Final Report
Neal's Landfill Fish Assessment
Rec'd Feb 19, 2002
LIST OF FIGURES
Figure 1: Fish Sampling Locations
Figure 2: Fillet Homogenization and Processing Procedure
Figure 3: 2001 Mean PCB Concentrations in Fish Tissue
Figure 4: Comparison of Mean PCB Concentrations in Fish Tissue 1998 & 2001
Figure 5: Fish Types Sampled
Figure 6: PCBs in Fish
1.0 INTRODUCTION
From August 7-9, 2001 the Response, Engineering, and Analytical Contract (REAC) provided
technical support to the United States Environmental Protection Agency/Environmental Response
Team Center (U.S. EPA/ERTC) and the U.S. EPA Region V to evaluate the concentration of
polychlorinated biphenyls (PCBs) in fish species associated with the Neal's Landfill Site. This
document summarizes the results of the chemical analyses of fish tissue samples representing target
species commonly found in the watershed surrounding Bloomington, Indiana (IN).
2.0 SITE DESCRIPTION AND HISTORY
The Neal's Landfill site occupies nearly 18 acres of property approximately three miles west of
Bloomington, IN, and is surrounded by farms and woodlands. Several residents are located within
a half-mile of the landfill and use private wells for drinking water. The landfill accepted industrial
and municipal wastes from 1949 until 1972. From 1962 until 1970, Westinghouse Electric
Corporation, currently doing business as CBS Corporation, disposed of waste electrical equipment
and parts in the landfill, including electrical capacitors containing PCBs, PCB-contaminated
capacitor insulation material, rags, and filter clay. Several springs are located at the foot of the
landfill and feed Conard's Branch, a tributary to Richland Creek. As a result, sediment and fish in
Conard's Branch and Richland Creek were contaminated with PCBs. In 1985, Westinghouse, the
U.S. EPA, the city of Bloomington, Monroe County, and the Indiana State Board of Health signed
a Consent Decree (CD) that required Westinghouse to perform interim control measures, to construct
an incinerator, and to incinerate PCB contaminated materials from six sites in and near Bloomington,
including Neal's Landfill. During the early 1990s, the state of Indiana passed a number of laws that
initially delayed and ultimately blocked the construction of the incinerator remedy required by the
1985 CD. Beginning in 1994, the parties associated with the CD began to explore alternative
remedies for the PCB sites (McLaren/Hart-ChemRisk 1998).
During 1987, interim control measures at the site were implemented by Westinghouse which
included: removal of visible capacitors and stained soils, installation of a two foot thick clay cap
overprimary landfill areas, installation of a locked chain-link security fence around the site, removal
of sediments and creek banks along the entire 4,500 foot length of Conard's Branch, and installation
of a collection and activated carbon treatment system to treat PCB-contaminated spring water.
Sampling of creek sediment, water, and fish in Richland Creek demonstrated that the spring water
treatment plant has abated the PCB loading to Richland Creek. In February 1996, the parties
submitted a schedule to the federal court that identified the specific steps needed to select alternative remedies for each of the six sites subject to the 1985 CD. In February 1997, the parties submitted an amended schedule to the court. The court stated through a judicial order that the landfill must be remediated by December 31, 2000. Future decisions by the U.S. EPA on expansion of the water
treatment plant and sediment removal will be completed in 2002 and are not included in the judicial
order. On March 29,1999, the U.S. EPA signed a Record of Decision (ROD) Amendment for Neal's
Landfill which addresses the source of the contamination.
On April 19,1999 the first phase of the cleanup at Neal's Landfill was initiated by CBS Corporation.
A total of 41,747 tons (83,495,000 pounds) of contaminated material with greater than 500 parts per
million (ppm) PCBs was excavated and shipped offsite to a landfill capable of accepting PCB
contaminated material. A total of 4,119 capacitors containing PCBs and weighing 484,624 pounds
were excavated and shipped off-site to a commercial, permitted incinerator. Approximately 90,000
cubic yards of landfill material were consolidated to reduce the landfill size from 18 acres to 10
acres. And 29,000 tires were excavated and shredded on-site and placed under the landfill cap. The
final landfill cap met the permeability requirements of less than I x 10-7 centimeters per second.
Areas outside the fence met the cleanup criteria with an average PCB concentration of 3.3 ppm. The
groundwater and surface water monitoring plans for Neal's Landfill are currently under development
(McLaren/Hart-ChemRisk 1998).
3.0 METHODS
Fish were sampled in conjunction with the Indiana Department of Environmental Management's
(IDEM) annual monitoring program. Fish were collected from a total of six locations: five locations on Richland Creek and one location on Conard's Branch (Figure 1). A total of 8 omnivorous fish
and 5 forage fish samples were targeted for collection from each location. Additionally, 6 upper
trophic level fish were targeted for collection from the 3 most downstream locations. A total of 87
fish samples were collected, of which 55 were analyzed as whole body, 18 were analyzed as fillet
composites, and the remainder were archived in the REAC biology laboratory, Edison, New Jersey
(NJ). Table 1 lists the species and number of fish caught at each location. Targeted omnivorous fish
included green sunfish (Lepomis cyanellus) and longear sunfish (Lepomis megalotis) while targeted
forage species included creek chub (Semotilus atromaculatus), white sucker (Catostomus
commersoni), and carp (Cyprinus carpio). The targeted upper-trophic level species were largemouth
bass (Micropterus salmoides) and rock bass (Ambloplites rupestris), but spotted bass (Micropterus
punctulatus) was substituted if either of the target species was not available.
At each location, field personnel attempted to catch eight omnivorous fish and five forage fish using
electro-shocking equipment administered by IDEM personnel. In addition, six upper-trophic level
fish (3 fish of 2 species), were collected from the three most downstream locations (i.e. 4,5, and 6).
At Locations 6, 5, and 4 a barge electro-shocker was used while a backpack electro-shocker was used
to collect the fish at Locations 3, 2, and 1.
3.1 Sampling Locations
The sampling locations were as follows (Figure 1):
Location 1: Downstream of the landfill on Conard's Branch. In the same vicinity as the
former CBS #1 sampling location.
Location 2: The most upstream location on Richland Creek. In the same vicinity as the
former CBS #2 sampling location.
Location 3: Approximately 3 miles downstream of Location 2 on Richland Creek, north
of Hendricksville. IDEM upstream annual monitoring location.
Location 4: Downstream of Location 3 on Richland Creek, south of Hendricksville and
upstream of the Route 43 bridge. In the same vicinity as the former CBS
#3 sampling location.
Location 5: Downstream of Location 4 on Richland Creek. IDEM midpoint annual
monitoring location.
Location 6: Downstream of Location 5 on Richland Creek both upstream and
downstream of the Rt 54 bridge. IDEM downstream annual monitoring
location.
3.2 Fish Tissue Sampling
Fish were processed according to REAC Standard Operating Procedure (SOP) # 2039 Fish
Handling and Processing. Each fish was weighed, measured for total length, and observed
both externally and internally for obvious abnormalities (e.g., neoplasms, parasites, etc.).
This information was recorded on fish data sheets (Appendix A). Forage fish and
omnivorous fish were analyzed using whole bodies. Since the mass of some individual
forage and omnivorous fish were not sufficient to meet analytical requirements, several
similarly-sized smaller bodied fish of the same species were composited. The required
number of samples per location was then comprised of both individual fish and composited
fish with no more than 3 composites (of the 13 total samples) per location. Fillet samples
were taken from the larger, upper-trophic level species. Due to the small body size of many
of these fish, fillets from both sides of the fish were utilized to obtain sufficient tissue mass
for analytical requirements. Fish were assigned sample numbers, wrapped in aluminum foil,
and placed in plastic bags. The sample number, and requested analyses were recorded on
Chain of Custody forms. The samples were shipped on dry ice, via overnight delivery, to
the REAC laboratory in Edison, NJ for homogenization.
All fish samples were homogenizedaccording to REAC SOP #1820 Tissue Homogenization
Procedure. Fish targeted for whole body contaminant determinations were homogenized
individually or in composites of same species and approximate size, depending upon the
mass of the fish relative to sample mass requirements for analyses. The fillet samples were
homogenized using the following procedure (Figure 2 is a graphic description of the fillet
sample handling). Each individual fillet was homogenized separately. Two aliquots were
removed from each fillet homogenate (e.g., Fillet #1 = Aliquot 1 and 2; Fillet #2 = Aliquot
3 and 4; Fillet #3 = Aliquot 5 and 6). Aliquots 1, 3, and 5 were combined and analyzed for
PCB and co-planar PCB congeners using a soxhlet extraction. Aliquots 2, 4, and 6 were
combined and analyzed for PCBs and co-planar PCB congeners using a micro-extraction
technique, as part of a separate technical evaluation. This yielded a total of 1 composite
homogenate sample to be analyzed per species per location, for this effort, (e.g., one for
PCBs using a soxhlet extraction and one for co-planar PCB congeners using soxhlet
extraction). Once homogenized, the fish tissue samples were prepared for shipping to the
subcontract laboratory per REAC SOP #2004, Sample Packaging and Shipment.
The fish sample extracts for Aroclor analysis were cleaned up using a sulfuric acid clean-up
procedure. Extracts for the congener analyses were cleaned up using a silica followed by
acid followed by florosil sample clean-ups.
Due to an error in the analyses directions given to the analytical laboratory, lipid
determinations were not conducted on the initial sample homogenates sent for PCB
determinations. Upon realization that the lipid determinations had not been conducted the
error was identified. Inventory of homogenized tissue which remained was made; fish
samples had sufficient mass remaining were send to the laboratory for total lipid
determinations. These samples were stored at -20° Celsius freezer for approximately 12
weeks after homogenization.
3.3 Sampling Equipment Decontamination
Non-dedicated sampling equipment was decontaminated prior and subsequent to sampling
each location per REAC SOP #2006, SamplingEquipmentDecontamination. The following
numerical sequence was used:
1 physical removal
2 nonphosphate detergent wash (Liquinox)
3 potable water rinse
4 distilled/deionized water rinse
5 10% nitric acid rinse
6 distilled water rinse
7 acetone rinse
8 distilled water rinse
9 air dry
3.4 Standard Operating Procedures
Sample documentation was completed per the following REAC SOPs:
- REAC SOP #2002, Sample Documentation
- REAC SOP #4005, Chain of Custody Procedures
- REAC SOP #4001, Logbook Documentation
Sample packaging and shipment was conducted in accordance with the following REAC
SOP:
- REAC SOP #2004, Sample Packaging and Shipment
3.5 Waste Disposal
Investigation derived waste (e.g., personal protective equipment) was disposed of in
accordance with all state and federal regulations. All samples were maintained per the work
plan.
4.0 RESULTS
Fish tissues (whole body and fillet) were analyzed for PCBs, co-planar PCB congeners, and percent
lipids. If a PCB concentration was not detected (ND) for a given sample, it was assumed to be
present at one tenth of its method detection limit (MDL) and this value was used to calculate mean
concentrations. The justification for the use of one tenth the MDL is as follows: for "clean samples"
the analytical procedure would identify a compound as present with a "J-ed" value, estimated
concentration, down to one fifth of the MDL. For averaging purposes one half the detection limit
is frequently used when non detect values exist in a data set. Therefore, if no signal for the
compound was found (non-detect), one half of the value which would have been reported as being
present (one fifth of the MDL) was used (one half of one fifth is one tenth). The final validated
analytical results can be found in Appendix B.
4.1 Results for PCBs in Fish Tissue
Aroclor 1248 was the only PCB detected in fish tissue and was detected at all locations
except Location 6 (Table 2). Concentrations of Aroclor 1248 ranged from below the method
detection limit (MDL) for all fish at Location 6 to 9.0 mg/kg in green sunfish at Location
1. Concentrations of Aroclor 1248 decreased from upstream to downstream locations. Fish
from Location 1 had the highest concentrations of Aroclor 1248, ranging from 5.9 to 9.0
mg/kg.
Means PCB concentrations were calculated for each location and for each species at every
location (Table 3). The mean location concentrations also decreased from upstream to
downstream locations (Figure 3). The mean Aroclor 1248 concentrations ranged from
0.01mg/kg (Location 6) to 6.93 (Location 1).
4.2 Results for Co-planar PCB Congeners
Fish tissue collected from Richland Creek and Conard's Branch was analyzed for co-planar
PCB congeners. Congener analysis was not performed on the data collected in 1998 and,
therefore, no comparisons between sampling events can be made. PCB congener
concentrations followed the same trend as the Aroclor 1248 concentrations. Analytical
results are summarized in Tables 4a,4b,4c, 4d,
& 4e.
4.3 Results for Lipids in Fish Tissue
Due to a technical error, total lipids were analyzed from fish tissue samples stored in a -20ø
Celsius freezer approximately 12 weeks after homogenization. Given this lapse in time, (i.e.
between homogenization and lipid analysis) total lipid results should be viewed with caution
as lipid degradation may have occurred resulting in lower total lipid concentrations (Table
2). These lipid concentrations were also used to calculate lipid normalized data. Lipid
normalized data was calculated by dividing the PCB concentration into the percent lipid
concentration and dividing by 100. As such, lipid normalized PCB concentrations should
be viewed with caution. In addition, lipid normalized data comparisons between locations
could not be performed as some samples did not have enough leftover tissue homogenate
to analyze for lipids.
5.0 DISCUSSION
A non-statistical comparison was done between mean PCB concentrations from the 2001
sampling event and mean PCB concentrations from the 1998 CBS sampling event (Table 3). Data
for the 1998 CBS sampling event can be found in Appendix C. Means were calculated for each
location. Data was only available for Locations 1, 2, and 3 from the 1998 sampling event and,
therefore only these locations were compared to the 2001 data. For this comparison, it is assumed
that Locations 1,2, and 4 from the 2001 sampling event are comparable to Locations 1,2, and 3 from
the 1998 sampling event.
A graph was created to visually compare the data from 1998 and 2001 (Figure 4). Apparent
differences in the data can be discerned from this graphical representation. For example, mean PCB
concentrations from Location 1 appear to be greater in 1998 than in 2001. Mean concentrations of
PCBs from Location 2 appear to be slightly greater in 2001 than 1998. Mean PCB concentrations
from Location 3 appear to be greater in 1998 than mean PCB concentrations from Location 4 in
2001.
6.0 CONCLUSIONS
Concentrations of PCBs are still present within the Conards's Branch and Richland Creek watershed.
Concentrations of PCBs are highest at the location closest to the landfill (Location 1). Using a non-
statistical comparison, the concentrations of PCBs appear to decrease from upstream to downstream
locations.
7.0 REFERENCES
McLaren/Hart-Chem Risk. 1998. "Neal's Landfill Stream Investigation Field Sampling Plan".
Prepared for CBS Corporation. Pittsburgh, PA.
|
