2001 Capping Pilot Study

Executive Summary


This report describes the pilot study conducted by Alcoa Inc. (Alcoa) within the lower Grasse River, Massena, New York to evaluate whether the polychlorinated biphenyl (PCB)-containing sediments can be covered by a clean sediment cap. Capping is being evaluated as a means to reduce PCB concentrations within the lower Grasse River surface sediments, water, and biota. The pilot study was conceived as a way to obtain the site-specific data needed to compare this remedial option to others. Important questions in this regard include how successfully a cap can be applied, and what placement methodology works best. The study involved the placement of a clean layer of capping materials over a 750-foot stretch of the River using several different cap designs and placement techniques.

The pilot study conformed to a Work Plan (Alcoa, 2001) that was reviewed by the United States Environmental Protection Agency (USEPA), the New York Sate Department of Environmental Conservation (NYSDEC), and the St. Regis Mohawk Tribe (SRMT). Capping activities as part of the pilot study were conducted between July 23 and October 9, 2001, although some post-capping monitoring activities continued into November, 2001. Long-term monitoring will continue into 2002.

Project Objectives

  • Alternative cap placement techniques (alone or in combination):
    • Surface and subsurface placement via mechanical clamshell
    • Subsurface placement via tremie pumping
    • Surface placement via pneumatic broadcasting (bentonite only)
  • Alternative cap materials (alone or in combination):
    • 1:1 sand/topsoil mixture
    • Granulated bentonite
    • AquaBlokTM (a commercial, clay-gravel composite)

The metrics for evaluation included the following:

  • Cap coverage effectiveness (including the ability to cap steep side slopes, and the extent of particle size fractionation of cap material during placement);
  • Extent of potential entrainment of underlying contaminants into cap materials during placement;
  • Water column impacts during placement;
  • Cost; and
  • Recolonization of sediment by benthic organisms.

Site Description

The capping pilot study area consists of an approximately 750-foot long by 400-foot wide stretch of the River, and has:

  • Steep side slopes (30 to 50%) (thus minimal adjacent wetlands);
  • Relatively flat bottom;
  • Water depths (excluding side slopes) averaging about 16 feet;
  • Low erosion potential (water velocities during tests ranged from 0.02 to 0.80 ft/sec);
  • Bottom sediments, ranging from 1 to 6 feet in depth, composed primarily of silt, sand, and organic matter;
  • PCB concentrations in surficial sediments on the order of 10 mg/kg; and
  • Minimal presence of boulders, cobbles, or debris on the sediment bed.

During the capping period, the flow in the River was generally low, averaging 237 cfs (range 78 to 765 cfs).


The 7-acre site was divided into four cells, and the project was divided into two phases. The first phase, designed to screen a number of capping materials and application methods, was conducted in Test Cell #1 (the Test Cell), which was divided into five subcells. The second phase, conducted in Pilot Cells #2, #3 and #4 (the Pilot Cells), was designed to evaluate, under operating conditions approximating a full-scale project, the material/application combinations considered most promising based on the Phase 1 work. Table ES-1 presents a listing of the key features of the capping conducted in each cell. In four treatments, the cap was applied in two or three lifts; a single lift was used in the other four treatments. Target cap thickness (considering all lifts in a cell) ranged from 0.75 to 2 feet, with 1 foot being the most common. The Test Cell was aligned along the north shore of the River. The Pilot Cells were aligned along the south shore. The site extended from bank to bank, and upon completion of the project a cap was in-place over the entire site except for nearshore vegetated areas (per the Work Plan), and nearshore areas blocked by overhanging trees.

Capping was conducted with an in-water equipment barge, usually carrying an 80-ton crane outfitted with a 2.5-yd mechanical clamshell bucket. Capping materials, prepared at an on-shore staging area, were placed on a separate barge. A key element of the capping was accurate horizontal control of the bucket using a combination of global positioning systems (GPS) and the Windows Offshore Positioning Software (WINOPS). Vertical control was maintained by the crane operator using markings on the lowering cable. The clamshell bucket was opened at the water surface when using the surface application technique or at a predetermined depth below the water surface when using the subsurface application technique.

The principal capping material used was a 1:1 mix of locally obtained sand and topsoil. The mixture had a total organic carbon content averaging about 0.7% (range ND to 1.8%). Other capping materials included granulated bentonite and AquaBlokTM, a commercial clay-gravel composite. All capping materials were tested/analyzed for a broad range of physical and chemical properties prior to use in the study.

During nearly all capping activities, an in-River silt curtain containment system was used along the perimeter of the cell or subcell being capped. Silt curtains were selected because they have the ability to reduce the migration of cap materials downstream and to adjacent cells during placement without unacceptably restricting the flow of the River. The placement of the silt curtains was adjusted during the program so that one side of the River always remained open for boat traffic and fish movement. Silt curtains were not used for the capping of a small, centerline wedge area that was found to have been blocked by the curtains.

Monitoring was conducted prior to, during, and immediately following capping activities in order to address each of the objectives listed above. Table ES-2 presents a summary of these activities. Of note is an extensive water quality monitoring program that included sampling at

  1. Upstream and downstream
  2. locations;
  3. In-cell locations; and
  4. Locations adjacent to each cell (just outside silt curtain).
A total of approximately 900 water samples and 490 sediment samples were collected and analyzed during the study. The results of water quality monitoring during capping were continually compared to a set of "corrective action triggers". Results exceeding these triggers could have resulted in suspension or modification of capping activities, however, no trigger levels were ever exceeded.


The pilot study was successful and provided reliable data to allow evaluation of the project objectives listed above. Each objective is discussed below.

Cap Placement Techniques

  • Placement of dry, bulk capping material by clamshell was demonstrated to be more effective, cost efficient, and successful in achieving environmental objectives than was placement of slurried capping material using a tremie pumping system.
  • Pneumatic broadcasting of granular bentonite was unsuccessful as a means of material placement. This method/material combination should be eliminated from further consideration for the Grasse River.
  • Clamshell and tremie application techniques were able to reliably meet cap thickness targets. Reliability was dependent on the use of the WINOPS system and GPS technology. Crane operator experience was also a significant factor.
  • Tests using surface release and subsurface release of the 1:1 sand/topsoil mixture indicated that small differences may exist in the associated water quality impacts (surface release better) and uniformity of grain size distribution in the cap (subsurface release better). These differences are not significant enough to prefer one method over the other. Both surface and subsurface applications of the sand/topsoil mixture yielded good cap coverage and met target cap thicknesses.
  • Average placement rates for the 1:1 sand/topsoil mix via clamshell ranged from 43 to 50 cubic yards/hour (cy/hr) for the Test Cell, and from 58 to 66 cy/hr for Pilot Cells #2 through #4. These rates were not significantly dependant on the placement method: e.g., surface vs. subsurface; one lift vs. two lifts. Expressed on an area/time basis, the placement rates with the sand/topsoil mix (via clamshell) were 650 — 1,750 sqft/hr and 560 — 1,200 sqft/hr for the Test Cell and Pilot Cells #2 through #4, respectively. The average placement rates for AquaBlokTM alone were 20 cy/hr and 2,330 sqft/hr. For the combination of AquaBlokTM plus sand/topsoil in Test Cell Subcell #1D, the combined placement rates were 31 cy/hr and 1,000 sqft/hr.

Cap Materials

  • The 1:1 sand/topsoil mixture was found to be optimal considering the combination of logistics, production rates, cap coverage and unit costs ($/sqft). In addition, the sand/topsoil mix is expected to provide a reasonably natural sediment bed conducive to the re-establishment of pelagic and benthic aquatic ecosystems.
  • The TOC content of the in-place cap was variable (non-detect to 1.78%; average 0.71%), but sufficient to significantly attenuate any PCB migration through the cap.
  • Clamshell capping with AquaBlokTM, although involving some additional costs, was found feasible.

Cap Coverage Effectiveness

  • Average cap thickness met (or exceeded) the target thickness (1 or 2 feet) for all cells using clamshell capping with the sand/topsoil mix. The subsurface clamshell application of AquaBlokTM and sand/topsoil (in 2 lifts) did not achieve the target cap thickness (9 — 10 inches), due to variability of the sand/topsoil lift as well as to the inoperability of the WINOPS system during application in this cell.
  • The uniformity of cap thickness varied from cell to cell. In general, it was acceptable except for Subcell #1D (AquaBlokTM plus sand/topsoil). For Pilot Cells #2 through #4, where application was most representative of a full-scale project, the standard deviation of the cap thickness measurements, expressed as a percentage of the average thickness, averaged 27%. A very few of the depth-measurement locations showed no significant cap material; most of these locations were adjacent to the steep side slopes or the silt curtains.
  • Target cap thickness was not met on the steep side slopes. The best coverage was achieved in Pilot Cells #2 and #4 where 2 to 12 inches of material remained in place after application. No PCBs were detected in this material.
  • The centerline silt curtain was found to block small areas of the sediment bed resulting in areas near the curtain with little or no cap. This was remedied by centerline wedge capping after the silt curtains had been removed.
  • Fractionation of the cap material during placement does not appear to be a significant issue. Vertical profiles of TOC and grain size in the caps indicated no significant loss of fine-grained material or separation by grain size. Subsurface application resulted in a more uniform grain size distribution than did the surface application, but the difference was small.

Entrainment of PCBs Into Cap Materials

  • None of the application techniques resulted in significant PCB entrainment into the cap. PCB levels typically were near or below detection limits (commonly 0.06 mg/kg). The one exception was Subcell #1C2 (tremie application) where PCB concentrations of a few tenths of a ppm were commonly found in the cap (3 out of 4 cores). Potential explanations for this include on-shore contamination of the capping mixture, inadvertent contact between the tremie diffuser head and the River bottom, and bursts of air from the tremie pump.

For the Pilot Cells specifically, PCBs were non-detect in 95% (153 of 161) of the cap material samples analyzed from 21 cores. When detected, PCBs were less than 1 mg/kg (except one sample at 1.51 mg/kg), and were almost exclusively seen at the interval just above the native sediments.

Considering all cells, PCBs were detected in cap material just above the cap-sediment interface (0- to 2-inch sample interval) in several locations. Detailed discussion of PCB entrainment is presented in Sections 5.2 and 6.2.

  • The PCB entrainment data suggest a typical cap/sediment mixing zone of about 2 inches.

Water Quality Impacts

  • Water quality impacts associated with the Capping Pilot program were found to be negligible. Most important, PCB concentrations measured at the station downstream of the capping area were non-detect (< 50 ng/L) in all samples. Average PCB levels measured inside and adjacent to the test cells during cap placement were generally near or below the detection limit and were within the range of water column samples historically analyzed from a monitoring station near this area (7 — 240 ng/L). Water quality data for the three Pilot Cells are summarized in Figure ES-1.
  • Water column TSS and turbidity were also not significantly elevated by the capping. Levels were elevated inside the cell undergoing capping, but levels returned to baseline a short distance outside of the silt curtain (see Figure ES-1). Downstream levels of TSS and turbidity were slightly higher than those observed upstream, with average increases of < 1 mg/L and 1-2 NTU, respectively. Where elevated levels of TSS and turbidity were seen, it was primarily due to clean material since elevations in PCB concentrations, if any, were significantly smaller.
  • The corrective action trigger levels for PCBs (2 : /L), TSS (25 mg/L over background) and turbidity (25 NTU over background) — at the downstream monitoring station - were never reached during this capping project.
  • Water quality monitoring results obtained during the centerline wedge capping — done without silt curtains - were comparable to the results obtained with silt curtains in place.
  • The installation and removal of the silt curtains had no significant effect on water quality.
  • Post-capping water quality monitoring in October/November, 2001 showed no residual effects of the capping.


  • Direct construction costs of the various capping applications were evaluated. (These costs are limited to those incurred by the contractor [and subcontractors] involved in the actual implementation of the capping activities, but excluding some fixed costs. Costs associated with mobilization, silt curtains, design, monitoring and project management are also excluded.) Test Cell results — evaluated only on a relative basis (using placement rates and $/sqft values) — showed Subcell #s 1A and 1B2 (simple sand/topsoil capping) to have the lowest costs. Subcell #s 1B1 (sand/topsoil plus bentonite) and 1D (AquaBlokTM plus sand/topsoil) had medium-high and medium costs, respectively. Subcell # 1C2 (tremie with bentonite/sand/topsoil mix) had the highest costs. The higher costs generally resulted from both material costs as well as additional materials handling requirements. Unit costs for Pilot Cells #2 and #3 — with 1-foot target caps — were similar and averaged $2.15/sqft. The unit cost for Pilot Cell #4 — with a 2-foot target cap — was $3.10/sqft. Thus, as would be expected, the application of a 2-foot cap involves substantially higher costs than a 1-foot cap.

Recolonization by Benthic Community

  • The results obtained from the post-capping benthic community analyses indicate that the cap provides a suitable habitat for benthic recolonization. A total of 17 benthic species, including 3 new species not identified in the pre-capping assessment, were observed in the pilot study area (Test Cell Subcell #1D and Pilot Cells #2, #3, and #4).

Monitoring Program

  • The monitoring program carried out during this capping pilot study was extensive. Based upon the good results obtained (e.g., achievement of cap thickness targets, negligible water quality impacts and negligible PCB entrainment in cap), it would be reasonable to employ a more limited monitoring program in any future full-scale capping program on the Grasse River.

Overall Conclusions

Results of the pilot study indicate that capping of PCB-containing sediments can be successfully implemented in the lower Grasse River. Several application methods and capping materials were evaluated. Optimal results were achieved with a 1:1 sand/topsoil capping material applied — at the water surface or subsurface - via a clamshell attached to a barge-mounted crane. This combination was capable of generating a cap: (1) of acceptable uniformity and thickness; (2) with no significant PCB entrainment from the in-place sediments; and (3) with no significant alteration of the cap material (i.e., TOC loss or grain size fractionation). A sophisticated clamshell positioning system (GPS/WINOPS), as well as crane operator experience, was found to be important to success.

The Capping Pilot was carried out with minimal impacts on the environment. Water quality impacts during capping were negligible. Nearshore aquatic vegetation areas were left undisturbed, and on-shore land disturbance (for the staging area) was minimal since a prior staging area was utilized. Results of the post-capping benthic community analyses also generally indicate that the cap provides suitable habitat for benthic recolonization.

The pilot study provided valuable operational information — including data on application rates and unit costs — that will allow a reliable evaluation of full-scale operational parameters.

For More Information

If you are interested in more detailed information related to the Grasse River other than what is included on this website, please visit one of the information repositories established for the project.