Abstract

Introduction

Description

Sampling methods

Suspended-solids transport during dredging

PCB concentration changes during dredging

PCB loading in the Fox River due to the dredging operation

PCB transport back into the river from the onshore-processing operation

Postdredging PCB concentration and loads

Adjusting water-column PCB concentrations to allow comparison with onshore-sample PCB data

Lessons learned

References

Acknowledgments and Information

A Mass-Balance Approach for Assessing PCB Movement During Remediation of a PCB-Contaminated Deposit on the Fox River, Wisconsin

PCB loading in the Fox River due to the dredging operation

Putting the PCB concentration increase in a useful context requires calculation of mass fluxes, such as the daily PCB load (expressed as mass) to the water due to dredging operations and the amount of PCB processed in the onshore operations. For a given sample-collection day, the PCB load due to dredging operations was calculated by multiplying daily streamflow by total PCB concentration (summation of all congeners, dissolved and particulate).

A daily net PCB load due to dredging (fig. 9) was computed by multiplying the difference between the downstream and upstream PCB concentrations on a given sample day by the daily discharge. Net PCB loads, in general, increased after November 15. This result is consistent with a change in operations: the dredge had been moved to an area of the deposit that contained higher PCB concentrations and was closer to the downstream transect (Blasland, Bouck, and Lee, Inc., 2000). Additionally, streamflow increased substantially after November 15 (fig. 10).

An initial overall PCB load estimate was calculated using the median daily PCB load for the two intervals (before November 16 and after November 15). The median was used, rather than the mean, because daily PCB loads were not normally distributed for either interval. The median daily PCB loads for each interval (42.4 gm and 364.3 gm, respectively) were multiplied by the median flows to provide initial load estimates of 3.2 kg and 10.9 kg, respectively.

Further PCB loading analyses examined relations with variables that were measured daily (dredge-slurry settled-fraction concentrations and supernatant concentrations). Concentrations of the material being removed from the deposit or the amount of dredging per day might be indicators as to how much PCB was transported downstream. Frequently monitored variables such as turbidity, streamflow, and stream depth also were examined. A usable regression relation could not be developed for the interval prior to November 16. For the dredging interval after November 15, however, a regression was developed in which four factors explained much of the variability (r² = 0.88): daily PCB concentration of the incoming slurry mixture (settled fraction (Set) and supernatant (Sup)), time spent dredging on a given day (T), and stream depth (D).

Daily PCB load (gm) = (-507.9)(D)-(1237.9)(T)-(5.69)(Set)+(32.5)(Sup)+293,703

The daily PCB loads resulting from this regression equation (fig. 9) were summed to arrive at a load of 13.7 kg for the post- November 15 dredging period. The standard error was 25 percent of the mean.

The final estimated PCB load (16.9 kg), combining the median based pre-November 16 load (3.2 kg) with the regression-based post-November 15 load (13.7 kg), was selected as a conservative approach.

An estimated PCB load entering into the dredged area from upstream (fig.1) was computed by applying the median daily upstream PCB concentration (51.4 ng/L) to the median daily flow (1,842 ft³/s) for the 106 days. The result was an estimated overall PCB load of 24.5 kg entering the deposit cross-section from upstream.

Congener distribution changed noticeably during dredging (fig. 11). Congeners 5/8, 4/10, and 6congeners that readily volatilize to the atmosphere-are noticeably less prevalent at the upstream site than at the downstream site. Air monitoring during remediation has shown that the river routinely volatilizes PCB to the atmosphere; thus depletion of these congeners is not surprising at the upstream site. Sub-surface sediments and pore waters that are newly exposed during the dredging may replenish these congeners, as is reflected by concentrations at the downstream site.

The Fox River Mass Balance Study (Steuer and others, 1995) estimated a PCB volatilization-to-advection ratio of 13 percent. Applying this ratio to the upstream PCB advection (20.9 kg; Aroclor*1242) yields an estimated 2.7 kg volatilization from the Fox River upstream from SMU 56/57 during the dredging period. Air monitoring at the shore-processing site indicated that between 0.3 to 4.9 kg of PCB volatilized from that facility during the 106 days of onshore processing (Blasland, Bouck, and Lee, Inc., 2000; David Grande, Wisconsin Department of Natural Resources, written comm., 2000).

To put into context the PCB input to the water column during the dredging operation (16.9 kg), one can consider PCB loading from the river with no dredging taking place. The monthly river PCB load is variable (fig. 12); in 1994-95, annual Fox River PCB loading (congener summation) was 186 kg/yr (D.W. Hall, U.S. Geological Survey, written commun., 1999).

U.S. Department of the Interior, U.S. Geological Survey
http://wi.water.usgs.gov/
Maintainer: Webmaster, djsulliv@usgs.gov
Last Update: Mon 22 Dec 2001