Highlights

General Comments

Comments on the 
Draft Feasibility
Study (FS)

Comments on the 
Draft Remedial
Investigation (RI)

References

Complete Report
for easy printing

About Donna Boreck

Donna Boreck's 
Resume (PDF)

Technical Assistance
Grant (TAG)

Area of concern: The 1ppm PCB target extraction level is not protective to human or ecologic health. The 1 ppm 'knee' has no health-based rationale to it. An assumption is made that the remaining PCB concentrations will be managed by natural recovery, an uncertain gamble.

Discussion: See the comments of Dr. Foran on www.foxriverwatch.com. 

Is it ethical to use the 1 ppm (1000 ppb) 'knee' philosophy to clean up a substance that poses risk to human health and to a wide range of wildlife both in and out of the Lower Fox River and Green Bay? 

Capping in Lower Fox River

Area of Concern: The capping, as proposed, does not effectively address the problem of PCB contamination in the Lower Fox River. Even if it is done right, capping is a high-maintenance temporary fix for a long-term problem.

Discussion: Capping in conjunction with natural recovery has been proposed as an alternative to the remediation actions proposed in the RI/FS. There are a number of weaknesses to using capping in the Lower Fox River. The following comments are based on what is presented in an article posted in the Green Bay Gazette on December 5^th, 2001:

• The article noted that several designs would be incorporated into the cap to be used in the river. From a WDNR publication on Capping dated July 2000, the requirements for capping include an area where "currents are no greater than 0.15 feet/second". Average current velocity for Zone 1 is reported at 0.25 feet/second. Other reaches of the river reportedly have a higher average velocity (0.37-1.23 feet /sec).

 
• WDNR (2000) also noted that capping should be done in areas where the maximum 100-year flood current is no greater than 0.7 feet/sec. I do not know what the 100-year flood current is for Zone 1, but average is 0.25 feet/sec (approximately 1/3 of the proposed 100-year maximum). If a 100-year discharge is 5-6X greater than the average discharge (as is true at the Rapide Croche gauging station) for Zone 1 (or any reach along the Lower Fox River), I question if Zone 1 or other reaches along the Lower Fox River meet the 0.7 feet/sec flood current requirement. 

 
• According to Dr. Reible, an expert on capping of contaminated sediment, the technique…"has never been tried in the United States on a scale like that proposed for the Fox River. I don't think there is any doubt this would be the biggest…". If this capping project is the biggest ever attempted in the United States, it is in the area of 'unproven technology'. There are risks and complications that may come up with 'super-sizing' any engineering technique and that is a risk that ultimately is handed to the people of Northeastern Wisconsin. Remediation of the large volume of PCBs present in the river and Bay requires the use of the Best Available Technology -- dredging and removal of the contaminated sediment.

 
• Capping requires maintenance and long-term monitoring. This is especially true in a riverine environment where the cap is exposed to failure from erosion or other river dynamics. Monitoring includes measurement of cap thickness, sediment chemistry, and observations of biological recolonization. Monitoring must include evaluation of zones in the cap that may need rebuilding. PCBs do not break down quickly. Cap monitoring would require significant expense and man-hours for an extended period of time (greater than 40-60 years projected by the panel) to guarantee the effectiveness of the cap. Removal of the PCBs from the river and bay is the most cost-effective means of improving health of future generations. 

 
• Capping ties the hands of residents and elected officials in a number of areas including land-planning, economics, and recreation. The action will continue to limit use of the river for centuries.

 
• The study recommends that capping be constructed to a standard that would enable it to withstand water forces three-times what would occur in a flood so severe it would only happen on an average of once per century. A 100-year-flood (a confusing term) is a flood that has a 1% chance of occurring every year. There is nothing magic about a flood event that happens to us only every 100 years. The same flood event could occur multiple times in five years. A larger flood event (250- to 500-year) could also occur in that same five years. Use of a term like a 100-year flood event gives a false sense of security. 

Area of Concern: Many of the findings in the RI/FS are based on modeling. Each step in the modeling process has its limitations. Making a decision to act based on model results is appropriate only if the decision takes into account the known limitations of the model. Making a decision not to act (i.e. proposed natural recovery and monitoring) based on model results is inappropriate. 

Discussion: The draft RI/FS is based mainly on the results of a chain of models run to determine potential levels or effects of different parameters. Even the best modeling has its weaknesses. For the draft RI/FS, each step in the modeling process has its limitations. The following are examples:
 

MODEL	UNCERTAINTY
Lower Fox River Sediment Bed Properties-Inverse Distance Weight Method.	The results are sensitive to what value is used for the radius of influence. A larger radius gives more coverage, but a smaller radius may show more accuracy for specific deposits. There may be an increase (or possible decrease?) in estimated PCB concentrations with the larger radius of influence.
GB-SED Green Bay Sediment Transport Model	The model does not include data to assess the impact of individual storm events or tributary loading. There is some uncertainty in specification of the settling velocity function, resuspension potential parameters and shoreline erosion. Settling function may not be applicable to all sites. Rususpension potential parameters were obtained from shaker studies on only one core. GB SED does not account for internal production of solids, an important component of the sediment mass balance in the bay.
Computation of Internal Solids Loads in Green Bay and Lower Fox River.	Listed areas of uncertainty include scarcity of data, representativeness of data, scaling of data, uncertainty in conversions and uncertainty in constants. Other areas of uncertainty include uncertainty in extrapolating data and in phosphorus loads. "Monte Carlo simulations show that parameter uncertainty and sample variability together imply a normal standard error of about 25% in predicted BIC (biotic carbon loading) for Green Bay and a normalized standard error of about 62% and 64% for predicted BSS (biotic suspended solids) in the Lower Fox River, above and below DePere Dam, respectively (Tech Memo 2c, 1999, p 26)".

Each step in the modeling process brings in a level of uncertainty. It is important to bring into focus the uncertainty when evaluating the results and making a decision to take action. But, with what is empirically known about PCB contamination in both the River and the Bay, deciding not to act based on any modeling results, is inappropriate. 

Modeling

Area of concern: In principal, the method used to determine the 1 ppm extraction level is not appropriate for a high risk contaminant like PCB. PCBs biomagnify in the environment and threaten human health and our ecology. Lower levels of PCBs (below the 1ppm "knee") do not equate to acceptable risk. The 1 ppm recovery goal seems more focused on minimizing volume or cost and less focused on what needs to be recovered to minimize risk to human health.

Discussion

In the RI/FS, a method was used to determine the volume of in-place PCB contaminated sediment and the percentage of the total deposit that would be extracted . According to the modeling, in many stretches of the river, a large volume of the PCB contaminated sediment will be recovered at the 1000 ppb (1 ppm) level. The work leads to the assumption that trying to extract the remaining in-place PCB contaminated sediment is not worthwhile. The 1 ppm value is what has been called the `knee' in recovering PCB. Below this knee, recovery of smaller amounts of PCBs would require digging out more volume of sediment in the river. The action will be more costly per pound or kilogram of PCB recovered and processed. 

The method is similar to a standard reserve study that would be done for industrial minerals, metals, and other materials that are to be extracted from the earth. A reserve study is cost-based and cost-driven and focuses on how much you can get per pound or ton of the commodity. You extract what you can for a profit and leave the remaining lower grade material in-place, either for future extraction when the mineral's price goes up or as waste

I see many parallels to how the PCBs have been handled. But, in this case. we are not dealing with a material that is beneficial to mankind. We are dealing with a pollutant that, even at lower levels, can biomagnify in the system. Despite its "lower volatility", PCBs can be mobile within the environment for a long time. The 1 ppm recovery goal seems more focused on minimizing volume or cost and less on what needs to be recovered for human health.

Modeling

Area of Concern: Recovery of any material or mineral is never 100%--Recovery of the PCB contaminated sediment will not be 100%. The recovery of minerals from a deposit can be 60% or less of the volume of mineral in-place. In Deposit N, approximately 80% of the PCBs were recovered and 20% remained in-place. Successful dredging of any reach in the river does not guarantee that all of the PCBs calculated to be in-place are taken out. To maximize recovery, remediation efforts need to focus on dredging and removing as much of the deposit as possible. Cutting off recovery at a set target level will further lower recovery of in-place PCBs.

Discussion: Recovery of any in-place substance in the earth is never 100%. A recovery factor for minerals can be 60% or lower. In Deposit N, approximately 80% of the mass of PCB (calculated to be in-place) was removed during dredging. Approximately 20% remained. The dredging operation hit hard rock at the base of the soft sediment which prevented them from digging further. The remaining PCBs are believed to be tied-up in sediment that has filled fractures in the rock.

In Deposit 56/57, a total of 6600 lbs of PCB were projected to be in-place. During the 1999 demonstration project and 2000 demobilization project, an estimated 1,441 pounds or 22% of the PCB sediment was recovered. The project was plagued by weather, large debris in the river and a number of other factors. 

Based on modeling, if contractors recover the 1000 ppb (1 ppm) sediment at the De Pere to Green Bay reach, they project they would recover 99 percent of the PCBs calculated to be in the reach. Yet, projected recovery may still be lower. If we use Deposit N's recovery factor, the actual recovery may be closer to 79%. 

They are planning to over-dredge the deposit as a safety. But, setting PCB recovery on the '1ppm knee' target concentration is aiming too high. Even if contractors over-dredge, the over-dredging may not be sufficient to actually recover the volume necessary to remove contamination at the 1 ppm target level. 

The presence of PCB is a human health and ecological health issue. We need to be conservative and shoot lower than the 1ppm target level to maximize both recovery of in-place PCBs and the long term benefit to human health and the environment.

PCBs Don't Move by What is Average

Area of Concern: Using average values in the RI/FS gives a false sense of security in visualizing PCB sediment transport or sediment deposition into Green Bay.

Discussion: In the report, average values were often quoted for wind speeds, currents, etc. Use of average values is scientifically valid in looking at modeling for steady state conditions. But steady state does not always reflect what is actually going on, sometimes daily. 

In the draft RI, the findings of Fitzgerald and Steuer (1996, RI p 2-5) were quoted. "…Based on the seasonal variations in PCB concentrations, it is estimated that more than 60% of the PCBs transported over the DePere Dam occurs during 20 percent of the year, when discharge is at its greatest." 

Average values for different parameters in Zone 1 are given below:

Zone 1- Lower Fox River -- average discharge. 4,300 cubic feet/sec (122 cu m/sec).

Zone 1-Lower Fox River -- average velocity. 0.25 ft/sec (0.08 m/sec). 

On October 14^th, real-time discharge varied from -5000 to +11,000 cfs at USGS Stream Gage 040851385 (Oil Tank Depot at Green Bay). The real-time data was taken over a time period from 1200 am to 830 am. During that same period, velocity varied from -0.62 to +1.07 feet/sec. The velocities recorded on the two Acoustic Velocity Meters (AVMs) exceeded 0.30 ft/sec 65% of the time during that 8 1/2 hour period. Higher flow levels were sustained for up to 2 hours and 45 minutes.

Over a 14 day period from Oct 28^th through November 10^th 2001, the velocity varied significantly with an average of 0.61 feet/sec recorded at the lower AVM. Discharge also varied significantly with an average discharge of greater than 6000 cu ft/ sec. On October 29^th, both upper and lower AVMs pegged near +2.0 and -2.0 feet/sec. Discharge reached 20,000 cu ft/sec. Reportedly, there was a significant wind storm during that time that affected the flow velocity during this late October-early November time period. 

The data sets are real-time and considered provisional, which means the values may have to be adjusted for some errors in equipment or other conditions. The data does reflect how much values can vary from 'average'. With both data sets, the value of 0.3 feet per second was surpassed often. During a 14 day time period, the average value for the lower AVM was 0.61 feet/sec which surpasses the erosion limit of 0.3 feet/sec in the Fox River. 

We cannot define the risk from PCBs by using average values. PCBs and other contaminants move in the system by events that are not 'average'.

Green Bay -- Remediation

Area of concern: Green Bay must be remediated and not left to natural recovery -- Green Bay is a dynamic system. Green Bay is not quiet water. Suspension and resuspension of sediments is possible and ongoing.

Discussion: From the RI/FS, leaving the PCBs in-place and allowing the Bay to recover naturally would eventually allow for dilution or breakdown of the PCBs to a level where they would no longer affect wildlife. Natural recovery also would include burying the PCBs with clean sediment sources from riverine and bank erosion and input from Lake Michigan. The theory only takes into account a part of the natural processes active in Green Bay. Green Bay is a dynamic system--it is not quiet. 

Currents -- Information on currents for Green Bay was limited. From the RI/FS:

Green Bay --- Current velocity east of Chambers Island to 1.1 ft/sec (0.35 m/sec)

Current velocity west of Chambers Island 0.4 -0.8 ft/sec (0.12-0.24 m/sec)

Current velocity Inner Bay to 0.4 ft/sec (0.12 m/sec)

• Currents near the bottom were well correlated with surface winds.

 
• At a 28 meter depth, they inferred from their results that local resuspension could occur approximately 20% of the time.

 
• The critical mean flow speed for resuspension at 0.7 meters (2.3 feet) above the bottom was approximately 0.18 m/sec (~0.6 ft/sec). This value was for a bottom of silty sand. 

Wind speeds -- Average wind speeds reported in the RI are at or below 19 mph . (specifically, wind velocity S-SW 37% of time 6-19 mph (10-30 km/hr)) These values agree well with other published sources. Again, `average' does not define risk of remobilizing PCBs into the water column. According to ESRI hazard maps (www.esri.com/hazards) for the Green Bay area, wind speeds in the area of 60-80 miles per hour have been reported to occur post 1970. On the Door Peninsula, one value has been reported within the 120-140 mph range. In the RI on page 3-43, the following research finding was quoted: "… Moderate to strong winds are the most important factor for bay sediment resuspension and occur, on average, every seven days on the Great Lakes. " 

Sediments -- In the Bay, sediments have a greater chance to settle and compact than in the Lower Fox River. PCB-bearing sediments are often very fine particles that tend to be cohesive once they settle in-place. Yet, once the fine sediments are resuspended, the particles can be transported by currents for a long distance before they settle again. According to research, even when fine particles flocculate, the flocs can be light, unstable and break apart easily, keeping the particles in suspension and transportable for long distances. 

Green Bay is a dynamic system that will continue to transport sediments and ultimately PCBs within the Bay, out to Lake Michigan, and back up the Lower Fox River. Natural recovery requires a quiet environment that will allow for continual burial of the contaminated sediments undisturbed by other factors. Green Bay does not offer that environment. As much as we would like the PCB contaminated sediment to 'stay put', it may not oblige. 

References

ESRI, Inc. Online Hazards Map -- Wind Hazards. Green Bay, WI area. 2001. At www.esri.com/hazards.

Lesht, Barry M. and Nathan Hausley. 1987. Near Bottom Currents and Suspended Sediment Concentration in SE Lake Michigan. Journal of Great Lakes Research ,13(3): 375-386).

Tsai, Cheng-Han, Sam Iacobellis and William Lick.1987. Flocculation of Fine-Grained Lake Sediments Due to a Uniform Shear Stress. Journal of Great Lakes Research 13 (2): 135-146.

Reineck and Singh, Depositional Sedimentary Environments. 2^nd Edition, Springer Verlag, 1980, 549pp.

Green Bay -- Remediation

Area of concern: Green Bay has a diverse bottom-dwelling community that will be affected by the PCBs. The RI reported a benthic community in Green Bay which included unidentified species of worms. A characteristic of worms as well as other bottom dwelling creatures is their tendency to burrow. The burrowing can turn over and disrupt the PCB contaminated layers. If the contaminated sediment is buried, bioturbation may bring buried contaminants to the surface. Lastly, the burrowers are a food source and can continue to bring PCB into the food chain long after the contaminated sediment is buried.

Discussion: If the currents don't stir up the bottom sediments, other factors may. The RI reported a benthic community of oligochaetes among other organisms. The oligochaetes were described as unidentified species of worm. According to Lauritsen and others (1985), in Green Bay, tubificids were the dominant oligochaetes found (usually >80%) at shallow depths of <40 m. The tubificid Limnodrilus hoffmeisteri --a eutrophic species abundant in Green Bay -- was suggested as an indicator for pollution. 

Another article by White and others (1987), quoted a maximum feeding depth of 3-6 cm for tubificids, although the organisms in densely populated areas may burrow deeper to compensate for lack of room. 

Even in quiet water, the presence of burrowers can turn over and disrupt the sediment (bioturbation) or otherwise provide access to lower buried layers of PCB contaminated sediment. The burrowers are a food source and can continue to bring PCB into the food chain long after the contaminated sediment is buried by clean sediment. 

The worms, especially Limnodrilus hoffmeisteri, maintain themselves in a high carbon contaminated environment. The most effective way of preventing spread of PCB contamination from sediment reworking and/or through consumption of the worms as a food source is to remove the source of the PCBs. 

References:

Lauritsen, Diane D, Samuel C, Mozley, and David S. White. 1985. Distribution of Oligochaetes in Lake Michigan and Comments on Their Use as Indices of Pollution.

Journal of Great Lakes Research 11 (1):67-76

White, David S., Patricia C. Klahr and John A. Robbins. 1987. Effects of Temperature and Density on Sediment Reworking by Stylodrilus Heringianus (Oligochaeta: Lumbriculidae). Journal of Great Lakes Research 13 (2): 147-156

Green Bay -- Remediation

Area of concern: Green Bay has a large volume of PCB in-place. The Bay will continue to be a source of contamination for an undetermined amount of time. The contamination includes exposure to wildlife and humans in the general area of Green Bay, transport of PCBs to Lake Michigan, and volatilization of PCBs into the atmosphere. Once in the atmosphere, the PCBs can be redeposited onto land and water far from Green Bay. Although the PCBs are in lower concentrations in the Bay, they are recoverable. If they are not recovered, the PCB contamination will spread.

Discussion: Although the Green Bay sediments are not as contaminated with PCBs as the Fox River, the sediments are and will continue to be a source of PCBs to Green Bay, Lake Michigan, and to Wisconsin and adjacent states. The sediments may be lower in PCBs than those slated for removal from the Lower Fox River. But, the sediments have enough PCBs in them to be a concern to wildlife in the area. The safe drinking water level for wildlife is 0.12 nannograms per liter (ng/L, 0.00012 ppb). Filtered water samples taken in Zone 3B (39 filtered water samples) varied from 0.5 ng/L to near 4 ng/L. All 39 samples exceeded the safe drinking water level for wildlife. These are low values when you compare with the higher values coming over DePere Dam, but they are still valid and they are still above the level of safety for wildlife. Analysis of the particulates filtered from the sampled water also resulted in values lower than the highest values reported in the Fox River. The values range to 7 ng/L. But, they still represent a mobile source of PCBs available for uptake or for transport. 

From the RI (Section 7.4, page 7-16), Green Bay gives away more PCBs than it receives from other sources. Green Bay transports an estimated 270 lbs of PCB annually into Lake Michigan. Green Bay's annual PCB contribution to the atmosphere through volatilization is estimated at from 287 to over 1000 lbs. According to USEPA (2001), PCB is recycled through the environment through volatilization from ground surfaces (water and soil) into the atmosphere followed by removal from the atmosphere by precipitation or by sedimentation. According to Atkinson (1996), the estimated residence time for particles in the atmosphere may be from ~5-15 days with more highly chlorinated arochlors in the atmosphere for up to ~30 days. 

The Lower Fox River supplied Green Bay with a substantial reserve of PCBs. But Green Bay is now an independent source of pollution. Contamination can be spread to Lake Michigan, via volatilization to distant areas and back to the area near the mouth of the Lower Fox River. Leaving Green Bay to natural recovery would be costly in terms of human health, local and regional environment and the future.

References:

USEPA. EPA Fact Sheet on PCB--Technical Drinking Water and Health Contaminant Specific Fact Sheet. Updated 12 April 2001. From www.epa.gov/OGWDW/dwh/t-soc/pcbs.html.

Atkinson, Roger. Atmospheric Chemistry of PCBs, PCDDs, and PCDFs in Chlorinated Organic Micropollutants. Ed.s R.E. Hester and R.M. Harrison Issues in Environmental Science and Technology No. 6. The Royal Society of Chemistry,. 1996, pp 53-72.

back to top