCRIPTION OF THE SELECTED REMEDY	1

V.   STATUTORY DETERMINATIONS 	3

PART II - DECISION SUMMARY	4

I.    SITE NAME, LOCATION, AND DESCRIPTION	4

n.   SITE HISTORY AND ENFORCEMENT ACTIVITIES	4


     A.   Site History	4
     B.    Site Enforcement Activities	8

III.   COMMUNITY PARTICIPATION	10

IV.   SCOPE AND ROLE OF RESPONSE ACTION  	11

V.   SITE CHARACTERISTICS	11
     A.   Surface Water 	11
     B.   Geology	12
     C.   Hydrogeology 	13
     D.   River Tide	14
     E.   Extent of Contamination	14

          1.    Building Area	15
          2.    Surface Water	16

                                ii

-------
            3.    Groundwater  	16
            4.    Surface Soils  .	20
            5.    Subsurface Soils	21
            6.    Sediments  	22
            7.    Fish and Other Living Organisms ("Biota")	27

VJ.   RISK ASSESSMENT		29
      A.    Human Health Risk Assessment	 30
      B.    Aquatic Ecological Risk Assessment	33
      C.    Terrestrial Ecological Risk Assessment	'.	37

VII.   DESCRIPTION OF ALTERNATIVES	39

VIII.  ANALYSIS OF ALTERNATIVES  	42
      A.    Overall Protection of Human Health and the Environment	44
      B.    Compliance with ARARs	44
      C.    Long-Term Effectiveness and Permanence	48
      D.    Reduction of Toxicity, Mobility, or Volume through Treatment	49
      E.    Short-Term Effectiveness	49
      F.    Implementability	49
      G.    Cost	50
      H.    State Acceptance	50
      I.     Community Acceptance	51

IX.   SELECTED REMEDY AND PERFORMANCE STANDARDS .	51
      A.    Soil/Sediment Excavation Performance Standards	 51
      B.    Soil/Sediment Disposal Performance Standards	53
      C.    Oil Collection System Performance Standards	54
      D.    Underground Storage Tank Performance Standard	54
      E.    Site Access/Use	.  . 55
      F.    Additional Investigation	55
      G.    Monitoring Requirements	55
      H.    Operation & Maintenance      	56
      I.     Five Year Review	56

X.    STATUTORY DETERMINATIONS	57
                                      in

-------
      A.    Overall Protection of Human Health and the Environment	57
      B.    Compliance with ARARs	57
      C.    Cost Effectiveness	57
      D.    Utilization of Permanent Solutions  and  Alternative  Treatment (or  Resource
            Recovery) Technologies to the Maximum Extent Practicable  	57
      E.    Preference for Treatment as a Principal Element	58

XH.   SIGNIFICANT CHANGES  	58
                   PART ID - RESPONSIVENESS SUMMARY

I.     ORAL COMMENTS FROM THE JULY 27, 1995 PUBLIC MEETING	60

D.    WRITTEN COMMENTS FROM THE SITE OWNER	65

      A. Human Health Risk Assessment	66
      B. Aquatic Ecological Risk Assessment  	84
      C. Terrestrial Ecological Risk Assessment	 . 98
      D. Non-Aqueous Phase Liquids (NAPL)  	 103
      E. Remedial Alternatives  	 114
      F. Legal and Historical Factual Issues	 117

III.    OTHER WRITTEN COMMENTS	 126

      A. PRP Group  	126
      B. DRBC  	 140
      C. PADEP	 143


                                  FIGURES

Figure 1 - Site Location  	5

Figure 2 - Site Proximity to Public Water Intakes	.6

Figure 3 - Site Plan 	7

Figure 4 - Total VOCs in Groundwater	18

Figure 5 - Total SVOCs in Groundwater	19
                                      IV

-------
Figure 6 - Soil Borings/Test Pit Locations with Total PCBs * 25 ppm	24

Figure 7 - Mudflat Sediment Sampling Locations with Total PCBs z  1 ppm	26

Figure 8 - River Sediment Sampling Locations with Total PCBs ;>  1 ppm	28


                                     TABLES

Table 1 - Total PCBs in Buildings	15

Table 2 - Groundwater Contaminant Concentrations  	17

Table 3 - Groundwater Seep Concentrations	20

Table 4 - Surface Soil Concentrations (Southern Portion)	21

Table 5 - Subsurface Soil Contaminants  	22

Table 6 - PCBs in Subsurface Soil (Southern Portion)  . . .	23

Table 7 - Total PCBs in Mud Flat Sediments .	25

Table 8 - Total PCBs in River Sediments	27

Table 9 - Sediment Contaminants	29

Table 10 - Summary of Human Health Cancer Risks	  31

Table 11 - Total PCBs in Sediment with Increased Distance
         from the Metal Bank Property	34

Table 12 - Hazard Quotients for Aquatic Receptors Exposed to Sediment	36

Table 13 - Groundwater Environmental Effects Quotients	38

Table 14 - Sediment Environmental Effects Quotients  	38

Table 15 - Summary of Costs	  50

-------
                         METAL BANK SUPERFUND SITE
                             RECORD OF DECISION

                             PART I - DECLARATION
I.      SITE NAME AND LOCATION

Metal Bank Superfund Site
Philadelphia, Pennsylvania

II.     STATEMENT OF BASIS AND PURPOSE

       This Record of Decision ("ROD") presents the final remedial action selected for the
Metal Bank Superfund Site ("Site"), located in northeastern Philadelphia, Pennsylvania.  This
remedial action was chosen in accordance with the Comprehensive Environmental Response,
Compensation, and Liability Act of 1980 ("CERCLA"), 42 U.S.C. §§ 9601 fiLsejj,, as amended
by the Superfund Amendments and Reauthorization Act of 1986 (SARA), and the National Oil
and Hazardous Substances Pollution Contingency Plan ("NCP"), 40 C.F.R. Part 300. This
decision document explains the factual and legal basis for selecting the remedial action and is
based on the Administrative Record for this Site. An index of documents for the Administrative
Record is included in Appendix A of the ROD.

       The Pennsylvania Department of Environmental Protection ("PADEP") has commented
on the selected remedy and the State's comments have been incorporated to the extent possible.
PADEP concurs with the selected remedy.

III.    ASSESSMENT OF THE SITE

       Pursuant to duly delegated authority, I hereby determine, pursuant to Section 106 of
CERCLA, 42 U.S.C. § 9606, that actual or threatened releases of hazardous substances from this
Site, as discussed in Section IV (Risk Assessment) of this ROD, if not addressed by
implementing the remedial action selected in this ROD, may present an imminent and substantial
endangerment to public health, welfare, or the environment.

IV.    DESCRIPTION OF THE SELECTED REMEDY

       The Environmental Protection Agency ("EPA"), in consultation with PADEP, has
selected the following remedial action for the Metal Bank Superfund Site.  This remedy
addresses contaminated soil, sediment, surface water, and groundwater at the Site and includes
the following components:

 •     Installation of an oil collection system consisting of a sheet pile wall around the southern
       and western perimeter of the property; interceptor trenches with oil-water separators and

                                         1

-------
sump pumps, or similar collection devices, inside the wall to prevent oil from migrating
to the Delaware River; discharge of collected groundwater to the Delaware River in
accordance with National Pollution Discharge Elimination System requirements; and off-
site disposal of collected oil in accordance with RCRA and TSCA requirements;

Installation of temporary cofferdams prior to soil/sediment excavation to minimize
transport of contamination into the Delaware River;

Excavation of contaminated soil within the Courtyard Area within two feet of the surface
where polychlorinated biphenyl ("PCB") concentrations exceed 10 ppm; excavation of
contaminated subsurface soil in the Southern Portion of the Metal Bank property where
PCB concentrations exceed 25 ppm; excavation of contaminated sediments within 100
feet of the Metal Bank property and within four feet of the surface of the river bed; and
excavation of sediments beyond 100 feet of the Metal Bank property which have PCB
concentrations exceeding 1 ppm if determined by EPA to be appropriate and feasible.

Disposal of contaminated soil and sediments that are not hazardous in the following
manner: (1) if PCB concentrations are less than 25 ppm, backfill material on the Southern
Portion of the property ; (2) if PCB concentrations are between 25 and 50 ppm, dispose in
a landfill permitted in accordance with the Resource Conservation and Recovery Act
("RCRA") Subtitle D or Pennsylvania Residual Waste Management Regulation
requirements; or (3) if PCB concentrations are 50 ppm or greater, dispose at a Toxic
Substances Control Act ("TSCA") landfill;

Disposal of contaminated soils and sediments that are hazardous in the following manner:
(1) if PCB concentrations are less than 50 ppm, dispose at a facility in compliance with
RCRA Subtitle C or Pennsylvania Hazardous Waste Management Regulations; or (2) if
PCB concentrations are 50 ppm or greater, dispose at a TSCA landfill;

Removal and disposal of the underground storage tank and its contents from the Southern
Portion of the property;

Backfilling of excavated areas in the Courtyard Area with clean soil, installation of a 12-
inch soil cover over the entire Courtyard Area, and establishment of an erosion-resistant
vegetative cover;

Backfilling of excavated areas in the Southern Portion of the property with excavated
soils and sediments with PCB concentrations less than 25 ppm, installation of a 24-inch
soil cover over the entire Southern Portion, and establishment of an erosion-resistant
vegetative cover;

Restriction of access by installing and maintaining a fence around the perimeter of the
Metal Bank property;

-------
       Posting signs prohibiting consumption offish caught in the Delaware River in the
       vicinity of the Site;

       Restrictions on the deed to the property to prevent future residential or agricultural use of
       the Site, use of the groundwater, and intrusive activities into the subsurface soils below
       the water table in the Southern Portion of the property;

       Additional investigation to determine whether dense non-aqueous phase liquids
       ("DNAPLs") are present at the Site and whether the storm sewer system in the vicinity of
       the Site is contaminated; and

       Monitoring of groundwater, the Delaware River, and the Baxter Water intake.
V.     STATUTORY DETERMINATIONS

       The selected remedy is protective of human health and the environment, complies with
Federal and State requirements that are legally applicable or relevant and appropriate to the
remedial action, and is cost-effective. The remedy utilizes permanent solutions and alternative
treatment technologies to the maximum extent practicable. This remedy does not employ
treatment as a principal element.

       Because this remedy will result in hazardous substances remaining on-site above health-
based levels, a review will be conducted within five years after initiation of the remedial action
to ensure that the remedy continues to provide adequate protection of human health and the
environment.
W. Michael McCabe           ^                      Date
Regional Administrator
Environmental Protection Agency Region III

-------
                         METAL BANK SUPERFUND SITE
                              RECORD OF DECISION

                          PART II - DECISION SUMMARY
I.     SITE NAME, LOCATION, AND DESCRIPTION

       The Metal Bank Superfund Site is a former scrap metal and transformer salvage facility
located at 7301 Milnor Street in an industrial area of northeastern Philadelphia, Pennsylvania
(see Figure 1).  The Site is bordered by Cottman Avenue on the west; Milnor Street on the north;
Safe Disposal Systems (an appliance recycling company) and Morris Iron & Steel Company (a
metal salvage yard) on the east; and the Delaware River on the south. To the west of Cottman
Avenue is St. Vincent's School, which is a day care center and a temporary shelter for abused
children.  A stormwater outfall owned by  the City of Philadelphia at the southern end of
Cottman Avenue empties into a mudflat area west of the Metal Bank property.  The Quaker
City Yacht Club is located west of the mudflat.  Tacony Warehouse, formerly known as U.S.
Army and Air Force Exchange Warehouse, is located approximately 1,000 feet south of the
Site and adjacent to the Quaker City Yacht Club.  A Philadelphia drinking water intake and
treatment plant are located 2.2 miles upriver of the Site (see Figure 2).

       The Metal Bank property includes  two areas: the southern area which consists of
approximately six acres of open area that was used for scrap metal recovery; and the northern
area which consists of six vacant brick and steel buildings, a courtyard, and a parking area
(see Figure 3).  The buildings are  locked and a 6-foot fence is maintained around the southern
area of the property except along the river to limit access.
II.    SITE HISTORY AND ENFORCEMENT ACTIVITIES

A.    Site History

      Beginning in 1962, the Site was used for scrap metal storage.  From approximately
1968 or 1969 until 1973, transformer salvage operations were conducted at the Site.
Transformer oil was drained on a concrete pad which was connected to an underground
storage tank. Spills of the oil and possibly a rupture of the underground storage tank caused
soil and groundwater contamination at the Site.  Between 1968 and 1972 copper wire may
have been burned at this Site to remove insulation.  From 1973 to 1985, storage of scrap metal
continued at the Site, but no transformer salvage operations have been conducted since 1973.

      L. Goldstein's Sons, Inc. ("Goldstein's") purchased the Site from H.K. Porter in
October of 1962. In the mid-1960's,  Goldstein's changed its name to The Metal Bank of
America, Inc.  On December 4, 1968, Metal Bank of America, Inc. entered into an agreement
                                         4

-------
                                                                               A    0^1000  2000
                                                                               N         «•"
PrapwM (tarn ft flgun H; USOS. 7.5 Mta* rowprvMb M«
 Figure 1:  Site Location

-------
MOO    «000    MOO     WOO
                                              BAXTER
                                            EMERGENCY
                                              INTAKE
                 SITE LOCATION
                                    NJAWC WELL
                                   HIGHLAND AVE.
                                       #13.27
                                                           Insurance fbte M«p'. F«dM«l Emwgwicy
                                                           Mvch 19, 1M2. Pwwl «420757-OOIB-O)
Figure 2:  Site Proximity to Public Water Intakes (In
miles) and Floodplaln. Metal Bank Supedund Site.

(SOURCE: U.S. D*pl o< Intorlof Q«o(oglc«l Surv«y. 7.6 Minute
    Topographic Mop. FranUord 4 B*vwly Quadranoto)

-------
                                                                      9103S1C  6/28/V9NK
                                                              COURTYARD
                  RIPRAP AREA
                                      ««s£   Southern Portion
                                              of the property
                                         (which includes 'hot spots', where PCB
                                         * concentrations exceed 25 ppm, see
                                               Figure 6 for locations)
                                                      NAPL AREA
                APPROX1A/A TE LOW
                WATER
                                                         UNDERGROUND
                                                         STORAGE TANK
             DELAWARE RIVER
              SEDIMENT AREA
NOTE:  ALL LOCATIONS AND BOUNDARIES
      ARE APPROXIMATE.
     90   180
    SCALE IN  FEET
360
.....  NAPL AREA BOUNDARY

—  SOUTHERN PORTION  OF
       THE SITE BOUNDARY

— —  RIVER SEDIMENT
       AREAS BOUNDARY
                                      REFERENCES:
                                      Rl FIGURE 2-8(UST)
                                      FS FIGURES 1-2. 2-1. 3-1
                                        Figure 3
                                        Site  Plan
                                     Metal Bank Site

-------
to sell its assets (except for its properties at Cottman Avenue and New State Road) to The
Union Corporation ("Union Corporation"). Union Corporation incorporated MBA, Inc. on
December 9, 1968, and assigned its rights under the asset purchase agreement to MBA, Inc.
The Metal Bank of America, Inc. liquidated in March of 1969, and its property, including the
Site, was distributed to its shareholders, Irvin and John Schorsch. After the liquidation of the
"old" Metal Bank of America, Inc., MBA, Inc. officially adopted that name ("Metal Bank")
on December 7, 1971. Union Corporation leased the Site from the Schorsches for its
subsidiary, the "new" Metal Bank, until 1980. In 1980, the Schorsches sold the Site and the
New State Road property to the Philadelphia Authority for Industrial Development ("PAID"),
and the Schorsches took back the mortgage.  PAID and Metal Bank entered into a twenty year
Installment Sale Agreement whereby Metal Bank purchased both the Site and the New State
Road property.  PAID assigned its rights to payment under this Agreement to the Schorsches.
Metal Bank changed its name to U.C.O.-M.B.A. Corporation in June 1985. Although PAID
continues to hold the title to the Site properties, Metal Bank has full possession of the premises
and is the equitable owner of the Site under the Installment Sale Agreement.

      Portions of the buildings on the premises were leased to various tenants: Lincoln Van
and Storage for 8 to 10 years, ending in 1985; an automobile body repair shop for four or five
years until approximately 1983; and Pennsylvania Alternator and Generator Exchange for
approximately 1 year, until 1983.  During the 1970's, Reisher Ford stored new cars on the
premises.

B.    Site Enforcement Activities

      The following chronology highlights various activities, including enforcement actions,
that have taken place at the Metal Bank property.

DAIE       EVENT

1950 - 1967  Approximately 15 feet of fill from unknown origins was gradually placed in the
             Delaware River to create the southern portion of the property.

1972         The United States Coast Guard ("USCG") investigated reports of oil seeps into
             the  Delaware River and concluded that the Metal Bank property was the source.
             Analyses using available technology did  not detect polychlorinated biphenyls
             ("PCBs") in the oil samples. Metal Bank of America, Inc. ("Metal Bank")
             performed various remedial actions based on USCG recommendations including
             cleaning up spilled oil and improving housekeeping. Metal Bank also reported
             that it had ceased all transformer salvaging activities.

1977         EPA retested  samples collected by USCG in 1973 using new procedures.  The
             new analyses disclosed the presence of PCBs at concentrations over 800 ppm.
             The USCG, EPA, PADEP,  the Army Corps of Engineers, the City of
             Philadelphia, the Fish and Wildlife Service, the National Oceanic and

                                          8

-------
             Atmospheric Administration ("NOAA"), the Delaware River Basin Commission
             ("DRBC"), and others inspected the Site. Following these inspections, EPA
             prepared recommendations for remediation of the PCB problem and requested
             that Metal Bank complete the work.  Metal Bank rejected EPA's
             recommendation and elected instead to perform additional studies.

1980         EPA filed suit in the District Court for the Eastern District of Pennsylvania for
             injunctive relief and costs against Metal Bank under the Resource Conservation
             and Recovery Act ("RCRA") and the Toxic Substances Control Act ("TSCA").
             During the litigation, Metal Bank's consultant designed a groundwater recovery
             and treatment system.  The system consisted of two recovery wells and several
             separation units which collected the oily materials.

1981         Metal Bank reported to PADEP in 1986 that the underground storage tank at the
             Site was drained, cleaned, and filled with concrete.

1983         EPA settled the suit with Metal Bank under a Stipulation that required Metal
             Bank to install and operate a groundwater recovery and treatment system until
             all recoverable oil was removed from the property.  The system consisted of
             two recovery wells, several oil  separation units, and several 55-gallon drums
             containing activated carbon to treat groundwater. The separated oily materials
             were collected in drums for disposal.  The treated groundwater was discharged
             into the City of Philadelphia sanitary sewer  system.

             The Site was placed on the National Priorities List based on a Hazard Ranking
             System ("MRS") score of 33.23.

1987         In December, EPA sent letters to individuals and companies notifying them that
             they were Potentially Responsible Parties ("PRPs") under CERCLA. EPA's
             allegations were based on invoices which indicate that the PRPs sent, either
             directly or through brokers, transformers and other electrical equipment to
             Metal Bank.

1989         On January 13th, Metal Bank notified EPA pursuant to the Stipulation that it
             intended to shut down  the oil recovery system, stating that all recoverable oil
             had been removed.

             In March, EPA collected samples from monitoring wells on the property which
             showed the continued existence of PCB-contaminated oil floating on the
             uppermost aquifer.

             On April 1st, due to the concern that PCB oil may have been burned at the Site,
             EPA conducted dioxin soil sampling at St. Vincent's School.  The soil samples
             did not demonstrate a health risk from dioxin.

-------
1991
On June 12th, the Court denied as moot a notion by the United States seeking
to prevent permanent shutdown of the recovery system, the United States having
agreed in substance with an order proposed by Metal Bank extending the period
for EPA to conduct sampling until August IS, 1989. Subsequently, the
recovery wells were permanently closed and the oil recovery system was
dismantled and removed.  The area was covered with fill and vegetated.  All
buildings and exterior grounds were reportedly cleared of trash and debris and
fences were repaired and are maintained by Metal Bank.

In August, EPA measured groundwater elevations in each of the monitoring
wells on the site to determine the thickness of the remaining oil layer. An oil
layer of at least 3 inches was measured in some portions of the property.

In June, EPA signed an Administrative Order by Consent with 10 PRPs to
conduct a Remedial Investigation ("RI") and Feasibility Study ("FS"). The Site
Owner declined to join the PRP Group which included the following utility
companies:
             Baltimore Gas & Electric Co.
             Consolidated Edison Co. of NY
             Delmarva Power & Light Co.
             Jersey Central Power & Light Co.
             Long Island Lighting Co.
                                        Metropolitan Edison Co.
                                        Philadelphia Electric Co.
                                        Potomac Electric Power Co.
                                        Public Service Electric & Gas Co.
                                        Virginia Electric & Power Co.
III.    COMMUNITY PARTICIPATION

       EPA has several public participation requirements that are defined in Sections
113(k)(2)(B), 117, and 121(f)(l)(G) of CERCLA, 42 U.S.C. §§ 9613(k)(2)(B), 9617, and
9621(f)(l)(G), respectively. The documents that EPA used to develop, evaluate, and select a
remedial alternative for the Metal Bank Site have been made available to the public in the
Administrative Record maintained at the Northeast Branch of the Free Library of Philadelphia,
located at 2228 Cottman Avenue, Philadelphia, and at the EPA, Region m, Philadelphia
Office.

       The RI/FS Reports and the Proposed Plan for the Site were released to the public in
July of 1995. The Proposed Plan described remedial alternatives being considered by EPA
and identified EPA's preferred alternative at that tune. The notice of availability of the
Proposed Plan and the Administrative Record was published in the Philadelphia Inquirer on
July 20, 1995. This notice also invited the public to a meeting on July 27, 1995, to discuss
the Proposed Plan with EPA. The public was encouraged to review the Proposed Plan and the
Administrative Record files and to submit comments on the proposed remedial alternatives to
EPA. The public comment period was initially scheduled to close on August 19, 1995,
allowing for the statutorily required 30-day period.  At the request of the Site Owner and the

                                          10

-------
PRP Group, EPA extended the comment period an additional 30 days until September 19,
1995. When the PRP Group requested additional time to complete a Treatability Study which
it began in response to the Proposed Plan, EPA extended the comment period 30 additional
days. The comment period, therefore, ended on October 18, 1995.

      EPA held the public meeting on July 27,  1995, at the Disston Recreation Center
located approximately V4 mile  northwest of the Site.  Attendees included members of the
Tacony Civic Association ("TCA"), local citizens, representatives of local businesses, the PRP
Group, and the Site Owner.  TCA members and local citizens asked questions and provided
comments during the meeting.  EPA representatives answered questions about conditions at the
Site and the  remedial alternatives under consideration.

      Responses to the comments received during the public comment period are included in
the Responsiveness Summary,  which is part of this ROD.
IV.    SCOPE AND ROLE OF RESPONSE ACTION

       The remedial action selected in this ROD is intended to address contamination in soils,
sediment, groundwater, and surface water impacted by the release of hazardous substances
from the Site. EPA does not contemplate further remedial action for the Site if the cleanup
requirements in this ROD are achieved.
V.     SITE CHARACTERISTICS

A.     Surface Water

       The Site is flat with a gentle slope towards the south and southeast. Surface water
from the Site drains towards the Delaware River.  The Delaware River provides drinking
water for the City of Philadelphia through the Baxter Water Intake and several municipalities
in New Jersey through the Delran Water Intake (see Figure 2). The river and its adjacent
estuary also serve as habitat to numerous aquatic species, including the endangered Shortnose
Sturgeon.

       Surface elevation at the Site is approximately 10 feet above Mean Sea Level ("MSL")
which is the expected height of water during a 100-year flood.  Most of the southern area is
within the 100-year floodplain (see Figure 2).

Baxter Water Intake:  The Samuel S. Baxter Water Treatment Plant (formerly known as
Torresdale until 1982) is located at 9001 State Road in the Torresdale section of Philadelphia.
The plant is operated by the Philadelphia Water Department. Although the plant and its water
supply intake are upstream of the Site, EPA concluded its presence was important based on the
possibility that PCBs from the Site may reach the intake when tides in the Delaware River

                                         11

-------
backwash against the river current. Dye studies have demonstrated that tides in the Delaware
River can transport detectable amounts of contaminants as far as 6 miles upstream from its
point of release.  The main intake, which has been in operation since 1966, is approximately
2.2 miles1 from the Site. An emergency intake, which has been in operation since the
completion of the plant in 1909, is located approximately 1.9 miles from the Site. The
emergency intake becomes operational for several days a year when the main intake is forced
to be shut off.  This occurs when freezing temperatures cause ice blockages at the intake gates
or when a raw water storage basin adjacent to the intake undergoes maintenance.

       The Baxter Water Intake obtains water directly from the Delaware River and supplies
drinking water to the eastern half of the City of Philadelphia as far west and north as the
Schuylkill River, bordering Roberts Avenue and Tabor Road. EPA estimates the population
served by the Baxter Water Intake to be over 1 million people.

       While the raw water from the river undergoes a series of standard water treatments
(such as chlorination, flocculation, sedimentation and filtration), the plant  does not have
special equipment to remove PCBs. Additionally, the plant  is not required to analyze for
PCBs as part of its routine  water quality control testing program.

Delran  Water Intake: This intake is approximately two miles north of the Site,  located across
from the Baxter Water Intake at the intersection of Taylor Lane and the Delaware River in
Delran, New Jersey.  As part of the Tri-County Water Supply Project being built to reduce
pumping and overburdening of New Jersey's groundwater aquifer, the intake will supply
additional drinking water for Burlington, Camden and Gloucester counties. The intake and a
new treatment plant have been in operation since the spring of 1996. Samples collected in
1989, during the design phase of the project, did not detect PCBs. Design studies also
demonstrated that floatation markers released into the Delaware River washed three miles
upstream.

B.     Geology

       Historical aerial photographs have shown that most of the southern area of the Metal
Bank property was part of the Delaware River prior to 1950. This area was gradually filled in
from 1950 to 1967.  Approximately 15 feet of fill underlies the Site. The  fill materials contain
pieces of brick, lumber, cloth, metal, and concrete along with natural earth materials (sand,
silt, gravel, etc.). The United States Department of Agriculture Natural Resource
Conservation Service has not classified the soils on the property and other similar areas in
Philadelphia County because identification of these reworked soils is not practical.
1    The Rl/FS, MRS, and other documents in the Administrative Record include various measures of the distance of
    the intake from the Site, including 1.2 miles.  Based on comments received during the public comment period, EPA
    further researched this issue and information on the emergency intake location.  The ROD includes the correct
    information.

                                           12

-------
                                                    Depth
                                                    (feet)

                                                    0-21

                                                    7-16

                                                    30-50


                                                    50
SITE GEOLOGY


 Material

 FILL

 WATER LEVEL

 ALTERNATING SAND &
 CLAY LAYERS

 BEDROCK
       The Site is located near the western edge of
the Atlantic Coastal Plain Physiographic Provence.
The fall line boundary between the Atlantic Coastal
Plain Physiographic Provence and the Piedmont
Physiographic Provence is less than one mile to the
northwest.  Basement rocks beneath the Site are
composed of a complex assemblage of Paleozoic
crystalline metamorphosed sedimentary and igneous
rocks.  These rocks outcrop at the fall line.  The
depth to the basement rocks increases towards the
east. Cretaceous sediments of the Raritan formation
are unconformably overlying the basement rocks. A
residual clay layer may separate the Raritan
Formation from the underlying bedrock.
       The Raritan formation is composed of nonmarine clays, sands, and gravels.  The
lowermost Farrington Sand member of the Raritan is the only member present beneath the
Site.  The Farrington Sand member consists of yellowish gray to pale yellowish brown coarse
sand and fine gravel which grade upward into medium to fine-grained sands. The thickness of
the Farrington Sand member has been mapped to be around fifteen feet at the Site. The
Farrington Sand member is continuous beneath the Delaware River. Cretaceous fluvial,
estuarine Van Sciver Lake beds of the Trenton Gravels unconformably overlie the Farrington
Sand member at the Site.  The Van Sciver Lake beds consist of interbedded quartz sand and
silts. This unit is between ten to fifteen feet thick at the Site. The Van Sciver Lake beds on the
Pennsylvania side of the Delaware River are separated from those in New Jersey by Holocene
clays beneath the river. The uppermost unit beneath the Site consists of recent fill material.

C.     Hydrogeology

       Generally, an unconsolidated and a bedrock aquifer are present at the Site.  The
unconsolidated sediments  of the Farrington Sand member of the Raritan Formation, the Van
Sciver Lake beds, and the fill material are in direct communication with each other at the Site.
None of the potentially confining clay layers of the Raritan Formation appear to be present.
No boreholes at the Site were drilled to bedrock. Therefore, the degree of connection between
the deeper bedrock aquifer and the overlying unconsolidated aquifer is unknown.  Depth to
groundwater beneath the Site varies from seven to sixteen feet. Shallow ground water at the
Site flows southward across the Site and discharges to the Delaware River.  Based on regional
well records, groundwater within the bedrock formation also flows towards the southeast.
Vertical flow directions and the degree and direction of underflow beneath the Delaware River
is unknown. The Raritan Formation is a major source of groundwater on both sides of the
river.

       The RI/FS estimated the range of hydraulic conductivity at the Site to be from
1.98x10^ to 9.53xlO'2 centimeters  per second ("cm/sec").  This range was attributed to the
                                          13

-------
different types of fill materials found on-site. Based on this hydraulic conductivity, the
theoretical discharge into the river is between 0.09 to 1.44 feet per day or 0.016 to 0.043
cubic feet per second ("cfs").

D.     River Tide

       Tidal monitoring was conducted during the RI/FS to determine  the degree of influence
tides in the Delaware River have on the water levels at the Site. The tidal cycle in the
Delaware River reaches maximum and minimum water levels every 12  hours. Water levels in
monitoring wells MW4 and MW5, which are located near the shoreline, showed fluctuations
similar to the tidal cycle.  The remaining wells showed more limited or no river influence.
The rate  of fluctuation is expected to be influenced in pan by the nature of fill material at the
well location and the distance from the shoreline.

£.     Extent of Contamination
       This section summarizes analytical data collected during the RI/FS and information
provided in or researched in response to comments received during the public comment
period. Historical data, which may not have been included in the RI/FS, have been used in
some instances to provide a more complete characterization of the extent of contamination at
the Site.  Documents have been included
in the Administrative Record, as
necessary, to support the use of these
data.
       Four areas of environmental
concern have been identified at the Site
(see Figure 3):

   •   the "Building Area," which
       includes the six buildings on the
       northern portion of the Metal
       Bank property;

   •   the "Courtyard," which is a open
       area adjacent to several of the
       buildings on the northern portion
       of the property;

   •   the "Southern Portion" of the
       property, which is an
       approximately six acre open area
       where the former transformer
       salvage operation was conducted;
     UNITS OF MEASUREMENT

To provide continuity and ease of comparison,
units of measurement in this document will be
presented in parts per million ("ppm") or parts
per billion ("ppb"), unless otherwise specified.
One ppm is the equivalent of 1000 ppb.
These units may differ from those used in the
various reports for the Site (e.g., RI/FS). Units
typically used in these reports are as follows:
MEDIA

solids (i.e., soils, sediments)
water and groundwater
concrete chips/dust
fish tissue
UNITS

mg/kg
mg/kg/1 00cm2
M9/9
Listed below are the respective conversion unit
equivalents.

ppm =  mg/kg, ^g/g, ng/mg, pg/ug, mg/L,
       //g/mL, or ng/uL

ppb =  ^g/kg, ng/g, pg/mg, ^g/L. ng/mL,
       pg/uL
                                           14

-------
       this area includes an
       underground storage tank
       buried inside a non-aqueous
       phase liquid area ("NAPL
       Area") where residual oil is
       suspected and where PCB
       "hot spots" have been
       identified; and

   •   the "River Sediment Area"
       along the shoreline of the
       Southern Portion of the
       property. This area includes
       the "Mudflat Area", the
       "Riprap Area", and the
       "Delaware River Sediment
       Area" shown in Figure 3.

       The maximum concentration
of contaminants observed for surface
water, groundwater, surface and
subsurface soils, and river sediments
has been summarized below. Tables
presenting these data also identify
the screening levels that were
considered in determining whether a
contaminant would be  included in
the quantitative risk assessment.
These tables are based on data
presented in the RI/FS and data from
split samples taken by EPA.  This
section also discusses data collected
in the buildings  and the results of biological monitoring at the Site.
         DEFINITIONS OF PCBs, PAHs & TPHs

Polychlorinated Blphenyls ("PCBs"): A class of 209
individual chlorinated hydrocarbon compounds ("congeners")
that contain a variable number of substituted chlorine atoms
on the biphenyl ring. PCBs are man-made chemicals and
classified as suspected human carcinogens based on
scientific data from laboratory animals.  PCBs are frequently
identified through commercial mixture designations
("Arodors") because it is less costly than analysis of
individual congeners.

Polynuclear or Polycyclic Aromatic Hydrocarbons
("PAHs"):  A group of Semi-Volatile Organic Compounds
("SVOCs") composed of two or more benzene rings. They
are commonly found at gas manufacturing plants, wood
preserving  sites that used creosote or petrochemical waste
sites. PAHs have also been found in fuel products such as
heating oil, diesel fuel and No. 6 fuel oil.  EPA has identified
16 PAHs as priority pollutants. Some of the four, five and six-
ring PAHs such as chrysene, benzo[a]pyrene and
indeno[1,2,3-cd]pyrene have been shown to be carcinogenic
to animals. The two and three-ring PAHs, such as
acenaphthene, acenaphthylene, anthracene, fluoranthene,
fluorene, phenanthrene, and pyrene have been observed to
have adverse non-carcinogenic health effects in animals.

Total Petroleum Hydrocarbons ('TPHs"):  A measurement
of the concentration of petroleum hydrocarbons in a given
amount of air, soil, or water. The term "total" is a misnomer,
in that procedures for quantifying hydrocarbons are not
capable of measuring all fractions of petroleum hydrocarbons
present. In addition, volatile hydrocarbons are usually lost
during the analysis while some non-petroleum hydrocarbons
may be accidentally included.
1.     Building Area

       The Building Area consists of six buildings.
Buildings 2, 3, and 4 are on the northernmost end of
the Site. Buildings 5, 6, and 7 comprise one large
structure and are immediately south of Building 4.
Building 7 is a large storage warehouse which contains
railroad tracks that may have transported the PCB
transformers, based on the location of PCBs in
elevated concentrations near the track.  Buildings 7
                     Table 1 - Total PCBs in Buildings
Building
2
5
6
7
Total PCBs (ppm)
9.7
12
19
372
                                            15

-------
and 2 were also used as part of the groundwater recovery and treatment system between 1983
and 1989. All that remains of Building 1, which was demolished in 1984, is an asphalt slab
that occupies a portion of the Courtyard area.  All the buildings are now empty and in
advanced stages of disrepair.  Buildings 3 and 4 were considered unsafe to enter.  Since
Buildings 3 and 4 were reportedly office structures and never used for PCB-related activities,
these buildings were not sampled.

       Twenty-five PCB  samples were taken inside Buildings 2,5,6 and 5 between 1991 and
1993.  Most samples were chips from stained areas on the building floors and on the masonry
walls where spills were suspected to have occurred. PCB contamination occurred randomly
throughout all the buildings. The highest level of PCBs (372 ppm) was found in a dust sample
(B7-5) collected under a railroad track in Building 7.

2.     Surface Water

       A limited attempt  was made during the RI to characterize the potential movement of
PCBs from the Metal Bank property into the Delaware River. In 1991, standing water
samples were collected from three locations in the Mudflat Area and analyzed for PCBs and
pesticides.  No PCBs were detected in these samples. Two pesticides, 4,4'-DDE and
Endosulfan I were detected at one location at concentrations of 0.02 and 0.05 ppb,
respectively.  Because of the limited surface water data available, EPA completed the risk
assessment for surface water at the Site by using groundwater contaminant concentrations
multiplied by a dilution factor of  3,100:1.  This dilution factor represents the surface water
concentration that would result if an average volume of groundwater discharged from the
Metal Bank property and  mixed with river water within 50-feet of the shoreline under low
flow conditions. Using this dilution factor, none of the Site-related contaminants exceeded
EPA's criteria for protection of freshwater aquatic life.

Stormwater Sewer.  No floor drains, pipes and other low spots were identified where PCBs
may have accumulated and been discharged into the stormwater sewer. The stormwater basin
located near monitoring well MW13 may be connected to a sewer line located under Cottman
Avenue,  which eventually discharges at the City's stormwater outfall and into the Mudflat
Area.  City maps do not show connections made by private parties before the 1980's.

3.     Groundwater

       In 1991, fifteen new monitoring wells were installed as part of the hydrogeologic
investigation for the RI. Seven wells were sampled hi 1991 and fifteen in 1992. While
analysis of each sample varied, the parameters measured generally included VOCs, SVOCs,
PCBs, pesticides and metals.  Table 2 presents the maximum groundwater concentration
detected  for contaminants used in the human health risk assessment and the  Region ID risk-
                                          16

-------
                                            Table 2 - Groundwater Contaminant Concentrations
based screening level for tap water2 as a
point of reference.

       PCBs were detected in groundwater
samples collected from monitoring wells
MW6 and MW7.  Concentrations of total
PCBs in MW6 were 12.3 and 25.6 ppb
during the 1991 and 1992 samplings,
respectively. Total PCBs of 1.3 ppb were
detected in MW7  during the 1992
sampling.

       The total concentrations of VOCs
and SVOCs  found at each monitoring well
during the 1992 sampling are presented in
Figures 4 and 5. These figures show the
possibility that some groundwater
contamination may reach the Site from
upgradient sources.  No monitoring wells
were located off-site in the upgradient
direction during the RI.

Non-Aqueous Phase Liquid (NAPL): In
1991, a sample was collected in the oil layer
found floating on the groundwater in MW6.
The PCB concentration in the oil sample
was  1,090 ppm. During the 1992 sampling
of MW6, the sampling technique used is likely to have disrupted any oil layer that may have
been present. Only 7 ppb was detected in the oil layer while EPA's split sample detected 183
ppb. A bailer rather than a tube and low-yielding pump was used to purge the wells. Use of the
bailer is likely to have dispersed any oil layer that may have been present and may have resulted
in lowering the PCB levels in the oil. Previous PCB concentrations measured in the oil layer
present at the Site  were 1,539 ppm in 1977 prior to the oil recovery operation and 1,540 ppm in
1989 when the oil recovery operation was being terminated.

PNAfLs:  PCB  levels in the oil layer samples from the Site have historically been greater than
those in water samples. When PCBs are released into the environment through transformer oil
spills, they usually bond to the oil and float on top of the water table because the specific gravity
of oil (0.8) is lighter than water (1.0). This oil layer is known as the Light Non- Aqueous Phase
Liquid ("LNAPL") layer.  However, the specific gravity of PCBs themselves are heavier than
Contaminant
Arsenic
Chromium
Total PCBs
Acenaphthene
Anthracene
Benz[a]anthracene
Benz[b]fluoranthene
Benzo[a]pyrene
Chrysene
Dibenz[ah]anthracene
Fluoranthene
Fluorene
lndeno[1 ,2,3-cdJpyrene
Naphthalene
Maximum
(ppb)
67
102
25.6
110
58
160
120
64
79
12
200
95
38
780
Screening
Level
(ppb)
0.038
180
0.0335
2,200
11.000
0.092
0.092
0.0092
9.2
0.0092
1,500
1.500
0.092
1.500
    The screening levels presented in the ROD are those used in the final risk assessment for the Site and are based
on the EPA Region III Risk-Based Concentration ("RBC") Table dated February 9,1995. Because the cancer slope
factor for PCBs has been revised, the current PCB screening levels from the RBC Table dated March 17,1997 are used.
                                           17

-------
EARTH
«w .

MW-2
42
V«M~
MW-6
NO

^-
f
/
/ „
/
£»
h* /
 Ox*     <%,
 ^v.^
        .Or
           2o,
   MW-3
            CROUNOWATER MONITORING WEU / PIEZOMETER
            VOC CONCENTRATION CONTOURS
            SAMPLING LOCATION No.
            TOTAL VOCS (ugA)
    NO       NOT DETECTED
   NOTE: DATA REFLECTS MAY/JUNE 1992 SAMPLING PROGRAM
                                                               90  180
                                                              SCALE IN  FEET
360
                       Figure 4 - Total VOCs in Groundwater

-------
E  A
                                                                                  09-76-94
 LEGEND
   MW-3
      	_•*. ^1000
            ^  "• ••iiui
                                                                                   MW-5
                  ,100
80
                        MONITORING WELL / PIEZOMETER
            SVOC CONCENTRATION CONTOURS
            SAMPUNG LOCATION No.
            TOTAL SVOCS
    NO       NOT DETECTED
    NT       NOT TESTED
    NOTE; DATA REFLECTS WAY/JUNE 1992 SAMPLING PROGRAM
        1 =180'
0    90   180
     ^      F"
    SCALE  IN FEET
                       UW-4
                                                                                   360
                        Figure 5 - Total SVOCs in Groundwater

-------
water (between 1.18 to 1.57) which may cause PCBs to sink in the aquifer until an
impermeable layer such as clay or a bedrock formation is encountered. This condition is
known as the Dense Non-Aqueous Phase Liquid ("DNAPL") layer. At this phase, gravity
dictates the flow of DNAPLs rather than the movement of oil.  During the subsurface soil
sampling performed during the RI, PCB concentrations were occasionally observed to be
increasing with depth (see Table 6). None of the monitoring wells constructed at the Site were
screened in the portion of the aquifer at the bedrock interface.  Therefore, the presence or
absence of a DNAPL layer was not confirmed.
Groundwater Seeps:  Samples were
collected from two locations in the river
bank along the southern portion of the
Site where seeps were observed.  These
samples were analyzed for PCBs,
pesticides, and several metals.  Table 3
presents the maximum concentrations
observed for contaminants that
exceeded Region in risk-based human
health screening levels for tap water.
4.
Surface Soils
                                   Table 3 - Groundwater Seep Concentrations
Contaminant
PCBs (Aroclor 1254)
4,4'-DDD
4.4'-DDE
Dieldrin
Manganese
Maximum
(ppb)
3.7
0.68
2^3
0.1
621
Screening
Level
(ppb)
0.73
0.28
0.2
0.0042
180
Courtyard:  In 1991, soil samples were collected and analyzed for PCBs in the Courtyard
near the entrance to Building 2 and near the railroad tracks entering Building 7. Two samples
were collected at each location-one from the top six inches and one from 6-24 inches.  PCBs
were detected in the top six inches at concentrations of 52 ppm near Building 2 and 140 ppm
near Building 7. At the lower depth, PCBs were not detected in Building 2 and were detected
at 6.5 ppm at Building 7.  In 1993, additional soil sampling was conducted in the Courtyard.
Samples were analyzed using a field screening technique with laboratory confirmation of 20%
of the samples.  Thirty-seven locations were sampled at a depth of 0 - 2 feet. An additional
sample was collected at a depth of 2 - 4 feet, where possible, at locations that had greater than
5 ppm of PCBs in the surface sample.  At 28 of the 37 locations, the field screening results
indicated that PCB concentrations were less than 5 ppm. Samples from four of these locations
were sent to the laboratory for analysis and concentrations were below detection limits for
three samples and 0.43 ppm for the fourth.  At two locations, the field screening results
indicated PCBs were present at concentrations between 5-10 ppm at a depth of 0 - 2 feet and
less than 5  ppm at a depth of 2 - 4 feet. Laboratory confirmation samples from the 0 - 2 foot
depth at these locations detected PCBs  at 0.46 and 8.1 ppm. Field screening results at the
remaining seven locations indicated that PCBs were present at greater than 50 ppm at the 0 - 2
foot  depth and at less than 5 ppm at depths of 2 - 4 feet. Two laboratory confirmation samples
from the 0  - 2 foot depth at these locations had PCB concentrations of 1.1 and 70 ppm.  One
confirmation sample from the 2 - 4 foot depth detected PCBs at 0.97 ppm. Based on the field
screening, an area approximately 100 x 50 feet immediately north of Buildings  5 and 7 was
delineated as potentially exceeding 50 ppm.
                                          20

-------
Southern Portion: In 1989, one to two feet of soil was placed over the Southern Portion of the
Site. As part of the RI, sixteen soil samples were collected from the 0 - 2 foot depth in this
portion of the property to characterize this cover material. All the samples were analyzed for
PCBs and pesticides and five samples were also analyzed for metals and SVOCs.  Table 4
presents the maximum soil concentrations observed and the Region m risk-based screening
level based on incidental ingestion of soil on a property being used for industrial purposes.
                                              Table 4 - Surface Soil Contaminants
                                                      (Southern Portion)
5.     Subsurface Soils

       In 1991, samples were collected
at various depth intervals from soil
borings at 23 locations. All samples
were analyzed for PCBs and one
sample from each boring was analyzed
for metals and SVOCs. Thirteen
samples were also analyzed for VOCs,
dioxins  and furans.  In 1993, samples
from various depths in eight additional
borings were collected. Samples from
five borings located in the vicinity of
borings B-17 and B-18 from the  1991
sampling were screened for PCBs in
the field and submitted for laboratory
analysis for lead and VOCs.  One
sample from each of these five borings
was also submitted for PCB laboratory
analysis. The three remaining borings
were located in the central and
northeastern portion of the Metal Bank
property.  All samples were submitted
for PCB, VOC and lead laboratory
analysis. One composite sample from
each boring was analyzed for SVOCs.
In August 1995, during the public comment period on the Proposed Plan, the PRP Group
requested the opportunity to excavate test pits and collect additional soil samples from the
southern portion of the Metal Bank property. EPA allowed the additional sampling and
samples were collected and analyzed for PCBs from 14 test pits.

      In addition to the above parameters, many samples were analyzed for total petroleum
hydrocarbons ("TPH").  Although high PCB levels were observed in oily samples, the TPH
measurements were not found to correlate well with the presence of PCBs.  The TPH data is
Contaminant
Arsenic
Beryllium
Chromium
Total PCBs
Acenaphthene
Anthracene
Benz[a]anthracene
Benz[b]fluoranthene
Benzo[a]pyrene
Chrysene
Fluoranthene
lndeno[1 ,2,3-cd]pyrene
Pyrene
Maximum
(ppm)
6.8
1.1
57.4
4.7
0.052
0.34
1.2
1.4
1
1.1
2.7
0.56
2.2
Screening
Level
(ppm)
3.3
1.3
10,000
2.86
120,000
610,000
7.8
. 7.8
0.78
780
82,000
7.8
61,000
                                         21

-------
useful in identifying areas where oil
may have been spilled, however,
specific analysis for PCBs is
necessary to quantify potential PCB
contamination. Table 5 presents the
maximum concentrations of other
contaminants detected in the
subsurface soils samples as well as
the Region in risk-based screening
levels associated with incidental
ingestion of soil from a property
being used for industrial purposes.
Table 6 summarizes the total PCB
concentrations detected in the soil
boring and test pit samples. Data in
parentheses are field screening
results. Figure 6 shows the locations
of the soil borings and test pits. The
locations that are underlined are those
that had at least one sample with a
concentration of 25  ppm or higher
total PCBs. The highest levels of
PCBs were found in the southwest
portion of the Metal Bank property.
This corresponds with the
groundwater results which were also
highest in this portion of the Site.
                                     Table 5 • Subsurface Soil Contaminants
6.
Sediments
Contaminant
Antimony
Arsenic
Beryllium
Cadmium
Chromium
Lead
Total PCBs
Acenaphthene
Anthracene
Benz[a]anthtacene
Benzo[b]fluoranthene
Benzo[k]fluoranthene
Benzo[a]pyrene
Chrysene
Dibenz[ah]anthracene
3,3'-dichlorobenzidine
Fluoranthene
Fluorene
lndeno[1 ,2,3-cd]pyrene
Naphthalene
Pyrene
2,3,7,8-TCDD (dioxin)
Maximum
(ppm)
123
21.1
8.2
31
2.730
227,000
230
53
77
140
110
110
140
130
8.2
37
220
93
83
270
280
0.004
Screening
Level
(ppm)
820
3.3
1.3
1.000
10.000
—
2.86
120.000
610.000
7.8
7.8
78
0.78
780
0.78
13
82,000
82,000
7.8
82.000
61,000
0.00004
       Sediment samples were
collected from different portions of
the aquatic environment surrounding
the Metal Bank property. Samples
were collected from the mud flat
immediately west of the property
which is exposed during low tide;
samples were collected from the rip rap bank along the west and southwestern perimeters of
the property; samples were collected from the river immediately beyond the low water line;
and samples were collected at depths ranging from the surface to ten feet from borings located
in the river at distances of up to 300 feet from the low water line.
                                          22

-------
Table 6 - PCBs in Subsurface Soils (Southern Portion)
Total
PCBs Depth
Boring (ppm) (feet)
B-1
B-2
B-3.
B-4
B-5
B-6
B-7
B-8
B-9
B-10
B-11
B-12
B-1 3
B-14
--
1.7
j|£28:' :
0.9
1.1
0.2
10.2
2.5
5.1
0.7
4.8
0.3
1
4.7
4.2
0.4
6.6
. 29
11.7
0.4
1
0.2
3.1
0.6
0.8
3.2
18.2
9.1
3
10.5
2
6
2.5
2
5.5
2
5.5
9.5
6
3
6.5
2-5
6.5
14.5
8
10
14
2.5
6
1
5.5
11
3
10.5
12.5
Total Total
PCBs Depth PCBs Depth
Boring (ppm) (feet) Boring (ppm) (feet)
B-1 5
B-1 6
B-1 7
B-1 8
B-1 9
B-20
B-21
B-22
B-23
B101
5
3.9
0.4
IN»^
NiS
3
0.1
1
3.2
7.5
3.6
6.3
13.7
3.6
0.4
p25":^
6.5
8.7
2.8
6.6
0.3
7
7.8
4.7
P«#
5.3
(<5)
8.5
5.5
11.5
3
6.5
12.5
8
10
12
14
3
6.5
6
8
16
3
6.5
14.5
6
8
14
3
8.5
10.5
6
8
10
B102 | 24.6
B103
B104
B105
B106
B107
TP-1
TP-4
Ppipj;
(5-10)
0.8
tf'rrf,* '•:.
K 1°*°)
||
fii|45"'V;:-
•if*-'- I" V- . :.'
Ifte
TP-8 | 2
TP-9 ^'35
10
8
11
15
10
12
14.
8
10
12
14
16
6
10
12
14
6.0
8.0
12.0
14.0
10
12
5
10
11.5
15
5
Total
PCBs De
Boring (ppm) (p
TP-12
TP-
T-1
TP-
-2
TP-
T-3
4.4
7.7
|p8 '
10.5
t> • •
£':57 '
ny ••
&230
riso
1-77
0.77
0.87
pth
pm)
5
10
5
10
15
5
10
12
5
15
Concentrations
potentially 2 25 ppm are
shaded.

                       23

-------
            J^
LfCfNO
    MMB UCMM {nut/on. KM)
    vsvrT uE*mi (wast !•«
    LCX^TIONWfTH>25PPMPC8s
      fstOO"
                            •-IVP-I TP-7

                              • 4*  ^»-«

                          }  V"7          /P-/^

                          I .   —, V"^E2   •*

                                 rt.77»-r-%r

                                 *  i.7 ^-i   ^e-JA«
                                                     V;
                                                TP-IJ ±&as2.\
                                                 *  *
    Figure 6 - Soil Boring/Test Pit Locations with Total PCBs 2 25 ppm

-------
                                             Table 7 - Total PCBs in Mudflat Sediments
Mudflat Sediment:  In 1991, sediment
samples were collected at twelve locations
in the mudflat and analyzed for PCBs,
pesticides, and SVOCs.  Samples were
again collected at six of the locations and
analyzed for PCBs (including PCS
congeners) during a subsequent benthic
organism survey. One additional sediment
sample was collected just below the low
water line at the downstream edge of the
mud flat and analyzed for PCBs, pesticides,
SVOCs,  and metals.  In  1993, 35 samples
were collected from the mud flat using a 60
x 40 foot grid pattern. All the samples were
field screened for PCBs.  Eight samples
were sent to a laboratory for PCB and
SVOC analysis.
Sample
MF-3
MF-5
MF-10
MF-11
MF103
MF104
MF106
MF107
MF108
MF109
MF112
PCBs
(ppm)
1.1
2.1
1.1
1.2
(1-5)
(>5)
(>5)
19.6
(1-5)
(1-5)
(1-5)
Sample
MF113
MF114
MF115
MF116
MF117
MF118
MF127
MF128
MF129
MF130

PCBs
(ppm)
(1-5)
(1-5)
(1-5)
(1-5)
(1-5)
d-5)
2.7
(1-5)
(1-5)
(1-5)

       Mudflat sediment samples with total
PCB concentrations greater than one ppm are listed in Table 7 along with the maximum
concentration observed at that location.  Figure 7 shows the mud flat sampling locations.
Those locations which had at least one sample with total PCB concentrations greater than one
ppm, the target cleanup level, are underlined.

Riprap Sediment:  Sediment samples were collected at five locations along the property shore
at or near the high-tide line.  The shoreline area is largely paved with stone,  brick, and
concrete riprap.  Samples were collected from three locations along the western property line
(RR-1, RR-2, and RR-3).  At two of these locations (RR-2 and RR-3), three  samples were
collected at six inch depth intervals (i.e., 0-6"; 6-12"; and 12-18").  Samples from these same
depth intervals were also collected at the southwest corner of the property (RR-4). A single
sample was collected from a location approximately midway along the southern property line
(RR-5). All samples were submitted for laboratory analysis for PCBs, pesticides, and SVOCs.
Two samples were also analyzed for dioxins and furans.

       PCBs were detected at two sampling locations along the western property line.  At RR-
2, which is located approximately midway along the western shoreline, total  PCB
concentrations of 2.2 and 2.34 ppm were detected at the 0-6" and 6-12" depths, respectively.
At RR-3, which is located midway between RR-2 and the southwest corner of the property
shoreline, total PCB concentrations of 6.5, 5.9, and 14 ppm were detected at the 0-6", 6-12",
and 12-18" depths, respectively.

River Sediment:  In 1991, sediment samples were collected from three transects along the
southern property shoreline.  Four additional transects were sampled in 1993. The number of
                                          25

-------
                    LEGEND
 ^MF-1 MUOFLAr SEDIMENT SAMPLE LOCATION
       MUOFLAT GRID SAMPLE LOCATION
     A  (JUNE. 1993)

    22  LOCATION WITH > 1 PPM PCB*
        r=to
-------
                                               Table 8 - Total PCBs in River Sediments
Sample
DR-8
MF110
MF111
MF131
MF132
MF133
MF134
MF137
MF138
MF139
PCBs
(ppm)
6.8
(1-5)
(1-5)
d-5)
(1-5)
2.3
(>5)
d-5)
(1-5)
(>5)
Sample
S5
S6
S7
S8
S9
S16
S20
S21
S22

PCBs
(ppm)
1.2
1.3
1.7
1.3
1.2
1.5
4.6
1.5
1.6

samples in these transects ranged from two
to four with the first samples being collected
near the low water line and the remaining
samples being collected further out into the
river. All 1991 transect samples were
analyzed for PCBs and pesticides.  Two
samples collected from the transect just
upstream of the Metal Bank property line
were analyzed for SVOCs and metals.
Samples from two of the 1993 transects were
analyzed for PCBs and SVOCs. Samples
from the other two transects were only
analyzed for SVOCs.

      Eight additional samples were
collected below the low water line along the
southwestern shoreline and south of the mud
flat area. These samples were field screened
for PCBs with 20% of the samples sent for
laboratory confirmation. In 1995, further
sampling was conducted to characterize PCB
levels in sediments extending further into the river and to evaluate the distribution of PCBs
with depth. Samples were collected at depth from 0 -10' at 26 boring locations. All samples
were submitted  for laboratory PCB analysis.

      Figure 8 shows the river sediment sampling locations associated with the various
sampling events.  Locations where PCBs were detected at 1 ppm or greater are underlined.
Table 8 presents the maximum PCB concentration found at each of these locations. Field
screening results are presented in parentheses.
SVOCs. Pesticides, and Metals:   Table 9 presents themaximum concentrations of SVOCs,
pesticides, and metals detected in sediment samples collected at the Site for contaminants that
exceed the Region m biological screening levels for the protection of sediment fauna.  These
screening levels provide a conservative guideline for the evaluation of sampling data at
Superfund sites and are based upon the lowest value from a combination of sources considered
to be protective of the most sensitive organisms in a medium.  Sources include peer reviewed
literature, regulatory agency criteria, and technical experts from federal agencies.

7.     Fish and Other Living Organisms ("Biota")

       In 1991, Asiatic clams (Corbicula) were collected at five sediment sampling locations in
the mud flat area. Total PCB concentrations in Corbicula tissue derived by summing the PCB
congener analytical results ranged from 0.23 to 1.03 ppm.
                                          27

-------
                           IECEHP

                         i *"     PROPOSED SEDIMENT SAUPUNC LOCATIONS
                                DELAWARE RIVER SEDIMENT SAMPUNG LOCATION (JUNE/OCT., 1991)
                                ADDITIONAL SEDIMENT SAMPLING LOCATION (JUNE, 1991)
                        Am«n-w DELAWARE RIVER SEDIMENT SAMPLING LOCATION (MAY/JUNE, 1993)
                        *"'"    MUDFLAT GRID SAMPLE LOCATION (JUNE. IWJ)
                          ;™    LOCATION WITH > 1 PPM PCBS
                                                                                               240
                                                           $lg
Figure 8 - River Seduofot Sampling Locationi With Total PCJU 2 1 ppm

-------
       In 1993, fish samples were
collected from the Delaware River
in areas near the shoreline where
water depths ranged from 3 to 20
feet.  Nine whole body catfish
samples were analyzed for PCBs
with results ranging from 1.08 to
4.0 ppm. Total PCB
concentrations in ten catfish fillet
samples ranged from 0.41 to 1.48
ppm. Thirteen whole body silvery
minnow samples had total PCBs
ranging from 0.55 to 2.8 ppm.  The
Region in risk-based screening
level for human consumption of
fish containing PCBs is 0.00041
ppm.

       The Commonwealth of
Pennsylvania presently has  an
advisory in effect warning against
the consumption of catfish, white
perch, and American eel from the
Delaware River due to elevated
concentrations of PCBs and
chlordane. The advisory extends
from Burlington Island to Tinicum
Island, and includes the Site area.
VI.    RISK ASSESSMENT

       Based on the data provided
in the Remedial Investigation,
analyses were conducted to estimate
the human health and
environmental hazards that could
result if contamination at the Site is
not cleaned up.  EPA completed the
human health risk assessment at
this Site.  To evaluate
environmental impacts, EPA, with
the aid of Fish and Wildlife Service
performed the terrestrial risk
                                             Table 9 - Sediment Contaminants
Contaminant
Arsenic
Cadmium
Copper
Lead
Mercury
Nickel
Silver
Zinc
4,4'-DDD
4,4-DDE
4,4'-DDT
2,4-Dimethylphenol
4-Methylphenol
Acenaphthene
Acenaphthylene
Anthracene
Benzo(a)anthracene
Benzo(a)pyrene
Benzo(b)fiuoranthene
Benzo(g,h,i)perylene
Bis(2-ethylhexyl)phthalate
Chrysene
Dibenz(a, h)anthracene
Fluoranthene
Fluorene
lndeno(1 ,2,3-cd)pyrene
Naphthalene
Phenanthrene
Pyrene
Maximum
(ppm)
290
10.8
695
2,030
0.554
35.2
490
1,110
0.31
0.34
0.016
19
26
67
11
69
60
46
38
19
1,900
17
1.6
27
76
66
53
67
130
Screening
Level
8.2
1.2
340
46.7
0.15
20.9
1
150
O.016
0.0022
0.00158
0.029
0.067
0.016
0.044
0.0853
0.261
0.43
3.2
0.67
1.3
0.384
0.00634
0.6
0.019
0.6
0.16
0.24
0.665
                                         29

-------
assessment, while NOAA conducted the aquatic risk assessment.

       Actual and threatened releases of hazardous substances from this Site, if not addressed
by implementing the response action selected in this ROD, may present an imminent and
substantial endangerment to public health, welfare, and the environment.

A.    Human Health Risk Assessment

       Contaminants which exceeded risk-based screening levels developed by EPA Region HI
were included in the Site-specific Baseline Human Health Risk Assessment.  The primary
contaminants posing a potential human health risk at the Site are PCBs. In addition, antimony,
arsenic, beryllium, cadmium, chromium, 3,3'-dichlorobenzidine, PAHs, and dioxin were
observed at levels of potential concern.  To assure that the assessment was protective of human
health overall, contaminants exceeding screening levels in one medium (e.g., soils, sediments,
groundwater) were assessed in all other media.

       EPA identifies potential human health risks by estimating a carcinogenic risk level
("cancer risk") and a noncarcinogenic hazard index ("HI"). The National Oil and Hazardous
Substances Pollution Contingency Plan ("NCP") establishes the acceptable range of
carcinogenic risk at Superfund sites to be from one excess case of cancer per 10,000 people
exposed to one excess case  per 1,000,000 people exposed. Expressed as scientific notation,
this carcinogenic risk range is between l.OE-04 and l.OE-06. Remedial action is warranted
when the total estimated cancer risk from site contaminants to potentially exposed individuals
exceeds l.OE-04. However, since EPA's cleanup goal is generally to reduce the risk to l.OE-
06 or less, EPA  also may take action where the risk is within the range between l.OE-04 and
l.OE-06.

       The potential for noncarcinogenic effects is evaluated by comparing an exposure level
over a specified time period (i.e., the chronic daily intake) with the toxicity of the contaminant
for a relatively similar duration of exposure (i.e., the reference dose).   The ratio of exposure
to toxicity is called a hazard quotient.  An HI is generated by adding the appropriate hazard
quotients for contaminants to  which a given population may reasonably be exposed. The NCP
also  states that sites should  not pose a health threat due to a noncarcinogenic, but otherwise
hazardous, chemical.  If the HI exceeds one (1.0), concern for the potential noncarcinogenic
health effects associated with  exposure to the chemicals may exist.  The HI identifies the
potential for the most sensitive individuals to be adversely affected by the noncarcinogenic
effects of chemicals. Generally, the greater the value of the HI above  1.0, the greater the
level of concern.

       The current zoning classification  for the Site and the adjacent properties is Least
Restricted Industrial which  allows any lawful industrial use. Areas immediately beyond these
properties are zoned as General Industrial, which include a broad range of industrial activities,
along with some commercial areas.  The nearest area zoned as Residential is approximately
one-half mile from the Site  beyond Interstate 95 and the railroad line, although some

                                          30

-------
residential uses exist nearer the Site, including the St. Vincent's School. Based on the type of
development that has occurred in the vicinity of the Site over the last decade, significant
changes in land use are not expected to occur.

       Based on the current and expected future use of the Site and surrounding areas, EPA
identified the groups of individuals who could be potentially exposed to contaminants from the
Site to include current and future off-site residents, future on-site industrial workers, future
on-site construction workers, and current and future recreational boaters who fish near the
Metal Bank property.  Table 10 summarizes the potential cancer risks to these groups of
individuals estimated in the final Baseline Human Health Risk Assessment.  The hazard index
for each group is less than one which indicates that Site contaminants do not pose an
unacceptable noncarcinogenic risk.
            TABLE 10:  SUMMARY OF HUMAN HEALTH CANCER RISKS
EXPOSURE
PATHWAYS
Ambient Air Inhalation
Soil Ingestion:
Courtyard surface soil
S. Portion surface soil
Building dust
Dermal Contact with
Ground Water
Dermal Contact with
LNAPL
Surface Water Ingestion
Dermal Contact with
Surface Water
Sediment Ingestion
Fish Ingestion
TOTAL
Off-site
Resident
2.02E-06
NA
NA
NA
NA
NA
NA
•»•••••* *••
NA
2.02E-06
Industrial
Worker
4.42E-07
(6.77E-05)
1.76E-05
2.55E-06
3.86E-06
NA
NA
NA
NA
	 NA
NA
(7.46E-05)
2.45E-05
Construction
Worker
1.63E-06
9.42E-06
6.37E-06
(6.00E-03)
1.56E-03
NA
NA
NA
NA
(6.01 E-03)
1.S8E-03
Recreational Boater
Adult
NA
NA
NA
NA
1.2E-10
3.41 E-07
3.17E-06
(2.62E-04)
6.79E-05
(2.65E-04)
7.14E-05
Child
NA
NA
NA
NA
5.97E-10
1.59E-07
7.24E-06
(1.53E-04)
3.96E-05
(1.60E-04)
4.61 E-05
NA - Not applicable
       Subsequent to completion of the final Baseline Risk Assessment for the Site, EPA
revised the cancer slope factor for PCBs from 7.7 kg-d/mg to 2.0 kg-d/mg.  Because the
                                          31

-------
potential risk from several exposure pathways resulted from PCB contamination, EPA
reevaluated these pathways and the total cancer risk to relevant groups of individuals using the
revised slope factor.  The exposure pathways which made a significant contribution to the total
cancer risk from PCB contamination include ingestion of Courtyard surface soil, dermal
contact with LNAPL, and ingestion of fish.  The cancer risks calculated using the revised PCB
slope factor associated with these pathways are underlined in Table 10 and the cancer risks
based on the PCB slope factor used for the 1995 Risk Assessment are presented in parentheses
above the underlined  values.  The resulting change in the total cancer risk using the new PCB
slope factor is also presented below the value from the final Baseline Human Health Risk
Assessment, which is in parentheses.

Off-Site Residents: The potential route of exposure for off-site residents is inhalation of dust
from Site surface soils. The potential cancer risk associated with this exposure pathway is
calculated to be 2.02E-06, which is within the acceptable risk range and near the target
cleanup goal of l.OE-06. The change in the PCB slope factor does not affect this calculation.
The people who would be potentially exposed would be adults and children living adjacent to
the Site, such as those at the St. Vincent's School.  No HI for noncarcinogenic risk was
calculated since none  of the Site-related contaminants have inhalation reference doses.

Future Industrial Workers: The potential cancer risk associated with future industrial
employees working at the Site on a regular basis is estimated to be 2.45E-05 using the revised
PCB slope factor, which is within the acceptable risk range, but exceeds the target cleanup
goal of l.OE-06.  The primary route of exposure is incidental ingestion of contaminated soil,
particularly soil in the Courtyard contaminated with PCBs. Incidental ingestion may occur
when an individual eats food with hands that have been in contact with contaminated soils.
Arsenic, beryllium, and chromium in surface soils in the Southern Portion of the property and
PCB contamination in dust from the Building Area also contributed to the cancer risk
associated with soil ingestion, but at substantially lower levels. Risk from exposure through
inhalation of ambient  air was calculated, but did not contribute significantly to the overall
cancer risk. EPA did not consider future ingestion of groundwater from the Site as a realistic
exposure pathway for future industrial workers since any future commercial tenants of the
property are likely to  use public water supplies. The HI for noncarcinogenic risk associated
with ingestion of surface soils in the Southern Portion of the Site is 5.64E-03, which is
substantially below the acceptable level of 1.0.  No HI values were calculated for the
remaining exposure pathways because appropriate reference doses are not available.

Future Construction Workers: The potential cancer risk to construction workers who spend one
year working at the Site is estimated to be 1.58E-03 using the revised PCB slope factor, which
exceeds the acceptable risk range.  This risk is primarily from workers coming into contact
with PCBs found in the oil layer floating on the groundwater at the Site. Other exposure
pathways  contributing to the risk are incidental ingestion of subsurface soil from the Southern
Portion of the property, inhalation of ambient air, and dermal contact with groundwater. As
with the industrial workers, ingestion of groundwater from the Site was not considered  a
viable exposure pathway. The HI for noncarcinogenic risks is 0.3, which is below the

                                           32

-------
acceptable level of 1.0.

Recreational Boaters: The potential cancer risks to recreational boaters are estimated to be
7.14E-05 and 4.61E-05 using the revised PCB slope factor for adults and children,
respectively. These risks are within the acceptable range, but exceed the target cleanup goal of
l.OE-06. The primary exposure route is consumption of fish caught in the Delaware River
while boating.  Other exposure pathways contributing to the overall cancer risk include
ingestion of and dermal contact with surface water and incidental ingestion of sediments.  The
HI values for recreational boaters are 1.07E-03 and 9.98E-03 for adults and children,
respectively, both of which are below the acceptable level of 1.0.

B.     Aquatic Ecological Risk Assessment

       The Metal Bank property is located adjacent to a three-acre tidal mudflat and the
Delaware River.  Both are sensitive freshwater ecological systems that provide permanent and
seasonal habitat for numerous species of anadromous, catadromous, estuarine, and freshwater
fish.  To assess the impact from various Site contaminants, the Aquatic Ecological Risk
Assessment ("Aquatic Assessment") for the Site identified potential receptors, contaminants of
concern, and exposure pathways. Potential risks to receptors from Site contaminants were
calculated in a manner similar to that used to estimate noncarcinogenic risk in the Baseline
Human Health  Risk Assessment.  Levels of contaminants found in media (e.g., sediment,
groundwater, surface water) to  which receptors could be exposed were compared to levels
considered to be toxic based on research literature. If the ratio between the contaminant level
observed and the  toxicity level exceeds 1.0, the potential for an adverse impact exists.

Receptors:  Receptors considered during the Aquatic Assessment to evaluate potential impacts
from the Site include shortnose sturgeon, channel catfish, silvery minnow, white perch, and
benthic invertebrates. The shortnose sturgeon has been designated an endangered species both
federally and by the Commonwealth of Pennsylvania. Shortnose sturgeon live their entire life
cycle  in the Delaware River and are known to be present in the river reach that includes the
Site during summer after spawning in upstream areas.

       Channel catfish serve as a representative of a benthic freshwater fish species that feed
on a variety of  prey types and are likely to live in the river near the Site for some time.
Channel catfish are also used as a surrogate species for the shortnose sturgeon.  Silvery
minnow are an  important forage fish species that feed in shallow water areas, such as the
mudflat next to the Site.  White perch are an anadromous fish species that is abundant near the
Site and recreationally important.

       Benthic  invertebrates, such as Asiatic clams (Corbiculaflumined), are present in the
mudflat and Delaware River at the Site.  The invertebrates were considered as a single group
in the Aquatic Assessment.  Benthic invertebrates may provide an important exposure pathway
for higher trophic level fish and wildlife species that feed in the intertidal and nearshore
Delaware River habitats.

                                           33

-------
Contaminants of Concern:  PCBs were identified as the contaminants of primary concern
because of their presence at elevated concentrations in groundwater, NAPL, and sediment.
Other contaminants were screened for consideration in the Aquatic Assessment using EPA's
chronic ambient water quality criteria for surface water and groundwater and Effects-Range
Low values compiled from research literature for sediment. Other contaminants of concern
included PAHs, phthalates, DDT and its metabolites DDE and DDD, and cadmium.

Exposure Pathways: The contaminant concentrations that receptors may contact were estimated
for pathways including surface water, direct exposure  to NAPL, and contact with sediment.
Exposure-point concentrations for surface water in the Delaware River which could result
from Site groundwater discharges were estimated using dilution factors calculated for a 15-
meter zone of the river adjacent to the Metal Bank property during average- and low-flow
conditions. Exposure-point concentrations for NAPL were estimated based on the
concentrations of PCBs, PAHs, and phthalates collected from monitoring well MW6 during
the 1991 sampling event.

       For sediments, the mean and upper 95 percent confidence limit values were determined
using one-half the detection limits for values below detection if at least one sample had
concentrations above the detection limit for that contaminant. PCB data from the January 1995
sediment sampling event were not used in the Aquatic Assessment. Most of these data were
collected in the farfield area and appear consistent with  earlier results. Because sediment
concentrations for PCBs, PAHs, and several phthalates decreased in concentration with
distance from the Metal Bank property boundary,  sediments were evaluated in three separate
zones:  samples from the riprap area; samples from the mudflat and Delaware River within 30
meters of the property boundary (referred to as nearfield samples); and samples from the
mudflat and Delaware River greater than 30 meters from the property boundary (referred to as
farfield samples). Table 11 presents the mean  and upper 95% confidence limits for total PCBs
normalized to dry weight which were calculated for these  three zones.  No similar gradient
was apparent in sediment concentrations of DDTs and  cadmium.  Therefore, the aquatic risk
evaluation for these contaminants was based on the combined values for all sediment samples.
                                                 Table 11 - Total PCBs in Sediment
                                                   with Increased Distance from
                                                      the Metal Bank Property
       PCB concentrations found in fish tissue
integrate all the above exposure pathways,
since fish accumulate most of their PCB body
burden through the food web. Whole body
concentrations in silvery minnow and channel
catfish collected in the fall of 1993 from the
Delaware River near the Metal Bank property
were used to estimate concentrations in other
fish species.

Toxicity:  Contaminant concentrations that a
receptor may be exposed to at the Site were
compared to concentrations considered to be toxic (i.e., toxicity reference values).  For
Area
Riprap
Nearfield
Farfield
Total PCBs (ppm)
Mean
5.9
3.8
0.87
Upper 95%
Conf. Limit
9.4
5.0
1.2
                                          34

-------
surface water, toxicity reference values were based on EPA's chronic ambient water quality
criteria for the protection of freshwater organisms. Because of the endangered species status of
the shortnose sturgeon, the limited amount of available chronic toxicity data for fish species,
and the lack of species-specific toxicity information, a safety factor of 100 was applied to the
toxicity reference value for sturgeon.  Toxicity reference values for NAPL were based on
EPA's acute ambient water quality criteria. No similar criteria are available for sediments.
Therefore, the sediment toxicity reference concentrations are based on research which
compiled a large number of studies on the adverse effects of sediment contaminants on aquatic
organisms and calculated the mean and lower tenth percentile concentrations when such effects
were observed.  The mean concentrations, referred to as the Effects Range-Medium or ER-M,
is the concentration at which effects are considered probable but not assured. The lower tenth
percentile concentration,  referred to as the Effects Range-Low or ER-L, is the concentrations
below which adverse biological effects would be unlikely.  If these reference values were not
available for a contaminant, the less conservative Apparent Effects Threshold ("AET") was
used. These concentrations represent levels at which some biological injury was always
observed in a study conducted in the Puget Sound. Toxicity reference values for PCBs in fish
tissue were also derived from a compilation of published studies linking adverse effects to PCB
tissue concentrations in a variety of fish species.   The lower tenth percentile and the mean of
the data set of tissue effects concentrations are used in the  Aquatic Assessment  as toxicity
reference values for PCBs in tissues for all fish species.

Risk Characterization:  To determine  the potential risk posed by a Site contaminant, the site-
specific exposure-point concentration  for a contaminant is compared to the appropriate toxicity
reference value. If the ratio of the exposure-point concentration to the toxicity  reference value
is greater than one (i.e., if the exposure-point concentration exceeds the toxicity reference
value), a potential risk is present. This ratio is referred to as the hazard quotient ("HQ").
Since exposure-point concentrations were developed using  both  the mean and upper 95th
percentile values and several toxicity reference values were used for some contaminants, a
range of HQ values was generated for various contaminants.  In general, if any HQ value
calculated was greater than one, a potential risk exists.  In order to fully consider all aspects
influencing the risk,  the frequency and duration an organism would be exposed to that media
was also considered.

Surface Water Risks:  Surface water exposure from groundwater discharge is reduced rapidly
with dilution from the Delaware River. Only PCBs were considered a contaminant of concern
for surface water.  Any exposure to PCB concentrations exceeding toxicity reference
concentrations would be confined to a 15-meter-wide band of the river next to the Metal Bank
property. The risk of adverse effects  is expected to be low. Under low-flow conditions,
which represents a worst-case scenario, PCB concentrations in this area of the Delaware River
were estimated to be between 1.34 and 1.97 nanograms per liter ("ng/l").  These
concentrations are considerably below the toxicity reference value of 100 ng/1 considered
protective of most aquatic organisms.   These concentrations slightly exceed the conservative
toxicity reference value of 1 ng/1 used for the shortnose sturgeon, indicating a possible risk.
However, the likelihood that individual shortnose sturgeon would remain in the exposure area

                                           35

-------
long enough to receive chronic exposure is very low.
NAPL Risks: Exposure to NAPL would occur if organisms directly ingest water from seeps
prior to dilution by the river.  The PCB concentration of 1,090 ppm in NAPL at monitoring
well MW6 greatly exceeds the toxicity reference of 2 ppb which is considered protective of
most aquatic organisms.  Individual PAH concentrations also exceeded toxicity reference
values by factors ranging from 20 to 4,500 and the bis(2-ethylhexyl)phthalate concentration of
590 ppm is 1,500 times greater than the toxicity reference value.  While direct seepage of oil
to the surface in the riprap or mudflat areas is likely to be highly localized, exposure to the
high concentrations of PCBs, PAHs, and phthalates in any NAPL seep is likely to result in
acute toxicity to organisms located in the path of these seeps. In addition, species with limited
mobility, such as clams,  that occupy the mudflat may experience long-term exposure to
elevated PCB concentrations from NAPL discharges.

Sediment Risks: Contact with sediment is the primary route of exposure for benthic organisms,
and subsequently the fish and other organisms that feed  on them, in the nearshore areas of the
Delaware River and mudflat near the Metal Bank property.  The range of hazard quotients
calculated for PCBs, PAHs, and bis(2-ethylhexyl)phthalate, di-n-butyl phthalate, di-n-octyl
ph thai ate, DDT, DDE, DDD, and cadmium are presented in Table 12. Shaded blocks in the
table indicate that the entire range of hazard quotients calculated for that contaminant  exceeds
the acceptable level of 1.0.  The highest values resulted from using an exposure-point
concentration based  on the upper 95% confidence limit of the data set and the ER-L (or lowest
AET if no ER-L value is available) for the contaminant.  The lowest hazard quotients were
calculated using the mean concentration and the ER-M.
           Table 12 - Hazard Quotients for Aquatic Receptors Exposed to Sediment
           Contaminant
                                             Hazard Quotient Range
Riprap Area
Total PCBs

Total PAHs

Bis(2-ethylhexyl)phthalate
Di-n-butyl phthalate

Di-n-octyl phthalate

DDT
                                     0.4-2.0
                                                Nearfield Area
                                                   1.0-3.8
                0.1-0.5

                0.2 - 4.0
                                                    Farfield Area
                             •^.B^SZ.4,

                              0.1-1.0

                              0.1-0.4
0-0.5

0-0.9
           DDE

           DDD

           Cadmium
                                                            :
                        f^ jgi'j; jfrjjff i'A^L**'!!'- ':'J: :y*& ":?*%*.&:£-*L.*-• i-., • <-.: •= ^^^j^f^^ • *«^ _._./_ L!V; L •_
                             SS^Si^^l^feft®^
                       t;-'f"-:^Slfl^^J5 -144:2:

                                          36

-------
       For total PCBs, total PAHs, and phthalates, the hazard quotients reflect the decreasing
pattern of contamination observed with distance from the Metal Bank property.  No spatial
pattern was observed in the concentrations of DDT, DDD, DDE, and cadmium based on the
limited data sets available.  Even though the hazard quotients for total PCBs decrease with
distance from the property, all values exceed the acceptable limit of 1.0. For total PAHs, the
entire range of hazard quotients exceeds 1.0 in the riprap and nearfield areas.  Hazard
quotients for bis(2-ethylhexyl) phthalate and di-n-octyl phthalate exceed the acceptable level of
1.0 for the entire range in the riprap area. The hazard quotients calculated for  DDT, DDE,
DDD, and cadmium using all the sediment data also exceed the acceptable  level for the entire
range of values.

Fish & Organism Tissue Risks: Concentrations of PCBs in the tissues of fish species represent
an integration of all exposure pathways. PCBs were found in Corbicula, channel catfish,  and
silvery minnows collected from the Delaware River next to the Metal Bank property. Because
invertebrates appear to be much less sensitive to PCB tissue residues than fish species, the
PCB concentrations in the Corbicula were not evaluated for potential toxicity due to tissue
residue effects.

       To estimate the potential risk to channel catfish, and indirectly to other fish species
including the shortnose sturgeon, the mean and upper 95% confidence limit of the PCB
concentrations found in whole body catfish samples (2.1 and 2.9 ppm, respectively) were
compared to the toxicity reference values. Hazard quotients ranged from 10.5 to 14.3 using
the toxicity reference value based on the lower tenth percentile of tissue effects concentrations
from compiled studies.  The hazard quotients calculated using the less conservative toxicity
reference value based on the mean  tissue effects concentrations were less than 1.0.  Similar
results were observed using the silvery minnow tissue data.  Using these species as
representative of other fish species with potential exposure to the nearshore PCB-contaminated
sediments or to nearshore food webs, the results suggest a potential risk of reproductive effects
in sensitive fish species. Although the shortnose sturgeon occupy a lower trophic level than the
channel catfish, the sturgeon may be particularly prone to accumulating and transferring high
concentrations of PCBs to their developing offspring due to their benthic feeding habit,
longevity (known to live over 60 years), late age of sexual maturity, and high lipid content of
their eggs. Thus, the potential risk to shortnose sturgeon resulting from accumulation of PCBs
from all exposure pathways at the Site may be greater than for other fish species because of
these life-history characteristics.

C.     Terrestrial Ecological Risk Assessment

       Potential risks to terrestrial  plants and animals from Site contaminants were estimated
by comparing the concentrations found in various media (i.e., soil, sediment, groundwater,
surface water) to toxicity values based on published criteria documents and literature reviews.
The approach used in the Terrestrial Ecological Risk Assessment ("Terrestrial Assessment")
does not focus on impacts to specific receptors because species-specific data was not collected

                                           37

-------
as part of the Remedial Investigation.  Therefore, the toxicity values are based on the most
conservative toxicity values available to provide the greatest degree of protection.
       The expected environmental
concentrations ("EEC") used in the
Terrestrial Assessment were
generally based on the upper 95%
percent confidence limit of the data
collected for each contaminant in
each medium. The ratio of the EEC
to the relevant toxicity value is
referred to as the environmental
effects quotient ("EEQ").  Any
contaminant with an EEQ that
exceeded 1.0 was included as a
contaminant of concern. The EEQs
for all contaminants found in a
particular medium were added.  Any
media with EEQs totaling more than
10 is considered to have a potentially
high adverse impact on terrestrial
receptors.

       The media evaluated in the
Terrestrial Assessment  include
ground water, soil, mudflat
sediments, and riprap sediments.
The findings are summarized below.

Groundwater.  Terrestrial receptors
may be exposed to contaminants in
groundwater directly through seeps
and springs and indirectly as a result
of contaminant transport to other
media such as surface water and
sediment. Table 13 lists the
groundwater contaminants, which
have potential to adversely impact
terrestrial receptors (i.e., have an
EEQ > 1), and their EEQs. The
total EEQ for groundwater is 978,
which exceeds the acceptable level of
10 and indicates a potential risk.
Table 13 - Groundwater Environmental Effects Quotients
Contaminant
Cadmium
Lead
Mercury
Chloride
Chromium (VI)
Copper

EEQ
3.61
115
251
1.25
26.9
7.47

Contaminant
Zinc
DDT
DDE
Endosulfan
Phenanthrene
Heptachlor
Total
EEQ
6.66
26.9
487
4.91
4.43
42.6
978
 Table 14 • Sediment Environmental Effects Quotients
Contaminant
Total PCBs
4,4'-DDD
4.4'-DDE
Acenaphthene
Anthracene
Benzo[a]anth racene
Benzo[a]pyrene
Chrysene
Dibenz[ah]anthracene
Fluoranthene
2-methylnaphthalene
Phenanthrene
Napthalene
Phenanthrene
Pyrene
Total
Sediment EEQs
Riprap
188
'—
—
1,600
2,910
950
419
510
290
462
6,750
1,670
706
2.430
-
18,885
Nearfield
100
235
—
—
—
3
3
2
38
2
2
—
—
—
3
388
Farfield
24
115
33
—
—
—
—
—
—
—
—
—
—
—
—
172
                                          38

-------
Soil:  Methyl ethyl ketone was the only soil contaminant with an EEQ greater than one.  This
contaminant was found in the subsurface soil in the Southern Portion of the property and had
an EEQ of 9.66.  Although this value indicates a potential risk, the risk is associated with soils
deep in the profile and would likely be an exposure route only for deeply rooted vegetation.

Sediment:  Contamination in the mudflat and riprap sediments are of concern in the Terrestrial
Assessment because of the potential for impact to benthic and intertidal zone organisms and
opportunistic birds and mammals that may use the area on either a diurnal"or seasonal basis.
Table 14 lists the sediment contaminants, which  have potential to adversely impact terrestrial
receptors (i.e., have an EEQ > 1), and their EEQs.  The sediments have been divided into the
same three areas used in the Aquatic Assessment: the riprap area,  nearfield sediments, and
farfield sediments. The total EEQs for these three areas are 18,885, 388, and 172,
respectively,  which all exceed the acceptable level of 10 and indicate a potential risk.

       The EEQs presented  in Table 14 differ slightly from those  presented in the Terrestrial
Assessment because the upper 95% confidence limits for total PCBs calculated in the Aquatic
Assessment for the three sediment areas were used as the EECs for consistency.  The
remaining EECs used throughout the Terrestrial Assessment were based on the data summary
used to prepare the 1994 Baseline Human Health Risk Assessment.  As noted previously, this
data summary did not provide a complete summary of the data collected at the Site during the
1991 and 1993 sampling events and was subsequently revised after completion of the  Terrestrial
Assessment.  However, the revised data summary did not significantly change the EECs used in
the Terrestrial Assessment, so the Terrestrial Assessment was not revised.
VII.   DESCRIPTION OF ALTERNATIVES

       The Feasibility Study for the Metal Bank Site screened possible approaches for cleaning
up the contamination at the Site. Based on this screening, five alternatives, including a
required no action alternative, were developed and evaluated. EPA modified one of these
alternatives as its preferred alternative in the Proposed Plan. Descriptions of each of these six
alternatives are presented below.

Alternative C-l:    No Action; Monitoring

Capital Cost:                      $      0
Annual  Operation and
  Maintenance ("O&M") Costs:    $  347,000 (years 1-2)
                                  $  87,000 (years 3-30)
Present Worth Cost:                $1,821,000
Implementation Time:                 3 weeks

       The NCP requires that a no action alternative be considered at every Site. This
alternative serves as a baseline  for the comparison of other alternatives. Although no cleanup

                                          39

-------
action would be performed under this alternative, monitoring of sediments, groundwater,
surface water and LNAPL would be performed for a period of 30 years.

Alternative C-5:    Limited Soil Excavation and Disposal; Sediment Excavation and
                    Consolidation; Oil Collection System; Impermeable Cap

Capital Cost:               $ 10,405,000
Annual O&M Costs:        $   455,000 (years 1-2)
                           $   195,000 (years 3-30)
Present Worth Cost:        $ 13,889,000
Implementation Time:          12 months

       Soil in Courtyard Area with PCB concentrations exceeding 10 ppm within two feet of
the surface (an estimated 600 cubic yards) would be excavated and disposed off-site at a
permitted facility.  An estimated 11,750 cubic yards of sediments with PCB concentrations
exceeding 1 ppm within approximately 100 feet of the Metal Bank property boundary would
be excavated and dewatered.  These sediments would then be consolidated with soil on the
Southern Portion of the property, if PCB concentrations are less than 25 ppm, or disposed off-
site at a permitted facility if PCB concentrations are greater than 25 ppm. Excavated areas
would be backfilled with clean material.  A temporary sheet pile cofferdam would be installed
prior to sediment excavation to keep water from the Delaware River out of the excavation
area. Water that has been removed from the sediments would be treated,  if necessary,  and
disposed on-site after the oil  collection system is installed. Riprap along the property
perimeter would be removed prior to excavation, decontaminated, and replaced along the
perimeter after excavation is completed.

       The underground storage tank in the Southern Portion of the property would be
removed.  Tank contents would be sampled and disposed  off-site at a permitted facility. The
tank would be decontaminated and disposed as scrap metal. Debris and soil in the vicinity of
the tank that exceed PCB concentrations of 50 ppm would be disposed off-site at a permitted
facility. Excavated sediment would be used to fill this area.

       An oil collection system  would be installed along the western and southern perimeter of
the Metal Bank property. The system would include a permanent sheet pile wall along the
mudflat and river shoreline.  Groundwater interceptor trenches would be excavated inside the
sheet pile wall and oil-water  separators and sump pumps,  or an equivalent system, would be
placed in the trenches to collect oil that seeps from the soil and groundwater on the Southern
Portion of the property.  Oil collected in the oil-water separator would be disposed off-site at a
permitted facility.

       Following completion of the above activities,  an impermeable cap would be installed
over the Southern Portion of the property.  Access to the property would be limited by fencing
the area and requiring deed restrictions to prevent improper use.  Long-term monitoring would be
performed and warning signs would be posted advising against consumption offish caught in the

                                          40

-------
Delaware River in the vicinity of the Site.

Alternative C-7:     Limited Soil Excavation and Disposal; Sediment Excavation and
                     Consolidation; Soil/Sediment Stabilization; Oil Collection System;
                     Impermeable Cap

Capital Cost:               $ 14,753,000
Annual O&M Costs:         $   451,000 (first two years)
                            $   190,000 (years 3-30)
Present Worth Cost:         $ 18,159,000
Implementation Time:           27 months

       Alternative C-7 includes all the components discussed above under Alternative C-5
along with an additional treatment component. An estimated 11,750 cubic yards of excavated
sediment would be consolidated with an estimated 10,000 cubic yards3 of soil from the
Southern Portion of the Metal Bank property which have PCB concentrations exceeding 25
ppm and treated using a stabilization process.  A treatability study would be required prior to
implementation to determine the appropriate mix of reagents necessary to immobilize the
contaminants.  Treatment would occur after the oil collection system had been installed.  The
stabilized material would be placed back on the Southern Portion of the property prior to
installation of the impermeable cap.

Alternative C-7A:   Soil Excavation and  Disposal; Sediment Excavation  and
                     Consolidation; Oil Collection System; Soil Cover

Capital Cost:               $14,499,000
Annual O&M Costs:         $   409,000 (years 1-2)
                            $   149,000 (years 3-30)
Present Worth Cost:         $ 17,273,000
Implementation Time:          27 months

       Alternative C-7A includes all the components of Alternative C-5 with the exception of
the impermeable cap.  Alternative C-7A would require the use of a soil cover rather than an
impermeable cap.   In addition, Alternative C-7A requires excavation and off-site disposal at a
permitted facility of contaminated soil from the Southern Portion of the Metal Bank property
3   For Alternatives C-7, C-8, and C-12, the FS estimated the volume of soil requiring remediation in the Southern
Portion of the Metal Bank property to be 71,000 cubic yards based on TPH concentrations. EPA has determined that
TPH measurements do not correlate with the presence of PCBs and, therefore, do not provide an adequate basis for
measuring protection of human health and the environment.  EPA has estimated the volume of soil requiring
remediation in the Southern Portion of the property to be approximately 10,000 cubic yards based on PCB data collected
at the Site. This volume represents soils which are estimated to have PCB concentrations exceeding 25 ppm. The cost
estimates for these alternatives have been revised from those presented in the FS to reflect this lower volume and
included in the Administrative Record for the Site.
                                            5
                                            41

-------
with PCB concentrations exceeding 25 ppm. Excavated areas would be backfilled with
excavated sediment and a 24 inch soil cover would be placed over the Southern Portion of the
property.  The soil would be graded and seeded to establish an erosion-resistent vegetative
cover. A similar soil cover 12 inches in depth would be placed over the soil in the Courtyard
area.

Alternative G-8:    Limited Soil Excavation and Disposal; Sediment Excavation and
                    Consolidation; Soil/Sediment Stabilization; Oil Collection System;
                    Impermeable Cap

Capital Cost:               $19,540,000
Annual O&M Costs:         $    451,000 (years 1 -2)
                           $    190,000 (years 3-30)
Present Worth Cost:         $22,946,000
Implementation Time:          34 months

       Alternative C-8 is the same as Alternative C-7 except that excavated soil and sediment
is treated using  a soil washing process rather than  a stabilization process. Soil washing is a
technology in which contaminated soil/sediment would be flushed with solvents to dissolve and
remove the contamination. The spent solvent would be recovered for continued treatment and
the concentrated contaminants would be disposed at a permitted facility.

Alternative C-12:    Soil Excavation and Disposal; Sediment Excavation and
                    Consolidation; Oil Collection System; Impermeable Cap

Capital Cost:               $15,917,000
Annual O&M Costs:         $   451,000 (years 1 -2)
                           $   190,000 (years 3-30)
Present Worth Cost:         $ 19,323,000
Implementation  Time:          21 months

       Alternative C-12 is the same as Alternative C-7A except for the soil cover component.
Alternative C-12 would not require any soil cover, but would require an impermeable cap over
the Southern Portion of the Metal Bank property following implementation of the  other
components of the remedy.
VIII.  ANALYSIS OF ALTERNATIVES

       Each of the six remedial alternatives summarized in this ROD, including the No Action
Alternative, has been evaluated against the nine evaluation criteria set forth in the NCP, 40
C.F.R. Section 300.430(e)(9). These nine criteria can be categorized into three groups:
threshold criteria, primary balancing criteria, and modifying criteria. A description of the
evaluation criteria is presented below. This section evaluates each alternative based on these

                                         42

-------
criteria.

Threshold Criteria:

  1.    Overall Protection of Human Health and the Environment addresses whether a remedy
       provides adequate protection and describes how risks are eliminated, reduced, or
       controlled.

  2.    Compliance with Applicable or Relevant and Appropriate Requirements (" ARARs")
       addresses whether a remedy will meet all of the applicable, or relevant and appropriate
       requirements of environmental statutes.

Primary Balancing Criteria:

  3.    Long-term Effectiveness refers to the ability of a remedy to maintain reliable protection
       of human health and the environment over time once cleanup goals are achieved.

  4.    Reduction ofToxicity, Mobility, or Volume through Treatment addresses the degree to
       which alternatives employ recycling or treatment that reduces toxicity, mobility, or
       volume of contaminants.

  5.    Short-term Effectiveness addresses the period of time needed to achieve protection and
       any adverse impacts on human health and environment that may be posed during the
       construction and implementation period until cleanup goals are achieved.

  6.    Implementability addresses the technical and administrative feasibility of a remedy,
       including the availability of materials and services needed to implement a particular
       option.

  7.    Cost includes estimated capital, operation and maintenance costs, and present worth
       costs.

Modifying Criteria:

  8.    State Acceptance indicates whether, based on its review of backup documents and the
       Proposed Plan, the State concurs with, opposes, or has no comment on the preferred
       alternative.

  9.    Community Acceptance includes assessments of issues and concerns the public may
       have regarding each alternative based on a review of public comments received on the
       Administrative Record and the Proposed Plan.
                                          43

-------
A.    Overall Protection of Human Health and the Environment

Soils:  Alternatives C-5, C-7, C-7A, C-8, and C-12 protect human health and the environment
by minimizing the potential for exposure to contamination in the Courtyard surface soils and
the subsurface soils in the Southern Portion of the Metal Bank property. These alternatives
also reduce the potential for migration of contamination to the nearby aquatic environment
thereby enhancing the overall level of protection. Alternative C-l is not protective of human
health or the environment because potentially unacceptable human health and environmental
risks from exposure to contaminated soils remain at the Site.

Sediments:  Alternatives C-5, C-7, C-7A, C-8 and C-12 protect the aquatic environment
currently impacted by Site contaminants by removing contaminated sediments from areas
within approximately 100 feet of the Metal Bank property and backfilling this area with clean
material.  Alternative C-l is not protective of the aquatic environment because potentially
unacceptable risks from exposure to contaminated sediments remain at the Site.

Oil/LNAPL Layer: Alternatives C-5,  C-7, C-7A, C-8, and C-12 protect human health and the
environment by installing an oil collection system along the  southern and western perimeter of
the Metal Bank property to passively remove remaining oil as it accumulates on the surface of
the shallow groundwater.  Direct contact with contaminants  in the floating oil layer is
prevented through a deed restriction prohibiting excavation that would encounter the water
table.  Migration of contaminants from the Southern Portion of the property into the adjacent
aquatic environment is prevented by the oil collection system and associated sheet pile wall
along the property perimeter. Alternative C-l is not protective because potentially
unacceptable human health and environmental risks remain at the Site from exposure to
contaminants in the oil layer which could occur during future construction activities or through
surface discharges to the aquatic environment.

       Overall, Alternatives C-5, C-7, C-7A,  C-8, and C-12 provide overall protection of
human health and the environment, but Alternative C-l does not. Therefore, since Alternative
C-l does not meet this threshold criterion, it will not be evaluated further in this comparative
analysis.

B.     Compliance with ARARs

       Section 121(d) of CERCLA requires that remedial actions at CERCLA sites at least
attain legally applicable or relevant and appropriate federal and State standards, requirements,
criteria, and limitations which are collectively referred to as  "ARARs", unless such ARARs
are waived under CERCLA Section 121(d)(4).

       Applicable requirements are those cleanup standards, standards of control, or other
substantive requirements, criteria, or limitations promulgated under federal environmental or
state environmental law or facility siting laws that specifically address hazardous substances,
pollutants, contaminants,  remedial actions, locations, or other circumstances at a CERCLA

                                           44

-------
site.  Relevant and appropriate requirements are those same requirements that, while not
directly applicable at a CERCLA site, address problems or situations sufficiently similar to
those encountered that their use is well suited to the particular site. Only those state standards
that are identified in a timely manner and are more stringent than federal requirements may be
applicable or relevant and appropriate.  On-site actions must comply with ARARs, but need
comply only with the substantive parts of those requirements. By contrast, off-site actions
must comply with legally applicable requirements, including both substantive and
administrative parts of those requirements.

       The ARARs for the alternatives developed in this ROD are listed below.

1.     Chemical-Specific ARARs

Soils: PADEP has identified Pennsylvania's Land Recycling and Environmental Standards Act
(Act 2) and its implementing regulations (Administration of the Land Recycling Program, 25 Pa.
Code Chapter 250), as an ARAR. Act 2 sets forth three alternative cleanup standards: (1)
background; (2) statewide health-based standards; and (3) site-specific risk-based standards.
EPA chose a site-specific risk-based approach to develop cleanup standards for PCBs at the Site.
EPA has generally endeavored to develop a consistent approach to PCB remediations at
Superfund sites, and has relied on EPA's Guidance on Remedial Actions for Superfitnd Sites with
PCB Contamination (EPA/540/G-90/007, August 1990). EPA has established the PCB cleanup
levels of 10 ppm for surface soil and 25 ppm for subsurface soil at the Site based, in part, on this
guidance. The more conservative level of 10 ppm is used for the surface soils because the
potential for exposure is greater. These established PCB standards are in line with the
recommended ranges in EPA's guidance document, which are designed to put the resultant risks
to humans within the acceptable risk range (between l.OE-04 and l.OE-06 for potential
carcinogens). Cleanup levels were not established for other contaminants present at the Site
because their contribution to the risks were substantially less than those attributed to PCBs.
Therefore, incidental removal of these contaminants found in soils being remediated because of
their PCB concentrations will likely result in acceptable risk levels for these contaminants as
well.  Alternatives C-5, C-7, C-7A, C-8 and C-12 all effectively limit potential exposure to PCB
contamination above the recommended cleanup levels either by isolating these soils with an
impermeable cap or clean cover of soil, treating or removing the soils, or both.

Sediments:  EPA has established the PCB cleanup level of 1 ppm for sediments again based,
in part,  on the above mentioned guidance. The 1 ppm standard is designed to be protective of
sensitive ecological receptors.  As shown in Table  11 earlier in this ROD, total PCB
concentrations in samples collected more than 100 feet from the Metal Bank property
boundary averaged less than 1 ppm (0.87 ppm) while total PCB concentrations in samples
collected less than 100 feet from the property and in the riprap area averaged 3.8 and 5.9 ppm,
respectively. Alternatives C-5,  C-7, C-7A, C-8, and C-12 all require  excavation of sediments
near the property with total PCB concentrations above 1 ppm and, therefore, achieve the
recommended cleanup level.
                                          45

-------
2.     Action-Specific ARARs

       Alternatives C-5, C-7, C-7A, C-8, and C-12 would comply with all ARARs associated
with the actions being taken. These action-specific ARARs are summarized below.

Excavation:  The substantive requirements of the Pennsylvania Erosion Control Regulations at
25 Pa. Code, Chapter 102 are applicable to earthmoving work performed on the Site including
soil and sediment excavation and earthmoving necessary to install impermeable caps and soil
covers.

Storage of Excavated Soil/Sediment and Recovered NAPL:  The Pennsylvania Residual Waste
Management ("PRWM") Regulations at 25 Pa. Code Chapter 299, Subchapter A (Standards
for Storage of Residual Waste) are applicable to the storage of excavated soil and sediment and
recovered NAPL at the Site. In addition, the Pennsylvania Hazardous Waste Management
("PHWM") Regulations at 25 Pa. Code Chapter 264, Subchapter I (Use and Management of
Containers) are relevant and appropriate for storage in containers of excavated soil and
sediment and recovered NAPL that are determined to be hazardous.  PHWM Regulations at 25
Pa. Code Chapter 264, Subchapter L (Waste Piles) are relevant and appropriate for storage of
excavated soil and sediment in piles.  The Toxic Substances Control Act ("TSCA")
Regulations at 40 C.F.R. § 761.65 are applicable to the storage of any excavated soil and
sediment and recovered NAPL with PCB  concentrations of 50 ppm or greater.

Transportation of Excavated Soil/Sediment and Recovered NAPL: The general requirements
(§ 299.211) of the PRWM Regulations  at 25  Pa. Code Chapter 299, Subchapter B are
applicable to the  transportation of excavated soil and sediment from the Site. If the soils,
sediments, and NAPL are determined to be hazardous under the PHWM  Regulations at
25 Pa. Code Chapter 261, Subchapter C,  the transporter of the waste  would be required to
meet the PHWM Regulations at 25 Pa.  Code Chapter 263.

Disposal of Excavated Soil/Sediment and Recovered NAPL: For soils, sediments, and NAPL
that are determined to be hazardous under the PHWM Regulations at 25 Pa. Code Chapter
261, Subchapter  C, disposal would be required at a landfill in compliance with the  PHWM
Regulations at 25 Pa. Code Chapter 264,  Subchapter N (Landfills) or 25  Pa. Code 265,
Subchapter N (Landfills). For soils, sediments, and NAPL with PCB concentrations of 50
ppm or greater, disposal would be required at a facility in compliance with the TSCA
Regulations at 40 C.F.R. § 761.60. Soils, sediment, and NAPL that are not hazardous and
have PCB concentrations of less than 50 ppm would be disposed of at a landfill in compliance
with the PRWM  Regulations at 25 Pa. Code Chapter 288 (Residual Waste Landfill).

Air Emissions: The substantive requirements  of the Pennsylvania Air Resources Regulations at
25 Pa. Code §§ 123.1 (prohibition of certain fugitive emissions), 123.2 (fugitive paniculate
matter), and 123.31 (limitations on odor emissions) are relevant and appropriate to remedial
activities such as excavation and materials handling.
                                          46

-------
Underground Storage Tank Removal: The substantive requirements of RCRA Regulations at
40 C.F.R. Part 280, Subpart G are relevant and appropriate to the removal of the underground
storage tank ("UST") at the Site. In addition, PHWM regulations at 25 Pa. Code Chapter
265, Subchapter J are applicable to the treatment of hazardous waste in the tank.

Discharges to Surface Water: The more stringent of the substantive requirements of the Clean
Water Act at 40 C.F.R.  Part 122 (National Pollutant Discharge Elimination System
("NPDES")) or the Pennsylvania Clean Streams Law at 25 Pa. Code Chapter 92 (NPDES
Program), and 25 Pa. Code Chapter 93 (Water Quality Standards) and 25 Pa. Code Chapter 16
(Toxic Substances Water Quality Criteria) would be applicable to discharges from the Site of
groundwater from the oil collection system, decontamination water, or water from the
sediment dewatering process. In addition, the Delaware River Basin Commission's Water
Quality Regulations, as amended through October 23,  1996, as they relate to the Delaware
River Basin Compact would be applicable to discharges from the Site of groundwater from the
oil collection system, decontamination water, or water from the sediment dewatering process.

On-Site Treatment of Soil/Sediment:  The PRWM Regulations at 25 Pa.  Code Chapter 299,
Subchapter A (Standards for Storage of Residual Waste) are applicable to on-site treatment of
non-hazardous soil/sediment, and the PHWM Regulations at 25 Pa. Code Chapter 264,
Subchapter I (Use and Management of Containers), Subchapter J (Tank  Systems), and
Subchapter L (Waste Piles) are relevant and appropriate to on-site treatment of soils/sediments
determined to be hazardous to the extent that containers, tanks, or waste piles are used to
stabilize or wash the excavated soil/sediment in Alternatives C-7 and C-8, respectively. For
any waste stream generated during the  treatment process that is hazardous under the PHWM
Regulations at 25 Pa. Code Chapter 261, Subchapter C, off-site disposal would  be required at
a facility in compliance with PHWM Regulations at 25 Pa. Code Chapter 264 or 265.

Impermeable Cap: The PRWM Regulations at 25 Pa. Code Chapter 288, Subchapter C,
would be relevant and appropriate in designing and constructing an impermeable cap in
Alternatives C-5, C-7, C-8, and C-12.   Relevant provisions include,  but are not limited to,
288.212 (access control), 288.234 (final cover and grading), 288.236 (revegetation), 288.237
(standards for successful revegetation), and 288.242 (soil erosion and sedimentation control).

Soil Cover: The PRWM Regulations at 25 Pa. Code §§ 288.234(d), (e), (f), (g) (final
grading), 288.236 (revegetation), 288.237 (standards for successful revegetation), and 288.242
(soil erosion and sedimentation control) would be relevant and appropriate to establishing the
erosion-resistant vegetative cover required under Alternative C-7A.

3.     Location-Specific ARARs

Floodplain:  Federal Executive Order 11988 on Protection of Floodplains and implementing
procedures at 40 C.F.R. Part 6, Appendix A require actions to avoid adverse effects, minimize
potential  harm, and restore and preserve natural and beneficial values of floodplains and are
applicable to cleanup actions at the Site, most of which will occur within the 100-year

                                          47

-------
floodplain.  The Pennsylvania Floodplain Management Act, 25 Pa. Code Chapter 106 is also
applicable to cleanup actions at the Site.

Wetlands:  Federal Executive Order 11990 on Protection of Wetlands and implementing
procedures at 40 C.F.R. Part 6, Appendix A require actions to avoid adverse effects, minimize
potential harm, and restore and preserve natural and beneficial values of wetlands and are
applicable to cleanup actions at the Site in wetland areas.  The Pennsylvania Dam Safety and
Encroachment Act,  25 Pa. Code Chapter 105 is also applicable to cleanup actions at the Site.

Endangered SpeciesfWildl(fe: The Federal Endangered Species Act Regulations at 50 C.F.R.
Parts 222 and 402 mandate the protection of fish and other species threatened with extinction
and are applicable to cleanup actions at the Site that could impact habitat used by the shortnose
sturgeon.

C.     Long-Term  Effectiveness and Permanence

       Alternatives  C-7A, C-8, and C-12 provide the greatest degree of long-term
effectiveness and permanence by  removing much of the contaminants from the Site that pose a
potential threat to human health and the environment.  Alternatives C-7A and C-12 remove the
contamination by transporting contaminated soil (and sediment if PCB concentrations exceed
25 ppm) off-site for disposal. Alternative C-8 removes the contaminants using the soil
washing treatment process. Concentrated contaminants washed from the soil and sediment are
then transported off-site.  Alternatives C-8 and C-12 prevent exposure to low level
contaminants remaining on the Southern Portion of the property through use of an
impermeable cap. This cap will also reduce migration of residual contaminants by reducing
infiltration of surface water.  While reduced contaminant migration typically  increases long-
term effectiveness and permanence, the use  of an impermeable cap in this case is likely  to
impede recovery of the NAPL through the oil collection system, thus reducing long-term
effectiveness and permanence. The soil cover required in Alternative C-7A,  on the other
hand, prevents exposure to low level contaminants remaining in both the Courtyard and the
Southern portion of the property, but will not impede collection of NAPL through the oil
collection system.

       Alternative C-7 provides somewhat less long-term effectiveness and permanence than
the above alternatives because the contaminants, while immobilized through treatment, remain
at the Site.  The permanence of this alternative depends upon the ability of the contaminants to
remain bound in a stabilized mass over time. Features of the Site including the variability of
the fill material on the Southern Portion of the property and the location of the property in the
floodplain increase the uncertainty regarding the long-term reliability of stabilizing the PCB
contamination.  Alternative C-7 also includes use of an impermeable cap which could impede
the collection of NAPL.

       The long-term effectiveness and permanence of Alternative C-5 is somewhat lower than
the above alternatives because the contaminants remain at the Site without any treatment to

                                          48

-------
limit mobility.  The impermeable cap will reduce infiltration of surface water and, therefore,
reduce migration.  However, contaminant migration associated with groundwater movement
could continue.

D.     Reduction of Toxicity, Mobility, or Volume through Treatment

       Alternatives C-7 and C-8 both utilize treatment technologies to address contaminated
soil in the Southern Portion of the Site.  Alternative C-7 uses a stabilization/solidification
process to reduce the mobility of the contaminants in the soil. Alternative C-8 uses soil
washing to remove the contaminants from the soil.  Concentrated contaminants recovered from
the spent solvents used in the process are then disposed off-site, most likely through off-site
incineration.  Therefore, Alternative C-8 would provide the greatest reduction in toxicity,
mobility, and volume through treatment.

       While Alternatives C-5,  C-7A, and C-12 do not treat the excavated  soils and sediments,
they do reduce the mobility of Site contaminants either through use of an impermeable cap or
through off-site disposal in a regulated landfill.

E.     Short-Term Effectiveness

       Alternative C-5 provides the greatest degree of short-term effectiveness because it
requires the least amount of time to implement and does not require as much excavation and
handling of contaminated material.  Alternatives C-7A and C-12 require more excavation as
well as off-site transportation of contaminated material, so the potential for adverse impacts
increases slightly.  However, the short-term impacts associated with excavation and
transportation can be readily addressed through use of the proper equipment and handling
techniques. Alternatives C-7 and C-8 require excavation of contaminated material and on-site
treatment.  The treatment processes require additional handling of the contaminated material
and, therefore, again increase the potential for adverse impacts. As with Alternatives C-7A
and C-12, however, these impacts can be readily mitigated by proper equipment and handling
techniques.

       All the alternatives have the potential to  have adverse short-term impacts on Site
workers conducting the remediation activities.  These short-term risks can be minimized
through use of the proper health and safety protocols.  Similarly, all the alternatives have the
potential for some short-term impacts to ecological receptors during sediment excavation.
Measures such as the use of cofferdams would minimize the off-site transport of contaminated
sediment during excavation.

F.     Implementability

       Implementability issues relating to excavation of sediments, removal of the
underground storage tank, excavation of soil, and installation of the oil collection system are
common to all the alternatives.  Equipment and services are readily available to implement

                                           49

-------
these components of the remedy.  Cofferdam installation becomes more difficult and expensive
with greater distance from the shoreline.  The location and specifications for installing the
temporary cofferdam will be determined during the remedial design.

       For Alternatives C-7A and C-12, transportation and off-site disposal of contaminated
soil can be readily implemented. Similarly, the equipment and materials needed to install the
soil cover in Alternative C-7A and the impermeable cap in Alternatives C-7, C-8, and C-12
are readily available.

       The treatment technologies used in Alternatives C-7 and C-8 will require treatability
studies during the remedial design to ensure that these technologies can be implemented given
the specific conditions at the Site. The stabilization treatment used in Alternative C-7 can be
difficult to implement if soils have a high oil and grease content (usually > 25%) or if the
matrix to be stabilized is  highly variable.  Similarly, variability in size of the materials to be
washed can create materials handling problems with the soil washing technology used in
Alternative C-8.

G.     Cost

       Table 15 provides the capital cost, the operation and maintenance ("O&M") cost, and
the present worth cost associated with Alternatives C-5, C-7, C-7A, C-8 and C-12.
Table 15 - Summary of Costs
Alternative
C-5
C-7
C-7A
C-8
C-12
Capital Costs
$10.405,000
$14,753,000
$14,499,000
$19,540,000
$15,917,000
Annual O&M Costs
$455,000 (years 1-2)
$195,000 (years 3-30)
$455,000 (years 1-2)
$195,000 (years 3-30)
$409.000 (years 1-2)
$149,000 (years 3-30)
$455,000 (years 1-2)
$195,000 (years 3-30)
$455,000 (years 1-2)
$195,000 (years 3-30)
Total Present
Worth Costs
$13,889,000
$18,159,000
$17.273,000
$22,946,000
$19,323,000
H.     State Acceptance

       PADEP has had the opportunity to review and comment on all the documents in the
Administrative Record and has participated in selecting the remedy for this Site.  PADEP's
comments on this ROD have been incorporated to the extent possible. The Commonwealth
                                          50

-------
has concurred with this ROD.

I.     Community Acceptance

       Local citizens and members of the Tacony Civic Association provided comments on the
Proposed Plan during the public meeting on July 27, 1995. The meeting transcript has been
included in the Administrative Record for the Site. Comments have also been summarized and
responses provided in the Responsiveness Summary in this ROD.  Generally, the comments
indicate the community's support for the selected remedy, Alternative C-7A. The community
also expressed concern about:  (1) the impact of the Site on the Baxter Water Intake; and (2)
the ability to use the Metal Bank property for recreational purposes such as a bicycle path or
residential development in the future.

       The PRP Group, which performed the RI/FS for the Site, had several concerns about
Alternative C-7A.  These comments and EPA's responses are also presented in the
Responsiveness Summary.

       The Site Owner provided the most extensive comments and challenged most of the
findings in the RI/FS and the Proposed Plan.  Major arguments  presented are: (1) the Site is
not the source of PCBs and other contamination; (2) if the Site  was the source, contamination
is not migrating from the Site; (3) if contamination was migrating from the  Site, there is no
risk because EPA's risk assessments are faulty; and (4) if there is a risk, the proposed
remedy is inefficient and  costly. The Site Owner's comments are summarized along with
EPA's responses in the Responsiveness  Summary.
IX.    SELECTED REMEDY AND PERFORMANCE STANDARDS

       Based on the evaluation of the nine criteria, EPA is selecting Alternative C-7A as the
remedy for this Site. The components of this remedy and the performance standards that must
be attained during remedy implementation are presented in this section.

A.     Soil/Sediment Excavation Performance Standards

  1.    Soils  in the Courtyard Area which have PCB concentrations exceeding 10 ppm and are
       within two feet of the surface shall be excavated. The volume of soil to be excavated is
       estimated to be 600 cubic yards based on existing information.  The extent of
       excavation shall be further defined during the remedial design and approved by EPA.

 2.    Soils  in the Southern Portion of the Metal Bank property which have PCB
       concentrations exceeding 25 ppm shall be excavated.  The volume of soil to be
       excavated is estimated to be 10,000 cubic yards based on existing information.  The
       extent of excavation shall be further defined during the remedial design and approved
       by EPA.  Excavated soils that are saturated with water shall be stockpiled on-site and

                                         51

-------
      dewatered by gravity.  Accumulated water shall be managed on-site to allow soil and
      sediment particles to settle and then disposed in accordance with Pennsylvania
      Hazardous Waste Management Regulations if hazardous, TSCA Regulations if PCB
      concentration exceeds 50 ppm, or discharged in accordance with State and federal
      NPDES regulations if not hazardous.

3.    Sediments within 100 feet of the Metal Bank property and within four feet of the
      surface of the river bed shall be excavated.  Sediments beyond 100 feet of the Metal
      Bank property which have PCB concentrations  exceeding 1 ppm shall be excavated if
      EPA determines during remedial design that such removal would be both appropriate
      and feasible. Excavation of these sediments shall be performed after completion of soil
      excavation activities in the Southern Portion of the property.  The extent of excavation
      shall be further defined during the remedial design and approved by EPA. Excavated
      sediments shall be stockpiled on-site and dewatered by gravity.  Accumulated water
      shall be managed on-site to allow  soil and sediment particles to settle and disposed in
      accordance with the Pennsylvania Hazardous Waste Management Regulations if
      hazardous, TSCA Regulations if PCB concentration exceeds 50 ppm, or discharged in
      accordance with State and federal  NPDES regulations if not hazardous.

4.    Prior to excavation of soil or sediment, temporary cofferdams shall be installed at the
      outer edge of the sediment excavation boundary to dewater the sediment area so
      excavation can be accomplished.  The  exact locations and specifications for the
      cofferdams, the method for sediment removal, and the process for handling water in
      the excavation and sediment storage areas shall be determined during the remedial
      design and approved by EPA. Care shall be taken during the dewatering process to
      protect fish and other aquatic organisms, especially the shortnose sturgeon, which may
      be present behind the cofferdams. Upon completion of the remedial action, the
      cofferdams shall also be removed  in a manner that minimizes turbidity.

5.    All equipment used during excavation of contaminated soil shall be decontaminated
      before entering uncontaminated  areas.  The design and specifications for the
      decontamination facilities shall be approved by  EPA as pan of the remedial design.
      Any wastewater generated from Site decontamination activities shall be disposed in
      accordance with Pennsylvania Hazardous Waste Management Regulations if hazardous,
      TSCA Regulations if PCB concentration exceeds 50 ppm, or discharged in accordance
      with State and federal NPDES regulations if not hazardous.

6.    Excavated sediment areas and soil areas in the Courtyard shall be backfilled with clean
      fill. Excavated soil areas in the Southern Portion of the property shall be backfilled
      with (a) clean fill; and/or (b) excavated sediments or Courtyard soils which have PCB
      concentrations of less than 25 ppm.  Following excavation and backfilling, the
      Courtyard shall be covered  with 12 inches of clean soil and the Southern Portion of the
      property shall be covered with 24 inches of clean soil.  The riverward side of the
      Southern Portion shall be armored with large riprap or an equivalent system, to prevent

                                         52

-------
       undercutting of the riverbank during flooding.  The soil cover shall be graded to
       promote drainage and vegetated to provide an erosion-resistant surface.  Tape markers
       shall be installed to indicate the boundary between clean soil cover and Site soils.

 7.    Additional sampling and analysis of soil and sediment shall be performed prior to
       excavation to further characterize the extent of contamination. Sampling locations and
       analytical methods shall be determined during the remedial design and approved by
       EPA. Sampling and analysis shall also be performed after excavation has been
       completed to confirm that cleanup levels set forth in the performance standards have
       been achieved. Methods for determining that the cleanup levels have been reached
       shall be finalized during remedial design and approved by EPA based on EPA 230/02-
       89-042, Methods for Evaluating the Attainment of Cleanup Standards. Vol I.

 8.    Excavated soil and sediment shall be temporarily staged on-site in accordance with 25
       Pa. Code Chapter 264, Subchapter L (Waste Piles), if material can be staged in an area
       of existing contamination.  If soil and sediment will be staged in a clean area, the waste
       material and soil  shall be temporarily staged in containers in accordance with 25 Pa.
       Code Chapter 264,  Subchapter I (Use and Management of Containers). Any soil and
       sediment with PCB concentrations exceeding 50 ppm that will be temporarily staged
       on-site before offsite disposal shall be stored in accordance with the TSCA Regulations
       at 40 C.F.R. § 761.65 (Storage for Disposal).

B.     Soil/Sediment Disposal Performance Standards

 1.    Excavated soil and sediment shall be tested to determine if the soil and/or sediment is
       hazardous, as defined in 25 Pa. Code Chapter 261, Subchapter C (hereafter referred to
       as "hazardous"),  and tested as well for asbestos and PCBs. Contaminated soils and
       sediments that are not hazardous and do not exceed PCB concentrations of 25 ppm
       shall be used as backfill on the Southern Portion of the property.  Contaminated soil
       and sediment that are not hazardous and have PCB concentration between 25 and 50
       ppm shall be disposed of off-site at a permitted RCRA  Subtitle D or Pennsylvania
       Residual Waste landfill.

 2.    Excavated soil and sediment that are hazardous, but have PCB concentration between
       25 and 50 ppm, shall be treated and/or disposed of off-site at a permitted RCRA
       Subtitle C or Pennsylvania Hazardous Waste facility.

 3.    Excavated soil and sediment that have PCB concentrations exceeding 50 ppm shall be
       disposed off-site in a TSCA landfill  in compliance with 40 C.F.R.  § 761.60.

 4.    Transportation of soil and sediment from the Site shall  be performed in accordance
       with the general requirements (§ 299.211) of the Pennsylvania Residual Waste
       Management Regulations at 25 Pa. Code Chapter 299,  Subchapter B (Standards for
       Collecting and Transporting of Residual Waste).  If soil and sediment is determined to

                                          53

-------
       be hazardous, this material shall be transported in accordance with the Pennsylvania
       Hazardous Waste Management Regulations at 25 Pa. Code Chapter 263, and federal
       regulations at 49 C.F.R. Parts 107, and 171-179.

  5.   Oversized materials such as boulders in the Riprap Area and debris from the Southern
       Portion of the property shall be decontaminated, using steamed cleaning or other
       equivalent method, in order to reduce PCB concentrations on its surface areas.  Any
       wastewater generated from Site decontamination activities shall be disposed in
       accordance with Pennsylvania Hazardous Waste Management regulations if hazardous,
       TSCA Regulations if PCB concentration exceeds SO ppm, or discharged in accordance
       with State and federal NPDES regulations if not hazardous.

C.     Oil Collection System Performance Standards

  1.    Prior to excavation of soil from the Southern Portion of the property, a sheet pile wall
       shall be installed around the southern and western perimeter of the property adjacent to
       the Delaware River to prevent erosion of fill materials into the river and facilitate
       installation of the oil collection system.

  2.    Prior to excavation of soil from the Southern Portion of the property, an oil collection
       system consisting of interceptor trenches, oil-water separators, and sump pumps, or
       similar collection devices, shall be installed to collect oil  floating on the shallow
       groundwater at the Site to prevent PCBs in oil from getting into the Delaware River.

  3.    Collected oil shall be disposed off-site in accordance with Pennsylvania Residual Waste
       Management regulations if PCB concentrations are less than 50 ppm or in accordance
       with TSCA regulations if PCB concentrations are 50 ppm or greater. Incidental
       groundwater collected in the system shall be discharged to the Delaware River if
       contaminant concentrations are within NPDES discharge  limits to be established in
       consultation with PADEP and DRBC during the remedial design process and approved
       by EPA.  Groundwater exceeding effluent discharge limits shall be treated and/or
       disposed off-site in accordance with state and federal NPDES or Pennsylvania Residual
       Waste Management Regulation requirements, as appropriate.

D.   .  Underground Storage Tank Performance Standard

  1.    Prior to excavation of soil in the Southern Portion of the property, the underground
       storage tank in this area shall be removed. The tank's contents shall be sampled  and
       analyzed to determine the appropriate method of off-site disposal. If the contents are
       determined to be hazardous, the material shall be disposed at a facility in compliance
       with the Pennsylvania Hazardous Waste Management Regulations. If non-hazardous,
       the contents shall be disposed at a facility in compliance with  Pennsylvania Residual
       Waste Management Regulations. After the contents have been removed, the tank shall
       be decontaminated and recycled as scrap metal or disposed off-site.

                                          54

-------
£.     Site Access/Use

  1.    Fence: A six-foot fence shall be installed and maintained along the entire property
       boundary, including the river. The final specifications shall be determined during the
       remedial design process and approved by EPA.  The existing fence can be used if it
       meets the final design requirements.

  2.    Signs:  Signs shall be posted along the property boundary, including the river
       shoreline, warning individuals not to eat fish caught in the vicinity of the Site.  The
       warning signs shall read:

                           WARNING...DO NOT EAT FISH
                    Fish caught in this area contain high levels of PCBs
                                PCBs may cause cancer

  3.    Deed Restrictions:  Deed restrictions shall be placed on the Metal Bank property
       prohibiting (1) the use of the land for residential and agriculture purposes; (2) the use
       of on-site groundwater for domestic purposes, including drinking water; and (3)
       excavation at depths that encounter the water table, contaminated soils and/or sediments
       beneath the soil cover, or compromise the stability of soil-covered areas along the
       Delaware River.

F.     Additional Investigation

  1.    DNAPLs: Discrete subsurface soil samples shall be collected to determine if PCBs are
       migrating downward. Soil borings for the DNAPL investigation shall include, but not
       be limited to, locations where excavation is required in the Southern Portion of the
       property. Samples shall be collected at various depths down to the top of bedrock or
       any confining clay layers. The number and frequency of samples shall be determined
       during the remedial design and approved by EPA.

  2.    Stormwater Sewer: The stormwater sewer adjacent to  the Site shall be investigated to
       determine if  Site contamination is entering the river via this conduit.  The investigation
       shall include documenting historical drainage paths from the Site, present Site contours
       and all Site connections into the City sewer system. The investigation shall also
       include collecting PCB samples from on-site drains and from the stormwater sewer.
       Any contamination found in the Site sewer system before it connects into the City's
       sewer that exceeds 1 ppm PCB shall be disposed in accordance with Section DC.B.
       above.

G.     Monitoring Requirements

  1.    Groundwater: A groundwater monitoring program shall be implemented to evaluate the

                                          55

-------
       effectiveness of the remedy in reducing the concentration of PCBs and other
       contaminants in groundwater.  In addition, a soil monitoring program shall be
       developed during the Remedial Design to monitor the soil cover for evidence of the
       upward migration of contaminants in the groundwater cause by flooding conditions that
       may raise the water table. These programs may include installation of new monitoring
       wells. The exact locations, frequency and duration of sampling, analytical parameters
       and methods to be used shall be determined during remedial design and approved by
       EPA.  At a minimum, the groundwater monitoring program shall include analysis of
       quarterly samples of two existing monitoring wells (MW6, MW7) for metals,
       pesticides, PCBs, SVOCs, and VOCs.  The monitoring programs shall be evaluated
       after three years to determine if changes are appropriate.

 2.    Delaware River:  A monitoring program shall be developed during the remedial design
       to evaluate the conditions of aquatic species hi the Delaware River at the Site before,
       during and after the remedy has been implemented.  This program shall include
       chemical and biological monitoring and shall be evaluated by EPA after five years to
       determine if changes are appropriate.

 3.    Baxter Water Intake: The influent at the Baxter Water Intake (both at the main and
       emergency intakes)  shall be sampled quarterly for PCBs during the remedial design and
       remedial action and for one year following completion of the remedial action at the
       Site.  If drinking water from the influent exceeds the Safe Drinking Water Act
       Maximum Contaminant Limit for PCBs, a confirmatory sample shall be taken
       immediately and officials at the Baxter facility shall be notified.

H.     Operation & Maintenance

 1.    Oil Collection System:  O&M of the oil collection system, including the sheet pile wall,
       interceptor trenches, oil-water separators, and sump pumps, or other collection devices
       installed, shall be routinely performed to ensure proper functioning of the  system.
       Details of the required O&M shall be determined during the remedial design and
       approved by EPA.  At a minimum, the permanent sheet pile wall shall be periodically
       inspected for corrosion. Appropriate measures, including replacement, shall be taken
       to avoid breach of its structural integrity.  O&M of the interceptor trenches shall
       ensure proper inflow, while O&M for the oil-water separators and sump pumps shall
       ensure proper removal of contaminated oil/NAPL.  Maintenance shall include
       replacement of components as necessary to ensure proper functioning.

I.      Five Year Review

 1.    Since the selected remedy results in hazardous substances remaining on-site, a five-year
       review of Site conditions shall be performed by EPA as required by Section 121 (c) of
       CERCLA.
                                          56

-------
X.     STATUTORY DETERMINATIONS

       This remedy satisfies the remedy selection requirements of CERCLA and the NCP.
The remedy is expected to be protective of human health and the environment, to comply with
ARARs, to be cost-effective, and to utilize permanent solutions and alternative treatment
technologies to the maximum extent practicable.  Because contaminated materials will be
transported off-site for landfilling at permitted facilities, the remedy does not meet the
statutory preference for treatment as a principal element of the remedy for soils and sediments.
The following is a discussion of how the selected remedial action addresses these statutory
requirements:

A.     Overall Protection of Human Health and the Environment

       The selected remedy (Alternative C-7A) provides protection of human health and the
environment by removing contaminated soils that pose an unacceptable risk for
industrial/commercial use of the Site.  The remedy provides further protection for individuals
using the Metal Bank property recreationally by providing a cover of clean soil over the
Southern Portion of the property and the Courtyard Area and requiring a vegetative cover to
prevent erosion.  The remedy also protects human health and the environment by removing
contaminated sediments that pose an unacceptable risk to ecological receptors. Removal of the
underground storage tank and installation of the oil collection system prevent contaminant
migration and ensure overall protection of human health and the environment in the future.
Additional investigations and monitoring requirements will ensure that all pathways for PCS
migration are addressed by the selected remedy.

B.     Compliance with ARARs

       The selected remedy shall attain all action-, location-, and chemical-specific applicable
or relevant and appropriate requirements for the Site. Section IX (Selected Remedy and
Performance Standards) identifies specific ARARs that shall be attained.

C.     Cost Effectiveness

       EPA has determined that the selected remedy most effectively addresses all
contaminated media while minimizing costs. The estimated present worth cost of the selected
remedy is $17,273,000.  Other alternatives were either less expensive but less effective, or
more expensive, but unable to offer a greater degree of protection.

D.     Utilization of Permanent Solutions and Alternative Treatment (or Resource
       Recovery) Technologies to the Maximum Extent Practicable

       EPA has determined that the selected remedy represents the maximum extent to which
permanent solutions and treatment technologies can be utilized in a cost-effective manner at the
Site. Excavation and off-site disposal of contaminated materials represents a permanent

                                          57

-------
solution. Although technologies to treat soils and sediments were considered in several
alternatives, these technologies would not achieve greater overall protection and may provide
less long-term effectiveness and permanence because of the soil matrix present in the Southern
Portion of the Metal Bank property and the location of this property within the 100-year
floodplain. Off-site treatment may be required to dispose of oil and contaminated water
collected in the oil collection system.

E.     Preference for Treatment as a Principal Element

       The selected remedy does not employ treatment as a principal element. Containment of
contaminated soils and sediments in an off-site landfill provides adequate overall remedial
protection.  The treatment technologies considered would not have provided any greater
overall protection and may provide less long-term effectiveness and permanence because of the
soil matrix present in the Southern Portion of the Metal Bank property and the location of this
property within the 100-year floodplain.  Some treatment may be required in the final
disposition of the contaminants from the oil collection system.
XII.   SIGNIFICANT CHANGES

       The Proposed Plan identified Alternative C-7A as EPA1 s preferred alternative. EPA
reviewed all written and oral comments submitted. Upon review of these comments EPA
determined that no significant changes to the remedy were necessary.  EPA has included
performance standards which detail how the selected remedy will be implemented.  Several
issues which have been clarified in the ROD in response to comments received during the
public comment period are summarized below.

  ••    The actual extent of soil and sediment excavation in various areas of the Site will be
       determined during the remedial design.

  •    The cleanup levels for soils and sediments are based on PCBs because PCBs presented
       the primary risk. However, EPA has determined that all sediments within 100 feet of
       the Metal Bank property line shall be excavated to a depth of four (4) feet from the
       river bed surface.  Incidental removal of other contaminants, such as PAHs in
       sediments, along with PCBs will occur within this area, and the cleanup is expected to
       achieve acceptable overall protection of human health and the environment (see Table
       12).  EPA has also eliminated TPH as a cleanup standard since there are no cleanup
       criteria for this contaminant and no correlation has been observed between the levels of
       TPHs and PCBs.

  •    EPA has included periodical sampling of the Baxter Water Intake to ensure that
       drinking water supplies for Philadelphia do not exceed acceptable levels for PCBs.

  •    EPA is requiring that 12 inches of clean soil be placed over the Courtyard Area and 24

                                          58

-------
inches of clean soil be placed over the Southern Portion of the property to ensure that
incidental use of the Site for recreational purposes will not result in an unacceptable
human health risk. A 12-inch cover is required in the Courtyard Area because PCBs
concentrations will be reduced to 10 ppm in the top two feet of soil after remediation.
An additional 12 inches of cover is required in the Southern Portion of the property
because PCB concentrations will only be reduced to 25 ppm following remediation.
                                    59

-------
                             RECORD OF DECISION
                         METAL BANK SUPERFUND SITE

                     PART ED - RESPONSIVENESS SUMMARY

      Comments raised during the public comment period on the Proposed Plan and the
Administrative Record for the Metal Bank Site are summarized in this Responsiveness
Summary. The comment period opened with a notice of availability published in the
Philadelphia Inquirer on July 20,  1995 and closed, after two requests for extension were
granted, on October 18, 1995.

      Oral comments were presented at the Proposed Plan Public Meeting held on July 27,
1995. These comments and EPA's responses are presented in Section I of the Responsiveness
Summary. A transcript of the public meeting has been included in the Administrative Record
(see pages AR302311 to AR302375).

      EPA received written comments from four entities during the public comment period:
the Delaware River Basin Commission ("DRBC"),  PADEP,  the Site Owner, and the PRP
Group.  The most extensive comments were received from the Site Owner.  These comments
are summarized along with EPA's responses in Section n of the Responsiveness Summary.
Comments from DRBC, PADEP,  and the PRP Group are summarized and answered in Section
DI of the Responsiveness Summary. Issues raised by the commentors in Section HI  which
have been previously addressed in Section TJ will not be repeated. All the written comments
received have been included in the Administrative Record for the Site.

I.     ORAL COMMENTS FROM THE JULY 27, 1995  PUBLIC MEETING

      On July 27, 1995, EPA hosted a public meeting to answer questions concerning the
Proposed Plan at the Disston Recreation Center, located approximately V£ mile northwest of
the Site, on 1511 Disston Street. Attendees included members of the Tacony Civic
Association ("TCA"), citizens, local businesses, representatives of the PRP group, and
representatives of the Site Owner.   Only the TCA, citizens and local businesses asked
questions and provided  comments  during the meeting.  The meeting lasted from approximately
7:30 to 9:15 p.m.

1.     A TCA member remarked  that there are no copies of the Administrative Record at the
      Northeast ("NE") Philadelphia Library  on Cottman Avenue.

      Response:  EPA had provided documents to the library to be placed in the Site
      Administrative Record. Apparently, the NE Library had not yet placed the documents
      on the shelves at the time of the public meeting.  At the TCA's request, EPA placed an
      additional set of documents at the Tacony Library and the Holmesburg Branch of the
      Philadelphia Library  to increase public availability of the Site Administrative Record.
      The Administrative Record also remains available at the NE Library and in the EPA

                                       60

-------
       Region in Office in Philadelphia. (Some library branches may be currently closed for
       renovation.)

2.     The president of the TCA remarked that EPA's deed restriction that assumes the Site
       may only be used for industries is flawed. Since industries are unlikely to move back
       into the area, agencies such as GreenSpace and the Pennsylvania Environmental
       Commission, will probably convert areas along the waterfront for public recreation.
       The Site could realistically be used for residential and condominium development in the
       future, such as that which has occurred along Delaware Avenue in Center City
       Philadelphia.

       Response: The ROD requires the establishment of deed restrictions prohibiting
       residential or agricultural use of the property, as well as use of the groundwater and any
       intrusive activities into the subsurface soils below the water table in the Southern Portion
       of the property. The Site and adjacent properties are currently zoned for industrial use.
       Any deviations must be approved by the Philadelphia Planning Commission. Present
       laws also require residential structures within a floodplain area, such as the Site, to
       have its lowest floor and basements elevated to or above the flood water level.  A
       majority of the southern portion of the Site encroaches into the 100-year floodplain.

              If residential development were to be approved in the future, EPA's remedy
       documented in the ROD would be protective of human health as long as proper
       measures were taken during construction and subsequent use of the property to prevent
       contaminants  in subsurface soil from being brought to the surface.  EPA's remedy
       requires that 12 inches of clean soil be placed and maintained on the surface of the Site
       to prevent any direct contact with Site contaminants. Any residential development at
       the Site would be connected to a public water supply. To ensure the safety of future
       residents, any use of groundwater at the Site will be prohibited. By taking these
       precautions, health risks associated with Site contamination should remain within
       acceptable levels for any future residents.  During a 100-year flood, the possibility
       does exist that residual oil containing PCBs may fluctuate with the water level and
       contaminate clean surface soils.  Although remote, this possibility should be considered
       in any future plans to rezone the Site for residential use.

3.     The TCA indicated there are conceptual plans to put a bike path with a 50-foot right-of-
       way along the shoreline of the  Site.  This stretch of potential park property would
       eventually connect Center City to Pennypack Park.

       Response:  EPA's remedy includes placing 12 inches of clean soil in the Courtyard
       and 24-inches in the Southern Portion of the Site so that recreational use of the Site's
       surface area would not pose a health risk to people.  As an additional precaution, the
       ROD  requires that tape markers be installed to indicate the boundary between the clean
       soil and the original Site soils to prevent the possibility of recontamination of the clean
       surface soil during any future construction activities.

                                           61

-------
4.     A representative of a local business asked when the cleanup would actually occur.

       Response:  After the ROD is issued, two additional steps are required before actual
       cleanup at the Site will begin.  First, EPA will try to reach an agreement with the
       potentially responsible parties at the Site for their performance of the cleanup.  EPA
       strives to negotiate and finalize an agreement within 4 months.  The next step is
       completion  of a detailed engineering design that the contractor performing the cleanup
       can follow to ensure the remedy is implemented properly.   Preparation of the design
       usually requires 12-24 months.  Therefore, actual construction at the Site could begin
       in as early as 16 months and should begin within 2V£ years of the issuance of the ROD
       unless unexpected delays are encountered during the negotiations or the design process.
       EPA has experienced delays in the cleanup process at the Site, the most recent being a
       delay in issuing the ROD itself. This delay resulted from the large volume of
       comments EPA received on the Proposed Plan and supporting documents in the
       Administrative Record.  EPA typically issues a ROD within six months of issuance of a
       Proposed Plan.  However, an additional 23 months was required at the Metal Bank Site
       because EPA wanted to thoroughly review and respond to all public comments.

5.     A TCA member stated that he sat in a meeting with the previous EPA project manager
       about three  years ago, when legal proceedings were being initiated to force the
       responsible  parties to put up funding to find out what needed to be done  to clean up the
       Site.  Nothing more  was heard until this public meeting.  From EPA's description of
       the cleanup timeline, it seems another 2 to 5 years will pass while corporate and
       government lawyers argue and more expenses will be incurred. Meanwhile the Site is
       still left unattended.

       Response:  Although the legal  process associated with cleaning up Superfund sites can
       be time consuming, EPA strives to have those panics who are most closely associated
       with the contamination problems at a site perform the cleanup, rather than rely on the
       federal trust fund created through  tax dollars.  EPA does ensure that any immediate
       threats posed by Superfund sites are addressed quickly. Fortunately, contamination at
       the Metal Bank Site does not pose any immediate threat to the community. The
       measures required in the ROD are necessary to ensure  that the potential risks associated
       with long-term exposure to Site contaminants are addressed.

             Although it has taken a longer time than usual from publication of the Proposed
       Plan to issuance of the ROD at this Site, as discussed previously in Comment #4, from
       this point, EPA should have more control over the timeframe. Upon issuance of this
       ROD, EPA will send Special Notice Letters ("SNLs") to all PRPs inviting them to
       negotiate a Consent Decree in which they agree to clean up the Site and pay EPA's past
       costs associated with the Site. If EPA receives a good  faith offer from any of the
       PRPs, the PRPs have 120 days from the issuance of the SNLs to negotiate a Consent
       Decree.  If significant progress  is  not made, EPA may  then issue an Order requiring
       the PRPs to clean up the Site.

                                          62

-------
6.     The TCA remarked that EPA has been aware of oil slicks into the river since 1977 and
       knew about the leaking underground storage tank ("tank") three years ago.  The TCA
       noted that during the Remedial Investigation, EPA just stopped when a concrete slab
       obstructed further excavation of the tank, so the tank is still in the ground and
       continues to release contaminants. The TCA asserts that EPA should remove the tank
       as its first priority and then build its groundwater collection system.

       Response:  The purpose of the Remedial Investigation was to characterize the nature
       and extent of contamination at the Site.  EPA does not require that cleanup actions be
       performed during the investigation process unless contamination is encountered which
       poses an immediate threat to human health and the environment.  According to
       available records, the tank at the Metal Bank Site was emptied and backfilled in 1986.
       When a test pit was excavated in the area of the tank during the investigation, a
       concrete slab was encountered. Since the tank did not pose an immediate threat, EPA
       did not require that the concrete slab or the tank be removed as part of the
       investigation.  EPA is requiring excavation and off-site disposal of the tank as part of
       the remedy selected in the ROD.  Although EPA agrees that removal of the tank is
       important, the ROD requires  that the leachate collection system be installed first as a
       precautionary measure. If the tank is not empty and ruptures during excavation, the
       leachate collection system will ensure that the released material is captured before it
       leaves the Site.

7.     A TCA member remarked that the intake for the Baxter Water Treatment Plant, which
       supplies the city's drinking water, is less than a quarter of a mile from the Site and may
       be at risk.  The TCA member expressed concern that PCBs may pose a threat to
       residents who drink the water.

       Response: In general, PCBs do not readily dissolve in water and tend to bind to soil
       and sediment particles.  The Remedial Investigation did not directly evaluate whether
       PCBs from the Site could pose a threat to the Baxter public water supply  because the
       Baxter water intake is located 2.2 miles upstream from the Site. An emergency water
       intake is located approximately 1.9 miles upstream from the Site. The Remedial
       Investigation does provide information which supports the conclusion that PCBs from
       the Site do not  pose a health risk to public water supply users.  As part of the Aquatic
       Ecological Risk Assessment, the discharge of PCBs in groundwater from the Site into
       the Delaware River was conservatively estimated to result in surface water PCB
       concentrations  of between 1.3 and 2.0 nanograms per liter ("ng/1").  These
       concentrations  are well below the Safe Drinking Water Act Maximum Contaminant
       Level of 500 ng/1 (or 0.5 Mg/0-  Therefore, if river water at the Site does not pose a
       human health threat, any site-related PCB contamination two miles upstream at the
       Baxter water supply intake would not be expected to pose a threat to public water
       supply users.

             Because the Philadelphia Water Department is not required to analyze for PCBs

                                          63

-------
       as part of its routine water testing program, the ROD requires sampling the Baxter
       water intakes for PCBs as a precautionary measure to ensure that levels are within the
       Safe Drinking Water Act Maximum Contaminant Limit.

8.     A TCA member inquired if EPA tested for asbestos since a scrap metal yard
       immediately north of the Site (Northern Metal) extensively removed asbestos from
       dismantled ships.  The member also indicated most of the PCBs brought to the Site
       originated from Northern Metal.

       Response:  The Remedial Investigation did not analyze for asbestos as a potential
       contaminant of concern at the Site. Asbestos could be present at the Site  from the
       removal activities at the adjacent scrap metal yard. EPA also has documentation (see
       AR304076) that the Metal Bank State Road facility had directed asbestos  wastes to the
       Site.  While no additional  information concerning  the fate of this waste is available, the
       quantity of asbestos wastes does not appear to be large.  Additional sampling will be
       required as  the cleanup is implemented to determine the disposal location for excavated
       soil.  Asbestos will be included as part of the analysis to ensure that proper disposal
       occurs.

             EPA does not have documentation to support the commentor's allegation that
       Northern Metal was a source of the PCBs found at the Site.  If information to support
       this allegation becomes available, EPA will consider Northern Metal as potentially
       liable for cleanup costs at the Site.

9.     A local resident of Tacony and a member the TCA remarked that children are fishing
       near the Site because there are no signs warning that fish in the Delaware River are
       contaminated.

       Response:  The ROD requires that signs be posted and maintained along the Metal
       Bank property warning individuals not to eat fish caught in this portion of the Delaware
       River because of PCB contamination. PADEP currently has a public health advisory in
       effect warning people not to eat channel catfish, American eels, white perch, and blue
       crabs in the portion of the Delaware River adjacent to Philadelphia and flowing
       downstream to the Pennsylvania/Delaware state line.  These species were found to
       contain levels of PCBs and chlordane in excess of federal guidelines.

10.    A representative of Morris Iron & Steel inquired about hazards posed to his employees
       during EPA's cleanup.

       Response:  The ROD requires that measures to control dust be implemented during
       the excavation activities.  Air monitoring is also required to ensure that no
       unacceptable releases of Site contaminants occur.  The control measures and air
       monitoring will ensure that individuals in the vicinity of the Site are not exposed to any
       hazards during the Site cleanup.

                                          64

-------
11.    The TCA inquired if EPA drilled through the contaminated river bed layer and tested
       the groundwater aquifer underneath.  The TCA  requested confirmation that the PCBs
       are only mobile through the oil.

       Response:  Contamination at the Site originated primarily from spills and/or storage of
       PCB-bearing oil recovered from electrical transformers.  A PCB-contaminated oil
       layer floating on the water table at the Site was the focus of a prior cleanup action by
       Metal Bank of America, Inc.  Therefore, the groundwater investigation conducted
       during the Remedial Investigation was concerned with characterizing contamination
       present in the upper portion of the shallow unconsolidated aquifer at the Site.
       Monitoring  wells were generally drilled to a depth of 15 - 20 feet with a 10-foot
       screened interval across the water table.  No wells were drilled into the underlying
       bedrock aquifer.

             While PCBs bound to transformer oil (i.e., LNAPL) float on the water table,
       PCBs themselves are denser than water (i.e., DNAPL) and, if not associated with oil,
       would continue to move down through the aquifer until a confining layer (e.g., clay or
       bedrock) is reached. Therefore, a possibility exists that PCBs could be present in
       lower portions of the shallow unconsolidated aquifer or in the bedrock aquifer beneath
       the Site. The ROD requires further groundwater investigation during the remedial
       design process to determine the full extent of PCB contamination.  If PCBs are found
       to be present as DNAPLs at the Site,  additional actions could be required to address
       this concern. Depending on the nature and scope of these additional actions, EPA
       would either issue an Explanation of Significant  Differences to this ROD or  issue a
       Proposed Plan for public comment to amend the  ROD.

12.    The TCA requested clarification on the sequence of how the remedy will be  installed.

       Response:  The elements of the remedy selected in the ROD will be performed in the
       following sequence: (1) installation of the temporary  sheet pile cofferdam; (2)
       installation of the leachate collection system; (3)  removal of the underground storage
       tank and contaminated soil; and (4) removal of sediments. Section DC (Selected
       Remedy and Performance Standards) provides details on the required elements of the
       cleanup.

II.     WRITTEN COMMENTS FROM THE SITE OWNER

       The Site Owner submitted extensive comments on the documents in the Administrative
Record supporting the Proposed Plan, as well as comments on the Proposed Plan itself.  These
comments were presented  in 22 volumes and have been included in the Administrative
Record. EPA has summarized in this section the substantive issues raised by the Site Owner
concerning information which  served as the basis for EPA's decision.  Comments by the Site
Owner that were editorial in nature (i.e., suggestions for including additional information in
various reports or drawing  different interpretations of the information presented) are addressed

                                         65

-------
only to the extent that they raise a question concerning EPA's basis for making its decision.
Comments from the Site Owner have been grouped into six categories: Human Health Risk
Assessment, Aquatic Ecological Risk Assessment, Terrestrial Ecological Risk Assessment,
Non-Aqueous Phase Liquids ("NAPL"), Remedial Alternatives, and Legal/ Factual Issues.

A.     Human Health Risk Assessment

       Unless otherwise specified, all PCB cancer risk calculations are based on the original
PCB slope factor of 7.7 kg-d/mg rather than the revised PCB slope factor of 2.0kg-d/mg.
However, for all scenarios that posed an unacceptable risk previously, the corrected risk levels
based on the revised slope factor, have been included in brackets ("{}").  Section VI. A of the
ROD and Comment #23 in this Section provide the final cancer risk values.

1.      The Site Owner contends that EPA's revised Baseline Human Health Risk Assessment
       dated April 28,1995 ("1995 Risk Assessment"), is invalid because EPA does not explain
       what data was used to generate the exposure point concentration or how contaminants of
       concern were selected.

       Response:  Upon review of the data used to generate the exposure point concentrations
       for the Baseline Human Health Risk Assessment dated January 27,1994 ("1994 Risk
       Assessment"), EPA identified several instances where data collected at the Site had not
       been included (e.g., PCB surface soil data from Courtyard, dioxin data from subsurface
       soil and sediments). EPA asked the PRP Group to review the dataset to ensure all data
       were included and to recalculate the exposure point concentrations.  EPA's toxicologist
       for the Site reviewed the dataset resubmitted by the PRP Group along  with the new
       exposure point concentrations and then recalculated the risk at the Site in the 1995 Risk
       Assessment.

             In response to concerns raised by the Site Owner, EPA again verified that the
       dataset used in the 1995 Risk Assessment was complete by searching EPA records and
       recalculating the exposure point concentrations for several key exposure pathways.
       These include exposure to PCBs in the Courtyard surface soil, to oil or LNAPL in
       monitoring well MW-6, and to catfish fillets. This review demonstrated that the exposure
       point concentrations used in the 1995 Risk Assessment were correct. The results of this
       review are presented in the Weston/SATA Technical Review of Risk Assessment
       ("Technical Review") dated June 4,1997, which has been included in  the Administrative
       Record (see AR304100). The raw data used to calculate the exposure  point
       concentrations for these pathways are presented hi Appendix D of the  Verification Report
       (see AR304121).  The detection limits for data used in the 1995 Risk Assessment are
       verified in Attachment B of the report (see AR304112).

       In general, the following procedures were followed during data handling. When samples
       had field or laboratory duplicates, the original result was used. For media such as soil
       and sediment where the duplicates are not homogeneous samples, a duplicate result may

                                         66

-------
       be used if contamination is present at greater levels in the duplicate than in the original
       sample. If a contaminant of concern was detected above the analytical quantitation limit,
       the result was used in the dataset, even if some question about the analytical quality of the
       result existed. When the result was below the analytical quantitation limit, the result was
       considered to be half the value of the quantitation limit and used in the dataset.  The
       exposure point concentrations were based on the log of the upper 95% confidence limit
       ("UCL") of the dataset rather than the normal UCL because the data matched the log
       normal distribution when screened using the Shapiro-Wilks Test for Normality1. This
       screening was verified for the selected pathways in Attachment C of the Technical
       Review (see AR304116). If a log UCL could not be calculated because of the nature of
       the dataset, the normal UCL was used. The above procedures are consistent with EPA's
       standard protocols for handling data collected for risk assessment purposes.  EPA uses a
       conservative approach to ensure that the risk assessment does not underestimate the risks
       associated with contamination at a site.

              EPA determined the contaminants of concern at the Site by following the EPA
       Region III guidance, Selecting Exposure Routes and Contaminants of Concern by Risk-
       Based Screening (EPA/903/R-93-001, January 1993) (see Appendix 4 of the 1995 Risk
       Assessment; AR301688). The 1995 Risk Assessment used the version of the Risk-Based
       Concentration ("RBC") table referenced in the above guidance which was current at the
       time of the assessment. This RBC table was dated February 9,1995 and is included as
       Attachment E of the Technical Review (see AR304174). The slope factors presented in
       Table 1 of the 1995 Risk Assessment (see AR301575) matched this RBC table.

2.     The Site Owner contends that EPA used inappropriate data to generate the PCB exposure
       point concentration used to calculate risk to industrial workers from ingestion of the
       Courtyard soils (see AR301578).

       Response: The exposure point concentration for PCBs was calculated by adding either
       the log  UCL of the dataset, or the maximum observed concentration, whichever was less,
       for the PCB aroclors that were detected in the soil samples collected at the Site. The
       values added include 140.0 ppm, which is the maximum concentration observed for
       Aroclor 1248, and 5.4 and 5.7 ppm, which are the log UCLs for Aroclors 1254 and 1260,
       respectively (see AR301619). The sum of these values is 151.1 ppm, which is the
       reasonable maximum exposure ("RME") concentration used in the risk calculation. This
       value was verified in Table A-3 of the Technical Review (see AR304108). The RME is
       defined as the highest exposure that could reasonably be expected to occur for a given
       exposure pathway at a site.

             The data used to calculate the log UCLs for Aroclors 1254 and 1260 were the
             See Supplemental Guidance to Risk Assessment Guidance for Superfund: Calculating the
             Concentration Term; US EPA - OSWER; Publication 9285.7-081; May 1992.

                                          67

-------
       laboratory results for the surface soil samples collected in the Courtyard (fee AR301660).
       The detection limits for some samples were elevated because of the dilution required to
       measure the PCBs present in the sample (e.g., Sample CY-136-0-C required a 200-fold
       dilution to measure the concentration of Aroclor 1248; this raised the detection limit of
       the other Aroclors in that sample by 200). However, the RME for the Courtyard surface
       soils results primarily from the observed PCB concentration of 140 ppm, not from
       manipulation of data below analytical quantitation limits.

3.     The Site Owner contends that the exposure assumptions used to estimate risk in the
       Courtyard soil were not realistic since contamination is only found in a small area of the
       Courtyard soil. The Site Owner asserts that EPA assumed that a worker would be
       employed full time in the Courtyard and that the entire Courtyard soil area is
       contaminated.

       Response:  In calculating the risk to industrial workers in the Courtyard soil area, EPA
       assumed that a worker could be exposed to the Courtyard surface soil for 250 days per
       year for a period of 25 years (see AR301578). EPA also assumed that  the average
       amount of soil that could be ingested daily is approximately 16.7 milligrams. These
       assumptions would result in a worker ingesting approximately a teaspoon of
       contaminated soil every 21A years. EPA typically uses an ingestion rate of 50 milligrams
       for industrial workers. A lower rate was used in the risk assessment for this Site because
       EPA recognized that a worker in this area could reasonably be expected to have less
       contact with contaminated soil.  EPA did not  assume that a worker would spend his entire
       workday in the Courtyard soil, nor did EPA assume that the entire Courtyard soil area
       was contaminated.

4.     The Site Owner contends that there is no basis for taking an action for  the Courtyard soil
       or soil from any other area of the Site. The Site Owner contends that the 10 ppm surface
       soil cleanup level results from improperly applied Agency policy and is not based on
       health risk.  The Site Owner further states that the Food and Drug Administration
       ("FDA") allows 10 ppm of PCBs in boxes used to package cereal.

       Response:  EPA's decision on the need to take action at a site is based on the overall
       potential for individuals who are exposed to contaminants at the site to experience
       unacceptable health effects. To determine the overall impact, EPA evaluates the various
       ways an individual may be exposed to Site contaminants and adds the risk associated
       with each pathway to calculate the total risk to the individual.  If the total risk to
       individuals who could reasonably be exposed to contaminants from the site is less than
       1 .OE-04 for carcinogenic contaminants, EPA would generally not require action
       (assuming the human health risks from noncarcinogenic contaminants  are acceptable and
       no adverse environmental impacts are present).  At the Metal Bank Site, the Risk
       Assessment identifies groups of individuals (i.e., future construction workers and
       recreational boaters) who could  be exposed to contaminants from the Site and potentially
       experience a total risk above 1 .OE-04.  (However, based on the revised PCB slope factor,

                                          68

-------
       the risk to recreational boaters no longer exceeds l.OE-04.) In addition, the Aquatic
       Ecological Risk Assessment identifies potential adverse impacts to aquatic organisms
       from the Site. Therefore, EPA's decision to take action at this Site is warranted.

             When action is warranted, Section 300.430(e)(2)(i)(2) of the NCP states that a
       l.OE-06 cancer risk level shall be the point of departure for determining remediation
       goals when other applicable or relevant and appropriate requirements are not available.
       Although Pennsylvania's Land Recycling and Environmental Remediation Standards Act
       (Act 2) provides standards for PCBs in soil, EPA has established the soil standards based
       on more stringent risk-based EPA guidance. EPA has estimated the carcinogenic risk to
       industrial workers from the surface soil in the Courtyard soil to be 6.77E-05 which does
       not achieve the 1 .OE-06 remediation goal. The Superfund program has adopted an
       approach to clean up PCBs that relies heavily on the PCB Spill Cleanup Policy published
       in 40 C.F.R. 761.120 -761.139 on April 2,1987. In August 1990, EPA issued OSWER
       Directive No. 9355.4-01 FS, A Guide on Remedial Actions at Superfund Sites with PCB
       Contamination, which recommends an action level of 10 - 25 ppm for triggering cleanup
       of PCB spills in industrial areas. Cleanup in areas where access is not restricted must
       achieve 10 ppm while cleanup in restricted access areas must achieve 25 ppm. Because
       the PCB contamination in the Courtyard soil area is found in surface soils  which
       individuals could encounter more readily than the contaminated subsurface soil in the
       southern portion of the Site, the ROD appropriately requires a PCB cleanup level of 10
       ppm for the Courtyard soils.

             FDA has established a temporary tolerance for residues of PCBs of 10 ppm in
       paper food-packaging material intended for or used with human food (21 C.F.R.
       109.30(a)(9)). FDA has established this tolerance because of concern that PCBs in paper
       food-packaging material may migrate to the packaged food and then be consumed.
       EPA's cleanup level, on the other hand, protects against direct consumption of PCBs
       through inadvertent ingestion of soil. A direct comparison of EPA's cleanup level to
       FDA's tolerance level can only be made if FDA's tolerance level is concerned with direct
       consumption of PCBs through inadvertent ingestion of the paper food-packaging
       material.  Since individuals are probably less likely to inadvertently eat paper food-
       packaging material than  they are to place hands or other objects which have contaminated
       soil  on them in their mouths, FDA's tolerance levels could be considered more
       conservative. Because these two action levels have been developed for very different
       purposes,  EPA considers any comparison of the two to be speculative and  not a
       reasonable basis for establishing a cleanup level at a Superfund site.

5.      The Site Owner contends that arsenic, beryllium, cadmium, and chromium should have
       been excluded from the risk assessment for exposure to soil because the Remedial
       Investigation indicates that the concentrations detected were within background ranges.
       The Site Owner further asserts that dioxin contamination found at the Site  should have
       been excluded because the two rip-rap sediment samples  show only one 2,3,7,8-
       containing ruran congener and the octachloro furan and probably represent background

                                          69

-------
       contamination.

       Response:   In general, a remedial investigation will attempt to determine the
       concentrations of inorganic contaminants naturally occurring near a site and, if possible,
       anthropogenic (i.e., manmade) concentrations from nearby point and nonpoint sources.
       Because both natural and anthropogenic concentrations are difficult to determine in an
       urban, industrialized setting, the Remedial Investigation for the Metal Bank Site did not
       attempt to quantify this information.

             In the absence of site-specific natural background concentrations, EPA
       recommends the use of literature values to approximate natural conditions. The Remedial
       Investigation correctly states that the levels of most inorganic contaminants found at the
       Site were within the background ranges found in research literature. This is not,
       however, a sufficient basis for eliminating these contaminants from the risk assessment.
       The range of values found in the literature varies widely. To avoid underestimating the
       potential site-related risk, EPA required that the inorganic contaminants be included in
       the risk assessment.  EPA acknowledges the uncertainty associated with possible
       background concentrations in the Risk Assessment (see AR301571). This approach is
       consistent with Section 5.7 of EPA's Risk Assessment Guidance for Superfund, Volume I,
       Human Health Evaluation Manual (Part A) (EPA/540/1-89/002, December 1989). This
       guidance does not mandate that contaminants present within background ranges be
       eliminated from the risk assessment.

             Similarly, EPA does not recommend excluding contaminants such as dioxin
       which are present at a site from the risk assessment unless site-specific information is
       available to conclusively show that the contaminants are from specific anthropogenic
       sources. Providing conclusive information of this nature would have been difficult, if not
       impossible, at the Metal Bank Site and was not  required. Therefore, EPA included the
       dioxin data in the risk assessment process to avoid underestimating the potential site-
       related risk.  Again, this approach is consistent with EPA risk assessment guidance.

             If the chemicals mentioned by the Site Owner had been eliminated from the risk
       assessment, the overall health risks posed by the Site would not have changed
       significantly. Therefore, the inclusion of these substances in the risk assessment did not
       impact EPA's final remedy selection decision.

6.      The Site Owner contends that the estimated risk to future industrial workers from
       ingestion of surface soil outside the Courtyard is invalid because the exposure point
       concentration for PCBs cannot be verified.

       Response:  EPA used the same method to calculate the PCB exposure point
       concentration for surface soils outside the Courtyard (i.e., those in the southern portion of
       the Site) as it did for surface soils in the Courtyard (see responses to comments #2 and #3
       above). The values used are 0.290 ppm, which  is the maximum concentration of Aroclor

                                          70

-------
       1248 detected, and 0.995 and 0.982 ppm, which are the log UCLs for Aroclors 1254 and
       1260, respectively (see AR301622). The sum of these values is 2.27 ppm, which is the
       RME concentration used in the risk calculation.

             PCBs found in the surface soils outside the Courtyard contribute approximately
       one percent of the total potential health risk to future industrial workers posed by the Site
       (i.e., a cancer risk of 1.02E-06 out of total risk of 7.46E-05 {2.45E-05}; see AR301579,
       AR301598) and, therefore, did not impact EPA's final remedy  selection decision.

7.     The Site Owner contends that risk to construction workers from exposure to subsurface
       soils outside the Courtyard has been overestimated because of discrepancies in the
       calculation of the exposure point concentrations for PCBs, benzo(a)pyrene, and
       dibenz(a,h)anthracene.

       Response: As previously indicated, exposure point concentrations were calculated by
       using the log UCL of the dataset for the chemical or the maximum observed
       concentration, whichever is less. In the case of PCBs, the appropriate values for each
       aroclor detected were added to determine the RME concentration used in the risk
       calculation. Specifically, the following log UCL values were used to determine the RME
       concentration for  PCBs:  1.24 ppm for Aroclor 1221; 1.35 ppm for Aroclor 1242; 1.43
       ppm for Aroclor 1248; 2.38 ppm for Aroclor 1254; and 4.0 ppm for Aroclor 1260 (see
       AR301624). The maximum concentration detected for Aroclor 1232 (0.89 ppm) was
       used because it was less than the log UCL. Aroclor 1016 was not included because it was
       not detected in any subsurface soil sample. The total of the aroclor values is 11.3 ppm,
       which is the RME concentration used in the risk assessment. The RME concentrations
       used for benzo(a)pyrene and dibenz(a,h)anthracene are 54.0 ppm  and 8.2 ppm,
       respectively, which represent the log UCL of each data set.

             The combined risk from PCBs, benzo(a)pyrene, and dibenz(a,h)anthracene in
       subsurface soils outside the Courtyard represents less than one tenth of one percent of the
       total potential health risk posed by the Site to future construction  workers (i.e., a cancer
       risk of 6.08E-07 for PCBs, 2.75 E-06 for benzo(a)pyrene, and 4.18E-07 for dibenz(a,h)-
       anthracene out of a total risk of 6.01E-03  {1.58E-03}; see AR301582, AR301598).
       Therefore, these contaminants did not impact EPA's final remedy selection decision.

8.     The Site Owner contends that EPA did not explain why it changed the exposure point
       concentration for dioxin in subsurface soils from 4.35 ppm in the  1994 Risk Assessment
       to 4.35 ppb in the 1995 Risk Assessment. The Site Owner noted similar changes in the
       exposure point concentration for dioxin in dust and sediments.

       Response:  When EPA verified selected exposure point concentrations prior to
       conducting the 1994 Risk Assessment, the Agency determined that the appropriate value
       for dioxin was 4.35 ppb (see AR301686). However, when EPA used this value in the
       1994 Risk Assessment, incorrect units were used (i.e., ppm were used rather than ppb).

                                          71

-------
       The PR? Group detected this error during its review of EPA's 1994 Risk Assessment (see
       AR301266), including the exposure point verification in Appendix 3. When EPA
       prepared the revised risk assessment in 1995, EPA used the ppb units for the dioxin
       exposure point concentration in all appropriate calculations. (Note: Additional problems
       identified with the dioxin dataset for the rip-rap sediments and the resulting exposure
       point concentration are discussed in the response to Comment #18 below.)

9.      The Site Owner contends that EPA overestimated the risk to future construction workers
       breathing dust from the Site because EPA evaluated chromium as the more toxic
       chromium VI form even though chromium VI quickly reverts to the less toxic chromium
       III form in the environment.

       Response:  In the absence of analysis for hexavalent (VI) chromium, EPA assumed both
       hexavalent and trivalent (III) chromium are present at the concentration of the total
       chromium detected. This assumption is highly conservative and, therefore, the most
       protective of human health and the environment. The potential presence of hexavalent
       chromium in the dust at the Site represents less than one tenth of one percent of the total
       potential health risk posed by the Site to future construction workers (i.e., a cancer risk of
       1.53E-06 out of a total risk of 6.01E-03 {L58E-03}; see AR301583, AR301598).
       Therefore, this contaminant did not impact EPA's final remedy selection decision.

10.    The Site Owner contends that EPA's exposure assumptions relating to a future
       construction worker's dermal contact with groundwater is unrealistic. EPA assumes that
       10 percent of a construction worker's body could be exposed to groundwater at the Site
       for an average of one hour a day for 25 days over a one-year period.

       Response:  EPA's exposure assumptions are intended to be conservative and reflect the
       reasonable maximum exposure that could be encountered at the Site to ensure that the
       potential health risks from contaminants are not underestimated. EPA assumes that a
       construction worker could be exposed to groundwater at the Site for a total of 25 hours.
       Although the risk assessment assumes this total exposure results from contact with the
       groundwater for one hour a day on 25 different days within one year, the total exposure
       could also result from longer daily exposures for fewer days (e.g., exposure for five hours
       on five days).  Exposure to groundwater would most likely occur if future activities such
       as excavation or drilling are performed to construct a foundation or some other subsurface
       structure at the Site. The water table was encountered at depths of 8 - 15 feet below the
       ground surface in the 15 monitoring wells installed at the Site with moist soil encountered
       at somewhat shallower depths.

             Using EPA's assumptions, risk from dermal contact with groundwater only
       contributed one tenth of one percent of the total risk to future construction workers at the
       Site (i.e., a cancer risk of 6.37E-06 out of a total risk of 6.01E-03 {1.58E-03}; see
       AR301584,  AR301598).  Therefore, risk from this exposure pathway did not impact
       EPA's final  remedy selection decision.

                                          72

-------
11.    The Site Owner contends that EPA overestimated the risk to future construction workers
       from dermal contact with Site groundwater by including aroclors that were not detected
       in the exposure point concentrations for PCBs. The Site Owner further contends that
       EPA failed to explain discrepancies in the groundwater data summary in Table 4-12 of
       Appendix 2 to the 1994 Risk Assessment (see AR300704) or verify that appropriate data
       were used to calculate the PCB exposure point concentration.
       Response:  The only groundwater exposure pathway considered in the risk assessment
       for the Site is dermal contact by future construction workers. As stated in the previous
       response, this pathway only contributes one tenth of one percent of the total potential risk
       posed by the Site to future construction workers.  Therefore, when EPA revised the risk
       assessment in 1995, EPA did not revise the risk calculations associated with the
       groundwater exposure pathway. The Site Owner is correct that when the risk was
       calculated in the 1994 Risk Assessment, the exposure point concentration was based upon
       data for all aroclors, including those that were not detected in any groundwater samples at
       the Site. This resulted in a highly conservative exposure point concentration. However,
       even with this conservative value, the potential risk associated with exposure to
       groundwater was minimal.

             The Site Owner is also correct that some discrepancies exist in the data
       summarized in Table 4-12 in Appendix 2 to the 1994 Risk Assessment.  The table
       presents two sets of data for PCB aroclors. The log UCLs for the second set were.used to
       calculate the exposure point concentration. These values were based upon 24
       groundwater samples collected during 1991 and 1992 at the Site. However, the data
       presented in the RI indicates that a total of 30 groundwater samples were analyzed for
       PCBs (see AR300493 and AR300569). In response to the Site Owner's concern, EPA
       recalculated the exposure point concentrations using the complete dataset and only those
       aroclors that were detected in at least one groundwater sample (i.e., Aroclors 1242 and
       1260). The revised concentration is 2.08E-03 mg/1, rather than 9.42E-03  mg/1 used in
       both the 1994 and 1995 Risk Assessments, and the revised lifetime cancer risk associated
       with the revised PCB concentration is 2.87E-07, rather than 1.30E-06.  This reduces the
       potential risk associated with the groundwater exposure pathway from 6.37E-06 to 5.07E-
       06 which results in a 2/100ths of one percent reduction in the total potential risk to future
       construction workers from Site contaminants.

12.     The Site Owner contends that PCBs should not be included as a groundwater contaminant
       of concern because only monitoring well MW-6 routinely shows  PCBs.

       Response:  The risk assessment for a site includes any site-related contaminant that has
       been detected in groundwater monitoring wells designed to characterize the site.  To
       exclude a contaminant actually detected at a site could result in the underestimation of the
       risk associated with the site. PCBs were detected in groundwater at the Metal Bank Site
       in monitoring well MW-6 in samples collected both in 1991 and 1992 and in MW-7 in

                                          73

-------
13.
14.
the 1992 sample.  The dataset used to calculate the groundwater exposure point
concentration for the risk assessment included these detected values along with values
equal to one-half the analytical quantitation limit when PCBs were not detected above
this level.  This standard approach provides a conservative estimate of potential risk from
Site groundwater. As indicated in the previous two responses, the potential human
health risk calculated using this approach indicates that dermal contact with groundwater
does not pose a significant risk to future construction workers at the Site.

The Site Owner contends that the concentrations of benzo(a)pyrene and
dibenz(a,h)anthracene used to calculate risk to future construction workers from dermal
contact with Site groundwater cannot be verified because the data summarized in Table 4-
12 of the 1994 Risk Assessment (see AR300706) is not consistent with the data
summarized in Table 4-26 of the Remedial Investigation Report (see AR300228).

Response: The Site Owner is correct in noting that discrepancies exist between the data
summary tables in the 1994 Risk Assessment and the Remedial Investigation Report. As
indicated in the response to Comment #11 above, EPA did not revise the risk calculations
for the groundwater exposure pathway in  the 1995 Risk Assessment because the overall
risk from this pathway was not significant.  However, since the Site Owner questioned
several of the risk calculations, EPA has reviewed the data and recalculated these values.
The exposure point concentrations and the lifetime cancer risks for future construction
workers exposed to benzo(a)pyrene and dibenz(a,h)anthracene in groundwater at the Site
are presented in the table below. Both the values used in the 1994 Risk Assessment and
the revised values are included for comparison.
Contaminant
Benzo(a) pyrene
Dibenz(a,h)anthracene
Exposure Point
Concentration 0/g/l)
1994 value
7.19
6.51
Revised
28.4
12
Lifetime Cancer Risk
1994 value
1.58E-06
3.23E-06
Revised
6.2E-06
6.0E-06
These revised values, along with the revised PCB values noted in Comment #11 above,
actually increase the potential risk associated with the groundwater exposure pathway
from 6.37E-06 to 1.28E-05. Therefore, the net result of recalculating the risk values for
the contaminants questioned by Site Owner is an increase in the total potential risk to
future construction workers from 6.01 E-03 to 6.02E-03.

The Site Owner points out that both the 1994 and 1995 Risk Assessments incorrectly
state, in the next to last sentence on page 5, that construction workers would contact the
"floating free product" found at monitoring well 5 at the Site.
                                          74

-------
       Response:  The Site Owner is correct. This sentence actually refers to the floating free
       product that was found at monitoring well 6. Table 11 in both the 1994 and 1995 Risk
       Assessments correctly indicate that the LNAPL was found in MW-6.

15.    The Site Owner contends that the estimated risk to future construction workers from
       dermal contact to LNAPL at monitoring well MW-6 is invalid because the exposure point
       concentration used cannot be verified and is not representative of current Site conditions.
       The Site Owner asserts that data from the 1992 sampling should have been used rather
       than data from the 1991 sampling.

       Response: Table 4-29 of the Remedial Investigation Report summarizes PCB data
       collected in monitoring wells at the Site during the 1991 and 1992 sampling events (see
       AR300243).  Samples labeled MW6-P are those that contained floating petroleum
       product (i.e.,  LNAPL). During the October 1991  sampling, a discrete LNAPL layer was
       encountered and sampled.  The PCB concentrations contained in this LNAPL sample are
       430,000 ppb of Aroclor 1248 and 660,000 ppb of Aroclor 1260. Table 4-29 also reports
       the results of a second sample labeled as MW6-P seven months later (May  1992) which
       contained 4 ppb of Aroclor 1232 and 3 ppb of Aroclor 1242. This sample was collected
       after the monitoring well had been purged and only droplets of LNAPL were observed in
       the groundwater sample, rather than a discrete LNAPL layer.  Although labeled as an
       LNAPL sample, this sample better represents the  groundwater concentrations of PCBs in
       MW-6, rather than LNAPL concentrations.

             In calculating the potential risk associated with dermal contact with LNAPL, EPA
       used the sum of the aroclor values observed in the 1991 sample as the RME concentration
       (i.e., 1,090 ppm). The PCB concentrations observed in this sample are consistent with
       those historically found in the oil layer at the Site. Total PCB concentrations  observed in
       the oil layer at the Site in 1977 and 1989 were 1,539 ppm and 1,540 ppm, respectively.
       Although the prior cleanup action at the Site removed much of the oil layer that could be
       recovered through active pumping, an oil layer has still been observed at the Site as
       reflected by the 1991 sample results2. Therefore, the 1991 results more appropriately
       represent a reasonable maximum concentration to which future construction workers
       could be exposed and were appropriately used in the risk assessment for the Site. EPA
       reviewed the analytical data for the 1991 sample and the risk calculations based on this
       sample to ensure their accuracy. The results are presented in Appendix D of the
       Technical Review, which is included in the Administrative Record (see AR304122).

16.    The Site Owner contends that the exposure point concentration for benzo(a)pyrene used
       to estimate risk to children from ingesting river sediment while recreational boating near
             Further discussion of the presence of an oil layer at the Site can be found later in the
             Responsiveness Summary in the responses to Comment #9 in Section B (Aquatic Ecological Risk
             Assessment) and Comments #1 - 7 in Section D (NAPL).

                                          75

-------
      the Site cannot be verified.

      Response: The RME concentration for benzo(a)pyrene used in this risk calculation was
      the maximum observed concentration of 17 ppm. This concentration was used because it
      is lower than the log UCL of the river sediment dataset. The log UCL was calculated
      using the data presented in Table 4-20 of the Remedial Investigation Report (see
      AR300214). The statistical calculations are summarized in the Appendix 2 to the 1995
      Risk Assessment (see AR301632).

17.    The Site Owner contends that the river sediment sample which has the maximum
      benzo(a)pyrene concentration (US2) is located upstream of the Site and should not be
      included in the risk calculations (see Figure 2-11 in the Remedial Investigation Report;
      AR300070). The Site Owner further contends that if this sampling location is included,
      data from the two samples collected at this location should have been averaged, rather
      than using the higher of the two results.

      Response: River sediment samples US1 and US2 were two discrete samples collected
      immediately upstream of the Metal Bank property boundary. Although these samples are
      labeled upstream, they can be used to characterize Site conditions because of the tidal
      influences of the Delaware River.  Benzo(a)pyrene was detected in US1 and US2 at
      concentrations of 2.8 and 17.0 ppm, respectively. Benzo(a)pyrene was also detected in
      seven of the eight additional river sediment samples collected downstream of US1 and
      US2.  The river sediment concentrations observed moving downstream from US1/US2
      are 12 ppm at TRANS1-HW; 0.93 ppm at TRANS1-LW; 2.8 ppm at TRANS2-HW; 0.33
      ppm at TRANS2-LW; 3.3 ppm at MF-119 and 7.9 ppm at MF-138 (see AR300214).
      These data indicate that benzo(a)pyrene is present in river sediments along the Metal
      Bank  property boundary. Although other sources could be contributing to this
      contamination, benzo(a)pyrene was observed in samples collected from the Site soils and
      groundwater as well as from the river sediments.  Therefore, benzo(a)pyrene
      concentrations found in all of the river sediment samples collected as part of the
      Remedial Investigation were appropriate to include in the risk assessment calculations.

             Even if the benzo(a)pyrene concentration in US2 sediment sample is not used in
      the risk assessment, as the Site Owner suggests, the change in the overall potential risk to
      recreational boaters would be insignificant. The RME for the river sediments would be
      revised from 17.0 ppm to 12.0 ppm since the latter is the maximum concentration
      observed in the sediment if US2 is excluded.  The reductions that result in the lifetime
      cancer risk for both adult and child recreational boaters are presented in the table below.
      No change occurs in the total lifetime cancer risk to adult recreational boaters and the risk
      to child boaters decreases from 1.60E-04 to 1.59E-04 {4.61E-05}.
                                         76

-------
Lifetime Cancer Risk to
Recreational Boaters
Risk from Benzo(a)pyrene in
River Sediment
Total Risk from River
Sediments
Total Risk from All Pathways
1995 Risk Assessment
Adult
7.49E-07
1.37E-06
2.65E-04
{7.14E-05}
Child
1.75E-06
3.20E-06
1.60E-04
{4.61E-05}
Revised Risk
(excluding US2)
Adult
5.29E-07
1.15E-06
2.65E-04
{=7.14E-05}
Child
1.23E-06
2.68E-06
1.59E-04
{=4.61E-05}
18.    The Site Owner contends that the dioxin exposure point concentration used to estimate
      risk to children and adults from ingesting sediment from the rip-rap area is incorrect
      because of improper addition of the data and improper conversion of data units.

      Response:  The Site Owner is correct in noting errors in the calculation of the exposure
      point concentration used to estimate risk to recreational boaters who inadvertently place
      their hands or other objects soiled with contaminated rip-rap sediment in their mouths.
      However, these errors have no impact on the overall potential risk to recreational boaters.
      The data summary prepared by the PRP Group for the rip-rap sediment samples and
      included as Table 4-9 in Appendix 2 of the 1994 Risk Assessment (see AR300697)
      contained several errors. The maximum values identified for each dioxin and furan
      compound did not match the data presented in the Remedial Investigation Report (see
      AR300523; AR300527). In addition, the summary indicated that the units for the dioxin
      and ruran data were Mg/kg when the actual units were ng/kg (nanograms per kilogram).

             To calculate the dioxin exposure point concentration for the rip-rap sediment data,
      EPA converted the maximum value as reported in Table 4-9 for each dioxin and furan
      compound to its equivalent 2,3,7,8-TCDD value (the most toxic form of dioxin). The
      sum of these values was used as the RME concentration in the risk assessment.  EPA
      made an additional error in the risk calculation in the 1994 Risk Assessment by
      inadvertently failing to convert the dioxin RME concentration to units of mg/kg before
      using it in the risk calculation.

             Several of the above errors  were identified and corrected in the 1995 Risk
      Assessment.  However, since the potential risk posed by this exposure pathway was not
      significant, EPA did not completely review and recalculate the exposure point
      concentration for dioxin in the rip-rap sediments.  In response to the Site Owner's
      concerns, EPA has revisited these calculations. Sediment samples from two locations
      (RR-3 and RR-4) were analyzed for dioxins and furans during the Remedial
                                          77

-------
       Investigation. A field duplicate sample was also collected and analyzed at RR-3.  EPA
       included the results from the field duplicate sample in the risk assessment because the
       dioxin and furan concentrations in this sample were significantly higher than those in the
       other two samples. Because the risk assessment focuses on the reasonable maximum
       exposure, EPA considers this conservative approach appropriate.  The revised data
       summary for the rip-rap sediment included in the 1995 Risk Assessment (see
       AR3016410) correctly identifies the maximum concentration observed for each dioxin
       and furan compound.

             If, as the Site Owner suggests, the concentrations from the field duplicate had
       been averaged with the other sample collected at RR-3 or if the RR-3 data had been
       completely excluded, the results of the risk assessment would be the same.  EPA
       calculated revised dioxin exposure point concentrations using the maximum dioxin and
       furan concentrations observed in the rip-rap sediment and using the data from only RR-4.
       The revised values are 1.66E-06 mg/kg and 7.35E-08 mg/kg, respectively.  Potential risks
       to recreational boaters using the dioxin RME from the  1995 Risk Assessment and the two
       revised dioxin RME concentrations are summarized below. Using the revised RMEs
       results in no change to the total potential risk to recreational boaters from the Site.
Lifetime Cancer
Risk to
Recreational
Boaters
Risk from dioxin in
Rip-Rap Sediment
Total Risk from
Rip-Rap Sediments
Total Risk from
All Pathways
Dioxin RME
in 1995
Risk Assessment
(5.91E-05)
Adult
1.19E-08
7.09E-07
2.65E-04
{7.14E-05}
Child
2.78E-08
1.65E-06
1.60E-04
{4.61E-05}
Revised Dioxin RME
All data (1.66E-06)
Adult
3.34E-10
6.97E-07
2.65E-04
Child
7.81E-10
1.63E-06
1.60E-04
RR-4 only (7 J5E-08)
Adult
1.48E-11
6.97E-07
2.65E-04
Child
3.46E-1 1
1.63E-06
1.60E-Q4
19.    The Site Owner contends that the PCB concentrations found in fish samples collected
      near the Site reflect background contamination and cannot be linked to the Site. The Site
      Owner further contends that the levels of PCBs detected in the edible fish samples (i.e.,
      fillets) are below the allowable level (2 ppm) established by the FDA for public
      consumption. Lastly, the Site Owner asserts that the exposure assumptions for fish
      consumption used in the risk assessment are inappropriate because the Pennsylvania Fish
      and Boat Commission has issued a fish consumption advisory indicating that channel
      catfish, white perch, and the American eel should not be eaten if caught in the Delaware
                                          78

-------
River between Yardley and the Pennsylvania/Delaware state line because of PCB
contamination.

Response: In characterizing the potential risk to recreational boaters from Site
contaminants, EPA acknowledges in the risk assessment that sources other than the Site
could have contributed substantially to the PCB levels in the fish fillets used to estimate
the risk (see AR301571). However, since PCBs are a known Site contaminant and
releases to the river have occurred, fish feeding on the PCB-contaminated sediments at
the Site are also likely to have accumulated some site-related PCBs in their tissue. EPA's
risk assessment process does not require that the Site be the sole source of contamination
before an exposure pathway can be included.

       EPA's assumptions used to estimate the potential risk to recreational boaters from
eating fish from the river are reasonable. To evaluate the reasonable maximum exposure,
EPA assumed that boaters would catch fish while boating on the river ten days during the
year and eat one meal from the fish caught during each trip. Adults would eat
approximately 200 grams (7 ounces) offish at each meal and children would eat
approximately half that amount.  The level of PCBs estimated to be in the fish (1.27
ppm) was derived from the levels found in the channel catfish fillet samples collected in
the river at the Site. The potential lifetime cancer risk resulting from this exposure is
estimated to be 2.62E-04 {6.79E-05} for adults and 1.53E-04 {3.96E-05} for children
which is outside the acceptable risk range of 1 .OE-04 to 1 .OE-06. The fish fillet analytical
results, the exposure point concentration, and the risk calculations were verified as being
accurate during preparation of this Responsiveness Summary.  The results are presented
in Appendix A of the Technical Review, which is included in the Administrative Record
(5eeAR304106).

       FDA's criteria for establishing temporary tolerances for residual PCBs in fish are
different than those used by EPA to make site-specific risk management decisions at
Superrund sites. EPA has followed its standard procedures for estimating the long-term
risk from potential exposure to Site contaminants. Based on reasonable assumptions
about the maximum exposure that recreational boaters could experience, EPA has
determined that the Site contributes to an unacceptable level of potential risk.

       EPA does not consider the presence of a fish consumption advisory on the
Delaware River as an adequate basis for not assessing the potential risk to recreational
boaters from eating fish containing a site-related contaminant.  One of the purposes of
addressing contamination from sources such as the Metal Bank Site is to eventually
reduce levels of PCB contamination in the Delaware River so that a fish consumption
advisory is not necessary.  Excluding the potential risk from eating fish because a
consumption advisory, which cannot be readily enforced, exists is not appropriate.
EPA's assumption that individuals may continue to consume fish caught in the Delaware
River is reasonable.
                                   79

-------
20.    The Site Owner questions why exposure point concentrations for surface water used to
       evaluate recreational exposure were based on values modeled from groundwater data
       rather than the actual surface water sample.  The Site Owner also questions which wells
       or groundwater data points were used to calculate surface water exposure point
       concentrations.

       Response:  The only surface water samples collected during the Remedial Investigation
       were those from ponded water at two locations in the mud flats.  These samples would
       not be representative of contaminant concentrations in the Delaware River at the Site. No
       samples of flowing surface water from the river were collected.  Therefore, EPA
       estimated surface water concentrations in the river by applying a dilution factor of
       4,300:1 to the contaminant concentrations found in the groundwater at the  Site. The basis
       for this dilution factor is included in Appendix 6 of the 1994 Risk Assessments (see
       AR300798).

             The groundwater concentrations used to estimate surface water values are the
       same as those developed to calculate the potential risk to construction workers from
       dermal contact with groundwater.  The RME concentrations for this exposure pathway
       are presented in Table 10 of the 1994 Risk Assessment (see AR300646) and are based on
       the groundwater data summary in Table 4-12 of Appendix 2 (see AR300704). The
       exposure point concentrations and risk estimates for recreational boaters exposed to
       surface water did not change in the 1995 Risk Assessment. As noted in the responses to
       Comments 11 and 13 above, errors have been identified in Table 4-12.  However, using
       the revised exposure point concentrations to estimate risk to recreational boaters from
       surface water would not result in any significant change to the overall risk  to boaters from
       Site contaminants.

21.    The Site Owner contends that EPA's risk assessment did not adequately explain why
       particular exposure scenarios were used or rejected or why particular pathways were
       selected to evaluate each exposure scenario. The Site Owner also contends that EPA did
       not adequately document the basis for the exposure assumptions used in the risk
       calculations.

       Response:  Section 3.3 of the Risk Assessment (see AR301566) identifies the four
       populations evaluated as potentially at risk from site-related contaminants and the
       pathways by which exposure could occur.  Section 3.4.2.1 of the Risk Assessment (see
       AR301567) also explains why several exposure scenarios were not included. The four
       groups of individuals who reasonably could be exposed to Site contaminants are off-site
       residents, recreational boaters on the Delaware River, and future industrial and
       construction workers at the Site. The various exposure pathways by which Site
       contaminants could reach one or more of these groups include breathing air containing
       dust from the Site, swallowing soil, surface water or sediments or eating fish
       contaminated with Site contaminants, and direct skin contact with contaminants in
       surface water, groundwater or the oil layer present at the  Site.  While the above

                                          80

-------
       referenced sections of the Risk Assessment do not repeat the physical characteristics of
       the Site and the surrounding areas to explain why these exposure scenarios are
       appropriate, this information is included in the Remedial Investigation Report, as well as
       the discussion of local land use provided in Appendix 5 of the Risk Assessment (see
       AR301713). The industrial nature of the Site, the presence of residential areas in the
       vicinity of the Site, and the presence of the Delaware River immediately adjacent to the
       Site justify the selection of these exposure scenarios.

22.    The Site Owner contends that the uncertainty analysis in the risk assessment is inadequate
       because it does not discuss the uncertainty associated with data evaluation and exposure
       assessment issues presented in the above comments.

       Response:   EPA's uncertainty analysis in the risk assessment focuses primarily on
       issues that have the greatest potential to result in an over- or underestimation of health
       risk from Site contaminants. Neither the data evaluation nor exposure assessment issues
       identified by the Site Owner result in significant changes in the risk assessment
       conclusions and, therefore, were not discussed in the uncertainty analysis.

23.    The Site Owner contends that EPA used outdated toxicological indices for PCBs, arsenic,
       beryllium, cadmium, and hexavalent chromium in estimated risks at the Site.

       Response: The toxicological indices used in the 1995 Risk Assessment were those being
       used at the time the assessment was completed. Subsequent to that time, only the
       carcinogenic oral slope factor for PCBs has changed from 7.7 kg-d/mg to 2.0 kg-d/mg.
       Because PCBs are the primary contaminant of concern at the Site, EPA reevaluated the
       potential risks for exposure pathways that had lifetime cancer risks greater than 1 .OE-05
       in the 1995 Risk Assessment. PCBs were the only  contaminants contributing to the
       potential risk in these pathways, which include exposure to PCBs in the Courtyard soils,
       exposure to PCBs in LNAPL, and exposure to PCBs in fish tissue. The decrease in the
       lifetime cancer risk for these pathways resulting from the revised PCB slope factor are
       summarized in the table below.  The calculations are presented in Appendix A of the
       Technical Review, which is included in the Administrative Record (see AR304110).
Exposure Pathway
Ingestion of Courtyard surface soil
Dermal contact with LNAPL
Lifetime Cancer Risk
PCB Slope
Factor
of 7.7 kg-d/mg
6.77E-05
6.00E-03
PCB Slope
Factor of 2.0
kg-d/mg
1.76E-05
1.56E-03
                                          81

-------
Ingestion offish by adult
Ingestion offish by child
2.62E-04
1.53E-04
6.79E-05
3.96E-05
             The revised total lifetime cancer risk for future industrial workers, future
       construction workers and recreational boaters using the above revised risk values are
       presented in the following table. While the revision in the PCB slope factor decreases the
       lifetime cancer risk to these populations, the potential risk to future construction workers
       remains outside EPA's acceptable risk range of 1 .OE-04 to 1 .OE-06.  In addition to
       potential risk to human health, the Site also poses potential risk to aquatic organisms.
       Therefore, the actions required in the Record of Decision remain appropriate.
Exposed Population
Future Industrial Workers
Future Construction Workers
Adult Recreational Boaters
Child Recreational Boaters
Lifetime Cancer Risk
1995
Risk Assessment
7.46E-05
6.01 E-03
2.65E-04
1.60E-04
Revised
PCB Slope Factor
2.45E-05
1.58E-03
7.14E-05
4.61 E-05
24.    The Site Owner contends that EPA did not prepare the risk assessment in accordance with
       EPA's Risk Assessment Guidance for Superfitnd, Volume 1, Human Health Evaluation
       Manual (Part A) (EPA/540/1-89/002) and did not include a sufficient level of
       information to explain its conclusions.

       Response:  EPA has reviewed the risk assessment calculations to ensure that the
       formulas and exposure assumptions conform with EPA guidance.  Although the Site
       Owner may disagree with many of the assumptions EPA made in conducting the risk
       assessment, EPA's decisions are consistent with Agency guidance. Similarly, EPA has
       provided adequate documentation  to support the conclusions reached in the risk
       assessment.  While the Site Owner's requests for additional information, clarifications,
       or other  editorial changes would improve the risk assessment if a revised version was
       being prepared, EPA has determined that the existing document adequately evaluates
       the risks to human health from the Site and that the additional expenditure of funds to
       revise this document is not necessary. Comments that raise substantive issues that
       could alter the findings of the risk assessment or EPA's remedy selection decision have
       been summarized and addressed above.  Editorial comments are not specifically
       addressed in  this Responsiveness Summary, but have been included in the
                                          82

-------
       Administrative Record for the Site.

25.    The Site Owner estimates there would be 15 truck accidents if EPA were to dispose of
       the contaminated soils off-site (to Model City, New York).  No calculations were
       provided to support this conclusion.  Although there were no discussions on fatalities in
       the Site Owner's original comments, one of the accompanying attachments included a
       1990 affidavit in which their consultant stated that "people (1 to 2) would die hi traffic
       accidents" if contaminated soil were to be transported to the nearest disposal facility in
       Alabama.  No calculations were provided to substantiate this statement.

             In two (2) documents submitted after the close of the comment period, the Site
       Owner added that truck accidents will result in 2.7 deaths per 100 million miles
       traveled  by large trucks.  Using two different estimates for the volume of soil to be
       disposed (approximately 9,000 yd3 according to EPA and approximately 31,000 yd3
       according to the PRP Group), the resulting  number of deaths due to truck accidents
       ranged from 0.010476 to 0.036726 deaths (see AR304056). The Site Owner  concluded
       that cleaning up the Site would entail a greater risk (1.0476E-02 and 3.6126E-02, see
       AR304386) than not (6.77E-05 and 6E-03,  see ROD Table 10).

       Response:  EPA does not normally respond to comments submitted after the comment
       period, but since the Site Owner stated, during a briefing for EPA on the significant
       issues raised in their 22 volumes of comments, that truck accidents would result in
       deaths, EPA will treat this new comment as if it was part of the original truck accident
       comment.

             EPA believes the Site Owner's comment concerning potential deaths from truck
       accidents is not relevant because:

             (a)  It is statistically improper to equate the number of potential deaths due to
             truck accidents to the predicted risk  of cancer from site-related exposures.  The
             risk due to truck accidents is based on statistics from a 1994 National Highway
             Traffic Safety Administration study which estimated that 2.7 deaths may occur
             for every 100 million miles traveled by large trucks. The risk calculations
             appearing in the Baseline Risk Assessment estimate the probability of an
             individual developing cancer from exposure to site-related chemicals; without
             remediation, these risks will remain. The Site Owner's estimate of 0.01476 to
             0.036726 deaths due to truck accidents indicates that less than one death is
             anticipated from the activity in question, while EPA's excess cancer risk
             prediction of 6E-03 indicates that for every 1000 exposed  individuals,  six may
             develop cancer as a result of contamination at the Site; and

             (b) transportation of the contaminated soil is not restricted to truck traffic.
             Other options such as railways or waterways may be considered. The choice of
             a Disposal facility is  also not restricted to Model City, New York.  Regardless

                                          83

-------
              of what the accident, death or injury statistics may be, all workers would be
              required to diligently follow the protocols of a Health and Safety Plan  in order
              to minimize accidents. ROD Section IX.B.4 specifically requires that all
              transferral of hazardous wastes will be performed in accordance with RCRA
              requirements as defined in 40 C.F.R. Parts 262 and 263, and 49 C.F.R. Parts
              107, and 171-179.  EPA believes notification combined with coordination with
              all communities affected  by the transportation routes will result in an accident
              free project.

B.     Aquatic Ecological Risk Assessment

 1.     The Site Owner contends that the Aquatic Ecological Risk Assessment ("Aquatic
       Assessment") does not recognize that degradation of the Delaware River in the vicinity
       of the Site is attributable to many sources.  The Site Owner further contends that the
       Aquatic Assessment does not demonstrate that site-related contaminants are directly
       responsible for any degradation.  The Site Owner asserts that no evidence is presented
       which shows conditions of aquatic populations at the Site to be any different from or
       worse than those at other areas of the Delaware River.

       Response: The Aquatic Assessment for the Site does not attempt to evaluate the area
       as ecologically pristine.  The assessment focuses on site-specific contaminants and their
       potential impact on aquatic receptors in the vicinity of the Site. EPA recognizes that
       other sources may also contribute to the presence of these contaminants in the  river
       near the Site. However, EPA is not required to demonstrate that the Site is the sole
       source of contamination contributing to environmental degradation in order to  require
       cleanup actions. As the Aquatic Assessment demonstrates, the Metal Bank property
       has historically made a substantial contribution to the decline of aquatic habitat in the
       river at the Site.

              The contaminants found at concentrations above the screening levels used to
       indicate potential concern are summarized in Tables 2, 3, 4, 5, and 9 of the ROD. In
       comparing the contaminants found in the subsurface soils at the Site with those found in
       the sediments,  elevated concentrations of the following contaminants were found in
       both media: 4,4'-DDD, 4,4'-DDE, PCBs (Aroclors 1248, 1254, and  1260), 2,4-
       Dimethylphenol, benzo(a)anthracene, benzo(a)pyrene, benzo(b)fluoranthene, bis(2-
       ethylhexyOphthalate, chrysene, indeno(l,2,3-cd)pyrene, naphthalene,  arsenic,
       cadmium, copper, lead, mercury, nickel, and zinc. Many of the contaminants found in
       elevated concentrations in the groundwater at the Site were also found in the sediments.
       The Aquatic Assessment has, therefore, adequately demonstrated the potential  link
       between site-related contaminants and degradation of the aquatic ecosystem.

              As to the Site Owner's concern about how the aquatic environment near the Site
       compares to other  areas of the Delaware River, this is not a matter that is of
       importance in conducting an Aquatic Assessment. The purpose of the Aquatic

                                          84

-------
       Assessment is to evaluate the potential impact that site-related contaminants may
       presently be having on the environment. As indicated above, the assessment is not
       intended to conclusively demonstrate that contaminants from the Site are responsible
       for degradation; nor is the assessment intended to evaluate the health of the
       environment in comparison to other non-site-related areas or evaluate the relative health
       of the environment over tinier  Whether other areas of the Delaware River are equally
       or more degraded than the portion at the Site or whether the health of the river in the
       vicinity of the Site has improved over the last 70 years is not  relevant to evaluating the
       current impact of site-related contamination and would be beyond the scope of any
       Aquatic Assessment for the Site.

2.     The Site Owner contends that the shortnose sturgeon should not have been included in
       the Aquatic Assessment as a potential receptor because it has  never been documented as
       present in the river near the Site and does not use this portion of the river as a
       spawning or nursery area.  The Site Owner further contends that the exposure pathway
       for the shortnose sturgeon is incomplete because the foraging  habitat  of the sturgeon
       does not overlap with the mudflats adjacent to the Site.

       Response:  Shortnose sturgeon were captured within about one kilometer both
       upstream and downstream of the Site during a tagging study conducted between 1983
       and 1987. This study  concluded that after spawning from late March through April,
       the shortnose sturgeon travel rapidly downstream into the tidal portion of the river near
       Philadelphia where they remain through the end of May. Some individuals may remain
       in the area through the summer and fall.3 The issue of whether or not shortnose
       sturgeon actually spawn at the Site is irrelevant. Tables 5-2 and 5-3 of the Aquatic
       Assessment illustrate that PCBs have been shown in numerous studies to elicit a wide
       range of toxic effects on a variety of species. Many of these studies have demonstrated
       that exposure to PCBs can result in reproductive impairment in fish.  For example,
       studies have associated increased tissue concentrations of PCBs in eggs and ovaries of
       fish with reduced egg hatchability, reduced fry survival, increased larval mortality, and
       physical abnormalities. Because eggs have high lipid content, PCBs have a tendency  to
       accumulate in the eggs and can be passed on to offspring by the adult females.  The
       actual dose encountered by the  offspring will be the result of exposures encountered by
       the adult females prior to spawning. Therefore, the shortnose sturgeon does not have
       to spawn at the Site for the offspring to  be impacted. As to the shortnose sturgeon's
       foraging area, the mudflat area at the Site is likely  to be a feeding habitat because the
       shortnose sturgeon prefer shallow areas  where they can feed indiscriminately on
       mollusks, polychaetes, and other benthic organisms.  Data from benthic
       macroinvertebrates sampling conducted  in 1991 indicate that prey of the shortnose
       sturgeon are present in the mudflats adjacent to the Site.
             See the O'Herron, et al, 1993, reference on page 7-7 of the Aquatic Ecological Risk Assessment

                                          85

-------
3.     The Site Owner contends that the channel catfish does not serve as an appropriate
       surrogate for evaluating potential risks to the shortnose sturgeon because the home
       range of the channel catfish is very small relative to the sturgeon, the channel catfish
       does not migrate to spawning grounds as do the sturgeon, the catfish have higher lipid
       content than the shortnose sturgeon, and the catfish feed higher in the food chain than
       does the shortnose sturgeon.

       Response:   The receptors of concern used to evaluate risks in the Aquatic Assessment
       were selected based on the need to evaluate species that (1) represent different trophic
       levels and food web pathways, (2) are likely to be present near the Site for at least part
       of the spring-summer-fall feeding period, and (3) have available data. The shortnose
       sturgeon, a Federal and Commonwealth of Pennsylvania designated endangered
       species, is a receptor of primary importance because the sturgeon lives its entire life
       cycle in the Delaware River and is known to be present in the river adjacent to the Site
       during the summer and after spawning in upstream areas.  Channel  catfish are also a
       receptor of concern because they serve as a representative of benthic freshwater fish
       species that feed on a variety of prey types and are likely to be resident in the river
       next to the Site for some time.  Because only limited data are available to evaluate
       impacts of site-related contaminants on the shortnose sturgeon, the channel catfish is
       also used as a surrogate species for the shortnose sturgeon.

              The Aquatic Assessment recognizes the uncertainties associated with using the
       channel catfish as a surrogate for the shortnose sturgeon (see page 6-11 of the Aquatic
       Assessment). However, contrary to the Site Owner's concern that risk to the shortnose
       sturgeon may be overestimated, several factors would support the opposite conclusion.
       The shortnose sturgeon may be particularly vulnerable to the accumulation of PCBs
       due to their benthic feeding habit, their longevity, and relatively late age of maturity.
       Shortnose sturgeon are known to live as long as 67 years. The fact that female
       sturgeon may spawn only once every several years may be an additional factor leading
       to increased accumulation of PCBs,  since transfer of PCBs from the females to their
       eggs  is thought to be an important mechanism in reducing the PCB body burden in
       female fish.  When these factors are balanced  with those raised by the Site Owner, the
       channel catfish is considered to be a good candidate overall for use as a surrogate for
       the shortnose sturgeon.

4.     The Site Owner contends that if the shortnose sturgeon is evaluated in the Aquatic
       Assessment, the actual PCB concentrations observed in six shortnose sturgeon collected
       32 kilometers upstream from the Site during a 1982 New Jersey study should be used
       to estimate risk rather than the higher estimated concentrations based on channel catfish
       data.

       Response:  No information is available to indicate whether the six shortnose sturgeon
       collected in the 1982 study were exposed to contaminants in the vicinity of the Site.
       Because of the difficulty in accounting for the uncertainty associated with using these

                                           86

-------
       data, the Aquatic Assessment used the more conservative, but certain, approach of
       estimating PCB concentrations in the shortnose sturgeon based on a surrogate species, the
       channel catfish, collected in the vicinity of the Site.

5.     The Site Owner contends that the Aquatic Assessment should have used a sediment-
       dwelling species such as Corbicula fluminea (Asiatic clams) to evaluate risk from Site
       contaminants through the sediment exposure pathway. The Site Owner further
       contends that the abundance of benthic invertebrates in the mudflats as quantified
       during the Remedial Investigation demonstrates that the sediments have no adverse
       effects on benthic invertebrates.  The Site Owner asserts that the benthic invertebrate
       survey data should have been statistically compared to contaminants of concern in the
       sediment at the same stations and that such an analysis would show that benthic
       invertebrate populations are not being adversely impacted.

       Response:  The Aquatic Assessment evaluated benthic invertebrates, which are
       sediment-dwelling species, as receptors of concern at the Site. This evaluation focused
       on the importance of these receptors as an exposure pathway to higher trophic-level
       fish rather than on a quantitative assessment of how toxic site-related contaminants are
       to these organisms.  The limited information available in the literature suggests that
       benthic invertebrates can accumulate high concentrations of PCBs in their tissue
       without adverse effects, although invertebrates may experience direct  toxicity from
       exposure to contaminated sediments (see page 6-13 of the Aquatic Assessment).  Tissue
       data from Corbicula were used to indicate the bioavailability of PCBs at the Site.
       Corbicula serve as an excellent model species for determining the bioavailabiliry of
       contaminants in surface water and sediments because they accumulate contaminants by
       the same mechanisms as similar native species.

              Although the Aquatic Assessment did not focus on the direct adverse impacts of
       Site contaminants on benthic invertebrates, the assessment did discuss the results of
       qualitative and quantitative benthic invertebrates surveys conducted at twelve and six
       sampling locations, respectively, in the mudflat area. A comparison of total PCB
       concentrations in the sediments from the May 1991 sampling and the qualitative
       diversity and density data from the July 1991 survey was conducted and the Aquatic
       Assessment states that no apparent correlation exists. The Assessment also explains
       that this result is not surprising since the sediment sampling and benthic survey were
       conducted at different times, thus raising uncertainty as to whether the locations used in
       the two events were the same.  The data did, however, show a correlation indicating
       that species diversity increased with greater distance from the western Site boundary.
       This result suggests that benthic productivity in the mudflat in the immediate vicinity of
       the Site may be reduced.

6.     The Site Owner contends that the Aquatic Assessment does not demonstrate that PCBs
       in field samples of fish are from the Site. The Site Owner contends that fish tissue
       data, when analyzed appropriately, demonstrates that PCB levels in fish near the Site

                                           87

-------
       are among the lowest over the 200 mile length of the Delaware River.

       Response:  The purpose of the Aquatic Assessment is to characterize potential risk to
       aquatic organisms by comparing the levels of contamination in sediment and surface
       water at the Site which the organisms may be exposed to (i.e., exposure point
       concentrations) with levels considered to cause adverse effects in the organisms (i.e.,
       toxicity reference values).  The Aquatic Assessment also compares concentrations of
       contaminants found directly in the tissue of fish collected at the Site with levels
       associated with adverse effects to provide an additional basis for evaluating potential
       risks. The  Aquatic Assessment concludes that exposure of aquatic organisms to PCBs,
       PAHs, phthalates, DDT and its metabolites, and cadmium in the sediment at the Site
       poses a potential hazard to aquatic organisms (see pages 6-4 and 6-5 of the Aquatic
       Assessment for actual hazard quotients). The assessment further finds that the
       accumulated levels of PCB in the tissue of silvery minnows and channel catfish
       collected at the Site could result in adverse impacts.  EPA does not assert that all the
       contaminants found in the sediment and fish come from the Site (see response to
       Comment #1 above). EPA does draw the reasonable inference that when contaminants
       found in the soil and groundwater on the Metal Bank property are also found in
       elevated concentrations in  sediment and fish collected adjacent to the property, the soil
       and groundwater are likely sources.  The Aquatic Assessment provides ample support
       for this inference.  The purpose of the Aquatic Assessment is not to show that aquatic
       populations at the Site are different or worse than other areas of the Delaware River.

7.     The Site Owner contends that, at a minimum, PCB concentrations in sediment at or
       adjacent to the Metal Bank property should be compared with concentrations in similar
       samples upgradient of the Site,  and only that amount which is greater than background
       should be evaluated.

       Response:  When  calculating potential risks at Superfund sites, EPA will generally
       only eliminate a particular contaminant from the evaluation if that contaminant is
       shown to occur naturally in the area at levels at or above the concentration detected at
       the Site.  Since organic contaminants are not naturally occurring, they are rarely
       eliminated from the risk assessment process  because of the extreme difficulty in
       conclusively showing that the contaminant found at the site is from a source other than
       the Site.  This is particularly true at the Metal Bank Site. Because the Delaware River
       is tidal at the Site, sediment samples upgradient of the Site's influence would be located
       at a significant distance from the Site. Any PCBs found in sediment samples collected
       at these upgradient locations would be from  unknown sources and determining whether
       these sources were  contributing to the PCB levels found at the Site would be extremely
       difficult and beyond the scope of the Site investigation.  In reality, any relationship
       between PCB concentrations at a distant upgradient station and the Site  would be
       questionable. Therefore, subtracting the upgradient PCB concentrations from
       concentrations found at the Site would likely result in a significant underestimation of
       site-related risk.

-------
8.     The Site Owner contends that the trend in the PCB sediment data from the Site
       discussed in the Aquatic Assessment incorrectly relied upon total PCB concentrations
       rather than congener-specific data. Even if total values are used, the Site Owner
       contends that the PCB sediment data does not show a trend of decreasing concentration
       with greater distance from the Site as indicated in Figure 4-3  of the Aquatic
       Assessment. The Site Owner points to Figures 4-16 and 4-17 in the Remedial
       Investigation Report to assert that an opposite trend is observed by simply plotting the
       data on a site map.  The Site Owner also states the EPA's trend is dependent upon one
       data point (MF-107) which should have been eliminated from the dataset as an outlier
       and is of questionable validity.

       Response:  For the most part, the PCB laboratory analyses conducted during the
       Remedial Investigation measured the concentrations of various PCB Aroclors present in
       the samples. There are actually 209 different PCB compounds, referred to as
       congeners, included under the general label of PCBs. Individual congeners having the
       same number of chlorine atoms are referred to as isomers.  Aroclors are mixtures of
       PCB congeners. For example, Aroclor 1242 is composed mainly of tri- and tetra-
       chlorinated biphenyls  with approximately 42% chlorine by weight, while Aroclor 1254
       is composed mainly of penta-, hexa, and hepta-chlorinated biphenyls with
       approximately 54% chlorine by weight.  At the time the investigation for this Site was
       completed, EPA typically required PCB analyses to measure the concentrations of
       various Aroclors present in a sample.  The concentrations of Aroclors in each sample
       can then be  added to derive a total PCB concentration. This is the  approach that was
       followed.

             Aroclor or total PCB data can be used to conduct trend analyses to show a
       contaminant gradient (i.e., to show that contaminant levels are increasing or decreasing
       with distance from a site). The data analysis in the Aquatic Assessment is intended
       only to show that such a gradient appears to exist.  The analysis is  not intended to
       conclusively determine that a point source related to the Site is present.  EPA agrees
       that more sophisticated analytical techniques would be needed to trace the potential
       movement of specific  PCB congeners in Site soil and groundwater into the aquatic
       environment.  Analytical methods and quality assurance protocols for congener-specific
       analyses are still in development.  In addition, this type of PCB analysis is more
       expensive, and the added value in terms of estimating potential risk is unknown.
       Therefore, this level of analysis is not typically conducted at Superfund sites.  Limited
       congener analysis was conducted during the Remedial Investigation at the Site and
       similarity in the congener composition between one groundwater sample collected on-
       site and sediment and  Asiatic clam samples collected adjacent to the Metal Bank
       property was observed.  The Site Owner had the opportunity to participate in
       performance of the investigation, but declined. If the Site Owner had participated and
       expressed a  desire to conduct a more detailed PCB analysis at the Site, EPA would
       have supported such an effort.
                                          89

-------
              The regression analysis conducted in the Aquatic Assessment and illustrated in
       Figure 4-3 used a standard statistical approach. The analysis evaluated whether the
       total PCB concentrations in sediments in the riprap and mudflat stations along the
       western property boundary increased towards the river.  Only sediment locations with
       laboratory results which were subject to quality control protocols were used in the
       analysis.  PCB concentrations for the aroclors detected in one or more sample
       (Aroclors 1248, 1254, and 1260) were added to determine the total PCB concentration.
       If one of these aroclors was not present at the analytical detection limit,  a concentration
       equal to one-half the detection limit was added in accordance with standard EPA
       protocol. As the Aquatic Assessment indicates, total PCB concentrations appear to
       increase gradually along the western property boundary, reaching a maximum
       concentration at MF-107 near the southern corner, and then drop off rapidly towards
       the Delaware River. The regression analysis conducted on sediment samples along the
       western boundary demonstrated a strong  correlation between total PCB concentration
       and distance from MF-107 (R2 of 0.71).  This correlation was even stronger when just
       the sediment samples immediately adjacent to the Site were used (R2 of 0.96).

              Contrary to the Site Owner's assertions, this correlation is consistent with the
       presentation of data illustrated in Figures 4-16 and 4-17 of the Remedial Investigation
       and is not dependent on the inclusion of MF-107 in the analysis.  An equally strong
       correlation would exist if the sediment data had been compared with distance from the
       sewer outfall.  In this case, the analysis would show an increasing trend in sediment
       concentrations with increasing distance from the sewer outfall. MF-107 is located
       closer to the more highly contaminated southwestern portion of the Site and would be
       expected to exhibit higher PCB concentrations.  Exclusion of this sampling point from
       the regression analysis would be clearly inappropriate.

9.     The Site Owner contends that prior remediation has  removed all recoverable oil (i.e.,
       NAPL) from the Site and that post-remediation sampling confirms that this exposure
       pathway is no longer complete.  Therefore, the Site  Owner asserts that evaluation of
       NAPL exposure should not have been considered in the Aquatic Assessment.

       Response:  In August 1981, Metal Bank of America,  Inc., began operating an oil
       recovery system consisting of several recovery wells, oil separation units, and 55-
       gallon drums containing activated carbon which was intended to remove all recoverable
       oil from the Site.  Collection of oil ceased in March 1989. EPA monitoring of on-site
       monitoring wells in March and August 1989  continued to show the presence of a
       floating layer of oil at the Site. An oil sample collected from a monitoring well located
       near the river on the western portion of the Site contained a total PCB concentration of
       1,540 ppm (see AR304232). No PCB analyses were performed in August 1989,
       however oil thickness observations indicated that despite eight years of oil recovery at
       the Site, a floating layer of oil was still present in  the subsurface with a thickness of up
       to 3 inches along the western portion of the Site (see AR304241).  Numerous oil seeps
       were also noted during the August 1989 monitoring  along the western embankment

                                          90

-------
       bordering the mudflat and extending approximately 100 feet upstream along the river
       adjacent to the property boundary.  The heaviest concentration of seeps appeared to be
       located at the water line during extreme low tide.  In 1991, sampling performed as part
       of the Remedial Investigation again demonstrated an oil layer with a PCB concentration
       of 1,090 mg/1 was present in monitoring well MW-6 located in the southwest portion of
       the Site. The maximum PCB concentration observed historically in the oil layer at the
       Site was 1,539 mg/1 in 1977 (see AR100034). A sample collected from MW-6 in 1992
       with a PCB concentration of 25.6 ppb was not representative of the oil layer because of
       the sampling technique used. Rather, this sample characterized the dissolved
       concentration of PCBs in the groundwater. (A limited attempt to characterize the
       impact of the oil layer on surface water during the Remedial Investigation resulted in
       the detection of PCBs in one groundwater seep sample at a concentration of 3.7 ppb.)

             Although the oil recovery system that operated at the Site is likely to have
       removed much of the NAPL that can be recovered through an active extraction system,
       this effort could not recover all the NAPL present at the Site.  Residual NAPL has
       been observed and can be expected to slowly continue to release from the soil into the
       groundwater at the Site. Because of the high concentrations of PCBs associated with
       the NAPL, even the release of small volumes from the subsurface soil into the
       groundwater and then into the surface water can pose adverse impacts to aquatic
       organisms. Therefore, the Aquatic Assessment appropriately included exposure to
       NAPL as a complete pathway in evaluating ecological risks. The acute toxicity value
       for PCBs used as the toxicity reference concentration in the Aquatic Assessment is 2.0
       Aig/1.  PCB concentrations observed in the NAPL sample collected in 1991  are over
       500,000 times greater than the acute toxicity value. As a result, small discharges of
       NAPL into surface water at the Site soils can pose a significant threat to aquatic
       organisms.

10.     The Site Owner contends  that PCBs were included as contaminants of concern in
       groundwater in the Aquatic Assessment after they had previously been eliminated using
       the established screening criteria.

       Response:  As explained  in the Aquatic Assessment, groundwater at the Site becomes a
       concern when it enters the Delaware River and becomes surface water.  Therefore, to
       determine which contaminants are of concern in the groundwater, the maximum
       concentration of each contaminant observed in the groundwater was divided by a
       dilution factor based on the estimated groundwater discharge rate and low-flow
       conditions in the Delaware River. If the  resulting value exceeded the ambient water
       quality criteria,  the contaminant was considered to be of concern.  When this screening
       was performed, all the contaminants were below the ambient criteria values.  However,
       because PCBs are contaminants of major concern at the Site, the Aquatic Assessment
       included an evaluation of the risk associated with exposure to surface water impacted
       by Site groundwater.  The results of the evaluation indicate that most aquatic species
       are not at risk from exposure to waterborne PCBs contributed by way of groundwater

                                         91

-------
       from the Site.  The Aquatic Assessment further concludes that it is highly unlikely that
       any individual shortnose sturgeon would spend sufficient time in the portion of the
       river adjacent to the Metal Bank property to be adversely impacted from PCB
       concentrations in the water resulting from the release of PCBs in the groundwater.

11.    The Site Owner contends that the Aquatic Assessment should have used EPA's ambient
       water quality criteria (" AWQC") of 0.014 ng/l as the toxicity reference concentration
       for PCBs in surface water rather than using the lowest documented maximum allowable
       toxicant concentration of 0.1 /zg/1 and then applying a 100-fold safety factor for the
       shortnose sturgeon (see page 5-2 of the Aquatic Assessment).  The Site Owner asserts
       that if a 100-fold safety factor is applied, it should be used with EPA's acute toxicity
       value of 2.0 ^g/1 because the shortnose sturgeon would be expected to be exposed to
       site-related contaminants for very short periods, if at all.

       Response:  As the Aquatic Assessment explains, EPA's AWQC for PCBs are based on
       "Final Residue Values," which were calculated to protect wildlife and human
       consumption of aquatic organisms contaminated with PCBs and are not based on
       toxicity to aquatic organisms. To determine an appropriate toxicity reference
       concentration which would be protective of the aquatic organisms themselves, a review
       was conducted of the published studies used to develop the AWQC which focused on
       chronic toxicity.  The highest concentration of PCBs observed which did not have an
       unacceptable effect on the  aquatic organism being tested was 0.1 ^g/1- The Aquatic
       Assessment further states that although a concentration of 0.1 ^g/1 may be protective of
       the majority of biota in the Delaware River in the vicinity of the Site, it may not be
       protective of sensitive  species. Because an endangered species, the shortnose sturgeon,
       is a potential receptor at the Site, application of an uncertainty factor is warranted.
       Based on literature research, an uncertainty factor of 100 is recommended when
       comparing between species within different orders of the same class.  Since the fish
       species used in the study generating the 0.1 ng/l PCB concentration (the  fathead
       minnow) is in the same class as the shortnose sturgeon, but represents a different order,
       the 100-fold uncertainty factor was used. To apply the 100-fold safety factor to the
       PCB concentration representing acute toxicity to freshwater aquatic life (i.e., 2.0 ^g/1)
       would result in a reference concentration of 0.02 A*g/l.  This level is higher  than the
       EPA AWQC for PCBs of 0.014 ^g/1 and would not represent a conservative estimate
       of potential risk to the shortnose sturgeon, an endangered species in the Delaware
       River.

12.     The Site Owner contends that extrapolation of the data in Tables 5-2 and 5-3 of the
       Aquatic Assessment to develop a toxicity reference concentration for PCBs in fish
       tissue is inappropriate because (1) not all endpoints included are ecologically relevant,
       (2) data from whole body, fillet and organs cannot be compared, and (3) adverse
       effects from field studies may be from contaminants or factors other than PCBs.  The
       Site Owner asserts that the  Aquatic Assessment overestimates risk by determining the
       lowest observed adverse effects level from data for all species and then applying an

                                          92

-------
       additional safety factor.  The Site Owner contends that the median, rather than the
       highly conservative 10th percentile of the tissue effects concentration should have been
       used to calculate the Hazard Quotients for the indicator species.

       Response:  The Aquatic Assessment used all the field and laboratory studies included
       in Table 5-2 to establish a toxicity reference concentration for PCBs in fish tissue
       because of the limited data available for receptor species and the apparent differences in
       sensitivity between species to exposure to PCBs.  The effects reported from the
       laboratory and field studies were all considered to represent adverse ecological effects.
       Most of the endpoints represented a measure of reproductive toxicity.  A weight of
       evidence approach was necessary because of the lack of specific information on
       reproductive toxicity  for the target receptor species.  The 10th percentile value for the
       tissue effects concentrations was intended to be highly conservative.  Use of the median
       value would not be protective of sensitive species. EPA recognizes that the approach
       used to develop the tissue toxicity  reference concentration for PCBs is conservative and
       the uncertainty analysis identifies many issues similar to those noted by the Site Owner.

13.    The Site Owner asserts that the Aquatic Assessment misrepresents the PCB fish tissue
       data collected at the Site and in other portions of the Delaware River by using data
       from both whole body and fillet samples to give the appearance that fish at the Site
       have higher levels of PCB contamination. The  Site owner further asserts that the
       Aquatic Assessment ignores naturally occurring factors such as dietary composition and
       breeding status that impact the fat  (i.e., lipid) content of fish and thus cause differences
       in PCB concentrations unrelated to source concentrations.

       Response:  The Aquatic Assessment  identifies  studies known to have generated PCB
       fish tissue data from various locations on the Delaware River and graphically depicts
       the concentrations detected in channel catfish on both a wet weight and lipid
       normalized basis see Table 4-10 and Figures 4-5 and 4-6 of the Aquatic Assessment).
       The purpose of including this information was to provide a general overview of the
       known tissue data from the Delaware River, including that collected as part of the
       investigation at the Metal Bank Site. No conclusions were drawn regarding the relative
       concentrations of PCBs in tissues from fish collected at the Site as compared to other
       portions of the river.  EPA agrees that comparisons of PCB concentrations in whole
       body and fillet samples such as those presented  in Figure 4-5 must be viewed
       cautiously.  Comparisons of data from these two types of samples can be made when
       the data are normalized based on lipid content as was done in Figure 4-6.  However,
       even in this instance,  care must be taken in comparing tissue concentrations from
       different years and different studies because variations in these concentrations may be
       due to temporal as opposed to spatial variations in the concentrations of the
       contaminants hi the environment or to variations in analytical methodologies. The
       naturally occurring factors noted by the Site Owner can also influence the variability in
       the data.  For these reasons, EPA does not recommend using this data to attempt to
       draw any conclusions concerning the relative concentrations of PCBs in fish at the Site

                                           93

-------
       in comparison to other portions of the river and the Aquatic Assessment does not do
       so.

14.    The Site Owner contends that the Aquatic Assessment inappropriately manipulated the
       bioaccumulation factors developed from PCB fish tissue and sediment data at the Site to
       conclude that silvery minnows are exposed, on average, to sediments containing the
       mean farfield concentration of PCBs. The Site Owner asserts that any number of
       factors such as the variability in sediment concentrations at the Site, the use of large
       feeding ranges, diet or non-equilibrium conditions could produce the resulting
       bioaccumulation factors.

       Response:  The Aquatic Assessment calculates bioaccumulation factors by dividing
       PCB concentrations in fish tissue from silvery minnows and channel catfish collected at
       the Site by the average sediment concentrations from samples collected within 30
       meters of the  Metal Bank property (near-field) and those beyond 30 meters (farfield).
       These bioaccumulation factors are then compared with values found in research
       literature in an effort to estimate where the fish which were collected may be feeding.
       The bioaccumulation factors calculated using the average sediment concentration from
       the nearfield are lower than the literature values.  When the average sediment
       concentration from the farfield is used in the calculation, the bioaccumulation factor for
       channel catfish is still below the range found in the literature, but the factor for the
       silvery minnow is within the literature range.  The Aquatic Assessment states that this
       suggests the silvery minnows collected at the Site, which have a relatively small range
       of movement, may have  been exposed to the farfield sediments. The factors suggested
       by the Site Owner could also explain the bioaccumulation factors observed, but do not
       discount the possibility suggested by the Aquatic Assessment.

15.    The Site Owner contends that the Aquatic Assessment employs the commonly accepted
       screening technique of comparing an environmental contaminant concentration to an
       ecotoxicity benchmark, but does not include a commonly accepted quantitative
       uncertainty analysis or the probabilities of adverse effects.

       Response:  A quantitative uncertainty analysis can only  be performed if a sufficient
       database exists to perform such an analysis, which was not the case at the Metal Bank
       Site.  The Aquatic Assessment included qualitative discussions of the factors
       contributing to uncertainty throughout the evaluation, including the final risk
       characterization section.

16.    The Site Owner contends that the sediment toxicity reference concentrations  were
       developed for marine organisms and, therefore, have questionable application to the
       Site.  The Site Owner asserts that more relevant benchmarks are readily  available and
       should be used.  The Site Owner further contends that use of the Effects Range-Low
       and Effects Range-Medium concentrations for developing the sediment toxicity
       reference concentrations is inconsistent with the use of equilibrium partitioning

                                           94

-------
       methods recommended in EPA guidance.

       Response:  The Aquatic Assessment used sediment benchmarks which were derived
       from a large number of studies of estuarine and coastal areas since similar information
       was not available for freshwater systems. The Delaware River at the Site is a tidal
       estuary, which is comparable to estuarine sites used to establish the sediment
       benchmarks.  As stated in the Aquatic Assessment, studies in the Long and Morgan4
       research included results from synoptically collected sediment chemistry and biological
       data, spiked sediment bioassays, interim equilibrium partitioning values, and Apparent
       Effects Threshold values. Using the effects concentrations observed or predicted by
       these various methods, the lower tenth percentile (Effects Range-Low or ER-L) and the
       median concentrations (Effects Range-Medium or ER-M) were calculated. The ER-L
       and ER-M concentrations were proposed as weight-of-evidence based guidelines for
       determining contaminant-specific sediment concentrations below which adverse
       biological effects would be unlikely (ER-L) and above which effects would be probable
       (ER-M).  Use of these values in establishing the sediment toxicity reference
       concentrations is appropriate and consistent with the approach used by EPA at  many
       Superfund sites.

17.    The Site Owner contends that toxicity reference concentrations for sediment and water
       were not consistently normalized to organic carbon content for contaminants such as
       PCBs which have high octanol-water partitioning coefficients.

       Response:  The Aquatic Assessment recognizes that environmental factors such as
       temperature, pH, salinity, water hardness, and organic content of soil or sediment can
       affect the toxicity of a  contaminant in the environment (see page 5-10 of the Aquatic
       Assessment, AR300977). In the case of sediment analyses, not normalizing based on
       organic content probably underestimates the risk. The median organic carbon content
       of the riprap sediments, which were the most contaminated, was 0.2%.  Since sediment
       samples included in the database used to derive the  sediment toxicity reference
       concentrations would likely have organic contents of 1-2%, not normalizing the data
       from the riprap sediments would underestimate the risks.

18.    The Site Owner contends that the silvery minnow should not have been selected as a
       receptor of concern since it is not listed as a "major species" in Table 2-1 of the
       Aquatic Assessment.  The Site Owner further contends that indicator species from a
             Long, E.R. and L.G. Morgan.  1991. The potential for biological effects of sediment-sorbed
             contaminants tested in the National Status and Trends Program. NOAA Technical Memorandum
             NOS OMA 52. Seattle: United States Department of Commerce, National Oceanic and
             Atmospheric Administration.  175 pp. + appendices.

             Long, E.R. and D.D. MacDonald.  1992. National Status and Trends Program Approach.  In:
             Sediment Classification Methods Compendium. EPA 823-R-92-006.

                                          95

-------
       minimum of three trophic levels should be included.

       Response:  The intent of the indicator species selection process was to choose species
       that consumed different types of food, so that movement of contaminants of concern
       from various environmental media (i.e., sediments and water) to primary consumers,
       secondary consumers, and so on, could be evaluated. The species selected as receptors
       of concern were shortnose sturgeon, channel catfish, silvery minnows, white perch,
       and benthic invertebrates (as represented by Asiatic clams). Shortnose sturgeon and
       channel catfish can both be considered to be opportunistic feeders, consuming whatever
       is available to them. White perch tend to consume plankton, oligochaetes,  benthic
       macroinvertebrates, aquatic insects and small fish.  Silvery minnows consume bottom
       ooze and algae, and Asiatic clams are filter feeders.  Using the species selected, it was
       possible to evaluate contaminant migration through the food chain at several trophic
       levels.   The risk assessment process does not require the use of indicator species from
       a particular number of trophic levels; nor does it require the use of species that have
       important commercial or recreational value.  Theoretically, a single species
       (representing a single trophic level) could be evaluated in a risk assessment, provided
       that any observed or predicted contaminant-related impacts to that species were of
       sufficient ecological significance to be used in the risk management decision-making
       process.

19.    The Site Owner contends that the Aquatic Assessment incorrectly states that the only
       surface water data available for the Site is from mudflat pools.  The Site Owner points
       to Table 4-5 in Appendix 3 of the 1994 Risk Assessment [for Human Health] (see
       AR300687) and Figure 4-1 of the Aquatic Assessment to indicate that Site-specific
       surface water data are available. The Site Owner asserts that these data should be used
       to evaluate risks associated with surface water rather than values estimated from
       groundwater.

       Response:  The Aquatic Assessment is correct in stating that the only surface water
       data available for the Site is from mudflat pools.  Figure 4-26 of the Remedial
       Investigation shows the two locations where samples of standing water found in the
       mudflat area were collected.  Two samples were  collected from each location.  These
       locations are also shown on Figure 4-1 of the Aquatic Assessment. The samples are
       labeled as MFA1, MFA2, MFB1, and MFB2.  The statistical summary on Table 4-5  of
       the 1994 Risk Assessment had incorrectly references these mudflat pool samples as
       river water data.  Figure 4-1 of the Aquatic Assessment also incorrectly identifies
       samples US1, US2, DR1 and DR2 as surface water samples.  These samples are
       actually sediment samples as shown in Figure 4-18 of the Remedial Investigation (see
       AR300215). Since no surface water samples were collected from the river flowing
       adjacent to the Site, EPA must rely on values estimated from the groundwater.

20.    The Site Owner contends that the Aquatic Assessment overestimated the potential risk
       from PCB concentrations in the Delaware River estimated from groundwater discharges

                                          96

-------
       from the Site.  The Site Owner asserts that a dilution factor based solely on the
       estimated average groundwater discharge rate (0.45 liters/second (1/s)) and the average
       river flow for the IS meter zone of dilution (5,295 1/s) should have been used. The
       Site Owner notes that these values result in a dilution factor of 11,688, not 5,295,
       which was the factor used in the Aquatic Assessment.  The Site Owner contends that
       estimating a dilution factor based on low flow in the river is inappropriate.  The Site
       Owner also asserts that site-wide average groundwater PCB concentrations should have
       been used rather than data from only MW-6 and MW-7 and questions why 15 meters
       was selected as the dilution zone.

       Response:  The Aquatic Assessment estimated potential PCB concentrations in the
       Delaware River resulting from contaminated groundwater discharging from the Site
       during  both average and low flow river conditions.  The estimated PCB concentration
       during  low flow provides a conservative estimate of potential risk to aquatic organisms.
       Even using highly conservative assumptions, the Aquatic Assessment concludes that
       most aquatic species are not at risk from PCBs in the surface water. While the hazard
       quotients for the shortnose sturgeon under low flow conditions indicate a slight
       potential for impacts, the Aquatic Assessment states that it is  highly unlikely that any
       individual sturgeon would spend the extended periods necessary to be  chronically
       exposed to the contaminants in the 15 meters of the river nearest to the Site.

              The Site Owner is correct in noting that the dilution factor for average river
       flow conditions is 11,688, rather than 5,295 which was used. However, the Aquatic
       Assessment already concluded that aquatic  organisms were not at risk from PCBs in
       surface water under average flow conditions using the lower dilution factor.  Similarly,
       shortnose sturgeon were not at risk using the hazard quotient resulting from the mean
       PCB concentration and only slightly at risk (hazard quotient of 1.2) using the PCB
       concentration at the upper confidence limit. By using the higher dilution factor and/or
       the less conservative assumptions suggested by the Site Owner, the hazard quotients for
       aquatic organisms and the shortnose sturgeon would all be less than one during average
       river flow.  The decision to use 15 meters as the dilution zone was based on best
       professional judgment since actual data relating to the dilution rate were not available.
       The dilution rate of 5,295 that was inadvertently used in the Aquatic Assessment would
       actually reflect a dilution zone of approximately 7 meters. As shown in the Aquatic
       Assessment, even using this more conservative assumption, the risk to aquatic
       organisms from PCBs  in surface water is minimal.

              Note that this discussion only reflects potential risks to aquatic organisms from
       exposure to PCB contamination in surface water.  Potential risks from exposure to
       sediment are evaluated separately.

21.     The Site Owner contends that the Hazard Quotients resulting  from exposure to surface
       water and sediment should have been calculated for each selected indicator species and
       that these values  should have been adjusted to reflect reasonable exposure durations for

                                          97

-------
       each species.

       Response: Toxicity reference concentrations were not available for each contaminant
       and species combination.  When this occurred, a toxicity reference concentration which
       was developed to be protective of most aquatic organisms or one derived from a
       toxicity test using a similar species was used.

             Data are not available to establish how long each indicator species may be
       exposed to contaminants at the Site.  The Aquatic Assessment notes when exposure
       durations may be unrealistically long. For example, the Aquatic Assessment notes that
       although the hazard quotient for exposure of the shortnose sturgeon to PCBs in surface
       water slightly exceeds the acceptable level under worst-case conditions, it is highly
       unlikely that any individual sturgeon would spend extended periods in the river at the
       Site sufficient to be at risk.

22.    The Site Owner makes numerous comments suggesting that the Aquatic Assessment
       should be revised to include additional information, correct citations, or make other
       editorial changes.

       Response:  EPA has considered all the comments provided by the Site Owner.  While
       incorporating  additional information or making editorial changes recommended in the
       Site Owner's comments may  improve the Aquatic Assessment if a revised version was
       being prepared, EPA has determined that the existing document adequately evaluates
       the risks to the aquatic ecosystem and that the additional expenditure of funds to revise
       this document is not necessary. Comments that raise substantive issues that could alter
       the findings of the Aquatic Assessment or EPA's remedy selection decision have been
       summarized and addressed above.  Editorial  comments are not specifically addressed in
       this Responsiveness Summary, but have been included in the Administrative Record for
       the Site.

C.     Terrestrial Ecological Risk Assessment

1.      The Site Owner contends that EPA did not use site-specific information, such as that
       summarized in the appendices of the Terrestrial Ecological Risk Assessment
       ("Terrestrial Assessment"), to characterize the onsite habitats, populations, and food
       sources. The Site Owner further contends that EPA failed to use realistic exposure
       duration given the minimal habitat and food available at the Site.

       Response:  EPA conducted a screening level ecological risk assessment (referred to as
       a Tier 1  assessment) at the Metal Bank Site.  This approach compares contaminant
       concentrations found in media (e.g., soil, surface water) to ecological benchmarks to
       determine if a potential for adverse impacts exists.  If a potential for adverse impacts is
       found, further site-specific characterization may be recommended to assess the
       likelihood of actual impact.  To conduct a site-specific ecological assessment, surveys

                                          98

-------
       are performed to determine the receptors present at the site and biological monitoring
       (e.g., bioassays, tissue analyses) is conducted along with the chemical analyses.
       Although the screening level ecological risk assessment identified the potential for
       adverse terrestrial impacts at the Metal Bank Site,  further site-specific characterization
       was not performed, primarily because the aquatic environment is considered to be the
       more sensitive system impacted by the Site and additional investigation was being
       performed as part of the Aquatic Assessment. The results of the screening level
       Terrestrial Assessment are presented in the ROD and used to support findings of the
       more detailed site-specific Aquatic Assessment.  The findings of the Terrestrial
       Assessment were not used, however, as a basis for the decision to require remedial
       action at the Site.

2.     The Site Owner contends that EPA failed to use terrestrial receptors to evaluate the
       terrestrial risk, but rather relied inappropriately  on aquatic receptors.  The Site Owner
       asserts that terrestrial indicator species should have been selected from the extensive
       species list provided in the appendices of the Terrestrial Assessment and from the
       biological data summarized in the Remedial Investigation.

       Response:  The species list provided in the appendices of the Terrestrial Assessment is
       a comprehensive list of species known to be present in the region and does not
       represent a site-specific survey of receptors likely to be present at the Metal Bank Site.
       EPA did not require collection of the detailed information needed to conduct a site-
       specific ecological risk assessment.  As discussed above, EPA used a screening level
       assessment to evaluate potential terrestrial impacts. The  soil was evaluated by
       comparing the soil chemistry data to ecological soil benchmarks. The intent of the
       Terrestrial Assessment was not to protect individual species, but to ensure that no
       species would be exposed to an unacceptable amount of contamination. This
       conservative assessment found that surface soils inside the Courtyard posed no risk and
       the soils outside the Courtyard indicate a potential  for risk from methyl ethyl  ketone in
       subsurface soil. The subsurface soil is not a major pathway to ecological receptors.

3.     The Site Owner contends that exposure to surface soil represents the  primary exposure
       pathway for terrestrial organisms at the Site and that the EPA dismisses this pathway as
       having no relevance rather than acknowledging there are no terrestrial risks because the
       surface soils are clean.

       Response:  EPA does not dismiss the surface soil pathway as having no relevance, but
       rather finds, as the Site Owner states, that the surface soils  inside the Courtyard do not
       pose a risk to ecological receptors.

4.     The Site Owner contends that to consider groundwater seeps as an exposure pathway,
       EPA should identify the circumstances whereby birds or  upland mammals would
       preferentially drink from seeps along the steep river banks and rip-rap and EPA should
       know the chemical quality of these seeps. The Site Owner asserts that, even if EPA

                                           99

-------
       could demonstrate that ground water was a complete exposure pathway, EPA erred in
       calculating the associated risk by not accounting for dilution that occurs when
       groundwater discharges into the Delaware River and by inappropriately using the
       freshwater quality criteria for aquatic organisms.

       Response: In conducting a screening level ecological assessment, conservative
       assumptions are used to ensure that the potential for adverse impacts is not
       underestimated.  Groundwater seeps at the Metal Bank Site could potentially be
       exposure pathways for terrestrial receptors.  Actual observation of terrestrial receptors
       using groundwater seeps and chemical data of contaminant concentrations found in
       seeps would provide the best information for conducting a site-specific ecological
       assessment. However, this level of information is not required to conduct a screening
       level assessment  as was done at the Metal Bank Site. Use of the groundwater data and
       freshwater quality criteria to provide a conservative estimate of potential risk to
       terrestrial receptors is appropriate for a screening level assessment.

5.     The Site Owner contends that exposure to subsurface soil and riprap sediments are not
       valid pathways for the Site and that EPA did not provide any evidence to support its
       assumption that terrestrial organisms will have direct contact with subsurface soils or
       rip-rap sediments.  The Site Owner asserts that, even if EPA can demonstrate that
       subsurface soil is a complete exposure pathway, EPA grossly overestimated risk
       because it compared soil concentrations directly with toxicity benchmarks even though
       soil concentrations do not represent the dose experienced by an organism.

       Response: As stated above, a screening level ecological assessment does not require
       the level of site-specific information described by the Site Owner. The screening
       approach makes use of conservative assumptions in order to avoid underestimating
       potential adverse impacts.  Terrestrial receptors have potential to be exposed to
       subsurface soil at the Site as a result of future Site activities that could bring these soils
       to the surface. Terrestrial receptors could also be exposed to the riprap sediments
       during low tide or through the food chain by feeding on other organisms exposed to
       these sediments.

6.     The Site Owner acknowledges that exposure to contaminants in the mudflats appears to
       be a plausible and complete pathway for wading birds,  but contends that EPA
       inappropriately extrapolated risk to terrestrial receptors from aquatic data.  The Site
       Owner questions  how the sediment concentrations used to evaluate risk were derived
       and asserts that the sediments are not acutely toxic, as the Terrestrial Assessment
       suggests, because the remedial investigation demonstrated that benthic invertebrates are
       present in the Site sediments.

       Response: As stated in the risk assessment, terrestrial and avian species are exposed to
       contaminants of concern via direct contact, ingestion, and through the  food chain.
       Terrestrial organisms consume water and  sediment in addition  to food. All organisms

                                          100

-------
       from the base of the food chain through the higher trophic levels are subject to
       exposure.  The presence of benthic organisms does not indicate that the Site is not
       causing adverse effects to ecological receptors since many benthic organisms can
       tolerate unfavorable conditions.

              The expected environmental concentrations used in the Terrestrial Assessment
       were generally based on the data summary, used to prepare the  1994 Baseline Human
       Health Risk Assessment.  As noted previously, this data summary did not provide a
       complete summary of the data collected at the Site during the 1991 and  1993 sampling
       events  and was subsequently revised after completion of the Terrestrial Assessment.
       However, the revised data summary  did not significantly change the expected
       environmental concentrations used in the Terrestrial Assessment, so the Terrestrial
       Assessment was not revised.

7.     The Site Owner contends that EPA did not follow its own guidance in selecting
       contaminants of potential concern at  the Site.  EPA's Risk Assessment Guidance for
       Superfund,  Volume II, Environmental Evaluation Manual, recommends  considering
       factors such as physical/chemical properties, biological and environmental
       concentrations, toxicity,  and potential ARARs as criteria.  The Site Owner asserts that
       EPA did not consider any of these factors.

       Response:  EPA Region Hi's draft 1994 guidance for performing ecological assessment
       was used to conduct the Terrestrial Assessment and is consistent with EPA guidance.
       Region Hi's guidance for selecting contaminants of potential  concern suggests
       comparing the contaminants found in each media to appropriate literature values. This
       is the approach that was used in  the screening ecological assessment performed at the
       Site.

 8.     The Site Owner contends that EPA used extreme rather than reasonable contaminant
       concentrations for the environmental exposure concentrations.  The Site Owner further
       contends that EPA failed to fully explain how these concentrations were derived and
       failed to provide sufficient information to reproduce these concentrations.

       Response:  EPA used the 95 % upper confidence limit of the arithmetic  mean of the
       dataset for each medium to determine the expected environmental concentration. This
       is the standard approach for establishing concentrations to be used in ecological risk
       assessments and is consistent with  EPA ecological risk assessment guidance.

9.     The Site Owner contends that EPA failed to use current toxicity databases and
       information pertinent to wildlife to assess the toxicity of Site contaminants and,
       therefore, selected completely inadequate ecotoxicity endpoints to evaluate risks.

       Response:  EPA Region III has established screening level concentrations to be used in
       ecological assessments to provide a conservative estimate of potential adverse impacts.

                                          101

-------
       These concentrations were used in the Terrestrial Assessment performed at the Metal
       Bank Site and a copy of the draft guidance listing these values has been included in the
       Administrative Record.

10.    The Site Owner contends that the ecotoxicity quotients presented in EPA's terrestrial
       risk assessment on page 13 cannot be validated because neither the toxicity endpoints
       nor the exposure concentrations have been presented.

       Response:  As discussed above, the expected exposure point concentrations are based
       on the 95 % upper confidence limit calculated in the data summary prepared for the
       1994 Baseline Human Health Risk Assessment. The toxicity endpoints were based on
       EPA Region HI recommended values for screening potential ecological risk. These
       values are based on literature research and intended to provide conservative estimates
       of potential adverse ecological impact.

11.    The Site Owner contends that while the Terrestrial Assessment performed  by HMM
       Associates, Inc., for the PRP Group is very conservative and  overestimates risk, it
       conforms with Superfund guidance  and concludes that there is a low likelihood of
       negative effects to terrestrial species at the Site.  The Site Owner asserts that this
       conclusion more accurately characterizes the terrestrial risks and should be used rather
       than EPA's Terrestrial Assessment.

       Response:  EPA's Terrestrial Assessment conforms with Superfund guidance for a
       screening level assessment and provides a conservative estimate of potential risk to
       terrestrial receptors at the Metal Bank Site.  The PRP  Group's ecological risk
       assessment inappropriately uses models in place of actual site-specific information  and
       is not consistent with EPA guidance for conducting ecological risk assessments.

12.    The Site Owner contends that EPA  failed to include a  discussion of the uncertainties
       associated with the terrestrial risk assessment and, therefore, misleads the reader into
       having too much confidence in the analysis.

       Response:  EPA documents the fact that the Terrestrial Assessment uses a  conservative
       screening level approach throughout the document.  A list of limitations is  also
       included on page 9 of the Terrestrial Assessment.

13.    The Site Owner provides numerous comments on the overall quality of the written
       document criticizing its organization, thoroughness, and objectivity. The Site Owner
       asserts that these deficiencies result in inadequate  documentation of the conclusions
       drawn in EPA's terrestrial  risk assessment.

       Response:  EPA prepared the terrestrial risk assessment in-house with Agency
       personnel experienced in the ecological risk  assessment process.  In preparing this
       document, as with any document, the Agency had to balance the time required to revise

                                          102

-------
       and improve the quality of the document against the need to move the cleanup process
       forward.  This tradeoff often results in documents that could still be improved upon
       from a quality standpoint. Fortunately, the review and comment process associated
       with public documents generated by the Agency effectively resolves any remaining
       ambiguities. EPA initially provided the terrestrial risk assessment to the PRP Group
       performing the RI/FS for review and comment.  The PRP Groups' comments and
       EPA's response are included in the Administrative Record to provide additional
       information on the terrestrial risks associated with the Site. EPA solicited further
       review and comment through the remedy selection process when it issued the Proposed
       Plan for public comment. Through the comments raised here and the Agency's
       responses, the remaining ambiguities in the terrestrial risk assessment have been
       addressed.  Through this iterative process, EPA has effectively completed  the risk
       assessment process and provided adequate documentation to support the actions
       required by this ROD.

D.     Non-Aqueous Phase Liquids (NAPL)

1.     The Site Owner contends that since there is no oil floating in the groundwater, there is
       no reason to remediate oil or PCBs. The Site Owner asserts that EPA's use of Photo #4
       of the Proposed Plan, depicting oil in a groundwater sample, is an irresponsible attempt
       to appeal to the publics' emotions rather than logic.

       Response:  EPA has presented analytical data that substantiates the presence of oil and
       PCBs  in the groundwater. Since PCBs cannot be visually  observed, a picture (Photo
       #4) of oil found floating in a groundwater monitoring  well is more descriptive.

2.     The Site Owner contends that even if EPA can prove there is oil in the groundwater,
       the measurements taken of the oil layers in monitoring wells historically and during the
       RI/FS were inaccurate and therefore cannot be relied on to substantiate  the presence of
       oil.  The Site Owner asserts  that actual oil thicknesses are  usually five (5) to ten (10)
       times less than their measured thickness.

       Response:  EPA is proposing to take an action at the Site because the concentrations of
       PCBs  pose a potential risk, not because a certain thickness of oil was detected.

3.     The Site Owner contends that even if EPA can prove there is oil, all historical data
       indicates that no more recoverable oil remains in the groundwater as follows: (a) its
       interpretation of a 1980 EPA-sponsored study (by Weston) that only 1,500 gallons of
       oil would be recovered;  (b)  various technical literature and journals cited oil recovery
       rates as normally between 10 to 20%; (c) its interpretation of oil recovery logs from
       1981 to 1989 (which includes an assumption that protective clothing and rags expended
       on the job contained a substantial amount of oil) suggests that 8,200 gallons of oil have
       been recovered from the  Site. Therefore, referring to another Weston report which
       theorized that 16,000 gallons remained on-site, a 20% recovery of that amount would

                                          103

-------
       be 3,200 gallons.  Therefore, the Site Owner concludes that the amount recovered
       (allegedly 8,200 gallons) is actually more than the amount that is theoretically possible
       (3,200 or 1,500 gallons, as described above).  [NOTE: Although the Site Owner did
       not organize his comments as a unified argument as EPA has done, EPA believes it has
       captured the essence of the comment and the Site Owner's conclusions.)

       Response:  In the Proposed Plan, EPA never attempted to estimate how much oil was
       or could be recovered. Also, EPA is not proposing to install more oil recovery wells
       on-site.  EPA realizes once contaminants such as oil have been released and diluted in
       the environment,  it is impossible to account for every drop.  However, despite the Site
       Owner's assertions concerning recovery  rates, Photo #4, as presented in the Proposed
       Plan, illustrates that oil can still be recovered from certain monitoring wells on-site.

             EPA believes that the Site Owner's estimate that 8,200 gallons of oil has been
       recovered from the Site is speculative because it was obtained by extrapolating
       measurements from the amount of oil that was diluted in groundwater.  As a result,
       each estimate may vary depending on the interpretations and assumptions  used.  Again,
       EPA relies on analytical data to demonstrate current conditions at the Site. The
       presence of oil  has been documented in monitoring well MW6 in 1991. The analytical
       results indicate that sample contains 1,060 ppm of PCBs.

4.     The Site Owner commented that there is no oil left because the oil recovery system
       installed by the Site Owner in 1981  ceased recovering oil by 1989.

       Response: The Site Owner's documentation (see Volume /, Book 10, Section 9.1.1,
       page 14) shows that in 1989, seven gallons of oil had been collected from the Site.
       However, the level of recovery is not relevant to EPA's decision in this document
       because EPA agrees that active recovery is no longer appropriate.  The selected remedy
       requires installation of a passive collection system that will allow the oil still remaining
       in the  soil to be captured and effectively  disposed of as it is gradually released into the
       groundwater.

5.     The Site Owner contends that there is very little oil remaining in the groundwater
       because RI Table 4-30 illustrates only low  concentrations of TPH were detected in the
       1991 groundwater samples. By 1992, only several wells (PW10, MW6 & MW7) had
       detected TPH.

       Response:  EPA does not consider TPH a  true quantitative measurement of oil in a
       sample.  As the definition box in ROD Section V.E.2 explains, TPH is a misleading
       terminology because not all petroleum compounds may be detected by the analysis
       while other non-petroleum compounds may be accidentally included. Nevertheless,
       EPA does not agree that RI Table 4-30 presents  evidence that oil is not present in the
       groundwater.  Table 4-30 only presents TPH results in the groundwater layer; the oil
       layer was never analyzed for TPH.  In addition, two of the 1992 groundwater samples

                                          104

-------
       had actually increased in TPH (i.e., MW6 and MW7 increased from 5.36 and 3.54
       ppm to 7.3 and 12.3 ppm, respectively).  One reason for this increase may have been
       due to the improper well purging action used (to be later explained in Comment #1)
       which resulted in diluting petroleum compounds in the oil layer and moving it into the
       groundwater phase.

6.     The Site Owner commented that any oil left within the Site is trapped in impermeable
       "pockets" and will not be going anywhere.  This condition was stable until EPA drilled
       a well through the  oil pockets and mobilized the PCBs. The reason why certain
       groundwater samples had an oil layer was because when EPA drilled a well through a
       dry region ("unsaturated zone") which contained oil pockets,  the drill pulled the oil
       down to a wet region ("saturated zone") below the water table.  Once in the saturated
       zone, the oil could move freely with groundwater. Since oil is lighter than water, it
       will float up towards the surface of the groundwater  in the wells.  This is where EPA
       makes its mistake by identifying it as an oil layer in the groundwater. PW10 is a
       perfect example of where impermeable oil pockets were encountered because it had a
       very low-yield during the pump tests.  (NOTE:  The Site Owner's argument concerning
       MW5 could not be understood).

       Response: EPA disagrees that contamination detected was due to trapped pockets of
       oil. EPA used proper well development and purging in accordance with EPA approved
       protocols, which would have removed all contaminants that may have been deposited
       during the installation of the wells.

             There were no profiles in the well boring logs,  including PW10 (see
       AR300360), which documented the presence of an impermeable layer (such as clay)
       which trapped the oil in small pockets.  On the contrary, the log  from PW10
       documents petroleum stains and porous fill materials (such as wood and concrete
       fragments) down from the surface, past the water table at 13.4 feet, to a non-
       impermeable layer  at 18  feet.  Nowhere was an impermeable  layer documented.

7.     The Site Owner commented that monitoring wells, especially  MW6, were not correctly
       purged prior to sampling in 1991.  Therefore all 1991 well sampling data, such as the
       1,090 ppm PCB in MW6, should be disregarded.  The samples taken in 1991 were
       actually oil left over from the construction of wells due to trapped oil.  As proof, the
       1992 samples contained very little or no oil because most of the oil had been removed.
       The proper way to  purge or develop a well would have been to aggressively bail,
       surge, or flush a well with air or water until  all  stagnant water has been removed. The
       Site Owner cites an EPA guidance document [EPA, 1987]5 as reference.

       Response: EPA has documentation that,  during the  1992 sampling event (which the
             EPA, 1987. Handbook: Groundwater (EPA/625/6-87/016); March 1987.

                                         105

-------
Site Owner's representative had also videotaped), the wells were not properly purged,
and, therefore, the samples are faulty. EPA has eliminated the 1992 groundwater data
from its risk assessment analysis. As discussed below, the absence of oil in the 1992
samples was the result of unacceptable purging techniques used by the PRP's
replacement contractor (HMM). The purging technique discussed in the cited reference
[EPA, 1987] does not conform to the approved RI workplan [BCM, 1991]6 written by
the PRP's original contractor (BCM) and will result in dilution of PCBs in the sample
collected. The approved RI workplan, to be discussed later, reflects methodology cited
in a more recent EPA guidance document [EPA, 1992]7, which was specifically
designed for sampling a floating oil layer (LNAPLs) in groundwater.

       The 1992 guidance document specifies (on page 7-9) the use of a low yield
pump and tubes held below the water surface for purging wells. The object (as stated
on page 7-8) is to ensure that "migration of water in the formation above the well
screen does not occur." In effect, if a more turbid method was used, oil and PCBs
from the top of the water table would have been drawn down into the wells and
resulted in an unrepresentative sample.  Furthermore, the guidance states: "When
sampling NAPLs, bailers should never be dropped into a well and should be removed
from the well in a manner that causes as little agitation of the sample as possible.  For
example, the bailer should not be removed in a jerky fashion or be  allowed to
continually bang against the well casing as it is raised..." [EPA, 1992 at page 7-13]

       In 1991, the PRP's original contractor (BCM) collected samples of the oil layer
in MW6.  The PCB concentration in one sample was 1,090 ppm. In  1992, the PRP's
replacement contractor (HMM) collected samples that only contained droplets of oil.
The PCB concentration was only 7 ppb.  EPA's  independent analysis of the 1992 split
sample was 183 ppb PCB (see AR303815).  EPA suspects the vast difference between
the 1991 and 1992 sample results may have been due to the different methods of
purging used by the different contractors. And the difference between the 7 and 183
ppb split samples may also have been due to the  different degrees to which PCBs were
dissolved in the groundwater.

       According to EPA's oversight contractor's (SAIC) notes dated June 3, 1992,
HMM collected samples from MW6 (see AR301902). Prior to collection, they had to
purge the well in accordance with the sampling plan.  The original BCM sampling plan
required a tube to be lowered through the oil layer into the bottom of the well and
pump water out at a low rate (< 1 gpm) in order to prevent drawdown of the oil.  Each
well was required to have 3 to 5 well volumes of water removed. On May 18th, after
6      BCM, 1991. RI/FS Field Sampling Plan; BCM Engineers; March 1991.

7      EPA, 1992.  RCRA Ground-water Monitoring: Draft Technical Guidance (EPA/530-R-93-
       001); November 1992.

                                   106

-------
       concluding that purging of the wells was too time consuming (at 0.5 gpm for an
       estimated 3 well volume of 14.5 gallons), HMM verbally requested a deviation by
       using bailers. This field variance was granted by EPA (see AR301876) but was never
       documented by HMM as a deviation from the approved workplan in the RI/FS. EPA
       had previously expressed and transcribed to the PRPs that all deviations from BCM's
       sampling plan must be identified and approved in writing (see AR304397). EPA took
       this position because samples taken using various techniques may result in
       inconsistencies that would be difficult to interpret later (see AR301868).  EPA also
       wanted to maintain consistency between sampling techniques used by two different PRP
       contractors.

             During the purging, SAIC observed that HMM's bailing technique was very
       aggressive and agitated the samples. For example, the bailer used at MW11 hit the
       bottom of the well whereby the sediments were stirred up into the water. The purge
       water drawn up with the bailer was black while previous water drawn up by the pump
       was light brown.  This bailing technique was used for most of the other wells.  This
       technique is unacceptable because the purging may have removed most of the PCBs
       (along with other contaminants) while diluting the oil layer in the wells. In turn,
       samples taken after the purging showed little oil and low PCB levels. This bailing
       technique was applied at MW7 (see AR301901), as well as MW6.

8.     The Site Owner commented that the presence of oil layers in the subsurface soils as
       suggested by EPA in Figure 2 of its Proposed Plan and Figure  1-5 [the correct citation
       should be Figures 2-1 or 3-1] in the Feasibility Study is not supported by the RI data
       for the following  reasons.  Based on a map (Figure 2, see AR302295) in the Proposed
       Plan, the area that EPA  claims to be saturated with oil (NAPL) is not.  Soil can only be
       considered saturated with oil if it is below the water table and its TPH concentrations
       are greater than 10,000 ppm. A table presented by the Site Owner (see Volume I, Book
       3, Table ESE-2, page 9) and several boring locations (B14, 15, 16, 17, 18, 22, and 23)
       were presented to illustrate that samples below the water table did not exceed 10,000
       ppm TPH. Since no soil below the water table exceeds 10,000 ppm TPH, there can be
       no NAPL.

       Response: EPA never suggested that soil in the NAPL Area must be saturated with
       water and oil. The Proposed Plan states on page 2 that "the NAPL Area is an area
       thought to contain residual oil" (see AR302274).  This does not mean an area has to be
       saturated with oil. The Site Owner failed to mention that most of the samples
       identified in Table ESE-2 were not tested below the water table. Therefore his
       statement that there are no "saturated" subsurface soils that exceeded 10,000 ppm TPH
       is unsubstantiated.

             The following table illustrates the locations, depth, and water level for samples
       that contradict the Site Owner's assertion that 10,000 ppm TPH has not been exceeded
       in the NAPL Area.

                                         107

-------
SAMPLE
LOCATION
B14
B15
B16
B17
B18
B22
B23
Max TPHs
Depth
(ppm)
3.0 1B^$«>.-&
8.5 J£i*12,800 -
5.5 ^33!&ryv':v
6C Ba^'r '_- < f9 Jl f\f\ --- ••'•
.5 HpVs47$4UU,'
tai-'-J • • '
8.0 ^15.800
8.5 |f^25;900: ;
Water
Table
(ft)
12.5
11.2
13.25
12.5
14
14
12.5
Table ESE-2 Cited
TPHs
Depth
12.5
(ppm)
4.080
10.5 |#?1 0,800, ~
11.5
12.5
1.416
4.160
ha;'*''" • •"•" - - -
14 ^^10.000
14
10.5
9.770
^13^00
Tested
Below
Water
Table?
No
Yes
No
No
No
No
No
Shading indicates samples where 10,000 ppm TPH was exceeded. Other locations
mentioned by the Site Owner. B16, SB-104, SB-105 were not tested below the Water Table
CWT). SB-101, SB-102. SB-103 were tested once below WT.
9.    The Site Owner commented that EPA's evidence for suggesting DNAPLs are present is
      not valid and EPA should not propose any DNAPL investigation.

      Response:  EPA has not unequivocally  concluded that DNAPLs are present.  EPA
      stated in the Proposed Plan that "the RI/FS provided evidence that there may also be a
      DNAPL" and profiles from borings SB-105 and SB-106 "show PCB levels that remain
      the same or potentially increase with depth below the water table." For that reason, the
      ROD requires additional sampling during Remedial Design to determine whether
      DNAPLs are present or absent.

10    The Site Owner commented that borings SB 105 & SB 106 do not support the possible
      existence of DNAPLs since they do not exhibit oil stains, petroleum or comparable
      odors (i.e., "heavy fish odor") below the water table or deeper (an exact depth is not
      given).  The Site Owner asserts that since the profile of SB106 does not contain any
      stains or odors down to its bottom depth at 16 feet, it should not be considered showing
      signs of DNAPLs.

      Response:  EPA did not rely on physical observations or odors as evidence of possible
      DNAPL. EPA used analytical results from soil collected from these borings which
      suggested PCBs may be increasing with depth.  However, a comparison of the
      observed TPH and PCB values in subsurface soils (see Exhibit 1, attached) illustrates
      that other borings throughout the RI have documented the presence of oil stains and
      odors deep  below the water table.  Many times, these borings (e.g.,  B23) documented
      the presence of oil stains or petroleum odors as deep as 8 feet below the water table.
                                        108

-------
11.   The Site Owner presented an untitled table (see Volume I, Book 1, Section 5.2.2, page
      79 of Site Owner's Comments) to illustrate the deficiencies of borings SB105 & SB 106
      which EPA cited as evidence of DNAPL in the Proposed Plan.  The Site Owner states
      that:  (a) EPA mixed unreliable field screening data with analytical data;  (b) EPA
      failed to cite that SB-106's detection of 10-50 ppm PCB at 12-14' was actually a "field
      test" rather than a "laboratory test" (the laboratory test did not detect any PCBs with a
      detection limit of 2.1 ppm Aroclor 1254); and (c) different aroclors were detected in
      SB105 and SB106 (Aroclors 1260 vs 1254).
                                             BORING PROFILE OF POTENTIAL DNAPLs
Response: (a)  The following
soil boring profile illustrates
that EPA did not rely on field
screening  data to indicate the
potential presence of DNAPLs
in SB105 & SB106.  The
presence of high PCB levels (12
& 15 ppm, as shaded in that
table) at the water table with the
absence of deeper test data
prompted  EPA's concern.
Other locations  such  as SB 102
and TP-T1, as presented and
shaded in  Exhibit  1, also
illustrate that not enough deeper
samples were tested (although
some borings, such as SB 105,
were drilled to as deep as 44
feet) to discount the possibility
that its PCB levels may exceed
the cleanup level of 25 ppm
below the  water table.
      (b)  Figure 3 of the Proposed Plan depicting sample locations and results for highest
      and lowest observed PCB contamination (see AR 302297) reminds the reader at note #4
      to "note that some of the data are screening results (data in italics); these results may
      not be comparable to analytical data."  The field test data in SB106 were also presented
      in italics.  Again, Exhibit 1 illustrates the presence of high PCB level (as shaded) at the
      water table followed by no tests below that depth.  Exhibit 1 (note #5) also cited that
      problematic detection limits as high as 11 ppm (see AR300603) were reported for non-
      detected PCB samples at the 10 feet depth of SB106.

      (c) As discussed in response to Comment #18 below, Aroclors are not an acceptable
      form of PCB finger printing.
Location
SB105




SB106




Reference:
Depth
(feet)
8.0
10.0
12.0
14.0
16.0
6.0
8.0
10.0
12.0
14.0
Exhibit 1.
PCB
(ppm)
(10-50)
(10-50)
3.7
pjjiid '
||^.:12.0 • •
1.8
ND
5.5
ND (10-50)
f£ 'I; •"'''I 5.0 '
attached. ND =
Profile
(feet)

stain @ 14-1 8
water @ 14.0

(Bottom 44.0)

no stains

water @ 14.0
(Bottom 16.0)
Not Detected.
                                         109

-------
             As stated previously, EPA never unequivocally concluded that DNAPLs are
       present.  The investigation during Remedial Design will attempt to determine if PCBs
       are actually increasing with depth.

12.    The Site Owner commented that given the measured concentrations of PCBs and the
       specific gravity of oil (0.8) found in subsurface soils at the Site, the oil would not be
       dense enough to form a DNAPL.  PCBs must be detected at 570,000 ppm in the
       subsurface soil for DNAPL conditions to exist, and the PCB levels in SB105 pale in
       comparison to the 570,000 ppm theoretical values. The  570,000 ppm PCB is derived
       from an algebraic equation which assumes the sum of the weights of oil and PCBs will
       be equal to the weight of DNAPL PCBs found on-site: A y0ii + B YPCBS = (A+B)
       YDNAPL; where "y" is density.

       Response:  EPA considers this equation to be theoretical and not representative of
       actual field conditions.

13.    The Site Owner presents a series of technical publications that discuss how oil is
       trapped below the water table (in soil pores) and how it can "smear" against the soil
       walls. This occurs when tides from the Delaware River fluctuate. EPA has mistaken
       the residual oil smears that are left in the lower depth for DNAPLs in the groundwater.

       Response:  EPA reviewed these technical publications. The conclusion that there are
       trapped oil pockets and that they can smear against previously uncontaminated soils has
       prompted new concerns for EPA.  Since the Site is located in a tidal and flood zone
       area, the water level fluctuations caused by these movements can liberate and move oil
       both in a horizontal and vertical direction.  The result is  that PCBs are mobilized with
       the oil and may contaminate a clean surface soil cover, as well as being released into
       the river.

14.    The Site Owner commented that EPA is fixated on the buried underground storage tank
       ("USX") and the belief that the tank is a source of contamination while the evidence
       supports the opposite conclusion.  The Site Owner provides testimony from former Site
       employees and a former PADEP employee as evidence that the tank has been cleaned.
       The Site Owner states that the UST location on Figure 2 is wrong.

       Response: The testimony provided were general observations from various parties that
       the tank did not appear to contain oil or that it was filled with concrete.  EPA does not
       rely on testimony alone to conclude whether an area is a source of contamination or
       not, especially where analytical data has been collected.  Samples collected around the
       tank location (i.e., MW6) have shown high levels of PCBs. The PRPs did not
       investigate the tank during the RI to determine whether or not the tank is clean.
       Therefore, the ROD requires that the tank, its contents and its surrounding soil be
       tested and remediated if its PCB levels exceed 25 ppm, consistent with the rest of the
       Southern Portion of the property.

                                         110

-------
             Figure 2 states that "all locations are approximate" and references RI Figure 2-8
      as the suspected location of the UST. EPA cannot change Figure 2 because the exact
      UST location is not known and the comment did not provide a corrected location.

15.    The Site Owner provided various publications that stated transformers disposed
      between 1968 and 1973 generally contained  mineral oil rather than PCBs.  The mineral
      oil, in turn, contained trichlorobenzene. Since none or very little trichlorobenzene was
      detected in the RI environmental samples, this could only mean that no oil from the
      original transformers had spread into the environment.

      Response: There is no documentation, much less actual samples, of the oil contained
      in historical transformers to determine if trichlorobenzene was a component of the oil.
      Historical sampling results, however, support EPA's conclusion that transformers
      drained at the  Site contained PCBs because sampling of the oil spills on the ground
      detected high levels of PCBs.

             Furthermore, according to the PCB regulations at 40 C.F.R. § 761.3
      (Definitions) published under Section 6(e) of the Toxic Substances Control Act
      CTSCA"), 15 U.S.C. Section 2605(e) and effective July 2, 1979, the definition for
      PCB-Contaminated Electrical Equipment includes the statement that "Oil-filled [i.e.,
      mineral oil] electrical equipment.  . . must be assumed to be PCB-Contaminated
      Electrical Equipment. The preamble elaborates on this definition as follows:

             Because of the widespread PCB contamination of transformers that were
             designed to use PCB-free mineral oil dielectric fluid, all such mineral oil
             dielectric fluid transformers must be assumed to be PCB-Contaminated
             Transformers, unless reasons exist to believe that a transformer was filled with
             greater than 500 ppm PCB fluid (in which case the assumption is that the
             transformer is a PCB Transformer).

      (See, Preamble to the final rule for Polychlorinated Biphenyls (PCBs) Manufacturing,
      Processing, Distribution in Commerce, and Use Prohibitions ("preamble"), 44 Fed.
      Reg.  31514 (May 31, 1979) at p. 31517.) According to the regulations, PCB-
      Contaminated  electrical equipment means any electrical equipment, including
      transformers, that contains 50 parts per million (ppm) or greater PCB, but less than
      500 ppm PCB. A PCB Transformer is any transformer that contains 500 ppm PCB or
      greater.  40 C.F.R. §761.3. In other words, all transformers, including those designed
      to be mineral oil transformers will be assumed to contain at least 50 ppm PCB.
      Additionally, according to the regulations, all transformers are presumed to contain
      PCBs unless specifically marked and certified otherwise. 40 C.F.R. §761.3.

             The trichlorobenzene data in the RI is not useful because only a few samples
      were analyzed for 1,2,4 trichlorobenzene. Of the typical samples analyzed, the oil
      phase of MW6P reported that trichlorobenzene was non-detectable at an extremely high

                                         111

-------
      detection level of 750 ppm (see AR300520).

16.   The Site Owner commented that even if the Site is a source of PCBs in the
      groundwater, there is no migration off-site because after the 1981-1989 remediation
      effort, PCB levels in the groundwater have been lowered to less than the 1983
      Stipulation performance standard of 200 ppb (barring the MW6 data).  Also by
      applying several conversion factors, the Site Owner concludes that less than 0.8 Ibs of
      PCBs are released from the Site every year, thereby negating any need for a
      containment wall to keep the oil and PCBs in.

      Response:  EPA disagrees that all PCBs within the Site have been contained when the
      highest level of the most recent PCB sampling (19.6 ppm) was observed immediately
      on the shorelines  (MF107), compared to less than 2 ppm at distances 300 feet into the
      river (S22). Regardless whether there is an annual theoretical PCB release of 0.8 Ibs,
      the analytical evidence still suggests that PCB contamination in the groundwater has not
      decreased much since 1977 (1,539 ppm in 1977 vs. 1,090 ppm in  1991).

17.   The Site Owner commented that despite EPA's best efforts, no oil seeps could be
      located during the RI or historically for sampling.  The sheen in Photo #2 of the
      Proposed Plan contains no PCBs.  EPA is also confusing residual bank oil sheens with
      active oil seeps.  Where EPA has identified contamination in the sediments, EPA now
      hypothesizes these occurrences are  due to "subtidal seeps" which is impossible because
      oil (specific gravity of 0.8) is lighter than water (specific gravity of 1.0) and therefore
      can only be found on the surface layers of the sediments.

      Response:  In the course of history of the on-site activity at the Site  (from the 1970's
      to 1990's), only a few days were spent looking and sampling for oil  seeps in the field
      (in 1990 and 1991).  One reason why oil-oozing seeps could not always be found was
      because tides from the river would  cover the shores daily. As a result, the shorelines
      would be exposed only for a short period.  However, oil sheens were observed on the
      river each time.  Of the times oil sheens were sampled  around the Site, the data
      demonstrate that they  contained high levels of PCBs. For example: (a) in 1977, 1,071
      ppm was detected in an oil sheen along the shoreline (see AR100034); and (b) in 1990,
      210 ppb was detected in water that  refilled a hole after  sediment containing 5.2 ppm
      was collected (see AR301796).

             The  water sample in Photo #2 of the Proposed Plan was collected because there
      was not enough oil sheen in the sample for analysis.  While these historical data were
      not included in the RI data set or used to evaluate various risk assessments, they do
      demonstrate that PCB contamination in oil emanating around the Site is extremely high.

             Despite the PRPs' failed efforts to locate actual  seeps in order to guide the RI
      sampling, sediments data demonstrates that PCBs have  contaminated both the surface
      and lower (or subtidal) layers.  One reason for this dispersion of PCBs may be because

                                          112

-------
       PCBs in the oil are being dissolved by the constant fluctuation and movement of
       groundwater caused by the river tides.  In any case, the analytical data show that PCB
       contamination is the highest along the property shorelines (i.e., 19.6 ppm at MF107);
       and PCB contamination has occurred on the surface as well as the deeper levels of the
       shoreline sediments (i.e., 6.5, 9.1, and 14 ppm at 0.5, 1, and 1.5 feet, respectively, at
       RR3).

             As to the argument that the seeps may be biological or due to off-site sources,
       there is no analytical data provided to support either suggestion.  The highest off-site
       PCB contamination in sediments suggested by the Site Owner was only 1.57 ppm (see
       AR303383).  In addition to not containing an exact location, this  alleged value is also
       minor compared to the highest PCB  (19.6 ppm) found adjacent to the Site's shorelines.

18.    The Site Owner commented that PCB Aroclors from the Site did  not match those found
       in the sediments.  Therefore oil on-site could not have been migrating into the river.
       For example, PCBs in most soil samples on the Site contained Aroclor 1260.  Only one
       of the 20 (1/20) sediment samples collected in 1991 around the Mudflat Area contained
       Aroclor  1260.  Meanwhile only one of the  14  (1/14) samples collected in the Delaware
       River Sediment Area contained Aroclor 1260.

       Response:  Aroclors cannot be used as a definite form of fingerprinting PCBs to
       determine the migration of PCBs from a particular source.  EPA  also notes that the
       most contaminated region was along the shoreline in the Riprap Area and the detectable
       PCBs there all contain Aroclor 1260.

             The following table is presented to illustrate how Aroclor  1260 was found in the
       most contaminated subsurface soil on-site (B17, TP-T2) as well as the most
       contaminated sediment samples off-site (MF107, RR3, DR8).
Highest PCBs levels observed in soils & sediments (ppm)

Aroclor 1248
Aroclor 1260
Total PCBs
TP-T1
(Soils)
ND (<30)
230
230
B17
(Soils)
ND (<5.3)
42
42
MF107
(Mudflat)
4.6
15
19.6
RR3C
(RipRap)
ND(<1.9)
14
14
DR8
(River)
4.6
2.2
6.8
19.    The Site Owner commented that the 10,000 ppm TPH cleanup level calculated in
      Appendix B of the FS is wrong.  The FS made an incorrect assumption as to the
      density of oil, and soil and oil retention values, while more realistic values would have
      yielded a 20,000 ppm cleanup level. However, using assumptions developed by the
      Site Owner's consultants, a more reasonable cleanup level of 30,000 to 100,000 ppm

                                         113

-------
       would result. The Site Owner further emphasized that EPA admitted its TPH cleanup
       level was flawed because there was no correlation between TPH and PCB values.

       Response:  The wide range of possible TPH cleanup levels was due to the equally
       diverse scenarios posed by the different consultants.  Although most of the assumptions
       made may hold true under laboratory conditions, they do not represent how PCBs may
       behave under actual field conditions. The Site Owner, for example, assumes: (a) oil
       residual saturation between 15 to 20%; (b) soil porosity between 2$ to 45%; and (c)
       soil dry density between 1.5 to 1.9 kg/L to obtain TPH cleanup levels ranging from
       15,789 to 42,353 ppm. EPA agrees that the TPH cleanup level developed by the PRPs
       to clean up PCBs is equally speculative and may be flawed. Therefore, EPA has
       eliminated it from the ROD (see response to PRP's Comment #5) and now requires
       cleanup of PCBs based on PCB values alone.

20.    The Site Owner states that EPA's Proposed Plan erroneously stated that TPH
       contamination up to 25,000 ppm in the soils was observed. According to Table 4-11,
       page 4-49 of the RI, the maximum TPH value was only 17,400 ppm.

       Response:  Table 4-11 hi Volume 1 of the RI  is  incorrect.  Backup data contained in
       Appendix C.I.b of RI Volume 2 (see AR300462) shows the highest TPH value to be
       25,900 ppm at soil boring B23 (8.5 feet depth).  Additional sampling by the PRPs
       during the public comment period yielded a higher TPH value of 66,100 ppm (see
       AR303891) in testpit sample TP-T2 (5.0 feet depth).

E.     Remedial Alternatives

1.      The Site Owner commented that soil washing is not technically and economically
       feasible at an urban fill site such as Metal Bank.

       Response:  Soil  washing is a proven technology  that has been used at other Superfund
       sites to remove PCB contamination from soil.  The materials handling issues associated
       with the nature of the fill material at the Metal Bank Site would be addressed during
       the remedial design process. A treatability study would be required to ensure that soil
       washing is feasible given the site-specific conditions.  Although the nature of the fill
       material at the Site may require additional materials handling measures, this concern
       does not render this technology infeasible for purposes of evaluating alternatives.
       However, EPA did not choose this technology because of its higher cost ($31.4
       million) compared to that of the proposed remedy ($17.2 million).

2.      The Site Owner commented that in-situ solidification/stabilization is neither technically
       nor economically feasible at the Metal Bank Site primarily because of the  impossibility
       of the use of the mixer in urban fill material.

       Response:  As with soil washing, in-situ solidification is a proven technology that has

                                         114

-------
       been used at other Superfund sites to treat contaminated soil. The materials handling
       issues associated with the urban fill at the Site would be addressed during the remedial
       design process and is not a sufficient basis for rendering this technology infeasible for
       purposes of evaluating alternatives.  A treatability study would be required to ensure
       that this treatment process works effectively to stabilize the PCB contamination in the
       fill matrix at the Site.  EPA has reviewed a preliminary treatability study submitted by
       the PRPs (see response to PRP's Comment #20) and concludes this technology would
       not be technically or economically advantageous based on the available information.

3.     The Site Owner commented that concerns associated with installation of the sheet pile
       wall along the southern and western borders of the Metal Bank property, including
       cost, time delays, and overall ability to obtain the permits necessary to install the sheet
       pile wall, dewater behind the pile wall, and discharge dewatering liquid back into the
       river, are not adequately considered.

       Response:  EPA recognizes that installation of the sheet pile wall will require careful
       design to ensure effective implementation. Use of sheet piling to prevent migration of
       Site contaminants into the Delaware River and to enable excavation of contaminated
       sediments is a feasible technology.  As to time delays associated with permitting,
       cleanup actions under Superfund are only required to meet the substantive requirements
       associated with the permitting process. On-site actions do not require completion of
       the administrative permitting process.

4.     The Site Owner commented that the conceptual design for the LNAPL collection
       system would not effectively collect floating product.

       Response:  The standards and specifications for design of a passive  system to collect
       PCB contaminated oil floating on the surface of the water table will be determined
       during the remedial design process.  The technology to implement a collection system
       of this nature is readily available and can be tailored to site-specific conditions. The
       conceptual model presented in the Feasibility Study may be altered during the remedial
       design process as additional site-specific information is obtained and  evaluated. The
       purpose of the conceptual model is simply to provide a general framework for this
       component of the remedy.

5.     The Site Owner commented that use of an impermeable cap over the  Southern Portion
       of the Metal Bank property is not necessary where excavated sediments are backfilled
       and maintenance of the existing permeable cap over the remaining area would be
       adequate.

       Response:  EPA agrees that an impermeable cap is not necessary at the Site if soils
       with elevated PCB concentrations (i.e., > 25 ppm in the Southern Portion of the Site
       and > 10 ppm in the Courtyard) are  removed.  The remedy selected in the ROD only
       requires that a soil cover be placed over the excavated areas.

                                           115

-------
6.     The Site Owner commented that probable recontamination of sediment in the tidal
       mudflat by PCBs from other off-site sources makes any remedial measures for the
       sediment ineffective.

       Response:  Data collected during the Remedial Investigation indicate that PCB
       concentrations in sediment decrease with distance from the Site. The mean PCB
       concentration in the farfield samples (samples  > 30 meters from the Metal Bank
       property) is 0.87 ppm which is below the PCB sediment cleanup level of 1.0 ppm.
       Therefore, recontamination above the 1.0 ppm cleanup level is not expected.

7.     The Site Owner recommends a Limited Further Action Alternative including the
       following components as an appropriate remedy for the Site:  property access
       restrictions formalizing existing conditions; verification of closure of the existing
       underground storage tank, formalize the existing institutional controls, maintenance of
       existing cover material, re-vegetation and maintenance of vegetation along the
       riprapped side slopes, and monitoring to confirm absence of LNAPL.

       Response:  The Limited Further Action Alternative proposed by the Site Owner would
       not be protective of human health and the environment because aquatic organisms
       would continue to be exposed to contaminated sediments and high levels of PCB found
       in the oil layer floating  on the water table could continue to act as a source of PCB
       contamination to the aquatic environment. This alternative also relies solely on access
       restrictions to prevent exposure to PCBs in the Courtyard surface soil. Because this
       alternative does not meet the threshold criterion of overall protection of human health
       and the environment, EPA cannot select this alternative as a remedy for the Site.

8.     The Site Owner also provides a second alternative which includes the components listed
       in the previous comment along with the following:  additional sampling followed by
       excavation and off-site disposal, if required, of the contaminated surface soils in the
       Courtyard Area; and additional sampling of the sediment of the mudflat areas followed
       by covering mudflat areas shown to require remediation with two feet of clean fill.

       Response:  While this alternative addresses potential exposure to the Courtyard surface
       soil, covering the mudflat habitat with two feet of clean fill is not an effective approach
       to providing overall protection of aquatic receptors.  The placement of two additional
       feet of fill material into the mudflat would itself be an adverse impact to this ecological
       resource.  In addition, the long-term effectiveness of this approach would rely on
       effective maintenance of this two foot cover.  While this alternative would be easier to
       implement and less costly, EPA's selected remedy requiring excavation of PCB-
       contaminated sediments provides a more permanent solution as preferred by the NCP.

9.     The Site Owner commented that the cost estimates for the alternatives considered are
       generally not explained, poorly justified, and overly conservative.  For example, the
       Site Owner asserts that  the use of disposal costs of $475 per ton is twice the typical

                                          116

-------
       hazardous waste landfilling cost.

       Response: EPA reevaluated the cost estimates for the alternatives based on a more
       realistic soil excavation volume and, in the case of off-site disposal, a more realistic
       disposal cost.  These revised costs are reflected in the ROD and supporting
       documentation for these revised estimates is provided in the Administrative Record.
       (see AR304198)

F.     Legal and Historical Factual Issues

       The Site Owner provided 40 pages of comments and 11 volumes of attachments
detailing its view of the history of the Site from the period of transformer recycling through
the present.  These comments are available for review in the Administrative Record for the
Site. The volumes of attachments are available for review upon request from EPA. Because
this comment is a continuous narrative, EPA is not responding in this document to every
statement, since much of it does not specifically address the remedy presented in the Proposed
Plan.  EPA has corrected the ROD where specific factual statements were in error as shown
by these comments and the references provided.  EPA does not generally consider it
appropriate to debate  points of disagreement with the Site Owner regarding historical events,
particularly with regard to negotiations between EPA and the Site Owner, Metal Bank.
However, a number of general concerns and issues were  gleaned from these comments, to
which EPA will respond below.

1.      The  Site Owner summarizes the ownership history of the Site from October 1962 to the
       present. (See  AR303675-303676)

       Response: EPA reviewed this information and has incorporated it into the ROD
       (Section n.A.  Site History), with the exception of one item.  The Site Owner stated
       that, "beginning with its incorporation in  1968, the 'new' Metal Bank leased both the
       Cottman and the New State Road properties until  1980." (See AR303676) EPA has a
       copy of a lease agreement  for the Cottman Avenue Site beginning June 1,  1972 for a
       period of 10 years between Union Corporation and Irvin G. and John B. Schorsch, the
       owners of the  Site from March 1969 until  September 30, 1980.  Although Metal Bank
       operated at the Site during this time period, the ROD reflects the fact that  Union
       Corporation held the lease.

2.      The  Site Owner provides a narrative of the history of recycling operations from 1968
       through July 1973.  (See AR303676-303679) EPA does not dispute this narrative,
       generally, but  notes as follows:

       (a) The Site Owner states that Metal Bank believed that the transformers contained
       plain mineral oil and not PCBs,  and that this belief was supported by a 1972 study
       performed by the Franklin Institute and by the fact that the recovered oil could be used
       as fuel for lead smelting at State Road, which would not be expected if the oil were

                                         117

-------
       contaminated with fire-resistant PCBs. (See AR303677)

       Response: While the 1972 Franklin Institute study concludes that "the composition of
       dielectric oils collected by Metal Bank from September 1971  through January 1972 was
       that of hydrocarbon mineral oils," and that "no sulfur, chlorine, or fluorine containing
       compounds were found," EPA notes that no PCBs were detected in samples of oil
       collected from Metal Bank by the United States Coast Guard  ("USCG") and analyzed
       during approximately the same time period (1972). However, when EPA retested the
       samples collected in 1972 by USCG using new procedures in 1977, PCBs at
       concentrations over 800 ppm were detected. The Site Owner also states that in 1977
       Metal Bank realized that the analytical methodology used in 1972 by the Franklin
       Institute to test for PCBs was outdated and inaccurate.  (See AR303680)

             In addition, even transformers designed to be mineral oil transformers are
       assumed to contain PCBs according to TSCA regulations. (See discussion of
       regulations in response to Comment #15 in Section D (NAPL), above.)

       (b)  The Site Owner noted that the suggestion in the RI that transformers were
       delivered to the Site by rail and handled in Buildings 2 and 7 is unsubstantiated by any
       witness testimony or depositions, and the testimony does not  indicate that any
       transformer activity occurred anywhere other than in the vicinity of the pad.  (See
       AR303677)

       Response: The statement in the RI and ROD that transformers may have been
       delivered by rail to Buildings 2 and 7 is based on the presence of high levels of PCB
       contamination in these areas and around the railroad track in Building 7. However, the
       PCBs in the area may also have come from the railroad transformers.

       (c)  The Site Owner states that "between 1968 and 1972, copper wire was burned to
       remove insulation at the Site in 'sputnik' type furnaces which were fired by natural
       gas."   (See AR303678) The Site Owner references exhibits accompanying the
       comments  [Vol. n, Bk. 8, Exh. 3(b), (e)].

       Response: EPA was unable to locate the attachments as specified, but believes the
       reference is to two letters from Site Owner's counsel dated May 6, 1987 and July 24,
       1987 [Vol. D, Bk. 9, Exh. 3(b), (c)]. In April of 1997, EPA conducted an interview of
       the former president and vice president of Metal Bank, both of whom had long-
       standing careers with the company and had worked on-site during the period in
       question.  These gentlemen did not clearly recall any burning of copper wire at the
       Site. EPA would appreciate any documentary evidence to support Counsel's
       statements.

3.     The Site Owner provides a narrative of the enforcement history at the Site beginning
       with the U.S. Coast Guard in August 1972 through September 1995.   (See AR303678-

                                         118

-------
303714) EPA does not contest the general factual accuracy of this narrative, with a
few exceptions, but certain specific issues are discussed below. EPA does not
necessarily agree with the Site Owner's characterization of the facts and the opinions
and motives imputed to EPA.

(a) The Site Owner's narrative summary of negotiations surrounding the United States'
1980 lawsuit under the Resource Conservation and Recovery Act ("RCRA"), the Toxic
Substances Control Act ("TSCA") and the Refuse Act, and the 1983 Stipulation (see
AR303682-303691) appears to be an effort to demonstrate that Metal Bank has done all
that was required under applicable statutes and as a result of these negotiations, and
that the EPA is not justified or authorized to require any further action.

Response: The Site Owner has previously petitioned the Court to enforce the
"settlement" reached in the 1983 Stipulation. EPA has argued that the Stipulation was
not intended to be a settlement of all claims brought in the 1980 lawsuit, since (1) there
is no language in the Stipulation stating that it is a settlement of all claims, and (2) the
Stipulation itself states that "Plaintiff agrees that the pending action shall be placed in
the suspense docket of the District Court." (See Volume n, Book 10, Exhibit 1,
AR303749A-Y, at  AR303749B)  The Court denied Defendant's Motion to Enforce
Settlement, stating  that the effect of the Stipulation was not to remove the case from the
Court's suspense docket,  and that until the Plaintiff (EPA) attempts to require any
further action on the part of the Defendant (Site Owner), the case will remain on the
suspense docket.

       The current action being taken under the authority of CERCLA is based on
sampling data retrieved subsequent to the shutdown of the oil recovery system which
was installed as a result of the  1980 litigation. This action is not based on the prior
historical evidence  but rather on subsequent analytical data.  Regardless of whether or
not the prior suit under RCRA, TSCA and the Refuse Act was settled, EPA is not
precluded from taking action under CERCLA to protect human health and the
environment from threatened releases of hazardous substances. The 1983 Stipulation
between the United States and Metal Bank and Union Corporation, its parent
corporation, required Metal Bank to operate the oil recovery system installed at the Site
until all "recoverable oil" was  obtained. "Recoverable oil" is defined as oil "which
can be recovered through use of oil-water separation technology when brought to the
surface" (see AR303749D). Even though Metal Bank complied with the requirements
of the Stipulation, that does not mean that no more hazardous substances,  including oil
bearing PCBs, are  present at the Site. It only means that using the technology applied
(and not even stretching that to the limit, e.g., by drilling alternative wells),
Defendants collected as much oil as they could.  EPA has the option to consider any
other technologies  that may work to remove the threat posed by the remaining
hazardous substances on the Site.

(b) The Site Owner states that "low level contamination of sediments  in the intertidal

                                   119

-------
       area had been observed by EPA as early as 1977 and the 1978 Weston Study."  (See
       AR303684)

       Response:  EPA reviewed the results of PCB analyses from samples taken in 1977 as
       provided in the 1978 Weston Study (see AR100034). While the majority of the
       sediment samples at various depths had PCB concentrations ranging from 4.3 to 22.1
       ppm, one sample of mudflat sediments taken 9/21/77 had a PCB concentration of 619
       ppm.  More recent sampling during the RI/FS showed lower PCB concentrations in the
       sediments; the highest level  of 19.6 ppm was found in the mudflats.  Nonetheless, these
       lower levels still resulted in Hazard Quotients for aquatic receptors far in excess of 1
       (the number at which no adverse impacts are expected to occur).  Although the
       numbers appear to be low, the risks associated with exposure  in the mudflat and river
       sediments, which are feeding grounds for many organisms, are great.

       (c)  The Site Owner recites negotiations history to demonstrate that EPA agreed that
       the actions prescribed in the Stipulation would adequately clean up the Site and that
       such actions represented a remedy that was "fully protective of human health and the
       environment." (See AR303689)

       Response:  These negotiations occurred prior to the sampling performed in the 1990's,
       which indicated that there was still contamination on the Site at levels of concern.

       (d)  The Site Owner notes that the 1983 Stipulation established 200 parts per billion
       ("ppb") PCBs as the standard not to be exceeded in groundwater.  (See AR303688)
       The Site Owner states that EPA confirmed that all recoverable oil had been removed
       and that PCBs in groundwater did not exceed 200 ppb in a letter dated September 16,
       1989.  (See AR303703)

       Response:  EPA's 9/16/89 letter does not specifically state that the PCBs in
       groundwater do not exceed 200 ppb; however, it does state that "EPA believes that
       Metal Bank has performed all of its duties and obligations under the Stipulation,
       because the system installed in 1981 is not capable of recovering any more oil." (See
       Volume E, Book 2, Exhibit 19, AR303728A)

             The highest PCB concentration detected in the "water phase" of the
       groundwater was 25.6 ppb.8  The highest PCB concentration in the "oil phase" of the
       groundwater was 1,090,000 ppb.  EPA believes that the 200 ppb standard provided in
       the Stipulation refers to the PCB concentration in the water phase. However,
       regardless of whether the 200 ppb standard applies to the water phase or the oil phase,
       8     The Aquatic Risk Assessment did not include a higher split sample result (183 ppb) in its
evaluation. The elevated PCB concentration in this sample may have been due to improper purging and
was discussed in Comment #7 of Section D: NAPL, above).

                                         120

-------
EPA is proposing to take an action at the Site because the concentrations of PCBs in
off-site sediments (i.e., 19,600 ppb) posed an unacceptable risk to aquatic organisms,
not because of the levels of PCBs in groundwater (see ROD Section VLB: Aquatic
Ecological Risk Assessment, Sediment Risks).  EPA is proposing to address the oil
phase of the groundwater through an oil interception and collection system, because its
high levels are the continuing source of off-site sediment contamination.   (See also
response to Comment #16, Section D: NAPL).

(e) The Site Owner states that, when no comments were received on the Stipulation,
the Department of Justice (DOJ) submitted the executed Stipulation "ending the ...
litigation." (S«? AR303689)

Response: See Response to 3A,  above. It is EPA's position that the Stipulation did
not purport to settle all claims brought by  the lawsuit. In any case, subsequent action
under CERCLA is not precluded by the previous suit, regardless of whether or not the
previous suit is concluded.

(f)  The Site Owner states that St. Vincent's is not a permanent residence for "orphans"
as stated in the Proposed Plan. According to Tim Paul, Director of Social Services,
the institution is actually a temporary (maximum 90 days) shelter for abused children
from ages 2 to 12 pending placement.

Response: EPA representatives also spoke with Mr. Paul and were informed that St.
Vincent's  is a temporary shelter for abused children and a day care center. (The
children benefitting from the shelter are "full-time temporary", while the children in
the day care are "part-time permanent".)  In addition, there are a number permanent
residents at the  school, including the grounds keepers and their families and ten of the
Sisters who care for the children. The ROD has been corrected to reflect this
information.

(g) The Site Owner cites a number of instances where it was excluded from meetings,
despite its efforts to cooperate with the PRP Group and EPA.  The Site Owner states:
"Despite Metal Bank's cooperation and its expressed desire to participate in the RI/FS
process to the extent consistent with its legal position, its representatives were
generally excluded from meetings between the Agency and the PRP Group."  (See
AR303696)

Response: EPA sent Special Notice letters to PRPs at the Site, inviting those parties to
negotiate to perform the RI/FS at the Site. Once the Site Owner declined to participate
in these negotiations, EPA was not  obligated to ensure its inclusion in all discussions
with those PRPs who agreed to perform the RI/FS.  EPA does not negotiate
performance of an RI/FS with any party who has not submitted a good faith offer to
undertake  or finance the RI/FS pursuant to section 122(e)(2)(B) of CERCLA, 42
U.S.C. § 9622(e)(2)(B). The Site Owner could not expect to influence the RI/FS

                                   121

-------
process and decision-making without having agreed to bear responsibility to perform or
finance the RI/FS.

(h) The Site Owner describes difficulties in the access negotiations with the PRP
Group and then states that, " EPA began working in concert with the PRP Group to
deprive Metal Bank of its bargaining position to provide [Site access and the Versar
data, including the resulting endangerment assessment] to the Group on appropriate
negotiated terms between private parties."  (See AR303698; see also AR303704)

Response: For purposes of determining the need for response, or choosing or taking a
response action,  or otherwise enforcing the provisions of CERCLA, EPA has the
authority to enter any facility in accordance with Section 104(e) of CERCLA, 42
U.S.C. § 9604(e), and the regulations at 40 C.F.R. § 300.400(d). EPA may, for the
purpose of access, designate  as its representative one or  more PRPs and their
contractors. 40 C.F.R. § 300.400(d)(3).  If access is not granted, or is conditioned in
any manner, EPA may issue  an order directing compliance with the request for access.
40 C.F.R.  § 300.400(d)(4)(I).  EPA typically includes a Paragraph in Consent
Agreements to perform work which states that, if the Respondents fail to obtain access
to perform the required activity onsite after reasonable efforts, then EPA will obtain
such access. In this case, the determination was made prior to execution of the
Administrative Order by Consent for the RI/FS that the PRP Group (Respondents) had
exercised all reasonable and best efforts to obtain site access without success.  EPA
negotiated an Access and Use Agreement with the Site Owner to permit EPA and its
authorized  representatives, the PRP Group, to perform the RI/FS at the Site.  It is in
the Agency's interest to get all  relevant data available on the Site. The EPA was not
favoring the PRP Group; it was just attempting to advance the remedial process to
protect public health and the  environment according to its mandate.

(i) The Site Owner stated that the United States' motion to alter substantive terms of
the Stipulation, to enlarge the time for EPA to complete its evaluation [of the oil
recovery system], and to enjoin sealing and closure of the wells as required by the
Stipulation, was denied by the Court as moot on June 12, 1989.  The reason this
occurred was that the Site Owner had proposed a form of order allowing for an
enlargement of time for EPA to complete its sampling and inspection of the oil
recovery system, and extending the time for Metal Bank to close the wells within a
specified time thereafter.  (See AR303701-303702)

Response:  EPA has  corrected the chronology of Site Enforcement Activities in the
ROD, Section II.  B., to reflect this sequence of events.

(j) The Site Owner states that implementation of the first Field Sampling Plan under
the Work Plan occurred between June 1991 and November 1991, and that the results of
these sampling activities confirmed the effectiveness of Metal Bank's oil recovery
action.  (See AR303708)

                                   122

-------
Response: During the 1991 sampling, levels as high as 1,090 ppm PCB in MW6 were
obtained.  This result does not confirm the effectiveness of Metal Bank's oil recovery
action.  (Site Owner commented that this well was not correctly purged prior to this
sampling event.  See Comment #7 and Response in Section D. (NAPL), above.)

(k) The Site Owner states that EPA denied its requests for an extension to the
comment period on the Proposed Plan beyond the 60 days granted, including an initial
30 day period, plus the minimum extension of 30 days.  Then, several days prior to the
conclusion of the 60 days, EPA extended the comment period for another 30 days when
Metal Bank's representatives had already hurriedly prepared comments to meet the
deadline.  The Site Owner states that this final extension was only posturing to
"minimize the consequences of [EPA's] arbitrary decision to deny the request in the
first instance," and that it showed that the earlier denial was groundless. (See
AR303711-303712)

Response: According to the NCP, EPA is only required to provide a single extension
of 30 days to the initial 30-day comment period on a proposed plan (see 40 C.F.R. §
300.430(f)(3)(C)). Although the NCP leaves open the possibility for longer extensions
to the comment period, in the  interest of expediting cleanups, Region ni very rarely
provides more than the minimum 30 day extension except in cases where the proposed
remedy is very complex.  In this case, the remedy is fairly straight forward, and
requires the use of proven technologies.

       Both the Site Owner and the PRP Group  had initially asked for longer
extensions, and both requests had initially been denied.  The PRP Group,  who had
performed the RI/FS, had proposed to perform an additional treatability study on the
soil, and when it became apparent that all of the results would not be available prior to
the end of the comment period, EPA provided an additional 30 day extension to the
comment period (for a total of 90 days) in order to permit the submittal of this new
information. The Site Owner  had requested an extension commensurate with the
approximately 8 months EPA took to review the RI/FS prior to issuance of the
Proposed Plan. Such an extension was not necessary or appropriate, since the RI/FS
was available to the Site Owner during this same period for review, and most of the
Site Owner's comments are with regard  to the RI/FS and historical data which had been
available to it prior to the release of the proposed plan for comment.

0)  The Site Owner states that it granted Site access to EPA and its representatives with
the understanding that Metal Bank would be provided with all information developed
during the RI/FS. Despite this understanding, the Site Owner states that it had to make
repeated requests for documents/information concerning the Site  exchanged between
EPA and the PRP Group or its consultants.  The Site Owner states that EPA failed to
abide by its agreement and its  statutory obligations under the  Freedom of Information
Act CFOIA").  (See AR303712-303713)
                                   123

-------
       Response:  Under CERCLA, EPA's right of access to the Site is not conditioned on
       any agreements to provide documents to the Site Owner (see 42 U.S.C. § 9604(e)(2)).
       While EPA may have agreed as a courtesy to keep Metal Bank informed, this was not a
       statutory obligation. The Site Owner's representatives have submitted a number of
       FOIA requests, including at least one request, dated November 30, 1995, to review the
       entire voluminous Site file, which EPA endeavored to respond to on a schedule
       negotiated with the Site Owner's representative. At this time, EPA believes that all
       releasable documents so requested have been provided.

       (m) The Site Owner states that it has repeatedly requested that EPA include various
       documents in the Administrative Record, and that EPA has ignored these requests.
       The information requested to be placed in the Record would have documented the
       implementation and completion of the remedy specified in the 1983 Stipulation, the
       regulatory history of the Site, various investigations performed by the PRP Group and
       EPA contractors, problems in the planning and implementation of the RI/FS, and the
       failure of EPA and the PRP Group to provide information to the Site Owner when
       requested.  (See AR303713)

       Response:  EPA does not consider the Administrative Record for a site to be complete
       until issuance of the ROD.  Although EPA considers all information submitted to
       determine whether it is relevant to the selection of the remedy and should be included
       in the Administrative Record, EPA is not required to put everything into the Record
       that the Site Owner, or any other party, requests be included. The NCP requires EPA
       to include and respond to significant comments received during the comment period
       and encourages EPA to consider and respond as appropriate to significant comments
       submitted prior to the public comment period. (See 40 C.F.R. § 300.815(b)) The
       Administrative Record generally should contain all documents containing factual
       information, data and analysis of the factual information, and data that may form  a
       basis for the selection of a response action.  (See 40 C.F.R. § 300.810(a))

             While the Site Owner may feel that the previous remedy and regulatory history
       at the Site are relevant to show its cooperation and level of investment, it is not
       particularly relevant to the determination of the proposed remedy, which is based on
       sampling subsequent to the remedy performed under the 1983 Stipulation.

4.     The Site Owner comments that EPA's contractor (Ecology and Environment, Inc.)
       incorrectly prepared a hazard ranking system ("HRS") model for the Site on July  21,
       1982 which scored  the Site at 33.23.  A crucial element that prompted the high score
       was the proximity of the Site to the nearest public water intake (the Baxter Water
       Intake formerly known as Torresdale), based on the theoretical possibility that PCBs
       from the Site may reach the intake. The Site owner alleges that due to the use of the
       erroneous distance of 1.2 miles instead of the actual 2.1 miles (or greater than 2 miles)
       reported in the RI, EPA incorrectly scored the Site and placed it on the Superfund
       National Priorities List ("NPL")  in 1983. If EPA had correctly  scored the Site

                                         124

-------
originally (at 24.79 MRS, based on the 2.1 mile distance to the intake), or reranked the
Site using the new MRS model, the Site would not have made the cutoff score of 28.5
and thus not be listed on the NPL.

Response: EPA has a longstanding policy that a site will not be rescored once it is
placed on the NPL, even if the scoring was based on incorrect facts (see 49 Fed. Reg.
37080-1 (September 21, 1984)). EPA believes that the listing process - including
scoring by EPA or the States, EPA quality assurance review, public comment on the
scoring (in which the  Site Owner participated), and EPA review of the comments -
provides adequate  safeguards against incorrect site scores.  Furthermore, the purpose
of the NPL is primarily informational, to serve as a tool for EPA to identify sites that
appear to present a significant risk to public health or the environment, for purposes of
deciding which sites to investigate fully and determine what response, if any, is
appropriate.  EPA believes that it is most consistent with that statutory purpose to cease
the costly and time-consuming efforts of site scoring once a site is on the NPL.  Once
the initial scoring effort is complete, the focus of EPA activity must be on investigating
sites in detail and determining the appropriate response.  If new information indicates
that a site does not present a significant threat to public health or the environment, the
site may meet one  of the EPA criteria for deletion regardless of the HRS score.  EPA
does not believe that the Metal Bank Site falls into this category; rather, EPA's
evaluation of current information about the Site indicates an unacceptable risk
warranting remediation.

       Nevertheless, regarding the original HRS score, EPA notes that while the
Baxter Plant's main intake is located over 2 miles from the Site, the plant's standby or
emergency intake is approximately 1.9 miles (see ROD Figure 2). Under the revised
HRS model, EPA evaluates standby intakes similarly to regular intakes if they are
regularly maintained and used annually (see 55 Fed.  Reg. 51613-4).  Although EPA is
not aware of any site scored on the basis of a standby intake using the original model,
EPA probably would have used the Baxter Plant's standby  intake for scoring purposes
using the original model, had EPA been aware of the distance to the standby intake.
Thus this Site's original score may still have been above the cutoff value.

       In addition, it is likely that the Metal Bank Site would score above the cutoff
value if it were to be scored under the revised HRS model.  Under the revised HRS,
the score for the Drinking Water Threat of the Surface Water Migration Pathway
would not depend on the exact distance from the Site  to the intake(s);  instead, the score
would depend only on factors such as the characteristics of the waste (e.g., type,
quantity, containment) and contaminants (e.g., toxicity), the likelihood of release, the
level of contamination at the intake(s) (e.g., actual versus potential), the dilution weight
of the water body (based on flow rate), and the population served by the intake(s) (over
1 million in this case). The revised model also places greater emphasis in the Surface
Water Migration Pathway on  the threats to human food chain organisms and to other
environmental receptors, including wetlands and various "sensitive environments".

                                   125

-------
      The score for the Human Food Chain Threat would be influenced by the existence of
      recreational fishing in the Delaware River near the Site, while the score for the
      Environmental Threat would be influenced by the presence of the mudflat and
      Delaware River sediment areas adjacent to the Site, which is believed to be habitat used
      by the shortnose sturgeon, a federally designated endangered species. Thus, while
      EPA did not formally score the Site under the revised HRS model, the aforementioned
      factors indicate that the score would very likely be over 28.5.
                                                                  fh'

5.    The Site Owner takes issue with EPA's statement in the Proposed Plan, regarding Site
      enforcement activities, that "due to the concern that PCB oil may have been burned at
      the Site, EPA conducted dioxin soil sampling at St. Vincent's School."  The Site
      Owner points out that it responded within several months to the Agency's contractor's
      recommendation to perform soil sampling, whereas it took EPA almost 2 years to
      collect soil samples, after having received Metal Bank's sampling data.  The Site
      Owner also quotes EPA's toxicologist as having said that it seemed "reasonable to
      accept [the Site Owner's] assertion that PCBs were not burned at the Site, since we
      have no other reliable information."  (See AR303717)

      Response:  EPA took soil samples in April of 1989 in order to be assured that no
      contaminants, including PCBs and dioxin, were present that would present a risk to the
      residents at St. Vincent's School.  Although EPA's sampling occurred subsequent to
      Metal Bank's, EPA would not have performed such sampling if it had not been
      concerned that contamination from the Site might be present at St. Vincent's School.
III.   OTHER WRITTEN COMMENTS

      In addition to the written comments received from the Site Owner, comments were
received from the PRP Group, DRBC, and PADEP.  These comments and EPA's responses
are presented below.

A.    PRP Group

      The PRP Group submitted approximately 27 pages of comments on the Proposed Plan
(see AR303757 to 303808), as well as four (4) volumes of a Treatability Study, in support of
their comments. These comments and the relevant excerpts from their Treatability Study (see
AR304273 to 304334) have been included hi the Administrative Record. EPA has
summarized the substantive issues of a comment if it was more than two (2) paragraphs long,
otherwise the comments were cited verbatim. Please refer to the respective Administrative
Record pages for the actual text of a comment.

1.    The PRP Group outlines the proposed remedy with 11 bullet items.  They list their
      objections to the following aspects of the proposed remedy: (a) cleanup standard of
      10,000 ppm TPH, because EPA had underestimated the volume of soils by three (3)

                                         126

-------
       fold and $9 million; and (b) monitoring of the groundwater, because the risk
       assessments did not demonstrate an unacceptable risk from groundwater.

              The PRP Group then recommends EPA reconsider a remedy similar to what
       they had previously suggested called ex-situ stabilization/solidification ("S/S") because
       their tests demonstrate the process to be effective. However, the PRP Group states that
       the S/S remedy should be limited to soils above the groundwater and be pilot-tested
       during Remedial Design to determine if it is actually cost-effective.

       Response:  Responses concerning the volume of contaminated soils,  the TPH cleanup
       standard, groundwater monitoring, and the S/S test results are addressed in Comments
       #4, 5, 18, and 20, respectively.

             EPA believes the 11 bullets do not accurately portray the components specified
       in EPA's proposed plan. The 7th bullet (river sediment removal), for example,
       attempts to limit the area in the river to be cleaned up to within 100 feet of the
       shoreline.  EPA's  proposed plan never specified a distance but rather relied on
       performance standards; in this case, the requirement is to reduce PCB levels in the
       river to less than 1 ppm. However, based on the aquatic life risk analysis presented in
       Table  12 of the ROD, EPA has determined that all the sediments within 100 feet of the
       shoreline shall be removed, as well as any additional PCB hot spots where the PCB
       concentration exceeds 1 ppm.  This will effectively remove the site-related
       contaminants of concern, including PCBs, PAHs, Bis(2-ethylhexyl)phthalate,  and Di-n-
       octyl phthalate. The other  contaminants in Table 12 with Hazard Quotients exceeding
       1, including DDT, DDE, DDD and Cadmium, appear throughout the river area
       sampled and do not show a strong relationship to the Site.  For a summary of the
       components of EPA's selected remedy, see Part I (Declaration page) of the ROD.

2.     The PRP Group reaffirms EPA's assumption that the Site will be limited to industrial
       use only.  Their contractor justified this statement in the RI/FS based on current use
       and zoning, and the lack of evidence to contrary.

       Response: EPA does not believe that, because no evidence exists, there is no
       possibility that the Site may be zoned for any other purpose. The current zoning for
       the area is industrial; however, even if this changes, the imposition of deed restrictions
       should prevent residential use of the Site. Also,  EPA has specified in Section IX. A.6
       of the  ROD that a  24-inch layer of clean soil shall be placed on top of the Southern
       Portion of the Site so that area can be used for recreational purposes.

3.     The Site is a Class m aquifer. The Proposed Plan makes the following statement on
       page 14 concerning groundwater at the Site:

             Since the aquifer beneath the Site is designated as a Class m aquifer, which is
             currently not a source of drinking water and will not likely be in the future,

                                         127

-------
              human health risks cannot be attributed to the groundwater contamination.

              For this reason, and due to the expected benefits of "hot spot" removal, "EPA
(.       proposes no groundwater remediation." The PRP Group supports EPA1 s conclusion
       that the aquifer at the Site should be designated as a Class in aquifer.  The aquifer
       meets the criteria for a Class m aquifer under both EPA guidance and the Pennsylvania
       Land Recycling and Environmental Remediation Standards Act ("PA Act"). See PA
       Act § 304(d)(l)(I). Since the areas surrounding the Site are likely to be used for
       industrial purposes, it  is not reasonable to assume that the groundwater could be used
       as a future source of drinking water.

       Response:  Although the groundwater immediately beneath the Site is not expected to
       be a source of drinking water, it discharges into the Delaware River and becomes
       surface water. The Delaware River serves as a source of drinking water for the City of
       Philadelphia and Southern New Jersey  (through the Baxter and Delran Intakes,
       respectively). Additionally, a New Jersey municipal well is within 1.8 miles of the
       Site.  For this reason,  EPA has detailed its Performance Standards in the ROD to
       include sampling of the Baxter Intake and the investigation of DNAPL contaminants,
       which may  migrate through the deeper aquifers below the Site into the river.

4.     The PRP Group contends EPA had underestimated the cost of the remedy by $11
       million because EPA assumed soils that were not sampled were clean and that the
       contaminated soil underneath clean soils can be "surgically" removed.  The PRPs also
       contend that EPA's assumption of the disposal cost ($188 per ton) was extremely low
       and the resulting cost of EPA's proposed remedy would be closer to $28 million.

       Response:  The varying estimates of the cost of the remedy, summarized in the table
       below, were due to the various volumes that each party assumed would  have to be
       removed, and the different assumptions as to the details of the remedy implementation.9
       The PRP's estimated cost of $28.7 million, provided in their comments  to the  Proposed
       Plan, was based on the assumption that all soils overlying the identified hot spots
       would  have to be removed and disposed of as well as the hot spots themselves. EPA
       believes it is appropriate to assume that contaminated hot spots would be replaced with
       excavated soils that do not exceed the cleanup level.  In addition, the PRP's estimated
       cost of $28.7 million included costs of removing TPH at concentrations in excess of
       10,000 ppm as well as PCBs; EPA has subsequently omitted the TPH cleanup
       standard10, and limited its cleanup standard to PCBs alone. The PRPs also provided an
              The PRPs had added new assumptions (i.e., adding a "Handling/Screening of Excavated Material"
              task and applying a 20:80 debris:soil ratio factor) to their original components, which makes a
              precise comparison impossible (see AR304034-035).

              EPA did not change the proposed cleanup standard because of cost; EPA made the change
              because PCBs did not have any correlation with TPHs (see AR302076).

                                          128

-------
estimate for volume of soils (10,840 CY) and cost of disposal ($19 million) based on
PCBs alone, which was closer to EPA's estimate in the Proposed Plan.
PARTY
ESTIMATING
PRPs for EPA
PRPs for EPA
EPA's Proposed
Plan
EPA's ROD
PRPs1
Comments
PRPs'
Comments
HOT
SPOTS
PCBs, TPH
PCBs, TPH
PCBs, TPH
PCBs
PCBs
PCBs, TPH
VOLUME
(CY)
8,564
8.584
9,942
10,000
10,840
31,500
COSTS
($ MILL)
$16.0
$16.9
$17.2
$17.3
$19.0
$28.7
ADMIN RECORD
REFERENCE & DATE
AR304040
(1/25/96)
AR304039
(1/25/96)
AR3002054-055
(July '95)
AR304203
(8/19/97)
AR303773&304019
(10/17/95 & 1/25/96)
AR303773 & 303806
(10/17/95)
COMMENTS
assume disposal =
$188/ton
assume disposal =
$255/ton
using 268,431 CF =
9941. 9 CY conversion
See below for volume
rationale
No backup calculation for
volume or cost
No backup calculation for
volume
       After receipt of the PRP's comments on the Proposed Plan, EPA required the
PRP Group to recalculate and document the volume of contaminated soils, using the
assumption that such soils could be surgically removed, in order to determine the
known hot spots that are above the proposed cleanup levels (>25  ppm PCB &
> 10,000 ppm TPH). The PRP Group's contractor obtained a value (8,584 CY, see
AR304023), which was not much different than EPA's original Proposed Plan volume
(9,942 CY). This result illustrates how the PRP's assumptions (31,500 CY) may have
included unnecessary removal of the entire Site and may have been overly cautious.
The PRP's provided two cost estimates for the recalculated volume: one based on
EPA's assumption of the disposal costs ($188 per ton), and one based on the PRP's
assumption of the disposal costs ($255 per ton).

       EPA has chosen 10,000 CY to represent the estimated volume of contaminated
soil that may need to be removed from the Site. EPA considers this estimate to be
reasonable because it is between the volumes (8,584 or 9,942 CY) estimated by the
PRPs and EPA, respectively, and the volume estimated by the PRPs for removal of
PCBs alone (10,840 CY) assuming that soils overlying the PCB hot spots would have
to be removed and disposed of as well. EPA believes this volume reflects the known
hot spots and contains enough margin for error that in the event any additional PCB
contamination has been overlooked, the resulting volume would still be within the
estimate.  By applying to this volume (10,000 CY) to the same assumptions provided in
the original RI/FS (with the exception of the cost of disposal, which the PRPs now
estimate to be $255 per ton), EPA estimates the cost of the remedy to be $17,272,567
(see AR304203). Therefore the ROD reflects the estimate of $17.3 million as the cost
of the remedy.
                                  129

-------
5.     The PRP Group contends that EPA should not select TPH as a cleanup standard
       because: (a) none of the risk assessments identified TPH as a chemical of concern; (b)
       there are no federal cleanup standards for TPH; (c) the proposed plan emphasized
       PCBs, not TPHs, as the chemical of concern; (d) the highest TPH contamination would
       be removed by the oil-water separator system and/ or by the hot spot excavation; and
       (e) the cost of removing TPH has been greatly underestimated.

       Response: EPA acknowledges that the TPH cleanup level is not appropriate and it has
       been eliminated.  However, EPA believes the estimated cost of removing TPH
       contamination is still accurate based on information supplied by the PRPs in the RI/FS
       (see Comment #4).

6.     The PRP Group argues that EPA did not follow the law or the terms of its Consent
       Agreement by failing to reject the PRP's remedy during the RI/FS.  If EPA had
       disapproved of the RI/FS, the PRP Group would have been given a chance to invoke
       dispute resolution and reach a negotiated settlement with the agency concerning the
       remedy (i.e., TPH cleanup standard, excavation versus S/S treatment).

       Response: CERCLA does not require that EPA must obtain the PRP's concurrence
       when deciding a remedy for a site. The development of the Proposed Plan and the
       Record of Decision is an inherent government function. The RI/FS gives the PRP
       Group a chance to obtain and present data in a manner that supports their conclusion.

             Issues concerning the elimination of TPH as a cleanup level have been
       addressed in the PRP's  Summarized Comments #'s 4 & 5. TPH has been eliminated as
       a cleanup standard.  Issues concerning the feasibility  of using  the  S/S technology to
       remediate the Site are addressed in the PRP's Comment #20.  EPA believes the
       Exhibits provided do not prove the technology will be successful.

7.     In the last paragraph on page two [of the Proposed Plan], EPA states that evaluation of
       the RI/FS provided evidence that DNAPLs exist on-site.  The PRP Group does not
       believe this is correct.  No site data provide evidence of DNAPL contamination.  The
       borings referenced by EPA to support its statement (BIOS and 106) are in (1) a tidally
       influenced zone subject to significant groundwater table fluctuations or (2) the area  that
       was  subject to hydraulic effects from the implementation of the owner's oil recovery
       system, which operated from 1983 -1989. The effects of tidal action and the oil
       recovery system caused the  "smearing" of oils in a manner that EPA appears,
       incorrectly, to have interpreted as presence of DNAPL. No other evidence of DNAPL
       contamination exists in  the RI/FS.

       Response:  EPA's Proposed Plan states that "the RI/FS provided evidence that there
       mav also be a DNAPL".  EPA never definitively said that there were DNAPLs (also
       see Site Owner's Comments #9, 11, and 13, NAPL).  The PRP Group's statement that
       the appearance of DNAPLs  may be due to  "smearing" may be valid; however, without

                                         130

-------
      actual samples (since no samples were analyzed beyond 14 feet), the existence of
      DNAPLs cannot be confirmed or discounted.  Accordingly the ROD requires sampling
      during RD to determine the presence or absence of DNAPLs.

8.    In the fourth paragraph on page five [of the Proposed Plan], the Agency states that,
      during the RI/FS in 1993, several anomalies were detected with a ground penetrating
      radar ("GPR"). This sentence gives the misleading impression that the anomalies were
      not evaluated. In fact, the geophysics investigation conducted by the Group included
      GPR and electromagnetic ("EM") studies to determine if the Site contained buried
      drums and/or tanks.  Of all the potential anomalies suggested by these studies, only
      seven were deemed to be of sufficient size/strength to warrant further analysis.  The
      seven anomalies were explored with test pits.  It was determined that no buried drums
      or tanks existed in these locations.                              i

             The only anomaly not fully investigated was the potential existence of a UST
      believed to lie beneath a concrete pad.  The PRP Group were not able to confirm by
      test pits the existence of a tank beneath the pad.

      Response:  EPA did not use the PRP's GPR data or subsequent investigations (i.e.,
      testpits) to formulate the proposed remedy. However, EPA notes that:

      (a)    4 of the areas containing suspicious anomalies were not investigated. These
             anomalies, as documented by RI Figure 2-8 (see AR300060), contain the
             following suspected sources: #2 (Metallic Object); #3B (Metallic Object); #8
             (Drum); and #9 (Metallic Object);

      (b)    all of the testpit logs contained in Appendix A of the RI (see AR300301 to
             300311) failed to record the dimensions of the testpits.  Photographs taken by
             EPA's oversight contractors (see AR301933 to 302045) show a majority of the
             pits were approximately 10 by 3 feet (the width of a backhoe shovel).  This is
             an insufficient area of coverage when most of the anomalies were more than 50
             feet in diameter; and

      (c)    both the Site Owner and the general public criticized EPA for not aggressively
             investigating the suspected underground storage tank.  EPA believes the PRP
             Group should have jack hammered past the concrete slab that blocked the UST
             or taken samples around its peripheries (especially at TP-1E where a field vapor
             detection instrument (HNu) peaked at 30 ppm, see AR300304).

             In summary, the GPR and its subsequent investigation was inconclusive and
      failed in its intended purpose: to locate potential sources of PCBs. Nevertheless, the
      issue of the GPR and its merit is moot because EPA does not discuss its results in the
      ROD.
                                         131

-------
 9.     The Proposed Plan states that samples of sediments taken at various depth along the
       Delaware River show concentrations of up to 19.6 ppm PCBs and 17,000 ppm TPH.
       The Group believes that this statement is misleading because it implies that
       contamination is more widespread than the test data indicate.  A review of the data
       shows that  the concentrations cited by EPA involve two riprap sampling locations taken
       at the edge of the site along the mudflats.  (19.6 ppm PCBs at MF107; 17,000 ppm
       TPH at RR-3.)  Other sediment data in the river and along the mudflats showed
       substantially lower concentrations (at least one order of magnitude Tower) in almost all
       other sampling locations. See RI figures 4-16, 4-17 (Exhibit 6).

              The Group requests that EPA revise this paragraph to state that PCB
       concentrations in river sediments are consistently less than 2 ppm, and that the highest
       values were immediately adjacent to the site.

       Response:  While the PRP Group's statement may be generally true, EPA notes that
       PCB concentrations greater than 2 ppm were found in the following areas not
       immediately adjacent to the Site (S23, MF133, MF127, MF5). EPA believes its
       statements  in the paragraph are more appropriate because:

              (a)  the purpose of the section of the Proposed Plan was to highlight the highest
              concentration detected in each media of the Site;

              (b) EPA had stated that "distribution and concentration appear to decrease with
              distance from the Site"; and

              (c) EPA also references a map, Figure 3 (which in turn references the PRP's
              Exhibit 6: RI Figures 4-16 & 4-17), whereby the readers can draw their own
              conclusions concerning the actual extent of PCB contamination in the river.

'10.    [Page 6, paragraph 4 of] the  Proposed Plan states that EPA's evaluation has identified
       PCBs as the major contaminant of concern. Further, the Proposed Plan states: "Other
       contaminants of concern at the Site include metals such as arsenic, beryllium,
       chromium, copper, lead  and  mercury; SVOCs such as Methylethyl Ketone and
       phthalates;  DDT-type pesticides; polynuclear aromatic hydrocarbons (PAHs); dioxins
       andfurans."  The PRP Group recognizes that these other substances were considered
       "contaminants of concern" in the development of the risk assessments by the Agency.
       However, since the risk  assessments determined that risks associated with almost all of
       these  chemicals were within acceptable ranges, the PRP Group believes that this
       statement is misleading to the reader. Further, EPA's own discussion on page 8 goes
       on to state that risks associated with arsenic, beryllium, chromium (paragraph 5), and
       dioxin and  furans (paragraph 6) were not of concern. The PRP Group accordingly
       requests that the Agency revise this section to eliminate all references to chemicals for
       which no remedial goals have been developed.
                                          132

-------
       Response: EPA believes the Proposed Plan should provide a discussion of how certain
       chemicals were eliminated as "contaminants of concern" through the risk assessment
       process. EPA considers the identification of all COCs important even though they may
       not pose serious risks or be driving the remedy;  consequently, they have been listed in
       conjunction with their screening level exceedances in the ROD. Section V.E.5. of the
       ROD, for example, lists the maximum concentration and screening level of metals in
       the subsurface soils. This will enable the party implementing the remedy to prepare for
       any potential problem associated with disposal or treatment of soils contaminated with
       chemicals other than PCBs (e.g., lead).

11.    The PRP Group does not agree with the use of EEQ and EEC values as measurements
       of acceptable or unacceptable risk [on page 9 of the Proposed Plan].  The PRP Group
       previously explained its position when it provided EPA with comments on the
       terrestrial risk assessment.  To summarize, the use of such values should be limited to
       screening chemicals for possible further evaluation.  EEQ and EEC values do not
       provide a quantitative measurement of risk.
      Response: Due to the limitation in time and money, the PRP Group did not provide a
      comprehensive study testing animals and other organisms for actual Site contaminants.
      As a result, EPA had to use values obtained from various literature sources (or EECs,
      expected environmental concentrations) to evaluate potential harm to organisms at the
      Site.  The comparison of this literature value to site-specific chemical concentrations
      resulted in the EEQ (environmental effects quotient).  This initial risk assessment
      process is used to identify which contaminants may pose risk to environmental
      receptors and may be used to guide the risk assessment process into the next phase
      (Tier 2 ERA).  However, EPA can decide whether or not to remediate a contaminant
      based on a wide variety of "risk management" factors.

12.    The Proposed Plan states that".. .the discharge of PCBs in the groundwater into the
      Delaware River will contribute to PCB contamination in the nearshore food web." The
      PRP Group does not believe that the data indicate that groundwater contains PCBs at
      concentrations of concern. The only groundwater sample that contained detectable
      PCBs (all others were below detectable concentrations) came from  a well with standing
      LNAPL product. This situation easily could bias the sample and result in the artificial
      detection of PCBs in a groundwater sample that actually measures PCBs associated
      with free phase oil, not dissolved PCBs.

      Response:  Although samples were collected from groundwater seeps at two locations
      (see ROD Section V.E.3.) which showed low levels of PCBs (3.7 ppb), the high
      sediment contamination along the shorelines indicates contaminated groundwater must
      have been discharged from underneath the Site. In this analysis, it is not appropriate to
      distinguish between LNAPLs and dissolved PCBs in the water phase since fish and
      other animals may ingest the LNAPL as well as the groundwater as it seeps from the
      Site.  Therefore, EPA maintains groundwater, as a whole, contributes to the "PCB

                                         133

-------
      contamination in the nearshore food web."

13.   The PRP Group are opposed to the collection of any additional sediment samples
      during Remedial Design because they believe the extent of contamination in the river
      was well defined, contrary to EPA's statements. They state that out of the SO samples
      that were taken during the RI, the only uncertainty exists at S7 and S9.  However, the
      contamination at S7 and S9 are inconsequential because the PCBs (1.71  to 1.2 ppm)
      have been seen to be decreasing towards the cleanup level (1 ppm). They also
      requested EPA to identify which data are lacking so that they may provide the
      necessary data before EPA selects a remedy.

             The PRP Group also disagrees with EPA's statement that data quality problems
      had rendered most of their results useless.  They cite the 1995 resampling effort, where
      26 locations were collected and the samples overlapped previous locations, as ideal.
      The PRP Group again disagreed with EPA's proposal that chemicals, in addition to
      PCBs and PAHs, should be monitored.

      Response:  While EPA agrees with the PRP Group that a dramatic decline in PCB
      contamination has been observed away from the Site's shoreline, EPA does not agree
      with the PRP Group's statement that the  1 ppm PCB contamination boundary is clearly
      defined. Of the 50 samples that were collected, most were either: (a) less than 100 feet
      from the shoreline into the Delaware River; (b) fragmented (DR8, no deeper tests); or
      (c) mismatched with field screening results (MF107).  While the 1995 samples had
      better detection limits, some deeper samples (S1,S3,S12,S19,S23) were  not analyzed,
      and only one sample (S2) had overlapped the previous river data.  Therefore, these data
      gaps would be more appropriately addressed in the Remedial Design.  EPA's
      determination to remove all sediment within 100 feet of the property line to a depth of
      4 feet from the riverbed surface will significantly reduce the amount of sampling during
      Remedial Design.

14.   [Page 13, paragraph 3 of]  the Proposed Plan calls for a PCB clean-up level of 10 ppm
      in surface soils in the courtyard.  To be consistent with the goal of 25 ppm for
      subsurface soils, the Group suggests that EPA define surface soil as soil within the first
      10 inches.  40 C.F.R. § 761.125(c)(4)(v) (1994) (PCB Spill Cleanup Policy); EPA,
      Guidance on Remedial Actions for Superfund Sites with PCB Contamination at 30
      (Aug. 1990). After excavation of the first 10 inches, the remediation goal should be
      the same as for other subsurface soils, i.e., 25 ppm.

      Response:  EPA has required removal of surface soils to 2 feet below surface because
      the deepest PCB contamination (70 ppm @ CY136) was observed at 2 feet below
      surface and this represents a small and manageable area. ROD Section DC.A.I. also
      requires additional sampling to be performed in order to determine the full extent of
      contamination because the samples collected during the RI contained several
      discrepancies (see explanation below). In determining the full extent of contamination,

                                         134

-------
      EPA will consider limiting the depth of confirmatory testing to 10-inches (the
      dimension suggested by the PRP Group) below the bottom of an excavated layer.
      Subsurface soils then can be defined as 10-inches below the bottom of an excavated
      layer and containing less than 25 ppm PCB.  Regardless of what the remedial design
      sampling shall entail, EPA will require all confirmatory samples to be analyzed in the
      laboratory.

             EPA observes that the RI data for the Courtyard (see AR300150 to 300153, RI
      Figures 4-3 to 4-5) was fragmented and not totally reliable for delineating the extent of
      contamination.  For example, two (2) locations were sampled in 1991; both contained
      PCB contamination above the cleanup level (52 ppm @ THIS & 140 ppm @ TB2S).
      In 1993, thirty-eight (38) locations were sampled in a grid pattern, however most
      samples were analyzed using field screening test kits; only one location (70 ppm @
      CY136) that exceeded the cleanup level was laboratory analyzed. Additionally, only
      six locations (CY119, 120, 132 -135) were sampled below 2 feet with only one
      location sent to the laboratory for verification analysis (0.97  ppm @ CY134).  In 1993,
      the areas around the 1991 hotspots (TB1S, TB2S) did not detect any PCB
      contamination but these locations (CYs-110, 116, 122, 128 and CYs-119, 126) were
      field screened, not laboratory analyzed. Of the 9 samples collected for laboratory
      verification, 2 samples (1.1 ppm vs. " >50" @ CY119 and 0.46 ppm vs. ' >5-< 10"
      @ CY133) did not match their field screening results.  Therefore, EPA concludes
      while field tests are useful for screening purposes, their results should not be relied
      upon to determine the exact area or depth of surface soil contamination.

15.    [On page 13, paragraph 6 of the Proposed Plan,]  the Agency has expressed concern
      regarding the existence of DNAPL. The Proposed Plan proposes a DNAPL collection
      system combined with an oil collection trench and oil/water separators.  The PRP
      Group does not believe that oil/water separators are required to collect any residual
      NAPL that may migrate towards the River after construction of the remedy.  Further,
      the Group does not agree that any treatment system should be developed for DNAPL
      because no data exist that indicate the presence of DNAPL.  Methods can be proposed
      during the design phase to deal with material that may find its way into and settle to the
      bottom of the trench.

      Response:  As explained in Comment #7, the presence of DNAPLs cannot be
      confirmed or discounted based on the PRP's data. Section DC.F.I pf the ROD allows
      for an investigation to determine if DNAPLs are present. If DNAPLs are present and
      do present a risk to human health and the environment, a system that allows the
      collection of DNAPL in the groundwater can be designed during the remedial design
      process.

16.    [On page 13, (last or) 7th paragraph of the Proposed Plan,] the Agency states that the
      depth and area! extent of contamination remain unknown and that further sampling will
      be required to determine the final placement of the temporary wall.  The Group

                                         135

-------
suggests that EPA select a maximum depth of 3-4 feet for any sediment excavation.
This is a depth at which, following replacement with clean fill, any residual
contamination would not be bioavailable.

       The Agency states further that removal would be implemented near the
shorelines where land-based excavation equipment can be utilized.  The Group agrees
with this approach and believes that this would be the least disruptive form of sediment
excavation.  The data support this because almost all the areas where PCBs exceed 1
ppm lie within 20-30 meters [66-98 feet] of the site. Any other approach to sediment
remediation would require evaluation beyond that already undertaken for the RI/FS.

Response:  The PRP Group is partially correct in stating that almost  all sediment areas
where PCBs exceed 1 ppm lie within 30 meters  (or 98 feet) from the  Site.  Based on a
statistical evaluation of the samples collected in  1991 and  1993, EPA calculated the
"mean" concentration of the samples more than 30  meters from the shoreline to be less
than 1 ppm (see Table 4-3 of the Aquatic  Risk Assessment, AR300951 or ROD Table
11). However, individual concentrations at some locations,  including locations of
samples taken subsequently in 1995, exceeded the proposed cleanup level of 1 ppm
(e.g., 1.2 ppm @ MF11,  1.3 ppm @ S6, and 4.59 ppm @ S20). Nevertheless, EPA
has determined that the removal of sediments within 100 feet of the shoreline will
probably address the majority of the PCBs in excess of the cleanup standard of 1  ppm.
The appropriateness and feasibility of removal of any PCB hot spots beyond 100  feet
will be determined by EPA during remedial design. Through the removal of the
sediments within 100 feet of the shoreline, the mean concentration of PCBs (3.8 ppm)
in the nearfield areas (a distance less than 30 meters (98 feet) from the shoreline) will
be reduced to levels equal to or below the  mean concentration (0.87 ppm) of the
farfield areas (a distance more than 98 feet from the shoreline).

       EPA has modified its proposed remedy to require removal of  all sediments
within 100 feet of the shoreline, regardless of the type of contaminant.  As Table 12  of
the ROD illustrates, several contaminants (Bis(2-ethylhexyl)phthalate, Di-n-octyl
phthalate, PAHs, DDT, DDE, ODD, and  Cadmium) in addition to PCBs pose an
unacceptable Aquatic risk. Bis(2-ethylhexyl)phthalate, Di-n-octyl phthalate, and PAHs
were deemed to be attributable to the Site because their  "Hazard Quotient" decreased
with distance from the Site. The Hazard Quotients  for these contaminants fell to  one
(1)  or less beyond 100 feet of the shoreline, indicating that contaminant levels beyond
100 feet would not contribute to the risk.  By contrast, the Hazard Quotients for DDT,
DDE, DDD and Cadmium did not decrease with distance from the Site,  and EPA
determined that these contaminants were not likely to be Site-related. Since EPA has
observed that the location of PCBs above 1 ppm may not always correlate with elevated
levels of other contaminants that pose a threat (i.e., Total PAHs = 926.5 ppm while
PCBs = "non-detect" at RR4A, see AR304395  & 300203),  EPA has concluded that
removal of PCB contamination alone will not result in the removal of other Site-related
contaminants that contribute to the risk.  Therefore, EPA  has determined that removal

                                   136

-------
       of all sediments within 100 feet of the shoreline would probably be more cost effective
       than sampling and determining the exact locations where elevated levels of the
       individual contaminants require remediation.

              EPA has also determined that sediment excavation depth may be limited to
       within 4 feet.11  EPA agrees with the PRP's rationale that coverages as deep as 3 to 4
       feet are sufficient to prevent contaminated sediments from being "bioavailable" to
       aquatic creatures feeding in the area.

              Finally, the lateral extent of contamination along the shoreline, including the
       areas near the City's storm water outfall and Morris Iron & Steel (see ROD Figure 3,
       near the northwest and southeast corner of the Southern Portion of the property), has
       not been defined.  This, combined with the possibility that it may not be feasible to
       excavate sediments in the river beyond 100 feet of the shoreline, has prompted EPA to
       require additional sampling during the Remedial Design. Since the PRPs did  not
       provide documentation during the RI/FS to substantiate their concerns regarding
       excavation difficulties,  additional engineering measurements (i.e., soil and river current
       data) will also be required  as part of the Remedial Design to determine where the final
       placement of the temporary wall should be.

17.    [Page 14, paragraph 3 of] the Proposed Plan states that, to further delineate hot spots
       within the southern portion of the site, EPA proposes to re-sample the area in a
       thorough and methodical grid pattern. The Group believes that, with the addition of
       data collected from test pits in August of 1995, adequate data exist to limit the
       methodical grid pattern to currently known and expected areas designated as hot spots.
       The Group suggests that there'is no need comprehensively to re-sample the entire
       southern portion of the site.

       Response:  EPA believes that a methodical grid sampling of the Southern Portion of
       the Site is  necessary to accurately identify hot spot areas that  must be removed.

18.    [On page 14, last paragraph of the Proposed Plan,] the Agency suggests that a
       groundwater monitoring program be implemented that would include monitoring of
       chemicals  sampled for in the RI/FS, including  PCBs, TAL VOCs/SVOCs, TAL metals
       and  groundwater chemistry parameters.  The PRP Group believes that there is no
       reason to sample groundwater or  liquids in the collection system for parameters other
       than PCBs.  The remedy is designed only to limit the migration of PCBs away from the
       site  into the environment.  Therefore, sampling for chemicals not part of the remedial
       goals/action is neither necessary nor reasonable. Furthermore, given that the  Site is
       11    EPA acknowledges that some hot spots within 100 feet of the shoreline (i.e., 4.59 ppm @ S20)
             may not be remediated under this new criteria because the PCB contamination appears below 4
             feet from the surface.

                                          137

-------
      located in an industrial area, it is reasonable to assume that groundwater may be
      affected by off-site sources.

      Response: The rationale for sampling of groundwater and liquids in the collection
      system has been addressed in Comment #12.  EPA is requiring the monitoring of
      chemicals in the groundwater as well as the collection system because:

             (a)  the Delaware River Basin Commission (DRBC) and Section 3.8 of the
             Delaware River Basin Compact (compact) require a permit application that
             records the impact of contaminants, in addition to PCBs, emanating from the
             Site, (see DRBC comments dated 9/27/96.) Although the party performing the
             remedy will not be required to apply for a permit, the substantive requirements
             of the law will have to be met, including the analysis of the impacts of
             contaminants emanating from the Site;

             (b)  as discussed in ROD Section V.E.2., the PRP Group did not obtain samples
             of groundwater being discharged into the surface water for the RI/FS.  As a
             result, EPA applied a dilution factor to the groundwater results to obtain a
             hypothetical surface water concentration for use in the risk assessments.
             Sampling of the groundwater on-site and in the collection systems compensates
             for that data gap and provides the necessary information for DRBC's
             evaluation.

19.    [On page 17, last paragraph of the Proposed Plan,] the Agency states that the remedial
      design must build in safeguards that would prevent a 100-year flood from entering the
      buildings and mobilizing unremediated PCBs.  The Group does not agree that flood
      waters would mobilize PCBs within the buildings to an extent that would result in
      unacceptable risks to human health or the environment. EPA has no record evidence
      that exposure from this pathway would present unacceptable risks.

      Response: Although there is no evidence, this issue was not investigated either. The
      PRPs were required to sample and document all low-lying drainage points where PCBs
      may have accumulated, as pan of the RI/FS.  Samples were collected in the Building
      Areas but no drainage pathways were ever documented. For this reason, ROD Section
      IX.F.2. requires the sampling and determination of all potential routes by which PCBs
      may have been discharged from the Site.

             Additionally, the PRPs (see Comment #7) and the Site Owner have commented
      that PCBs may have been "smeared" to the surface soil by the fluctuation of
      groundwater. If this is true, subsurface PCBs may also be "smeared" to the surface
      during a 100-year flood.  The Site Owner  (see Volume I, Book 3, Section 2.5, page 72
      of Site Owner's Comments) has estimated  a 100-year flood level will not exceed 10 feet
      above MSL or the surface of the Site.  Therefore EPA will not require physical
      barriers to safeguard inside the Building Area but will require a monitoring program to

                                         138

-------
       assure that there are no historical or potential pathways for PCBs to migrate off-site.

20.    [On page 19, paragraph 1 of the Proposed Plan,] the Agency states that alternatives to
       excavation involving treatment of contaminated soils have not proven successful at
       other sites.  The PRP Group does not agree with this summary statement.  S/S recently
       was shown to be successful in binding TPH and PCB contaminated soils at the Site
       (PRP Group's Exhibit 2). The attached summary of S/S results (PRP Group's Exhibit
       1) also describes numerous examples of successful S/S projects at similar sites.

       [The PRPs also referenced the same exhibits as part of Comment #6 (see AR303780) in
       which they claim the technology works because the treated solid mass:  (a) does not
       leak other chemicals (teachability); (b) is strong (compressive strength " > 50 psi");
       and (c) does not retain water (hydraulic conductivity " <  IxlO"5 cm/sec").]

       Response: Based on the Exhibits, there is no strong evidence that the treatment
       (Solidification/Stabilization) is effective in reducing the mobility of PCBs. The
       evidence provided by  the PRPs was inconclusive because:

             (a) before any treatment, the  highest PCB leachate concentration submitted  in
             Exhibit 1 was  " < 0.0004 ppm"; the lowest concentration after treatment was
             also < 0.0004 ppm.

             EPA has also noted that the most contaminated soil used in the test contained
             only 22 ppm PCB (Wastech, see AR304304). This information was found in a
             subsequent treatment vendors' reports submitted after the comment period.
             Considering the cleanup level is only 25 ppm, these soil samples were not
             representative  of the hotspots; and

             (b) none of the Superfund sites identified in Exhibit 2 provided a before and
             after comparison of PCB leachate levels.  The fundamental question still
             remains: "What were the leachate levels of PCBs  before and after treatment  to
             demonstrate mobility reduction?"

             Several of the  sites identified (Douglassville, Imperial Oil, and Portable
             Equipment Salvage Co.) were only field demonstrations and not actual
             remediations.  The most comparable site identified was Pepper's Steel but the
             groundwater data presented does not show what leachate concentration would
             have occurred  if the waste was not treated.

             Furthermore, their hydraulic conductivity result was deficient.  EPA's criteria
     ,  for hydraulic conductivity is generally " <  1x10* cm/sec" not"< IxlO"5 cm/sec."

21.    [On page 19, paragraph 2 of the Proposed Plan,] the Agency estimates the cost of its
       preferred alternative to be $17.1 million.  This cost estimate assumes that total PCB-

                                          139

-------
       and TPH-contaminated soil disposal costs would be approximately $2.4 million.  Data
       both contained in the RI and collected this year demonstrate that EPA's estimate is far
       too low.  The Group completed a sampling program during August 1995 designed to
       provide further data on the extent of PCB and TPH contaminated soil. As discussed
       earlier in these Comments, these data indicate that EPA has greatly underestimated the
       volume of soil that would be excavated to achieve the 25 ppm PCBs and  10,000 ppm
       TPH standards. The Group has recalculated the cost to implement EPA's Proposed
       Plan and estimates that cost to be approximately $28.7 million.

       Response: Issues concerning the cost of the remedy have been addressed in response
       to Comment #4.  EPA has determined $17.3 million is a valid cost estimate for the
       remedy.
22.    [Concerning Table 1 titled "Clean-up Levels", and] as discussed above, the Group
       recommends that EPA eliminate the TPH remediation goal.

       Response:  Issues concerning the elimination of TPH as a cleanup level have been
       addressed in Comment #5.  TPH has been eliminated as a cleanup standard.

B.     DRBC

The Delaware River Basin Commission (DRBC) submitted two (2) letters (see AR302412 and
303752) on separate occasions before the closing of the public comment period. The
comments contained in the two letters have been recounted verbatim and combined according
to the issues raised.

1.     We have reviewed the Proposed Plan for remediating the Metal Bank Superfund Site
       and have severe reservations about the adequacy of the proposed actions to protect
       resident fish and aquatic life in the  adjacent Delaware River. Currently three states
       (Delaware,  Pennsylvania and New  Jersey) have issued advisories on consuming
       anadromous and resident fish including striped bass, white perch and catfish due to
       elevated levels of polychlorinated biphenyls (PCBs) in their tissue.  The Commission is
       currently completing Phase I of a program to control toxic pollutants discharged to the
       tidal Delaware River between Trenton, NJ and Delaware Bay.  This phase  is focusing
       on point sources of toxic pollutants, specifically industrial and municipal waste
       treatment plant discharge. Monitoring conducted under the auspices of this program
       have not revealed these  sources to contain detectable levels of PCBs. Additional
       monitoring  is planned for the coming year to determine whether PCBs are present in
       these discharges at very low levels. Phase n of this program will focus on non-point
       sources of toxic pollutants including Superfund sites with special emphasis  on PCBs,
       DDT and its metabolites. In view of the current consumption advisories and proximity
       of the site to the river, this site must be considered a significant source of PCBs.  Data
       collected by the Commission in  1994 indicated that the highest concentration of PCBs

                                         140

-------
       in tissues of white perch and channel [catfish] occurred at the Tacony-Palmyra Bridge
       station approximately 1 mile from the site.

       Response:  EPA acknowledges DRBC's comment that, because (1) various studies
       performed by their office have not revealed PCBs to be present at the discharge point
       from industrial and municipal wastewater treatment plants in the area, (2) the level of
       PCB contamination in fish sampled near the Site is high, and (3) the Site is in close
       proximity to the river, the Site must be a significant source of PCBs.

             While EPA agrees that the Site is a significant source, EPA does not agree with
       DRBC's rationale for deriving this conclusion. First, DRBC did not provide EPA with
       any analytical data to substantiate the absence of PCB contamination from wastewater
       treatment plants. Next, as explained in the response to the Site Owner's comment (see
       Aquatic Risk Assessment, Comment #7), it would be extremely difficult to determine
       the degree of contamination by the Site of upgradient locations and vice versa because
       of the constant movement of tides in the river. EPA has surmised that the Site is a
       significant source because high PCB levels (i.e., 19.6 ppm) were observed  at the
       shorelines while PCB levels at approximately 300 feet  from the Site decrease to around
       1 ppm or less.

2.     We are concerned about the lack of sufficient data to characterize the spatial extent of
       contaminated sediments in the river, the location and PCB concentration in soil "hot
       spots" in  the southern portion of the site near the river, and the extent of non-aqueous
       phase liquids in the same area.  An appropriate remediation plan cannot be  selected
       until the extent of contamination of river sediment is determined and the effectiveness
       of any treatment system for the non-aqueous phase liquids is established.

             We have severe reservations about the adequacy of the proposed actions to
       protect resident fish and aquatic life in the adjacent Delaware River, and impact the
       current advisories on consuming anadromous and resident fish, including striped bass,
       white perch and catfish, which advisories have been issued due to elevated  levels of
       poly chlorinated biphenyls (PCBs) in their tissue.

       Response:  EPA understands that DRBC is concerned  that any remedy selection would
       be premature at this stage because the extent of contamination  has not been determined
       in:  (a) the river;  (b) the hot spots;  or (c) the NAPL area.  Without that information,
       an effective means to treat the NAPL cannot be obtained.

             DRBC is correct that the full extent of contamination has not been determined in
       the aforementioned areas. For this reason, the ROD requires additional extent of
       contamination and confirmation sampling of the river and the hot spot areas to assure
       that the cleanup levels are obtained. Any deviation from these requirements would
       have to be justified, for example, due  to the impracticability of engineering a cofferdam
       very far into the river.

                                          141

-------
             EPA believes the removal of the most contaminated sediments along the
       shorelines will serve as the first step towards reducing pollution from the Site.
       Subsequent measures such as the installation of the containment systems are used as
       backup systems and will eliminate the release of residual contamination if some hot
       spots are not identified during remedial design. EPA's risk assessments have shown
       that PCBs in the NAPL, rather than NAPLs,  pose a threat.  Therefore EPA's remedy
       is directed towards the removal of PCBs and not the NAPL and its related constituents.
       EPA has also included the investigation of DNAPLs to ascertain whether PCBs are
       migrating to the lower depth of the aquifer and whether they pose a threat.  If any
       portion of the remedy requires change based on new information, such as the discovery
       of DNAPLs, EPA may issue an amended ROD or an Explanation of Significant
       Differences ("ESD"), depending on the scope of the changes to the remedy.

3.     We are concerned that dewatering operations from the proposed installation of a sheet
       metal cofferdam and dredging in the mudflat and river may  mobilize PCBs associated
       with sediment further exacerbating fish tissue contamination in the Delaware Estuary.

       Response: The dredging of contaminated sediments is anticipated to be performed on a
       dry surface after the area behind  the temporary cofferdam has been dewatered. To
       further minimize any possibility of disturbing contaminated  sediments which may
       migrate into the river, ROD Section DC. A.4 requires the cofferdams to be installed and
       removed in a manner that minimizes turbidity. Additionally, the ROD requires care to
       be taken during the dewatering process to protect fish and other aquatic organisms
       which may be present behind the cofferdam.

4.     We also note no assessment of the potential impact on the City of Philadelphia drinking
       water intake just 2.5 miles upstream of the site. Past studies have documented
       upstream excursions of approximately seven miles during twice daily flood tides.

       Response: DRBC is correct that the impact of the Site on the city's drinking water
       was not  investigated.  As explained in the response to a Tacony Civic Association
       member's comment received during the July 1995 Public Meeting (see Public Meeting
       Comment #7), EPA estimates the discharge of PCBs into the Delaware River to be well
       below the Safe Drinking Water Act standards required for the city water supply.  EPA
       also realizes that all estimates are based on theoretical conditions.  Because it would be
       extremely difficult to determine the actual contribution of contaminants from the Site to
       distant off-site locations (see Aquatic Risk Assessment Comment #7), especially when
       these locations are affected by river tides, EPA has not required further assessments in
       the ROD.  However, since the city is not required to analyze for PCBs as part of its
       routine testing program, the ROD (Section DC.G.3) requires sampling for PCBs at the
       city's intake as a precautionary measure.

5.     It should be noted that the Commission is proposing a water quality criterion  for the
       protection of human health from  the ingestion of fish and water for  total PCBs of

                                         142

-------
       0.000044 ug/1. The Commonwealth of Pennsylvania's current criterion for total PCBs
       is also 0.000044 ug/1.  Given the lack of data on the contribution of PCBs from the site
       to the river, and any detailed assessment of the dilution of PCBs entering the river from
       the site, the selection of a remediation plan is premature.

       Response: Since DRBC submitted its comments, the proposed criterion was enacted
       on October 23, 1996 as part of the DRBC's Administrative Manual Part m ("Water
       Quality Regulations") and represents the most recent regulation governing the
       discharge of toxic pollutants into the Delaware River.

             The listed criterion (0.000044 ug/1 PCB) is not the actual level that discharges
       from a containment system, such as the oil-water separator, must attain.  Rather, it is
       the value after dilution by the river at which a toxic pollutant may have an acute or
       chronic impact on aquatic life. Therefore, the allowable contaminant level at the
       discharge point would be higher than the water quality criterion.

             To implement the new procedure, the policy allows for the mathematical
       modeling  of toxic pollutants under various hydrological conditions (such as river flow
       and tidal stage) to predict its effluent outcome  in the river. The Cornell Mixing Zone
       ("CORMIX") model is one such scenario which evaluates a wastewater effluent mixed
       in with river flow.  The ROD specifies that any party implementing the remedy must
       acquire the commission's approval on effluent limits prior to actual discharge.

6.     The U.S.  Environmental Protection Agency Region in and the Delaware River Basin
       Commission executed a Memorandum of Agreement ("MOA") in October 1991 with
       respect to Superfund sites located within EPA  Region HI that are subject to the
       jurisdiction of the Commission.  Our review of the proposed plan indicates  that the
       discharge and withdrawal facilities are covered by the MOA.

             [from September 5, 1995 letter] Further evaluation of the proposed plan
       indicates that the remedial action at this site will require an application for a docket
       under Section 3.8 of the Delaware River Basin Compact.  Please contact Mr. George
       Elias, Head, Project Review Branch for information on the application and review
       process.

       Response: EPA acknowledges that the MOA  is an ARAR and will reference that
       agreement in the ROD.  As explained in the previous response, any party implementing
       the remedy will acquire the Commission's approval on effluent limits prior to actual
       discharge.

C.     PADEP

1.     Building Area - As a general statement there needs to be more clarification and
       commitment to the future use of the property when determining cleanup levels.

                                         143

-------
       Statements related to the availability and safety of the buildings for use are too vague.
       Could EPA provide more details regarding the deed restrictions and how that could
       affect the future use of the property?  Contact should be made with the Delaware River
       Partnership and the City of Philadelphia regarding future use of the property and the
       possible "greenway" along the river.

       Response: The deed restrictions will prohibit residential and agricultural use and use
       of groundwater underneath the property.  EPA cannot make a definite commitment
       concerning future land use because the future use of a property can change.  EPA
       acknowledges, as a member of the Tacony Civic Association suggested during the
       Proposed Plan public meeting, that property along the riverfront may eventually be
       converted  to residential use.  While residential cleanup levels would be more
       protective, they would be more costly to implement since they would involve a larger
       portion of the Site.  EPA would be considered impractical if it were to impose a
       residential cleanup standard when the surrounding areas (with the exception of St.
       Vincent's School and Quaker City Marina) are industrial.  Therefore,  EPA has
       performed a risk assessment based on the scenario that future land use will be
       industrial.

             ROD Section IX.E.3 has identified the details of the deed restriction, which
       prohibits development of the land for residential or agricultural uses, and use of the
       groundwater.

             As requested, EPA has contacted the primary coordinator for the Delaware
       River Partnership and City of Philadelphia revitalization effort (Mr. Mark Me Wiggins
       of the GreenSpace Alliances). Our recent discussions with him indicate that the
       proposed bike path is only in the conceptual stages of planning; GreenSpace has
       undertaken an effort to identify owners of area properties in order to determine if right-
       of-way acquisitions are feasible.  Also Mr. McWiggins has indicated the  City has
       generally bypassed existing obstacles, such as the buildings, when installing its bike
       paths. Therefore the Building Area on-site may not need to be disturbed.

2.     Could the  site fence be constructed so that the building could be put into productive use
       independent of the Southern Portion of the Site?

       Response: A site fence can be installed to separate the buildings from the Southern
       Portion of the Site.  These plans can be developed as part of the Remedial Design or
       even later  as part of the Operation and Maintenance of the Site by any party
       implementing the change.

3.     Page 17 of the Proposed Plan discusses  the Spill Policy with regard to cleanup levels
       but goes on to state that the sampling performed was inappropriate for the Spill Policy.
       EPA should consider using the appropriate sampling techniques in order to determine
       cleanup levels.

                                          144

-------
       Response:  What EPA meant to say in the Proposed Plan was that the Spill Policy
       (specifically 40 C.F.R.  § 761.123) would have required a less conservative wipe test.
       The RI indicated that various types of samples (i.e., concrete chips, brick chips, wall
       paint, dust or debris) were collected in the Building Area.  Although the exact type of
       each sample was not clearly documented (see AR300040, 300146, and 300148), EPA
       treated all the analytical results as resulting from concrete chip samples that were
       ground up;  the highest being 372 ppm PCB. EPA then calculated the risk it posed to
       industrial workers based on  a scenario that on any given day, that worker would
       accidentally ingest a limited  amount of this dust as a result of working inside the
       buildings. EPA concluded that risk was only 3.86x10-6 (see AR301580) and, therefore
       acceptable.  The Spill Policy (specifically § 761.125(c)(3)(ii)), however, suggested
       PCBs on impervious surfaces should be cleaned to 0.01 ppm.  Since the floors in the
       building are not impervious  and absorb PCBs easily, EPA rationalized that highly
       contaminated samples such as the 372 ppm PCB in the debris would have been
       absorbed into the concrete chips and would be unlikely to be released.  Therefore, EPA
       concluded the Spill Policy's  cleanup level of 0.01 ppm did not apply here because
       analysis of the dust resulting from the chip sample tests was more conservative and did
       not yield any unacceptable risks.

4.     The fifth paragraph on page  18 [of the Proposed Plan] indicates that there is a concern
       regarding the mobility of the PCBs if the building material does not remain intact. If
       the building is eventually demolished, is there a concern as to how the demolition
       debris would be disposed? Is there any threat to a future construction worker
       performing  the demolition of the building?

       Response:  If demolition were to occur, the disposal performance standards would be
       similar to those specified for contaminated sediments and soils. ROD Section EX.B
       specifies the appropriate waste disposal facilities  for the following PCB concentrations:
       (a) non-hazardous & 25-50 ppm PCB, RCRA Subtitle D landfill;  (b) hazardous & 25-
       50 ppm PCB, RCRA Subtitle C facility; (c) > 50 ppm PCB, TSCA landfill.

             The  threat to future construction workers performing demolition would be
       similar to that of future industrial workers ingesting dust imbedded in the concrete of
       the building floor (3.86xlO~7).  There are no unacceptable human  health risks.

5.     River Sediment Areas - The bottom of page 17 of the Proposed Plan indicates that
       some portions of the site would be  10 feet underwater during a 100 year flood.
       Discussions on the permanent sheet pile wall and clean soil cover must address their
       effectiveness during flood conditions.  This should also be addressed in relation to
       Executive Order 11988 in the ARARs section.

       Response:  The Site will actually be at water level during a 100-year flood  according
      •to the Site Owner's calculations (see Volume I, Book 3, Section 2.5, page 72 of Site
       Owner's Comments). Although the RI/FS did not discuss the potential impacts of

                                          145

-------
       flooding conditions, ROD Section VHI.B.3 has cited Executive Order 11988 as a "to
       be considered" ARAR.  EPA will apply criteria specified in this executive order and
       attempt to minimize the impact of floods on the Site's contaminants and design the
       remedy in such a way that it does not interfere with the beneficial values of the
       floodplain in the area.

             The estimated degree of flooding should have no impact on the intended
       purpose of the sheet pile wall and clean cover soil.  The permanent sheet pile wall is to
       keep Site fill from sloughing into the river.  If properly designed, flood waters will
       surround the Site but not overburden the wall from behind. The  soil cover is to isolate
       people from residual contaminants underneath the Site. Since the Site is not expected
       to be under water, flood water should not be able to wash the soil cover away. EPA is
       more concerned about the possibility, as suggested by the Site Owner (see Comment
       #13, NAPL) and PRP Group (see Comment #19), that flooding will raise the water
       table and may cause residual contamination to float and "smear" to the clean surface.
       For that reason, EPA will also require the monitoring of the soil  cover as well as the
       subsurface soil.

6.     General Comment - The Proposed Plan contains many indecisive statements which tend
       to increase confusion about the plan.

       Two examples include:

             Pg. 13, 3rd paragraph, 1st sentence - EPA anticipates that some residual
       contamination may remain after remediation...., should say will remain.

             Pg 18, 4th paragraph -...effectiveness of the remedy if residual PCBs exist.

       Response: The ROD has eliminated all indecisive statements on whether residual
       contamination will remain.  The components of the remedy (e.g., leachate collection
       system) aid in the  containment of residual contamination left on-site after remediation.
                                         146

-------
Exhibit 1:    Observed TPH, PCB and potential DNAPLs in Subsurface Soils.
Location
Bl
(MW1)
B2


B3
(MW3)


B4
(MW4)
B5

B6


B7
(MW14)

B8


B9

BIO
(PI)
Bll
(MW11)



Depth
(feet)
3.0
10.5
2.0
6.0
6.0 Dup
12.0
2.5
6.0
12.0
14.0
2.0
5.5
2.0
5.5
9.5
2.5
6.0
12.0
12.0 Dup
3.0
6.5
8.5
10.5
10.5 Dup
2.5
6.0
6.0 Dup
12.0
14.0
14.0 Dup
3.0
6.5
10.5
14.5
6.0
8.0
8.0
10.0
14.0
2.5
6.0
10.0


TPH
(mam)
2,670
3.600
3,010
1.320
1,340
754
2,150
359
3,120
697
NT?
NT?
2.850
211
5,620
147
355
253
314
861
1.480
262
3.595
1,117
4,290
510
645
6.310
3,410
3,010
5,690
4,780
1.670
4,470
3,970
4.180
4.590
3.880
2,140
148
232
106


PCB
(DDm)
1.7
!Ez8.o :
0.95
ND
1.13
ND
0.24
ND
ND
ND
10.2
2.48
5.13
0.70
4.81
ND
0.33
ND
ND
1.07
4.65
ND
ND
ND
4.22
ND
ND
ND
ND
ND
ND
0.39
ND
6.65
ND
29.0
15.0
11.7
0.38
1.0
0.25
ND


Profile
(feet)
(Boom 16.5)
oil 6.5 - 13
wafer O 10.71 1

mere 11.3
(Booom 14.0)

wster 9 12.5
(Booom 18.0)
(Bourn 29.5)
water O 15.0 I

(Beam 17.5)
mm ffl 10.0 I

water® 12.0
(Booom 16.0)
water ® 10.0
(Boom 16.5)

water® 11.8
(Booom 18.0)

No water
(Booom 16.5)
(Bottom 18.0)
Water® 14.0
BUD 2.0- 1 6.75
mter ® 9.75
(Booom 18.0)


Location
B12
(MW12)

B13


B14



. B15
(P2)


B16

BI7

B18


B19
B20
B21
(MW1S)



Depth
(feet)
1.0
7.0
11.0
11.0 Dup
5.5
9.0
11.0
11.0 Dup
3.0
6.5
10.5
10.5 Dup
12.5
3.0
8.5
10.5
12.5
12.5 Dup
2.0
5.5
11.5
11.5 Dup
3.0
6.5
10.5
12.5
8.0
10.0
12.0
14.0
14.0 Dup
3.0
6.5
8.5
2.0
6.0
6.0 Dup
8.0
16.0
3.0
6.5
6.5 Dup
12.5
14.5
TPH
(DDQ)
266
ND
ND
6,890
922
510
4,740
6,890
12.400
863
6,570
4,730
4,080
857
12,800
10,800
8,940
6,650
10,850
13,190
1,416
1,436
11.500
17.400
2.630
4.160
6,480
9.130
8,780
8,090
10,000
NT?
NT?
NT?
NT?
NT?
NT?
NT?
NT?
5,655
1,402
1,402
6,429
2,369
PCB
(DVDI)
3.2
ND
ND
ND
0.6
ND
0.78
ND
3.2
ND
7.0
18.2
9.1
ND
5.0
ND
ND
ND
ND
3.85
0.4
ND
26.0
42.0
ND
3.0
0.14
0.97
3.2
7.54
7.45
3.6?
6.3?
ND?
ND?
6.1?
13.7?
3.6?
0.37?
25.0
6.5
3.6
ND
8.7
Profile
(feet)

(Booom 18.0)
water® 12 1


(Booom 16.0)
water® 12.5 1


(Booom 18.5)
water A 12.5
sain 2.5 • 12.5
water® 11.2
(Booom 20.0)
oil&suins®
(Bottom 17.5)
water® 13.25 1
ttaio4.5- 16.5
(Booom 26.5)
water® 12.5
PtTTff £ pain*
® 10 - 16
water O 14.0
(Booom 28.0)
stain 3.5- 19.5
(Bottom 20.0)
water 9 15.25 1
peuo& stains
® 4.5 • 20.75
(Bottom 26.0)
water® 16.0


water 913.25
(Bottom 20.5)
                                  Sheet 1 of 2
                                                                    Metal Bank ROD

-------
Exhibit 1:     Observed TPH, PCB and potential DNAPLs in Subsurface Soils.

Location
B22




B23




SB101


SB102



SB 103


SB 104


SB 105



SB106





Depth
(feet)
2.5
6.0
8.0
8.0 Dup
14.0
3.0
6.5
6.5 Dup
8.5?
10.5
6.0
8.0
10.0
6.0
8.0
10.0
12.0
8.0
11.0
15.0
10.0
12.0
14.0
8.0
10.0
12.0
14.0
16.0
6.0
8.0
10.0
12.0

14.0
TPH
(mm)
10.400
5,600
15,800
15.600
9,770
4.860
621
552
25,900
13,600
NT
3.500
NT
NT
NT
NT
2,100
NT
NT
870
NT
NT
3,100
NT
NT
NT
NT
NT
NT
NT
NT
2,800

NT
PCB
fonn)
ND
2.8
5.2
6.6
0.34
7.0
ND
ND
7.8
4.7
(70-50^
5.3
«5)
NT
fe^NT
p24.6
r4 NT
(70-50)
(5-10)
0.75
(10-50)
(10-50)
7.3
(70-50;
(10-50)
3.7
Ef.,i2.o •.
R 12.0
1.8
ND-
5.5
ND(70-
50)'
&i$jo Y- :.'
Profile
(feet)



(Booom 76.0)
water O 14.0

•aim O 2.5-21

(Booom 30.5)
water O 12.1 I
water & 7.0

(Booom 12.0) .
water 9 7.0 '
nain 10 -24 .

• (Booom 37.0)

water 0 14.5
(Bottom 34.0) '
OU08- 18
(Bourn 18.0)
water O 14.0

Bain 14 -18

water 9 14.0
(Bottom 44.0)

DO sains

(Boojxn 16.0)

water 9 14.0

Location
SB107




SB108




TP-1


TP-4


TP-8


TP-9


TP-12

TP-T1


TP-T2



Depth
(feet)
6.0
8.0
10.0
12.0
14.0
6.0
8.0
10.0
12.0
14.0
5.0
10.0
12.0
5.0
10.0
11.5
5.0
10.0
15.0
5.0
10.0
13.0
5.0
10.0
5.0
10.0
15.0
5.0
10.0
12.0

TPH
from)
NT
NT
NT
NT
NT
NT
NT
NT
NT
NT
1.400
7,000
24,000
52,000
12,000
28,000
180
27
1,210
3,040
630
380
1,730
2.190
28,900
13.300
52.800
66,100 j
51.200
16.000 1

PCB
fpprn)
1.8
1.4
ND
0.75
1.1
ND
ND
ND
ND
ND
<2
22.0
10.0
£36.0
r 45.0
£50.0
<2
<2
2.0
35.0
<2
<2
4.4
7.7
felOS.O
^10.5
• 57.0
E230.0
Nso.o
*~77.0

Profile
(feet)

water 4) 8.0
oo tarns

(Booon 16.0)
water 9 7.0

oo stuns

(Bottom 16.0)
Profile
Nat
Available
Profile
Not
Avaibble
Profile
Not
Available
Profile
N«
Available
Profile Not
Available
Profile
Not
Available
Profile
: Not
Available

                                                                 TP-7, TP-11, TP-14.TP-16, TP-18, TP-21 & TP-T3 not
                                                                          presented since PCBs are <2 ppm.
SOURCE: RI Vol 1 for Soil Boring ("BVSB") data; PRP
(Jordanger) 9/11/95 letter for Testpit (*TP*) data.
REFERENCES
Location
TPH
PCB
Profile
"TP'data
TABLE
2-2 to 2-5
4-11
4-10
2-2
2
FIGURE APPENDIX
2-5
c.i.c.n
c.i. c.n
A (Boring Logs)
1 to6
                                                            NOTES:
                                                            1.    ND = Not Detected; NT = Not Tested
                                                            2.    ? = Data inconsistency. Value in RI Volt but no backup data
                                                                 in RI Vol 2 (Appendix C.I. C.H) or vice-versa.
                                                            3.    (10-50) 'italics" indicate field screening values.
                                                            4.    SHADING indicates possible DNAPL.
                                                            5.    "superscript" <" indicate high detection limits (low: .-high-total).
                                                                 Example: (SB 106-8) =  4.2 ::8.6: :33.8indicateslowest& highest
                                                                 detection limits of individual Aroclcrs plus total PCBs are 4.2.
                                                                 8.6, and 33.8 ppm, respectively.
                                                            Other Examples:  (SB106-10)  =  5.2::11::37;  (SB106-12)  =
                                                            2.1::4.2::16.8.   (SB106-14) =  4.8::9.8::33.8;  (TP-T2-5)  =
                                                            30::30::410.
                                                            6.    1 « indicates occuraoce below profile.
                                                 Sheet 2 of 2
                                                                                                    Metal Bank ROD

-------