Final Report
Completion of the Removal Action for the Trickling Filter Area
Winston-Thomas
April 22, 1999
PSARA Technologies for CBS Corp.
1.0 INTRODUCTION
This document constitutes the final report for the Completion of the Removal Action for the trickling filter area at the Winston Thomas site in Bloomington, Indiana. Figure 1 shows the location of the Winston Thomas site. Figure 2 shows the trickling filter area within the Winston Thomas site. The work described in this report was performed during May, June, and July of 1998.
This report includes a descriptive narrative of work performed, along with associated figures, tables, and photographs. Additional records associated with this project, including air monitoring results, weekly progress reports, laboratory certificates of analysis, chain-of-custody documents, etc., are available for review and future reference at CBS Corporation in Pittsburgh, Pennsylvania.
2.0 BACKGROUND
2.1 GENERAL
The Winston Thomas wastewater treatment plant is owned by the City of Bloomington and was operated from 1933 until 1982, when the plant was closed. From 1958 to 1982, the Westinghouse facility in Bloomington used the sewer system associated with the Winston Thomas facility. During this time period, polychlorinated biphenyls (PCBs) from the manufacture of capacitors at the former Westinghouse facility were occasionally discharged to the treatment plant. Consequently, sludge, water, and waste materials associated with many of the Winston Thomas process units became impacted with PCBs and subject to remediation under a Consent Decree issued for the site in 1985. These units include the sludge digester tanks, sludge drying beds, tertiary lagoon, trickling filter, abandoned lagoon, and sludge piping associated with the digesters and drying beds. Since closing the Winston Thomas facility in 1982, the City of Bloomington has continued the maintenance of the former treatment plant site.
The Winston Thomas facility is located in Monroe County, approximately 2.5 miles south of downtown Bloomington, Indiana (Figure 1). The facility is bounded by Old State Road 37 South to the east, Gordon Pike to the south, Clear Creek to the west, and a business area to the north. The inactive facility covers an area of approximately 26 acres. Seventeen acres comprise the tertiary lagoon, which contains 2 to 5 ft of water. The facility is surrounded by a fence to discourage public access. In 1997, anaerobic digesters, three drying beds, sludge piping associated with the digesters and drying beds, an area east of the entrance road, and an area west of Clear Creek were remediated. In 1998, two areas at the Winston Thomas facility were remediated under the Consent Decree, including the abandoned lagoon and the trickling filter. In addition, remediation activities revere initiated at the tertiary lagoon. Remediation of the tertiary lagoon will be completed in 1999. Each of these areas is accessible via a network of gravel tributary roads tied tothe main entrance off West Gordon Pike at the south end of the facility.
The specific area addressed by this report is the trickling filter, which is located south of the sludge drying beds as shown in Figure 2. The scope of work for the trickling filter remediation project is described in the Work Plan dated May 5, 1998.
2.2 SITE DESCRIPTION
The trickling filter is a 280-ft by 245-ft subgrade concrete-walled structure that was filled with angular limestone gravel (2 to 3 in.) to an approximate depth of 7.5 ft below grade. The bottom of the structure includes a cast-in-place floor slab with two 18-in. by 30-in drainage tunnels, which trend north-south across the entire structure. Clay drainage tiles were installed over the entire floor surface and were connected into the drainage tunnels. A network of horizontal and vertical pipes was located within the bed of crushed stone. The horizontal piping included supply pipes as large as 30 inches in diameter, and the vertical piping included riser pipes and spray heads approximately 1 inch in diameter.
The trickling filter was used to provide biological treatment of vvastewater effluent from the primary settling tanks. Liquid waste was broadcast over the surface of the trickling filter through the horizontal supply piping and vertical riser/spray head assemblies. The wastewater percolated through the stone filterpack and clay drainage tiles, collected in the underdrain tunnels, and was finally discharged to the secondary clarifiers.
2.3 PREVIOUS SAMPLING ACTIVITIES
In 1982, Westinghouse collected samples of the drainage media within the trickling filter, including samples of the crushed stone (pulverized for laboratory analysis) and debris, which had accumulated inthe void spaces. In general, the results indicated that PCB content within the crushed stone was very low (approximately 0.1 ppm). The concentration of PCBs within the debris material ranged from 2 ppm to 146 ppm.
In 1997, CBS collected water samples within the manholes at both ends of the two drainage tunnels. The concentration of PCBs detected in the water samples ranged from 0.57 ppb to 2.4 ppb .
3.0 SUMMARY OF REMEDIATION ACTIVITIES
A Work Plan, dated May 5, 1998, was prepared to address removal of the filter media and cleaning ofthe exposed trickling filter surfaces. This plan was reviewed and approved by the Consent Decree parties in May 1998 prior to the start of excavation activities at the site. Health and safety, air monitoring, post-cleaning verification sampling, and transportation and disposal were conducted in accordance with the approved Work Plan.
All remediation activities were performed by Blasland Bouck & Lee (BB&L) under the direction of CBS and in cooperation with the site owner, the City of Bloomington. Project photographs showing remediation activities and processing equipment are presented in Appendix A.
3.1 SITE PREPARA TION
Prior to beginning remediation activities, site security and support facilities were arranged for the site. In addition, a processing area was constructed in an area immediately north and west of the trickling filter as shown in Figure 3. This area comprised a portion of the asphalt access road located along the west side of the trickling filter. The remainder of the processing area was constructed using a granular sub-base and bermed asphalt surface. Equipment located in the processing area included dry screen, conveyor, wetscreen, and water handling equipment. Space was also provided in this area for stockpiling waste materials prior to being loaded for offsite disposal. In addition, a temporary decontamination area was constructed as part of the processing area. The decontamination area was used for equipment and personnel decontamination, as well as for washing truck tires that passed through the material loading station.
3.2 REMEDIA TION OF THE TRICKLING FILTER
Remediation of the trickling filter was accomplished by removing the filter bed, which was separated into two size fractions for cleaning and reuse (oversize material) or disposal in a landfill (undersize material and small-diameter pipes), followed by pressure washing of exposed surfaces, concrete blocks, and large-diameter pipes. The following subsections describe the demonstration and startup testing of the process equipment and subsequent remediation efforts.
3.2.1 Demonstration and Startup Testing
Prior to removal of the filter media, two test pits, one in the southeast corner and one in the northwest corner of the trickling filter were dug using a backhoe. The test pit material was visually examined and determined to be relatively uniform in both size and surficial buildup. This material was stockpiled forfuture use in the demonstration and testing of the processing equipment.
The demonstration and startup testing program involved the processing of the drainage media under arange of operating conditions. The purpose of this exercise was to confirm that the remedial technology was effective in removing PCBs from the surface of the drainage media and to determine the full-scale operating parameters of the remedial process. This program was begun on June 1, 1998, using a feed rate of 100 tons per hot r and a screen size of 1 in. The water spray was set at 40 percent of full capacity. Two additional test runs were conducted on June 3, 1998, using processing rates of 150 and 200 tons per hour.
Based on analytical results from samples taken during the startup testing (described in Section 4.1), a feed rate of 200 tons per hour and a sieve size of 1 in. were selected as operating parameters for the oversize material.
3.2.2 Filter Bed and Small-Diameter Pipe Removal
Remediation activities were initiated on June 4, 1998. A track excavator was used to excavate materials from remote portions of the trickling filter and to place these materials in an intermediate stockpile located in the central portion of the trickling filter. A second track excavator then transferred these stockpiled materials into the feed hopper. The hopper included an inclined bar screen to reject any large material thatcould block or jam the processing equipment.
Material passing through the bar screen fell onto an inclined feed conveyor, which transferred the material onto a dry vibrating screen. Material retained on the vibrating screen (oversize material) was passed onto another transfer conveyor leading to a wet screening operation, where the material was rinsed. The clean material dropped off the wet screen and onto a radial stacking conveyor, which transferred the processed material to the stockpile area.
During the first 9 days of excavation and treatment, wipe sample results of the crushed stone showed an upward trend in PCB content. Pursuant to the Work Plan, the remediation process was modified before the wipe sample results exceeded the clean criterion. As part of the remedy, fresh water was added to the system and 5-micron filters were installed in place of the original 100-mesh bag filters. Later, 25-micronfilters were used instead of the 5-micron filters, which tended to clog with sediment and restrict flow. Prior to this event, samples of the recirculated process water were collected every 3 days to determine PCB content. Following this event, dirty process water was pumped pumped through two bag filters and directed to a 20,000-gal batch tank. Fresh water was then added to the system weekly.
The actual feed rate during remediation averaged approximately 150 tons per hour. The recirculated process w ater spray ranged from 50 to 100 8pm during washing of the oversize stone. These parameters were maintained during the production operation. A total of 23,851 tons of crushed stone was processed and stockpiled in the area of former drying bed #3.
Material that passed through the dry vibrating screeening (undersize material) dropped a transfer conveyor and was moved to a temporary waste stockpile area or placed directly into lined rolloff boxes; this material represented the "waste" portion of the material flow. All undersize material was determined to be non-Toxic Substances Control Act (TSCA) waste (i.e., below 50 ppm PCBs) and was removed and disposed of at the Southside Landfill in Indianapolis, Indiana. A total of 70 loads of debris and sediment totaling approximately 1,036 tons were disposed of as special waste at this landfill during the course of the project.
Small-diameter (6-in.) pipes from within the filter media were broken into approximate 3-ft lengths andstockpiled on untreated crushed stone within the trickling filter. Based on composite sediment sample results from inside these pipes (see Section 4.2), the small diameter pipes were discarded at a TSCA-permitted landfill in Belleville, Michigan, operated by Environmental Quality. A total of seven rolloffs containing approximately 1()9 tons of small-cliameter pipe were disposed of at this landfill.
Samples of the recirculated process water were collected every 3 days to determine PCB content.
After the first 9 days of excavation and treatment, a second 20,000-gal holding tank was added to theend of the system for batch treatment of excess water accumulation due to high storm water flow and for treatment and disposal of water at the end of the project. Figure 4 shows the configuration of the water treatment system.
Over the course of the project, a total of 363,000 gal of wastewater was discharged to the sanitary sewer in accordance with the City of Bloomington Utilities Industrial
Pretreatment Permit. Of this total, approximately 45,000 gal was process water and the remainder was storm water. The discharge limit for PCBs was 0.01 mg/L.
Upon completion of wastewater treatment activities, the spent bag filters and filter residue were disposedof as TSCA waste at the Environmental Quality Co. landfill in Belleville, Michigan; the exact quantities ofthe materials disposed are not known.
4.0 SUMMARY OF SAMPLING ACTIVITIES
All sampling activities were conducted by PSARA Technologies, Inc., in accordance w ith the approved Field Sampling Plan.
4.1 DRAINAGE MEDIA SAMPLING
4.1.1 Pre-Treatment Test Pit Sampling
As described previously, two test pits were dug in the trickling filter to provide material for use in the demonstration and testing of the processing equipment. In addition to visual observation, four grab samples of crushed stone exhibiting dark staining were collected at approximate 2-ft depth intervals to the bottom of the filter bed at each test pit location. These stones were sampled and analyzed to providea baseline for the subsequent demonstration and testing program. The entire surface of each stone was wiped with new gauze soaked in hexane, and the surface area was calculated using the formula fordetermining the area of a sphere (4pir2). ). The radius (r) was determined by measuring the maximumand minimum diameters of the stone, calculating the average diameter, and dividing the result by 2. The wipes were then submitted to an offsite laboratory for PCB analysis. Table 1 presents the results of the pre-treatment sampling. Baseline PCB concentrations ranged from non-detect to 1.5 ug/100 cm2.
4.1.2 Demonstration and Startup Test Sampling
Process optimization sampling was conducted after each demonstration test run by collecting fourrandom grab samples in accordance with the Field Sampling Plan. The treated stones were sampled andthe surface area calculated as described previously. Samples were analyzed for PCBs. Table 2 presents the results of the process optimization sampling. PCB concentrations were below laboratory detection limits for each set of conditions tested.
4.1.3 Daily Operational Sampling During Filter Bed Removal
To ensure that the remediation process was meeting the clean criterion, samples of the clean stone reveretaken daily from the output of the wash system. Five stones were randomly grabbed during each day'soperation and wipe sampled in accordance with the Field Sampling Plan. Samples were analyzed forPCBs. The analytical results from these five samples were averaged to obtain a representative number fo rthat day's operation. This daily average was compared to the project clean criterion of 10 ug/100 cm.Table 3 presents the analytical results and daily averages for the daily production run samples for the period from June 4 through June 22, 1998 All production runs achieved the cleanup criterion after the first pass.
4.1.4 Debris and Undersize Material Sampling
Smaller stones, organic < material, and other debris that passed through the vibrating screenswere stockpiled as a separate waste stream. This material was sampled by collecting grab samples periodically as the waste was generated.
Sampling During Demonstration and Startup Testing
Each of the batch tests conducted during the demonstration and startup phase of the project generated asmall quantity of debris. A single sample was collected from each batch operation. Upon completion of the demonstration and startup phase, one composite sample was prepared from the individual batch samples and submitted to the laboratory for analysis. Table 4 presents the laboratory analytical result. The PCB concentration in this sample was below the TSCA limit of 50 ppm.
Daily Operational Sampling
Debris waste was contained in lined rolloff boxes in the process area as the material came off the dry screening process. In accordance with the Work Plan, debris stockpiles comprising approximately 50 to 60 cu vd of material were to be represented by at least four grab samples. Since the debris retaste was contained in rolloff boxes instead of stockpiles, every fourth rolloff box was sampled (four rolloff boxeshold less than 60 cu yd of material). Once the fourth rolloff box was filled with debris, four grab samples were taken from the waste material contained in the box, composited into a single representative sample,and submitted for laboratory analysis. Table 4 presents the laboratory analytical results of the debris waste sampling. All waste materials generated by the screening process were determined to contain less than 50 ppm PCBs and thus classified as non-TSCA. Consequently, these materials were disposed as special waste at Southside Landfill
(see Section 3.2).
4.2 REMEDIATION AND VERIFICATION SAMPLING OF TRICKLING
FILTER SURFACES
Interior Concrete Surfaces
Following removal of the crushed stone and clay tiles from the trickling filter, the concrete walls, floor, and trenches were spray washed, and the resulting sediment was removed. Surface wipe samples were then collected from the interior concrete surfaces of the trickling filter using a 100-sq-cm template in accordance with the Field Sampling Plan. Locations were selected in the field by representatives of CBSand the City of Bloomington. Two representative samples were collected from each wall of the trickling filter. In addition, two samples were collected from each of the three sections of floor, and one sample was collected from the inside surface of each of the trenches in the floor of the trickling filter. Figure 5 shows the sample locations. The wipe samples were submitted for laboratory analysis for PCBs. Table 5 presents the laboratory analytical results of the surface wipe sampling. The wipe sample results were below the cleanup criterion of 10 ug/100 cm: for all sample locations.
4.2.2 Piping and Concrete Blocks
Piping removed from the trickling filter was also wipe sampled using a 100-sq-cm template as provided in the Field Sampling Plan. Small-diameter (6-in.) pipes from within the filter media were broken into approximate 3-ft lengths and stockpiled on untreated crushed stone within the trickling filter. Wipe samples were collected from the exterior surface of five lengths of pipe and from the interior surface of one section of ripe for PCB analysis. In addition, sediment and debris from inside these pipes were collected and homogenized in a stainless steel bowl. Two composite samples were created from this material and submitted for PCB analysis. Table 6 presents the laboratory analytical results. The wipe sample results ranged from 5.5 to 28 ug.100 cm. The composite sediment samples both exceeded the 50 ppm TSCA limit. Consequently the small-diameter pipes were disposed at the Belleville, Michigan, landfill (see Section 3.2).
Large-diameter (24-in.) pipes were spray washed and staged in 20-ft lengths on the floor of the trickling filter. Six exterior surface wipe samples and six interior surface wipe samples were collected from the pipes under the auspices of City of Bloomington officials. Table 6 presents the analytical results of the surface wipe sampling. PCB concentrations exceeded the cleanup criterion of 100 ug/100 cm on only one exterior surface (sample number WT-2715). This pipe was recleaned; subsequent wipe sample results met the cleanup criterion. The two large-diameter pipes penetrating the north wall of the trickling filterwere cleaned and wipe sampled in the manner described previously; these results were also below the cleanup criterion for PCBs.
The concrete blocks and slabs discovered beneath the filter media were spray washed find staged on the floor of the trickling filter. Surface wipe samples were collected from dark or discolored areas of four of these blocks and submitted for PCB analysis. Table 6 presents the laboratory analytical results. No detectable concentrations of PCBs were found on any of these blocks.
WASTEWATER SAMPLING
4.3.1 Storm Water Sampling
Accumulated storm water in the trickling filter was sampled on four occasions and analyzed for PCBs in accordance with the Field Sampling Plan. Table 7 presents the laboratory analytical results. All water samples were below the discharge limit of 0.01 mg/L of PCBs. With permission from the City of Bloomington, the water was pumped from the trickling filter sumps and discharged to the City sanitary sewer.
4.3.2 Process Water Sampling
During the first 9 days of excavation and treatment, samples of the recirculated process water were collected every 3 days to determine PCB content (sample nos. WT-2503 WT-2531, and WT-2557). Handling of the process water was later modified as described in Section 3.2. As part of the system modification, approximately 15,000 gal of recirculated wash water was sampled (sample no. WT-2648) and discharged to the City sewer on June 28, 1998. Wastewater samples were collected and analyzed for PCBs in accordance with the Field Sampling Plan. Table 7 presents the laboratory analytical results. At the end of the project, all remaining process water (approximately 30,000 gal) was discharged to the City
sewer with approval from the City of Bloomington. No additional sampling was required by the City prior to final discharge of the process water.
|
