Sediment Dredging at Superfund Megasites: Assessing the Effectiveness (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Appendix C Summary of Remedial Action Objectives, Cleanup Levels (Numerical Remedial Goals), and Their Achievement at Sediment-Dredging Sites Note: For additional details on sites, see Table 3-1. For abbreviations, see page 294. Site: Bayou Bonfouca, LA Stated Remedial Action Objectives (Related to Sediment Removal): Reduce or eliminate the potential for ingestion of carcinogens in groundwater, surface soils, and shellfish. Reduce or eliminate the direct contact threat posed by bayou sedi- ments and onsite surficial creosote waste deposits (EPA 1987). Stated Cleanup Levels: “Contaminated sediments will be excavated either to a depth of about 6 in. into the upper cohesive layer or until PAH contamination is less than 1,300 ppm” (EPA 1987). Dates of Remediation: 1994-1995. Were Cleanup Levels Achieved? No chemical confirmation samples immedi- ately after remedy. Later sampling1 met cleanup levels. Were Remedial Action Objectives Achieved? Partially confirmed.2 1997 sampling by Louisiana (CH2M Hill 2001); 2003 sampling by EPA (EPA 1 2003a); and 2006 sampling by EPA after Hurricane Katrina (CH2M Hill 2006). 274 

Appendix C 275 Comments and Lessons Learned: Advances in dredging technology highlighted ability to dredge sediment accurately. Importance and difficulty of characterizing contaminant sediment deposits accurately.3 Importance of backfilling. Difficulty of accessing data (see Chapter 4). Lack of planned post-dredging monitoring. Less stringent cleanup level (PAHs at 1,300 ppm). Site: Lavaca Bay, TX Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. Goals of this treatability study were as follows (Alcoa 2000): develop information to support the technical and economic evaluation of poten- tial remedial actions; evaluate the effectiveness of dredging equipment on re- moval of mercury impacted sediment in the study area; evaluate potential im- pacts of dredging on mercury mobilization and residual sediment concentrations; and, understand the impact that dredging mercury contaminated sediment may have on mercury levels in Bay biota. Stated Cleanup Levels: None given. Comments: Pilot study. Dates of Remediation: 1998. Were Cleanup Levels Achieved? Not applicable―no cleanup levels indicated. Were Remedial Action Objectives Achieved? Pilot-study goals apparently achieved; not expected to achieve long-term risk reduction. Comments and Lessons Learned: (1) No significant change in average surficial sediment contaminant concentrations after dredging. (2) Advantages of pilot study for describing results of large-scale dredging at this site. (3) Evaluation of residuals after each dredging pass provided useful information on generation and concentrations. Site: Black River, OH Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. The goal of the sediment remediation project was to remove PAH-contaminated sediment to eliminate liver tumors in resident brown bull- head populations (Zarull et al. 1999). 2 Fish sampling was conducted in 1996 and 1997 and resulted in lifting of fish- consumption advisory in 1998 (CH2M Hill 2001), but there is no indication that shell- fish sampling has been conducted. 3 “Detailed design investigations during the summer and fall of 1988 [post-ROD, pre-dredging] showed the volume of contaminated sediments to be approximately 150,000 cy, an increase of three times that estimated in the ROD. This dramatic vol- ume increase resulted in a cost estimate for the selected remedy rising from approxi- mately $55 million to about $150 million” (EPA 1990b). 

276 Sediment Dredging at Superfund Megasites Stated Cleanup Levels: No chemical specific cleanup levels given. “The primary cleanup target was the removal of sediment in the area of the former USS coke plant to ‘hard bottom,’ or the underlaying shale bedrock. No quantitative envi- ronmental targets or end points were established, although post-dredging sam- pling was required to test for remaining areas of elevated PAH concentrations” (Zarull et al. 1999). Dates of Remediation: 1989-1990. Were Cleanup Levels Achieved? No quantitative chemical targets, but appar- ently met operational targets (mass removal and dredging to bedrock). Were Remedial Action Objectives Achieved? Short-term risk increased, but long-term risk-reduction targets met. Comments and Lessons Learned: Need for monitoring of biota. Dredging effec- tive, although uncertain improvement over natural attenuation. Increase in fish tumors after implementation. Site: Outboard Marine Corporation (OMC)―Waukegan Harbor, IL Stated Remedial Action Objectives (Related to Sediment Removal): None given. Stated Cleanup Levels: “Sediments in excess of 50 ppm PCB will be removed from the harbor by hydraulic dredging” (EPA 1984). Dates of Remediation: 1991-1992. Were Cleanup Levels Achieved? No chemical confirmation samples immedi- ately after remediation. Remedy based on assumption that removal of fine- grained “muck” overlying glacial till would achieve cleanup levels. Later sam- pling4 met cleanup levels; current state unclear.5 Were Remedial Action Objectives Achieved? Remedial action objectives not defined. Fish-tissue concentration trends inconclusive (see text). Comments and Lessons Learned: Insufficient pre-dredging and during- dredging data on fish concentrations to make comparison with post-dredging data. Less stringent remedial-action trigger (sediments greater than 50 mg/kg removed to depth of clean underlying geologic stratum) than for other PCB cleanups. No chemical verification samples taken on sediment immediately after dredging. 4 Apparently, sampling in 1996 (EPA 1999) and in 2003 (ILDPH/ATSDR 2004) indi- cates PCBs in sediments at less than 50 mg/kg; however, sampling events not men- tioned in 5-year reviews or site summary submitted to committee. 5 EPA states “OMC Plant 2 is likely a continual source of PCBs to Waukegan Har- bor, thus further harbor sediment sampling and analysis is likely needed to confirm whether cleanup levels are still being met” (EPA 2002). 

Appendix C 277 Site: Commencement Bay―Head of Hylebos, Tacoma, WA Stated Remedial Action Objectives (Related to Sediment Removal): Achieve “acceptable sediment quality in a reasonable time frame.” Acceptable sediment quality is defined as “the absence of acute or chronic adverse effects on biological resources or significant human health risks.” Reasonable time frame was further defined to be a period of 10 years to allow for natural recovery (via sedimenta- tion) (EPA 1989c). Stated Cleanup Levels: SQOs were established in 1989 ROD, and PCB value was modified in 1997 ESD (EPA 1997b). Dates of Remediation: 2003-2006. Were Cleanup Levels Achieved? Cleanup levels met in all but one area; adjacent nearshore cap was extended to address this area. Were Remedial Action Objectives Achieved? Dredging operation recently completed; no long-term data. Comments and Lessons Learned: Ability to meet cleanup levels under favorable conditions. Sufficient sampling and reference cores aided effective site charac- terization and contributed to success of dredging. Capping of one area above CULs contributed to success. Site sediment characteristics (“soft black muck” over native material) permitted overdredging and visual characterization of con- taminated vs native material. Cost-plus-fee contract incentivized dredging team to implement BMPs. Pilot testing6 indicated extent and type of debris and issues related to dredging and dredge-material handling. Site: Commencement Bay―Sitcum Waterway, Tacoma, WA Stated Remedial Action Objectives (Related to Sediment Removal): Achieve “acceptable sediment quality in a reasonable time frame.” Acceptable sediment quality is defined as “the absence of acute or chronic adverse effects on biologic resources or significant human health risks.” Reasonable time frame was further defined to be a period of 10 years to allow for natural recovery (via sedimenta- tion) (EPA 1989c). Stated Cleanup Levels: SQOs were established in 1989 ROD, and PCB value modified in 1997 ESD (EPA 1997b). Dates of Remediation: 1993-1994. Were Cleanup Levels Achieved? Cleanup levels met.7 6 “The pilot program [260 cy removal] provided information on mechanical dredg- ing, offloading of barges to rail cars, and rail transportation of dredged material to and into an upland landfill” (Dalton, Olmsted & Fuglevand, Inc. 2006). 7 CULs met in all areas immediately after dredging (redredging was conducted in one subarea) exception one underpier area where natural recovery was able to achieve SQOs within allowed period (confirmed in 2003) (EPA 2006a [Commence- ment Bay–Sitcum Waterway, April 26, 2006]). 

278 Sediment Dredging at Superfund Megasites Were Remedial Action Objectives Achieved? Not confirmed through biologic sampling, but long-term monitoring has shown continued (10 years) compliance with effects-based and risk-based cleanup levels, which were accepted as surro- gates for biologic-effects testing. No further monitoring required by EPA at the site. Comments and Lessons Learned: Clearly distinguishable contaminated layer was valuable in achieving remedial action objectives. Combining cleanup with port redevelopment provided economies for overdredging, ensuring removal. Compliance evaluated on an area basis used probabilistic criteria (averages and upper confidence limits) that accommodated single-chemical, noncontiguous, low-level exceedances after dredging. Inclusion of natural recovery for low-level contamination in remedial options allowed reasonable response to undredged inventory in under-pier areas. Site: Duwamish Diagonal, Seattle WA Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy (Natural Resource Damage Settlement). Restore and replace natural resources within the lower Duwamish River and Elliott Bay that have been injured by releases of hazardous materials through remediation of con- taminated sediments in the vicinity of combined sewer overflows and storm drains, source control, and habitat restoration (U.S.A et al. v the City of Seattle, Consent Decree No. C90-395WD, December 23, 1991). Stated Cleanup Levels: No chemical specific cleanup levels given. Dredging performance criteria based on achieved specific elevation for cap placement (EcoChem Inc. 2005). Comments: Project viewed as source-control action to address natural resource damages at a CSO through a hot-spot cleanup, with remaining contamination addressed as part of Lower Duwamish Superfund site actions. Dates of Remediation: 2003-2004. Were Cleanup Levels Achieved? Dredging performed for cap placement; per- formance criteria based on elevation specification. Toxicity data used to define area to be dredged, although performance based on chemical criteria. Were Remedial Action Objectives Achieved? Unlikely. Cleanup performed as interim action before selection of remedy for LDW site. Comments and Lessons Learned: Lack of adherence to BMPs8 resulted in signifi- cant transport of contaminated sediment outside the dredge prism. Post- dredging monitoring showed increased concentrations of PCBs and other COCs 8 The sediment remediation project closure report indicated that “the most obvious problems were over-filling the dredge bucket and spilling material out of the bucket as it was moved to and from the barge.” Water quality monitoring also indicated ex- ceedances of turbidity-compliance criteria (EcoChem Inc. 2005). 

Appendix C 279 in adjacent areas, which required placement of additional thin layer of clean ma- terial. Biologic monitoring conducted as part of wider LDW site indicates in- creases in fish-tissue contaminant concentrations at project site (see text). Site: Puget Sound Naval Shipyard, Bremerton, WA Stated Remedial Action Objectives (Related to Sediment Removal): Reduce risks to subsistence fishers consuming seafood from Sinclair Inlet by reducing PCB concentrations in biologically active zone of sediment in marine Operable Unit, controlling shoreline erosion of contaminated fill material, and selectively removing high concentrations of mercury that were colocated with PCBs (EPA 2000c). Stated Cleanup Levels: PCBs at 0.023 mg/kg wet weight in fish tissue and at 3.0 mg/kg (OC normalized) in sediment on area-weighted average. Sediment reme- dial action objectives to be achieved within 10 years. No time frame for recovery of fish tissue (EPA 2006a [Puget Sound Naval Shipyard, May 15, 2006]). Comments: Action levels were defined to distinguish which technology would be implemented. Dredging occurred when sediments had PCBs above 12 mg/kg (OC normalized) or above 6 mg/kg (OC) when mercury was at over 3 mg/kg. Enhanced natural recovery (thin-layer placement) was applied where PCBs were at 6-12 mg/kg OC (EPA 2006a [Puget Sound Naval Shipyard, May 15, 2006]). Dates of Remediation: 2000-2004. Were Cleanup Levels Achieved? No immediate post-dredging sampling. Initial long-term monitoring shows cleanup levels not met. Were Remedial Action Objectives Achieved? Long-term monitoring shows sediment quality has not met interim target that would support achieving goals in desired 10-year period. Comments and Lessons Learned: Lack of adherence to BMPs. TSS exceedances. Fish9 and sediment10 contaminant concentrations did not decrease or increased after dredging. Importance of recognizing issues with entire dredging process train beforehand. 9 Fish (English sole) sampling in 2003 indicates that concentrations after dredging exceed the cleanup goal. Average PCB concentrations are similar to average pre- remediation concentrations documented by historical (1991-1997) monitoring. The average mercury concentration is slightly higher than in previous (1994) sampling (URS 2006). 10 The post-remediation area-weighted average PCB sediment concentration of 7.4- 13 mg/kg of organic carbon (OC) (90th percentile confidence interval) exceeds the pre- remediation action area-weighted average value of 7.8 mg/kg of OC calculated from data collected before remediation (URS 2006). 

280 Sediment Dredging at Superfund Megasites Site: Harbor Island―Lockheed Shipyard, Seattle, WA Stated Remedial Action Objectives (Related to Sediment Removal): “Reduce concentrations of hazardous substances to levels which will have no adverse effect on marine organisms by eliminating the exposure pathways associated with residual concentrations of these contaminants. . . . Restore the marine habi- tat to its most productive condition to the extent practicable. . . . Minimize or eliminate the potential for recontamination of the cap from groundwater. . . . Achieve adequate source control to prevent recontamination” (EPA 2006a [Har- bor Island Lockheed Shipyard Sediment OU, May 11, 2006]). Stated Cleanup Levels: Arsenic at 57 mg/kg dry weight, copper at 390 mg/kg dry weight, lead at 450 mg/kg dry weight, mercury at 0.41 mg/kg dry weight, zinc at 410 mg/kg dry weight, PCBs at 12 mg/kg organic carbon normalized, LPAHs at 370 mg/kg organic carbon normalized (low-molecular-weight polynu- clear aromatic hydrocarbons), HPAHs at 960 mg/kg organic carbon normalized (high-molecular-weight polynuclear aromatic hydrocarbons), and tributyltin at 76 mg/kg organic carbon normalized (EPA 1997c, 2003c, 2006a [Harbor Island Lockheed Shipyard Sediment OU, May 11, 2006]). Comments: Cleanup levels were based on Washington State Sediment Manage- ment Standards (Apparent Effects Thresholds); TBT cleanup level was developed on basis of site-specific data for protection of invertebrates. Area background concentrations of PCBs and mercury were allowed to modify boundary (but not cleanup level within boundary) and define acceptable levels of recontamination. Dates of Remediation: 2003-2004. Were Cleanup Levels Achieved? Cleanup levels not met for metals, PAHs, and PCBs in some open-water areas that were to be remediated through dredging only.11 “Enhanced natural recovery” (placement of 6 in. of sand on the sediment surface) used in some of these areas. No actions taken in two areas with single- chemical, low-level exceedances. Toe of slope at transition from dredging only to dredging and capping also did not meet CULs. This noncompliant area ad- dressed through overplacement of cap material. Were Remedial Action Objectives Achieved? No long-term data yet available on objective of protection and recovery of benthic community health. Comments and Lessons Learned: Debris affected schedule and cost. Extensive sediment characterization during dredging included progress cores to assess adequacy of dredge cuts. Use of test dredge or pilot dredge would have helped 11 The 2005 5-year review (EPA 2005c) stated that “a total of eight sediment samples were collected from the post-dredge surface of the channel area. . . . All analytical results were compared to the SQS [sediment quality standards] chemical criteria to evaluate compliance. . . . [F]rom eight samples, three samples exceeded the SQS for PCBs only. Three other samples . . . exceeded the SQS for a combination of COCs [chemicals of concern].” 

Appendix C 281 to characterize debris. Experienced contractors successfully completed sediment handling with careful site management and successfully reduced contaminant loss. Implementation of BMPs. Technologies (WINOPS) incorporated to permit successful dredge placement. Change in dredging contracting strategy from pro- duction-based to time-and-materials-based contributed to successful remedia- tion. Site: Harbor Island―Todd Shipyard, Seattle, WA Stated Remedial Action Objectives (Related to Sediment Removal): “Reduce concentrations of hazardous substances to levels which will have no adverse effect on marine organisms by eliminating the exposure pathways associated with residual concentrations of these contaminants. . . . Restore the marine habi- tat to its most productive condition to the extent practicable. . . . Minimize or eliminate the potential for recontamination of the cap from groundwater. . . . Achieve adequate source control to prevent recontamination” (EPA 2006a, [Har- bor Island Lockheed Shipyard Sediment OU, May 11, 2006]). Stated Cleanup Levels: Arsenic at 57 mg/kg dry weight, copper at 390 mg/kg dry weight, lead at 450 mg/kg dry weight, mercury at 0.41 mg/kg dry weight, zinc at 410 mg/kg dry weight, PCBs at 12 mg/kg organic carbon normalized, LPAHs at 370 mg/kg organic carbon normalized, HPAHs at 960 mg/kg organic carbon normalized, and tributyltin at 76 mg/kg organic carbon normalized (EPA 1997c, 2003d, 2006a [Harbor Island Lockheed Shipyard Sediment OU, May 11, 2006]). Comments: Cleanup levels were based on Washington State Sediment Manage- ment Standards (Apparent Effects Thresholds); TBT cleanup level was developed on basis of site-specific data for protection of invertebrates. Area background concentrations of PCBs and mercury were allowed to modify boundary (but not cleanup level within boundary) and define acceptable levels of recontamination. Dates of Remediation: 2004-2005. Were Cleanup Levels Achieved? Mercury and PAH cleanup levels not achieved at a few locations, but concentrations were below action levels12 and thus accept- able without additional remediation. Were Remedial Action Objectives Achieved? No long-term data yet available on objective of protection or /recovery of benthic community health. Comments and Lessons Learned: Extensive sediment characterization during dredging included progress cores to assess adequacy of dredge cuts. Experienced 12 EPA stated that “the average (mean) concentration and the upper 95% confidence level on the mean concentration for all COCs are less than SQS chemical criteria for all analytes. Based on this statistical evaluation, Todd and EPA have concluded that the post-dredge surface in all areas of the Site meets cleanup criteria” (EPA 2006a [Harbor Island Todd Shipyards Sediment Operable Unit, May 12, 2006]). 

282 Sediment Dredging at Superfund Megasites contractors successfully completed sediment handling with careful site manage- ment to reduce contaminant loss. Implementation of BMPs throughout process train. Technologies (WINOPS) incorporated to permit successful dredge place- ment. Use of dredging contractor as consultant during design phase and use of environmental performance-based contracting contributed to successful remedia- tion. Site: Cumberland Bay, NY Stated Remedial Action Objectives (Related to Sediment Removal): Mitigate the immediate threat to the environment posed by the PCB-contaminated sludge bed. Rapidly and significantly reduce human and environmental risks. Prevent further environmental degradation resulting from this known source of PCB contamination (NYSDEC 1997). Stated Cleanup Levels: None given.13 Comments: Entire PCB-contaminated sludge bed to be removed (NYSDEC, 1997). Dates of Remediation: 1999-2000. Were Cleanup Levels Achieved? No cleanup levels established. Were Remedial Action Objectives Achieved? Not determined; some residual contamination present. Comments and Lessons Learned: Hardpan, rocks, and gulleys inaccessible to hydraulic dredge created unfavorable conditions that required multiple dredge passes and hand-held diver dredging. High residuals after initial dredging; some contamination remained at termination of project. Lack of quantitative criteria. Inadequate sampling and characterization techniques limited initial understand- ing of full extent of contaminated materials. Site: Dupont―Christina River, DE Stated Remedial Action Objectives (Related to Sediment Removal): Prevent exposure to contaminated sediments (EPA 1993a). Stated Cleanup Levels: Original site-specific Cleanup Revised Site-specific Cleanup Contaminant Criteriaa Criteriab Zinc 5,600 ppm 3,000 ppm Lead 1,200 ppm 700 ppm Cadmium 60 ppm 20 ppm a From 1993 ROD (EPA 1993a). b Original cleanup values were lowered to eliminate need for extensive long-term monitoring program that was part of 1993 ROD (EPA 2005b). 13 From record of decision (NYSDEC 1997): “Question: What is the target level DEC hopes to achieve of remaining PCB contamination? Response: The NYSDEC has not set an action level. The goal of remediation is the removal of the sludge bed in its en- tirety.” 

Appendix C 283 Comments and Lessons Learned: These cleanup criteria were apparently used to delineate area for remediation. In practice, chemical analyses were not used to verify removal of contaminated sediments. Sediments were removed to the re- quired minimum depth of 2 ft or until underlying stratum was encountered (URS 1999). Dates of Remediation: 1999. Were Cleanup Levels Achieved? Not determined; no confirmation samples taken after dredging and backfilling. Removal targets (elevation) were met. Were Remedial Action Objectives Achieved? Not determined, no confirmation or long-term monitoring. Comments and Lessons Learned: Lack of chemical confirmation sampling and long-term monitoring is problematic.14 Dredging operation was based on re- moval, not on concentration. Need for source control and a reference site. Site: Fox River (OU 1), WI Stated Remedial Action Objectives (Related to Sediment Removal): Achieve, to the extent practicable, surface-water quality criteria throughout Lower Fox River and Green Bay. Protect humans who consume fish from exposure to COCs that exceed protective levels. Protect ecologic receptors from exposure to COCs above protective levels. Reduce transport of PCBs from Lower Fox River into Green Bay and Lake Michigan. Minimize downstream movement of PCBs during imple- mentation of remedy (WI DNR/EPA 2002). Stated Cleanup Levels: Dredge all sediment with PCBs at over 1 ppm or achieve a surface-weighted average concentration (SWAC) of 0.25 ppm (WI DNR/EPA 2002) (see Comments for explanation). Comments: If after dredging is completed for OU 1, sampling shows that the 1- ppm remedial action level (RAL) has not been achieved, a SWAC of 0.25 ppm may be used to assess effectiveness of PCB removal. If that SWAC has not been achieved, the remedy provides options to reduce risk further. The first option is additional dredging to ensure that all sediments with PCBs at over 1-ppm RAL are removed throughout the particular deposit. The second option is to place a sand cover on dredged areas to reduce surficial concentrations so that a SWAC of 0.25 ppm for OU 1 is achieved (WI DNR/EPA 2002). Dates of Remediation: 2004-present. Were Cleanup Levels Achieved? Dredging not complete; some subunits have not achieved desired cleanup levels. Were Remedial Action Objectives Achieved? Dredging not yet completed. 14 Second 5-year review for the site indicates that vegetative cover and vegetation species composition are monitored. However, there is no systematic monitoring of chemical concentrations in sediment, water, or biota and no evaluation of toxicity end points. 

284 Sediment Dredging at Superfund Megasites Comments and Lessons Learned: Baseline monitoring, although extensive, is not sufficient to inform long-term monitoring because it began after dredging had begun at the site. Thin layer of highly contaminated sediment and residuals have limited success at reaching 1 ppm. Heterogeneity of deposits creates difficulties in defining the dredge prism. ROD permits flexibility in achieving cleanup levels and stipulates additional actions (further dredging or capping) if dredging does not achieve results. Site: Fox River (Deposit N), WI Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. Demonstration-project objectives were environmental dredg- ing to remove contaminated sediment to specifications; protection of the river, local properties, and residents during sediment removal; safe transport and dis- posal of sediment; and maintenance of good local relations during the project (Foth and Van Dyke 2000). Stated Cleanup Levels: Average residual thickness no more than 3 in. in West Lobe and no more than 6 in. in East Lobe (Foth and Van Dyke 2000). Comments: Pilot study. Dates of Remediation: 1998-1999. Were Cleanup Levels Achieved? Target elevations met.15 Were Remedial Action Objectives Achieved? Pilot-project goals met. Not ex- pected to achieve long-term risk reduction. Comments and Lessons Learned: Pilot project indicated mass removal can be achieved. Release of PCBs during dredging.16 Bedrock limited ability to dredge 15 Project summary stated that “project specifications did not require either total removal of the sediment or removal to a specific PCB sediment concentration as these sediments rested on a fractured bedrock surface, preventing a dredge cut into a clean underlying layer” (Foth and Van Dyke 2000). 16 “During dredging in 1998 (Phase I), the upstream average reported PCB water column concentration was 3.2 ng/l compared to the average downstream PCB water column concentration of 11 ng/L. The variation between the upstream PCB water column concentration and the downstream PCB water column concentration meas- ured during dredging reflects an average increase downstream of 3.5 times the up- stream value. Similar water column PCB results were obtained during Phase II and III in the 1999 dredge season. For the 1999 dredge period, the average upstream PCB water column concentration was 14 ng/L compared to the average downstream PCB water column concentration of 24 ng/L. This variation represents an increase of 1.7 times the upstream reported value. It can be concluded from this data that dredging caused an increase in PCB concentrations downstream of the dredge site” (Malcolm Pirnie, Inc. and TAMS Consultants, Inc. 2004). 

Appendix C 285 completely, and residual layer was left. Post-dredging concentrations were simi- lar to that before dredging.17 Site: Fox River (SMU 56/57), WI Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. Demonstration-project objectives were to evaluate potential effects of large-scale dredging of PCB-contaminated sediments on the Fox River, to evaluate efficacy of large-scale dewatering and land disposal of PCB- contaminated sediments, and to evaluate potential costs of large-scale dredging, dewatering, and land disposal of PCB-contaminated sediments (Montgomery Watson 2001). Stated Cleanup Levels: 1999 action: to depths consistent with PCBs at 1 mg/kg of sediment or less (Montgomery Watson 2001). 2000 action: total PCBs at 1 mg/kg of sediment or 10 mg/kg with at least 6 in. of clean sand backfill (Fort James Cor- poration et al. 2001). Comments: Pilot study. Dates of Remediation: 1999-2000. Were Cleanup Levels Achieved? Cleanup levels not met in first season of dredg- ing (1999); met in 2000 and then backfilled. Were Remedial Action Objectives Achieved? Pilot-project goals were met; not expected to achieve long-term risk reduction. Comments and Lessons Learned: Residual mass and PCB concentrations re- sulted in redredging and backfilling. Difficulties experienced in dewatering and solids handling indicated value of pilot studies and of considering full train of treatment. Dredging released PCBs to water column despite silt curtain controls. Using turbidity as an indicator of PCB transport is insufficient.18 No sampling after sand backfilling, so final surface concentrations are not known. Site: Ketchikan Pulp Company, Ward Cove, AK Stated Remedial Action Objectives (Related to Sediment Removal): Reduce toxicity of surface sediments. 17 “Using the 1998 data, collected just prior to dredging, the pre-dredge average PCB sediment concentration in Deposit N was 16 ppm, with a maximum concentra- tion of 160 ppm. The post-dredge average PCB sediment concentration in Deposit N was 14 ppm, with a maximum of 130 ppm” (Foth and Van Dyke 2000). 18 “The TSS and PCB comparison (downstream minus upstream) illustrates that TSS is not a reliable indicator of PCB transport during a dredging operation. For ex- ample, from September 1 to October 6, a period of negative TSS loading (less at the downstream than at the upstream site), the PCB loading was positive. Thus, if one is to monitor PCB transport during a remediation operation, sole reliance on turbidity or TSS measurements is inadequate” (Steur 2000). 

286 Sediment Dredging at Superfund Megasites Enhance recolonization of surface sediments to support a healthy marine benthic infaunal community with multiple taxonomic groups (EPA 2000b). Stated Cleanup Levels: None given. Comments: Health of benthic communities is to be assessed through toxicity testing and benthic community analyses as part of long-term monitoring. Dates of Remediation: 2000-2001. Were Cleanup Levels Achieved? Success of cleanup to be determined by toxicity testing and benthic community analysis. Concentration-based cleanup levels not defined. Were Remedial Action Objectives Achieved? Monitoring is continuing; there is some initial success in reducing benthic toxicity, which was the desired objective. Comments and Lessons Learned: Effectiveness of backfilling in reducing toxicity was demonstrated in comparison with locations without backfilling.19 Site dem- onstrated use of toxicity assays and benthic community analyses as a useful indi- cator of ecologic improvement after remediation. Dredging and backfilling con- ducted in a small area compared to backfilling only and natural recovery areas. Site: Newport Naval Complex―McCallister Point Landfill, RI Stated Remedial Action Objectives (Related to Sediment Removal): Prevent human ingestion of shellfish impacted by sediments with COC concentrations exceeding cleanup levels. Prevent exposure of aquatic organisms to sediments with COC concentrations exceeding cleanup levels. Prevent avian-predator in- gestion of shellfish impacted by sediments with COC concentrations exceeding cleanup levels. Minimize migration of sediments with COC concentrations ex- ceeding selected PRGs to offshore areas and previously unaffected areas of Nar- ragansett Bay. Prevent washout of landfill debris into marine environment (EPA 2000d). Stated Cleanup Levels: Copper, 52.9 ppb in pore water; nickel, 33.7 ppb in pore water; anthracene, 513 ppb in sediment; chrysene, 1,767 ppb in sediment; fluo- rene, 203 ppb in sediment; total PCBs, 3,634 ppb in sediment (EPA 2000d). Dates of Remediation: 2001. Were Cleanup Levels Achieved? Cleanup levels confirmed analytically immedi- ate after dredging except when bedrock encountered. Long-term monitoring 19 “The 2004 [first long-term monitoring event] data indicated that conditions in the three thin-layer capping areas had generally improved, while those in three of the four natural recovery areas generally had not (the shallow natural recovery area was the exception)” (EPA 2006a, Ketchikan Pulp Company; April 26, 2006). 

Appendix C 287 indicates that sediment cleanup levels have been maintained although pore wa- ter exceedances and toxicity20 remain. Were Remedial Action Objectives Achieved? Although narrowly defined, re- medial action objectives based on exposure to sediment above cleanup levels apparently met. Long-term risk reduction inconclusive. Comments and Lessons Learned: Cleanup levels were met although verification sampling was insufficient in some locations (hitting bedrock was considered meeting values). Pore water exceedances of cleanup levels persist after dredging although source of contamination is not clear.21 Ability to dredge much of site from shore was advantageous. Incomplete recolonization by shellfish and sub- merged aquatic vegetation in near term (less than 5 years). Useful and compre- hensive range of pre-monitoring and post-monitoring metrics. Site: GM Central Foundry, St. Lawrence River, NY Stated Remedial Action Objectives (Related to Sediment Removal): Remedial action objectives are not specifically provided. EPA does state: “Hot spots in the St. Lawrence and Raquette rivers and Turtle creek will be dredged and excavated to remove PCBs. All PCB contaminated sediments in the hot spots will be re- moved given the technological limitations associated with dredging” (EPA 1991). Stated Cleanup Levels: St. Lawrence and Racquette Rivers: PCBs at 1 ppm (EPA 1991). Dates of Remediation: 1995. Were Cleanup Levels Achieved? Cleanup levels not met in St. Lawrence River, because of residuals, backfilling, and capping required in one area. Cleanup lev- els met in Racquette River. Were Remedial Action Objectives Achieved? Dredging alone unable to achieve cleanup levels, although combination remedy effectively reduced surface concen- trations. Comments and Lessons Learned: Intensive monitoring before, during, and im- mediately after dredging provided useful indications of dredging effect. Sheet- pile walls limited PCB release during dredging. Inability to eliminate residuals and possible increase in residual concentration due to contaminant retention within sheet-pile walls. Lack of sediment sampling for contamination since 20 “Sea urchin pore water toxicity results indicate toxicity at some stations within Groups 1, 2, 3, and 4 (TetraTech NUS, 2006).” (Groups are geographic areas with sam- pling stations; groups 1, 2, 3, and 4 are all the areas besides the reference area.) 21 The first long-term monitoring report for the site (Tetra Tech NUS 2006) states that “groundwater discharge may be responsible for elevated pore water metals con- centrations found at some of the near shore locations.” However, the committee notes that dredging may have influenced redox conditions leading to the dissolution of metals from residual contaminated materials. 

288 Sediment Dredging at Superfund Megasites dredging in 1995 eliminates insight into concentration changes over time. No apparent trend in fish concentrations after dredging. Site: Grasse River, NY (Non-Time-Critical Removal Action) Stated Remedial Action Objectives (Related to Sediment Removal): Project objectives were: (1) remove the most upstream major PCB source in the Grasse River; (2) eliminate a potential source of PCB exposure to biota; reduce potential long-term risks to human health and the environment; (3) provide valuable site- specific data for use in the Analysis of Alternatives for the study area (ALCOA 1995). Stated Cleanup Levels: None given. Comments: Non-Time-Critical Removal Action (NTCRA). Dates of Remediation: 1995. Were Cleanup Levels Achieved? None provided. Average sediment concentra- tions decreased.22 Were Remedial Action Objectives Achieved? Majority of sediment removed, but corresponding reductions in fish-tissue and water-column concentrations not observed.23 Comments and Lessons Learned: Debris and bedrock created operational diffi- culties. Site served as useful pilot study; substantial useful data were collected. Pilot improved site conceptual model. Much of targeted sediment mass was re- moved with substantial portion of PCBs in river system.24 Removal released PCBs to water column as evidenced by increased water concentrations and ac- cumulation by caged fish adjacent to work zone. Turbidity release did not corre- late with PCB release. Project demonstrated value of caged-fish studies. 22 According to BBL (1995): “Area A [the largest area] pre-NTCRA samples con- tained PCB concentrations ranging from non-detect (MDL varied) to 11,000 mg/kg, while post-removal PCB samples contained PCB concentrations ranging from 1.1 to 260 mg/kg. The arithmetic (geometric) average PCB concentration detected within area A (considering all depths) was reduced from 1,109 mg/kg (263 mg/kg) to 75 mg/kg (40 mg/kg), or about 93% (85%). Considering PCBs contained within the bioavailable zone (that is, top 12 in. of sediment), average arithmetic (geometric) PCB concentrations were reduced from 518 mg/kg (195 mg/kg) to 75 mg/kg (40 mg/kg) or approximately 86% (80%). At several locations, PCB concentrations in the residual sediment (all bioavailable) increased from pre-remediation bioavailable conditions.” In the smaller area B, the pre-NTCRA arithmetic (geometric) average sediment PCB concentration of 300 mg/kg (275 mg/kg) was decreased to 108 mg/kg (106 mg/kg) following dredging, approximately a 64% (62%) decrease (BBL 1995). 23 About 84% of the original volume of sediment was removed, leaving 550 cy of the 3,500-cy sediment that was targeted. Average thickness of sediment in area was reduced from about 22 in. to 4 in. (BBL 1995). 24 About 27% of PCB mass in Grasse River study area (BBL 1995). 

Appendix C 289 Site: Grasse River, NY Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. Demonstration project objectives were (1) evaluate dredging as a remedial option to reduce the potential risk that may be posed by future ice jam related sediment scour events by removing a targeted area of sediments with elevated PCB concentrations in an area of the river that is known to be subject to ice jam-related scour; (2) develop site-specific information related to dredging effectiveness, dredging residuals, dredging production rate, and sediment resus- pension that can be used in the development of the revised Analysis of Alterna- tives Report (Alcoa Inc. 2005). Stated Cleanup Levels: Not a CERCLA remedy. No chemical specific cleanup levels given. Remove all soft sediments, to the extent possible, from an approxi- mate 8-acre area of the main channel. To the extent practical, all soft sediments will be removed to hard bottom leaving a stable dredge face on the adjacent sediments (Alcoa Inc. 2005). Comments: Pilot study. Dates of Remediation: 2005. Were Cleanup Levels Achieved? No site-specific cleanup levels. Were Remedial Action Objectives Achieved? Stated goals were achieved. Not expected to achieve long-term risk reduction. Comments and Lessons Learned: Debris and bedrock limited dredging effec- tiveness. Limited control over backfilling created higher than expected cap con- centrations. No significant change in surface concentrations. Significant increase in some biota followed dredging. Site: Lake Jarnson, Sweden Stated Remedial Action Objectives (Related to Sediment Removal): Not a CERCLA remedy. Remediation goal was to substantially reduce transport of PCBs from lake sediments to lake water and downstream system to reduce PCB concentrations in biota (Fox River Group 1999). Stated Cleanup Levels: PCBs at maximum of 0.5 ppm and no more than 25% of remediated area at over 0.2 ppm (Bremle et al 1998b). Dates of Remediation: 1993-1994. Were Cleanup Levels Achieved? Cleanup levels achieved except in one loca- tion.25 25 After dredging was completed, 2.9 kg of PCB was calculated to be left in lake sediment. Sedimentary pool of PCB was distributed, and only one of 54 defined su- bareas exceeded 0.5 mg/kg dry weight. At commencement of remediation, this had been set as highest acceptable level to be left in sediment. Only 20% of sediment areas ended up with PCB higher than 0.2 mg/kg dry weight, which was better than reme- diation objective of 25% (Bremle et al. 1998b). 

290 Sediment Dredging at Superfund Megasites Were Remedial Action Objectives Achieved? Results indicate reductions in transport26 and fish-tissue concentrations.27 Comments and Lessons Learned: Ability to overdredge. Sediment concentration decline corresponds to water and fish-tissue concentration declines. Monitoring data seek to differentiate regional declines in background PCB concentrations from those resulting from remediation. Did not target or dredge near-shore PCBs, which are later implicated as a continuing source of PCBs for fish.28 Site: Manistique Harbor, MI Stated Remedial Action Objectives (Related to Sediment Removal): Reduce PCB concentrations in fish and water in the Manistique River and Harbor to lev- els that would not present an unacceptable human health or ecologic risk and would allow elimination of existing fish-consumption advisories. Maintain har- bor as a navigable waterway for commercial shipping, fishing boats, and recrea- tional watercraft. In general, restore river and harbor areas for use by deeper- draft vessels. Minimize need for future remedial action in area after completion of a non-time-critical action. Implement actions that would best contribute to efficient performance of any future remedial actions in the area. Achieve compli- ance consistent with federal and state ARARs for site. Comply with risk-based objectives defined by TERRA, Inc., as part of the risk assessment. Reduce, as much as practicable, the release of PCBs associated with particles and dissolved in the water to Lake Michigan (EPA 2006a [Manistique River and Harbor Site, May 10, 2006]). Stated Cleanup Levels: Initially, goal of action was to remove sediments with PCB concentrations greater than 10 ppm. Later, goal was modified to state that objective was 95% removal of total PCB mass and an average sediment concen- tration not over 10 ppm throughout sediment column (Weston 2002). Dates of Remediation: 1995-2000. 26 “After completed remedial activities in Lake Jarnsjon in spring 1995, PCB concen- trations had decreased at the outlet of the lake. In addition, the PCB transport during high flow in the early spring was lower than the year before” (Bremle et al. 1998b). 27 “Fish from all the locations in 1996 [after dredging] had lower PCB concentration than in 1991 [before dredging]. The most pronounced decrease was observed in the remediated lake, where levels in fish were halved. The main reason for the reduced levels was the remediation” (Bremle and Larsson 1998). 28 “Since some of the contaminated sediment was left in the lake after remediation and this sediment was mainly located in the most shallow, littoral areas, these sites constituted a probable source of PCB to zooplankton and fish. PCB still remaining in littoral sediment was probably the cause for a recorded gradient of PCB in fish from the lake and downstream the river” (Bremle and Larsson 1998). 

Appendix C 291 Were Cleanup Levels Achieved? Average cleanup concentration level was met; it is unclear whether mass-removal goal was met.29 Were Remedial Action Objectives Achieved? Progress toward remedial action objectives after deposition event. Comments and Lessons Learned: Initial cleanup levels not met. Poor initial characterization of wood debris and bedrock issues resulted in incomplete dredging and a longer dredging time frame. Dredging caused an initial increase in surface concentrations and no decrease in fish concentrations. Deposition due to dam removal and sand placement led to decreased surface PCB concentra- tions. Site: Reynolds Metals, St. Lawrence River, NY Stated Remedial Action Objectives (Related to Sediment Removal): Prevent human and biota contact with contaminated sediments. Reduce or prevent hu- man ingestion of fish caught from the St. Lawrence River. Reduce short-term effects on surface water and air expected as a result of remedial activities (EPA 2006c). Stated Cleanup Levels: PCBs, 1 ppm; PAHs, 10 ppm; TDBF, 1 ppb (EPA 1993b). Dates of Remediation: 2001. Were Cleanup Levels Achieved? Cleanup levels for PCBs were met after cap- ping;30 PAH cleanup levels not met; work continues.31 Were Remedial Action Objectives Achieved? Remedial action objectives not met. PAH remedial activities are ongoing. Comments and Lessons Learned: Residuals due to bedrock and dredging over boulders and cobbles. PCB concentrations used to indicate PAH contamination; however, there was a lack of concordance between PCB and PAH contamination. 29 “The range of removal efficiency stretches from 82% to 97% depending upon the many assumptions made in the calculation of the residual mass of PCBs. The outcome is highly dependant upon the specific gravity assumed for the in-situ sediments along with sediment volume estimates. The best case estimate indicates that the objective of 95% removal of PCBs from the AOC may have been met while the worst-case esti- mate indicates that the objective may not have been met.” (Weston 2002). 30 “Despite extensive dredging of the St. Lawrence River, the cleanup goals of 1 mg/kg PCBs, 10 mg/kg PAHs, and 1 μg/kg TDBFs were not achievable in all areas. As a result, a 0.75-acre, 15 cell area, containing a range of PCB concentrations from 11.1 mg/kg PCBs to 120.457 mg/kg, was capped with the first layer of a three-layer cap to achieve the cleanup goal [for PCBs]. The remaining exposed sediments average 0.8 mg/kg PCBs within the remaining 255 cells (21 acres), which is below the cleanup goal” (EPA 2006c). 31 “The remedial action activities in the remaining cells containing elevated levels of PAHs above the cleanup goal have not been fully implemented” (EPA 2006c). 

292 Sediment Dredging at Superfund Megasites Site: Marathon Battery, Hudson River, Cold Spring, NY Stated Remedial Action Objectives (Related to Sediment Removal): Reduce cadmium in sediments to protect aquatic organisms and protect human health. Reduce the transport of suspended sediments from east and west foundry coves and the pier area (EPA 1989a). Stated Cleanup Levels: Dredging to 1 ft. Comments: According to the Record of Decision (EPA 1989a), “The data com- piled for east foundry cove indicate that over 95% of the cadmium contamination is located in the upper layer (1 foot) of the sediments. Due to the nature of the dredging process, dredging to a specific action level (for example, 10, 100, or 250 mg/kg of cadmium) would be technically difficult, since these concentrations vary in the sediments by only a few inches of depth. Therefore, expectations are that by dredging the upper layer of contaminated sediments, 95% of the cad- mium contamination will be removed. Following remediation, it is anticipated that cadmium concentrations would not exceed 10 mg/kg in most of the dredged areas. . . . Sediment samples at and beneath the cold spring pier will be collected, analyzed, and evaluated to ascertain whether this area is a source of cadmium contamination. If, based upon this analysis, these sediments are determined to be a source, these sediments will be dredged to a depth of one foot.” Dates of Remediation: 1993-1995. Were Cleanup Levels Achieved? Cleanup levels met (dredging performance targets met). Were Remedial Action Objectives Achieved? Exposure measures of remedial action objectives met. Comments and Lessons Learned: Discrete contaminated layer. Low amounts of debris. Ability to overdredge.32 Site has useful post-dredging verification sam- pling and long-term monitoring data on sediments and biota that indicate bene- ficial effect of remedial activity. Site: New Bedford Harbor, MA Stated Remedial Action Objectives (Related to Sediment Removal): Pilot- project objectives were to significantly reduce PCB migration from hot-spot area sediment, which acts as a PCB source to the water column and to the remainder of the sediments in the harbor; to significantly reduce the amount of remaining PCB contamination that would need to be remediated to achieve overall harbor cleanup; to protect public health by preventing direct contact with hot-spot 32 “EFC [East Foundry Cove] and EFP [East Foundry Pond] bottom sediments con- sist of silts and clays with some sand. These sediments have a very low bearing capac- ity which extend to nearly 80 feet in depth. The river bottom sediments consist of silts and clays with varying amounts of sand” (EPA 2006a [Marathon Battery Superfund Site, May 10, 2006]). 

Appendix C 293 sediments; and to protect marine life by preventing direct contact with hot-spot area sediments. Stated Cleanup Levels: Short-term hot-spot goal, 4,000 ppm total PCBs (EPA 1990a); long-term hot-spot goal, 10 ppm total PCBs after additional remediation occurs (EPA 1998b). Comments: Owing to limited scope of hot-spot dredging action, EPA did not expect to achieve standards or levels of control associated with final cleanup lev- els (such as FDA PCB tolerance for fish tissue and water quality criterion). How- ever, the action was expected to comply with some ARARs, including compli- ance with RCRA facility regulations, Executive Order 11988 regarding protection of flood plains to the extent practicable, Executive Order 11990 regarding protec- tion of wetlands, and federal and state air standards during dredging and treat- ment of contaminated sediments (EPA 1990a). Dates of Remediation: Hot-spot removal, 1994-1995. Were Cleanup Levels Achieved? Operations appear to have achieved interim total-PCB cleanup level of 4,000 ppm (USACE 1995); removal was completed with minimal net transport of PCBs. Were Remedial Action Objectives Achieved? Hot-spot removal was not in- tended to meet long-term risk-reduction goals. Unclear whether hot-spot re- moval objectives were met. Comments and Lessons Learned: Usefulness of pilot studies. Value of EPA’s process modifications on the basis of increased PCB and hydrogen sulfide con- centrations in air during sediment handling. Indication of increased exposure to terrestrial plants and no increased exposure to aquatic biota during dredging. Challenge of relating contamination and dredging in a large harbor to human exposure and effects. Site: United Heckathorn, Richmond, CA Stated Remedial Action Objectives (Related to Sediment Removal): The ROD (EPA 1996b) states the clean up goal was based on a surface-water quality crite- rion for protection of human health from consumption of fish and bioaccumula- tion of DDT and dieldrin. DDT concentrations exceeded dieldrin concentrations by a factor of 10-100, so sediment remediation goals for both contaminants were based on DDT concentrations. Stated Cleanup Levels: Water column, DDT at 0.59 ng/L and dieldrin at 0.14 ng/L. Sediment, DDT at 590 μg/kg (dry weight) (EPA 1996b). Comments: Values are based on achieving a 10-6 lifetime excess cancer risk level. As described in the ROD (EPA 1996b), this value is lower than the chronic ma- rine aquatic-life criterion of DDT at 1 ng/L or dieldrin at 1.9 ng/L. Human health criteria were judged likely to be achieved if average sediment DDT concentration was below 0.59 mg/kg (dry weight), which is lower than the 1 mg/kg that would probably meet marine chronic water quality criteria (EPA 1996b). 

294 Sediment Dredging at Superfund Megasites Dates of Remediation: 1996-1997. Were Cleanup Levels Achieved? Cleanup levels apparently met immediately after dredging. Recontamination after dredging. Were Remedial Action Objectives Achieved? Little improvement in channel; remedial action objectives not met. Comments and Lessons Learned: Dredging was not effective in decreasing sediment or water concentrations, and biota concentrations did not decline to clean levels. Side slopes, piers, ship traffic, debris, and an outfall may have led to increased residual contamination. Difficulty in reaching agreement between par- ties. Conceptual site model was not sufficient to discern effect of dredging and likelihood of recontamination. Usefulness of deployed mussel studies to indicate effect of dredging.33 ABBREVIATIONS: PAH, polycyclic aromatic hydrocarbon(s); CERCLA, Com- prehensive Environmental Response, Compensation, and Liability Act; PCB, polychlorinated biphenyl; SQO, sediment quality objective; ROD, record of deci- sion; ESD, explanation of significant differences; CUL, cleanup level; BMP, best management practice; LDW, Lower Duwamish Waterway; COC, contaminants of concern; TSS, total suspended solids; TBT, tributyl tin; PRG, preliminary remediation goal; ARAR, applicable or relevant and appropriate requirements; TDBF, total dibenzofurans; RCRA, Resource Conservation and Recovery Act; DDT, dic. 33 “Year 1 biomonitoring showed that pesticide concentrations in the tissues of mussels exposed at the site were higher than those observed before remediation. Year 2 samples, collected in February 1999, showed tissue levels that were much reduced from Year 1 but still exceeded pre-remediation levels of DDT at Lauritzen Channel, Santa Fe Channel, and Richmond Inner Harbor Channel” (EPA 2006a [United Heckathorn Superfund Site, Richmond, CA, May 16, 2006]).