Successful Remediation by Dredging of PCB-Contaminated Sediments in Lake Järnsjön: Investigations, Considerations and Remedial Actions

Ellicott's Experience in Dredging PCB's

Successful Remediation by Dredging of PCB-Contaminated Sediments in Lake Järnsjön: Investigations, Considerations and Remedial Actions

Remediation of Lake Järnsjön in Emån (a river in southeastern Sweden) is part of the Swedish National Site Remediation Action Plan. Lake Järnsjön contained approx. 400 kg PCB that were slowly leaching out from the sediments into Emån. Without remedial action the sediments would have caused problems for many decades. Several investigations and studies were carried out to determine how and under what limitations cleanup could be performed. An alternative for the remediation was selected that included vacuum dredging within a protective barrier of geotextile screens. Dredged material was to be dewatered and disposed in a landfill. Remedial activities in Järnsjön could, however, lead to an additional burden on Emån. Restrictions on and protective measures against such releases during remediation of Järnsjön were therefore important issues. Dredging was carried out in 1993 and 1994. When dredging was completed, an amount of about 150 000 m³ had been dredged. According to the follow-up, 394 kg of PCB were removed, which constitutes about 97% of the estimated total amount of PCB in the lake. In the areas included in the remediation operation, about 2.9 kg of PCB were left. The major part of the remaining PCB (7.4 kg) is now located at or near the lakeside, an area which was not included in the remediation operation. In the landfill, heavily contaminated sediments were separated from slightly contaminated sediments by a geotextile screen. The reason was to make it possible to localize the heavily contaminated sediments in the future, when remedial techniques have evolved and it may be feasible to carry out more definitive action.

Thus far, the follow-up indicates that the landfill is behaving satisfactorily. No increase in PCB concentration has been found in groundwater under the landfill, compared to reference levels. After remediation the concentration of PCBs in the lake decreased from 8.6 ng/L in 1991 to 2.7 ng/L in 1996. PCB concentrations in fish from Järnsjön decreased from 34 ug/g extractable fat in 1991 to 16 ug/g in 1996.

Remediation resulted in lowered PCB concentrations in sediments, decreasing from 5 ug/g d.w. to 0.06 ug/g d.w. After the landfill had been closed PCB concentrations in the air were similar to background levels. The dioxin-like activity of the diaromatic fraction in sediment collected after dredging had about 1% of the activity of the predredged sediment, showing that remediation of the lake was successful. Compared with the levels in the 1993 samples of settling particular matter, the levels of diaromatic dioxin-like compounds in the 1996 samples were substantially lower. Two years after the remedial activities, the EROD activity was slightly induced in caged rainbow trout, but the fish showed no sign of histopathological changes.

INTRODUCTION
Remediation of the Lake Järnsjön in Emån is part of the Swedish National Site Remediation Action Plan ⁽¹⁾. It has been estimated that there are 22 000 sites (1998) in Sweden with contaminants hazardous to human health and the environment. The long-term goals of the action plan are to identify these sites, investigate and, if necessary, remediate them within a 40-yr period. Several hundred of these sites involve contaminated sediments. The most serious contaminants found in sediments are methyl-mercury and PCB; Downstream of a number of paper and pulp plants, wood fiber contaminated with mercury has accumulated in lakes and rivers. It has been regarded as difficult to remove these sediments without harming the ecosystem. To demonstrate that this can be done in an ecologically and economically sound way, the Lake Järnsjön Project was planned and implemented. This project aimed at extending knowledge about cleaning up contaminated sediments and at decreasing PCB exposure in the lake and downstream areas.

EMÅN: GENERAL DESCRIPTION
Emån is the largest watercourse in southeastern Sweden. It runs from an altitude of approximately 330 m and discharges its waters into the Baltic Sea. Emån is 220 km in length and its drainage basin, encompassing four counties and eleven municipalities, covers an area of approx. 4500 km² ⁽²⁾.

The upper part of the catchment area has many lakes, the lower part has none. The percentage of lakes in the entire catchment area is 5 to 6% ⁽³⁾. The absence of reservoirs in the lower part of the catchment area results in large variations in water flow: from 2 m s^-1 to 270 m s^-1. The mean average flow at the outlet into the Baltic Sea is 30 m s^-1 ⁽²⁾. These large variations in water flow cause problems, such as flooding during spring and lack of water for irrigation and industrial purposed during the summer months. Precipitation in Emån catchment area is relatively sparse, totaling some 700 mm/yr for the highland areas, and 500 mm/yr at the coast ⁽²⁾. The entire catchment area is situated within the southern coniferous forest region. Agricultural land is found primarily around lakes and in the river valley.

Biological Values
Emån is a valuable watercourse in southeastern Sweden. The main watercourse as well as several tributaries are of national interest for biological, cultural, and recreational reasons ⁽²⁾. The river is characterized by multiple rivulets and very picturesque meanders. The great variety of habitats in the catchment area provides a solid basis for considerable biodiversity, in terrestrial as well as aquatic environments. The river accommodates more than 30 fish species. A large population of sheatfish (Siluris glanis) is found in the river ⁽²⁾. Large and fast-growing sea-trout (Salmo trutta), are common in the parts of the river close to the outlet to the Baltic Sea ⁽²⁾. Sweden's southernmost population of char (Salvelinus alpinus) is found in a lake of the catchment area. Other valuable fish species in the river are asp (Aspius aspius), chub (Leuciscus cephalus), vimba (Vimba vimba), and spined loach (Colbitis taenia) ^(2).

The Emån catchment basin has a very high diversity of benthic organisms, some of them (11 species) very rare and listed as threatened species. A few such "red-listed" species are Margaritifera, Ibisia marginata, Myxas glutinosa and Rithrogena germanica. Certain parts of the river are home to more than 60 species of benthic organisms, some of which occur abundantly ⁽⁴⁾. A rare mammal species found in several places in Emån catchment area is the otter (Lutra lutra). Observations are most frequent in the upper parts of the catchment area where PCB concentrations are lowest.

Nutrients
The concentration of nitrogen is high in the main watercourse of Emån and moderate in most of its tributaries. The concentration has increased steadily over the past 30 years and is today 1.5 to 2 times the background concentration ⁽⁵⁾. Agriculture, cities, and atmospheric deposition are responsible for the largest portion of anthropogenic nitrogen. During 1994, Emån transported more than 1000 t of nitrogen to the Baltic Sea, where it contributed to eutrophication ⁽⁶⁾. The concentrations of phosphorus in the main watercourse and n the tributaries are low to moderate and, unlike those of nitrogen, have decreased over the past three decades. Today, the concentrations are close to the background concentration. This is due largely to the construction of sewage-treatment plants during the 1970s, and subsequent improvement of techniques, including chemically-mediated precipitation of phosphorus ⁽⁵⁾. During 1994, Emån transported some 22 t of phosphorus to the Baltic Sea ⁽⁶⁾.

Metals and Persistent Organic Compounds
Parts of Emån are heavily polluted by metals (Cd, Cu, Pb). Until it was shut down in the mid-1970s, a Ni/Cd battery factory discharged large amounts of nickel and cadmium into Emån. During this period, concentrations downstream were 10-20 times higher than upstream. Some years later measures were taken to reduce leaching from the slag deposited beside the river, which resulted in drastically decreased downstream concentrations. However, Emån is still one of the most cadmium-polluted watercourses in Sweden. Further decontamination of soil and buildings is planned ⁽²⁾.

Silveran is one of Emån's tributaries, and was formerly heavily polluted by lead from an accumulator factory. Over the years, large amounts of lead sedimented in a lake downstream. Today, this lake constitutes a point source of lead to Silveran through release from its sediments.

Paper and sulfite pulp have been manufactured in four papermills located in the catchment area for over more than 100 years. The tributaries Silveran and Pauliströmsan as well as certain stretches of Emån mainstream have been heavily affected by discharge of paper-pulp deposits containing among other compounds mercury, PCB and PCN. Mercury was used in production until 1968; PCB in open systems was prohibited in 1972. From that year on, sewage-treatment plants were built to reduce the burden of paper fiber on the river. In papermills situated alongside the main stream, PCB-containing self-copying paper was recycled in the production process. Large quantities of PCB-containing paper fibers sedimented in Järnsjön situated downstream.

Use of the River Water
The waters of Emån have, for many years, been used for several purposes. Most of the large lakes within the catchment basin are regulated. This control of the water was introduced at the beginning of the century for power production purposes. There are approximately 40 power stations and more than 100 dams within the catchment area ⁽²⁾. These power stations prevent migrating fish species, such as salmon (Salmo salar), from reaching their breeding grounds. Furthermore, the power stations exert a negative effect on the area's biota through large and rapid fluctuations in the water level. During a dry year, some 5700 ha are irrigated using approximately 2.6 million. M3 of water ⁽²⁾. Within the near future, the use of water will be coordinated by an Irrigation Association in order to ensure adequate water flow in the river during periods of drought.

Background

Identification of the Environmental Problem

Elevated levels of PCBs were measured in the early 1980s in foam at the mouth of Emån. High levels of PCBs were also measured in fish (140 mg/kg of fat). Through various processes, e.g. particle transport, the pollutants are carried downriver, finally reaching the Baltic Sea.

A large number of industrial enterprises are or have been situated along Emån and its tributaries. Before adequate pollution control was adopted at these plants, large quantities of pollutants were discharged into the watercourses. Knowledge and awareness of the risks of using certain preparations were also inadequate. Due to the use of recycled wastepaper containing PCBs in self-copying paper, large quantities of these substances were discharged into Emån with the former fiber effluents from a papermill. Large quantities of PCB-contaminated fibers had accumulated in Järnsjön (Fig. 6). There were about 400 kg of PCBs in about 150,000 m3 of sediments ⁽⁷⁾.

Figure 6. PCB distribution in Lake Järnsjön before remediation. Sediments in Lake Järnsjön, level o-40 cm. Divided into sectors, figures in mg kg^-1 d.w. and quantities in kg.

Hazard Assessment
The presence of PCBs in the water system threatened the values, which form the basis for the river's classification as a national interest area. Effects on stationary fish, like damage to internal organs, development and reproduction were expected. A sampling and analysis program for Emån has shown that the Järnsjön sediments were the primary source of ongoing discharges of PCBs to the river ^(8,9). Based on a yearly discharge of about 7 kg PCBs, the 400 kg in the sediments would have caused problems until at least the year 2060 ^(9,10). However, the gradual reduction in the amount of available PCBs meant that the discharge rate would probably decrease. The problem would therefore persist for a much longer period of time, although at a lower rate. For these reasons remediation of Järnsjön was necessary if continued damage was to be avoided.

Project Base
The Swedish Environmental Protection Agency was responsible for the project up to and including the detailed planning. After that a local municipality took responsibility for remedial action. This meant that the project organization changed during the course of the project.

From the start in the spring of 1990, up to and including detailed planning in August 1992, the cost amounted to USD 770,000. The total cost of the project was estimated at USD 6.4 mill.

Organization
The organization of the project was functionally divided into an organization characterized by a project principal; project manager; subproject leaders responsible for different areas such as project-planning, environmental monitoring, legal official permits, purchasing and contracting, administration and technical and environmental support. The project principal also functioned as the client's representative. The client's organization consisted of representatives of the financiers and the Kalmar County Administration, where the Swedish Environmental Protection Agency, the Kalmar County Administration and the municipality of Hultsfred had a primary function. The work of the project followed a specific manual and specific procedures that were gradually implemented into the organization.

Planning
A network plan created in a primary time schedule was applied within the project. The time schedule aimed at ensuring that established time goals should be satisfied. The project was carried out within the agreed time schedule.Cost estimates for the project were based on unit costs and quantities. The cost estimates were revised successively over a period of time, as the uncertainties decreased and the estimates became more accurate. Periodically, the project manager produced reports and predictions through a system involving checking off cost estimates and budget. The project was completed within the agreed financial frame.

Strategy
The decision taken in June 1991 was to clean-up Järnsjön, using suction dredging and disposing the dredged sediment by landfilling after mechanical dewatering. In a comparison with other possible alternatives this turned out to be the option that best satisfied both environmental and technical/economical requirements ⁽¹¹⁾.

Projects Involved
The feasibility of remediating Järnsjön and the conditions under which such actions could be carried out were examined in a number of studies and investigations.

Environmental scenarios
Environmental scenarios for various PCB exposure situations were studied and the results were fundamental for the design of the project ⁽¹²⁾. Remedial activities in Järnsjön could lead to an additional burden of dissolved PCBs and distribution of suspended matter contaminated with PCBs in Emån downstream of the lake.

Restrictions of such releases during remediation of Järnsjön had to be enforced. Even a long-term dredging procedure with a low release of PCBs can lead to an unacceptable transport of PCBs downstream of the lake.From an ecological viewpoint, based on the various situations described in scenarios, only two alternative remedial procedures could be considered. The first alternative was a diversion of Emån past Järnsjön via a new channel. The second alternative was conventional dredging of PCB-contaminated sediment in Järnsjön with the maximum allowable release to the Emån equivalent to a doubling of the present transport.

Methods for cleanup and processing of industrially-polluted sediments
Another study was carried out with the aim of describing various methods for cleanup and processing of industrially-polluted sediments ⁽¹³⁾. Remediation measures in aquatic environments place high demands on reliable cleanup methods not to disturb ecosystems close to or downstream of the area being cleaned up. Methods for in situ treatment, removal and processing were described in the study.

Enviro-geotechnical report
An enviro-geotechnical report that described the disposal of the PCB-contaminated sediments was published ⁽¹⁴⁾ as was a general inventory of possible landfill sites for the sediments.

Reference studies
Several reference studies were carried out in 1991 to analyze reference material, as a basis for future assessments of possible environmental disturbances in connection with the cleanup of Järnsjön. Furthermore, the studies provided a basis for the design of monitoring programs to be used during the cleanup, as we well as subsequent follow-up studies after the cleanup. The studies were primarily aimed at gathering background material for an assessment of the transport of suspended matter and PCBs in various waterflow situations. Furthermore, a uniform collection of fish and bottom fauna was carried out.

Studies of PCBs in water and fish have been carried out by the Department of Ecology at the University of Lund ⁽¹⁵⁾. The results showed that the PCB-contaminated sediments in Järnsjön govern the transport of PCBs in the Emån water system and uptake in fish.

Fish physiology studies have been carried out by the Department of Zoophysiology at the University of Gothenburg in cooperation with the Swedish Environmental Protection Agency, Section for Aquatic Toxicology ⁽¹⁶⁾. The studies focused on the liver's detoxification system, enzyme activity, liver and gonad size, etc. Cage tests on rainbow trout showed that exposure to the water in Järnsjön resulted in an induction of liver EROD (ethoxyresorufin-O-deethylase) activity and morphological damage such as skin lesions and degeneration of ventral fins. These changes can be caused by PCBs. In parallel with the fish physiology studies, the Department of Pathology at the University of Agricultural Sciences, Uppsala ⁽¹⁶⁾ has carried out histopathological studies of gills and liver as well as enzyme analyses in plasma.

Collection of suspended matter with sediment traps, analyses, and toxicity tests on this material were carried out by the Department of Zoology, University of Stockholm, and the Department of Environmental Toxicology, University of Uppsala ⁽¹⁷⁾.

The acute toxicity and sublethal effects included by bottom sediment were studied by the Swedish Environmental Research Institute, IVL ⁽¹⁸⁾. A study was carried out by the Swedish Environmental Protection Agency to determine the levels of polychlorinated dibenzo-p-dioxins and furans (PCDD/PCDF), Coplanar polychlorinated biphenyls (pPCB) and polychlorinated naphtalenes (PCN) in sediment, pike and water from different sites along the Emån ⁽¹⁹⁾. Bottom fauna studies were carried out by the Swedish Environmental Protection Agency ⁽²⁰⁾. A summation of the Emån's physical and chemical status has been compiled for the period 1987 to 1991 ⁽²¹⁾. Physical-chemical water surveys in the Emån upstream and downstream of Järnsjön were carried out during the summer and autumn of 1991 by the municipality of Vetlanda ⁽²²⁾. The average pH during the survey period was 7.2, alkalinity 0.42 meq L-1, color 50 mg Pt L-1, conductivity 13-15 mS m-1 and turbidity 1-2 FNU. Documentation of the vegetation in and around Järnsjön has been carried out by the University of Kalmar ⁽²³⁾.

REMEDIATION: PRINCIPLES AND CONSIDERATIONS

Dredging

Dredging was to be carried out with a suction dredger with an auger head designed specially by the Mud Cat Division of Ellicott International to reduce the suspension of material. Not more than 1% of the dredged material was allowed to be spilled. The dredging was to be carried out in the protection of geotextile screens (Fig. 9), in high contaminated areas i.e. the eastern part of the lake (Fig. 6). The protective screen was used to separate this area from the main river flow after the 1993 spring flood. Preliminary dredging was carried out in order to find out how large quantities of material were spilled and the subsequent PCB levels within the screen. In this way, a preliminary assessment could also be made of the consequences of dredging the western part of the lake without a protective screen, at the same time as the trial dredging provided an opportunity to calculate the residual quantities that may have to be post dredged within the screened-off eastern area.

Figure 9. Geotextile screen used to protect against PCB-contaminated stirred-up sediments while dredging the eastern part of Lake Järnsjön.

It was concluded that the western area of the lake could be cleaned up without a screen, the work was planned to be done in this way. The screen should then be left around the eastern area to allow sedimentation of suspended matter and for post-dredging. The screen should be left in place until late in the autumn, further reducing the risk that the PCB concentrations in the water reached high levels. It was to be opened at higher discharges.

Considerations and background studies before dredging
Before the decision was taken to suction-dredge under the protection of geotextile screens, other alternatives were considered. One possibility would have been to divert Emån past Järnsjön and either drain the lake and dig out the sediments or dredge in a completely isolated basin. The Emån may reach very high discharges, up to 100 m3/sec at Järnsjön. If a diversion canal had not been sized for high flow rates, it would have been necessary for the cleanup to be finished before the high flows came. If, on the other hand, the diversion canal had been sized for the large flows, the costs would have been high due, e.g. to the need for erosion protection. The cost of such a canal would have amounted to more than USD 1.3 mill. This estimate was based on field studies. Draining the lake was also judged to be difficult, since it is located in an area with glaciofluvial deposits, which permit high mobility of the groundwater. It would have been necessary to pump out part of the water from the lake. This water would probably have been contaminated by the sediments in the lake. It would, therefore, have been necessary to decontaminate this water.

Another alternative that was considered was covering the bottom of the lake with uncontaminated material. An advantage of this alternative was that there would no longer be any need for a landfill site. Providing that the cover material is not affected, a covered repository would have been created in the lake. The disadvantages with this alternative was a lack of experience of its use; the pollutants would remain in the watercourse; it is uncertain whether an effective layer of material could be deposited on the porous substrate constituted by the sediments. The water discharge through the lake is sometimes strong, which could affect the protective layer. A protective cover would make the already shallow lake (mean depth 1.5 m) even shallower. Moreover, if the covering alternative should fail, it would be considerably more difficult to suction dredge the lake in the future.

As far as the main alternative is concerned, studies have been conducted of current conditions in the lake, as well as what types of protective screens should be used and how they should be positioned. Model tests have been conducted at the Royal Institute of Technology, Stockholm. The current conditions at different discharges and water levels have been examined. Several positionings of the protective screens have been studied, also under different conditions of discharge and water level. Particular interest has been devoted to the effects of screens positioned to screen off the eastern or the western side of the lake. The model provided information on where the currents will flow and how strong they will be. At moderate discharges, about 10 m³ s^-1, which is the highest discharge at which dredging was intended to be conducted, no strong currents arise in the lake even if half the lake is closed off by screening.

Information obtained form the model was not sufficient to clarify the effects on erosion and sedimentation in the lake. Information about the bottom conditions and the sediment properties was also needed. Within the framework of a special study, sediment samples have been taken and studies with respect to particle size distribution, structure, density, ratio of organic/inorganic matter and other factors of importance for determining how the planned cleanup measures might affect sedimentation and erosion. An important question was to determine whether so much material might be released and transported out form the lake that the concentration of PCBs in the water could become more than twice as high as during normal conditions. The reference data needed have been provided b y the County Administration in Kalmar, on the basis PCB measurements performed in 1987 and 1991. These measurements had been carried out during about the same periods of the year as those during which the cleanup was planned to take place. A limit was set at 70 mg of PCBs/L in the lake water, not to be exceeded during six months of remediation.

Dewatering
Dewatering can be done mechanically or by natural drainage. The Järnsjön sediment was to be dewatered mechanically. With different types of equipment, the material should be given a dry solids (DS) content of ca. 35%, which was desirable to make the material manageable. The DS content is the first of two requirements that were made on the dewatering; the second concerns the quality of the water to be returned to the lake. The water returned to the lake may not contain more than 50 mg L^-1of suspended matter. This means that about 2 kg of PCBs may be returned to the lake with the return water. Together with the PCBs that may be released during dredging, about 2 kg, if it is assumed that the amount of PCBs released from the post-dredged eastern area is very small, the PCB released due to the cleanup work will be of the same size as the naturally occurring release of PCBs during the same period of time. This means that the total amount of PCBs released during the cleanup period may be doubled.

Disposal
The dewatered material (35% DS) was to be deposited on a site west of Järnsjön. The landfill was estimated to require about 5 ha of land. The landfill was to be established with its lowest bottom level well above the high-water level in Järnsjön. It would reach about 6.5 m above the level of the lake.

The main principle was to keep the water away from the landfill so that the throughflow of water acting as a transport medium would be small. The material was to be distributed so that the most contaminated spoils were situated in a special part of the landfill to be accessible if economically feasible methods for breaking down PCBs are developed at a later date. The least contaminated material was to be situated on a higher level within the landfill. The landfill was then to be built up with a sealing and cover layer so that infiltration of precipitation water would be low. The entire landfill was to be covered with 1.2 m of uncontaminated material from the site in order to further reduce emissions. After this the site was to be sown and otherwise restored as pastureland.

Considerations and background studies on disposal
Before a decision was taken to dispose of the sediments from Järnsjön by landfilling, the possibilities of using methods where the PCBs are broken down or separated were explored. This work was included in the study of possible cleanup methods. The work mainly entailed a review of the literature, plus a subsequent study trip to the US-Environmental Protection Agency. The conclusion that could be drawn after these studies was that while methods do exist for the final destruction of PCBs, for example by incineration, all of these methods were too costly to be used on the large volume of dredging spoils in question here. There would also have been technical and practical problems due to the large a mount of sediments to be handled.

Based on the fact that PCBs absorb organic particles, a property which can be exploited in disposing of sediments in an environmentally acceptable manner, a study was conducted of the material's leaching properties, etc. tests were made both on experimental fields and in the laboratory. The results indicated that around 1-10 g of PCBs will be leached from the landfill every year. Leaching from the landfill was estimated to be small, especially in comparison with the 3000-8000 g that was leached annually from the sediments directly into the water or Emån.

Project Implementation
Dredging was carried out with a Mud Cat™ dredger using an auger head specially designed to reduce the distribution of suspended particles (see picture below). Milman/Skanska operated the dredge. To control the operation, the dredger was equipped with an advanced positioning system and could work with high precision. In order to further reduce suspension of contaminated materials, disturbance by pipes, wires, service boats, anchoring, etc. was minimized by strict restrictions.

Mud Cat™ dredger uses an auger head specially designed to reduce the distribution of suspended particles

Dredging of the heavily-contaminated sediments within the enclosed area (eastern part of the lake) was carried out from May to November 1993. The total thickness of sediments to be dredged I this area varied from 0.4 to 1.6m and dredging was carried out in layers of 0.4m.

The western part of the lake was dredged in summer 1994. In this area, the sediments were of limited thickness and dredging could be limited to a single layer with a thickness of 0.4m. When dredging was completed, a contracted amount of about 120,000 m³ had been dredged. Including "overcut", the dredged amount was in reality close to 150,000 m³. According to the follow-up, 394 kg of PCB were cut, which constitutes about 97% of the estimated total amount of PCB prior to the dredging operation. About 80% of the dredged amount of PCB originated from the enclosed area in the eastern part of the lake containing the heavily contaminated material. In the areas included in the remediation operation, about 2.9 kg PCB was left. The major part of the remaining PCB (7.4 kg) was instead located at or near the lakeside, an area which was not included in the remediation process.

In the landfill, heavily contaminated sediments were separated from slightly contaminated sediments by a geotextile screen. The reason for this was to make it possible to locate the heavily contaminated sediments in the future, when remedial techniques have evolved and it may be feasible to carry out more definitive actions.

The topmost meter of the disposed sediments consists of the most fine-grained and least contaminated sediments from the lake. Calculations indicated that the hydraulic conductivity of these sediments, after consolidation for the pressure from the landfill cover, should be less than 5x10^-9 m s^-1. This is normally not low enough to serve as a sealing layer for landfills, but was regarded sufficient bearing in mind the calculated low leakage of PCB. Finally, the landfill is covered with a 1.2m protective cover of sand and gravel form the site. The total area occupied by the landfill is about 32,000 m³, the maximum height is 6.5m and the volume is about 90,000 m³. So far, the follow-up indicates that the landfill is behaving satisfactorily. No increase in PCB concentration has been measured in groundwater under the landfill.

ENVIRONMENTAL MONITORING

Monitoring

Remediation implies certain hazards for the environment and environmental monitoring is necessary to manage the operation. An accurate measurement system is also required so that the experience gained can be applied to future remediation operations. Environmental monitoring in the Järnsjön project can thus be divided into two parts. The base is the environmental monitoring of contract work including measurements for managing of the work and certifying that the claims given in the contract work and from the various authorities are fulfilled. The program for this part includes measurements for following turbidity, suspended solids and release of PCB to Emån and certain other measurements for monitoring the efficiency of the work. The extended monitoring includes measurements at more stations in Emån and measuring of other water quality parameters. Furthermore, measurements of PCB and metal content in groundwater and PCB content in the air during remediation are included.

The purpose of the reference stations at Flögen (a lake) and Melakvarn (running water) was to monitor background exposure and possible effects caused by this burden.

At Aspödammen (dam) and Kvillsfors (running water) stations, the aim was to monitor the exposure on Järnsjön and the effects this could give.

The purpose of the Järnsjön and Järnforsen (running water) stations was to monitor the emission of PCB from the lake sediments, to monitor the effects of protective measures and to direct the contract work so that guidance and limit values were maintained. In this way, basic data were obtained for assessment of the effects which the emission from the lake sediments could have on Järnsjön. The stations also aimed at monitoring the results of remediation.

The Blankan (dam) and Ammenas (running water) stations were used to monitor transport from Järnsjön and the burden and effects this caused on Emån.

The Grönskogssjön (lake) and Emsfors (running water) stations were used to monitor long-distance transport from Järnsjön and the burden this imposed on Emån, and the effects it resulted in. results from the station at Emsfors were also used to calculate the burden on the Baltic Sea from Emån.

Lake Järnsjön
The program for environmental monitoring of the contract work was developed according to contract conditions and permit requirements pursuant to environmental legislation.

During remediation the turbidity, suspended solids and PCB in Emån upstream from Järnsjön downstream were measured. The contribution of suspended solids and PCB from the discharge of the dewatering and water treatment plant was measured. The tributary Pauliströmsan discharges into Emån just downstream of Järnsjön, but upstream of the downstream station, which in turn is located beside a power station ca.1 km further downstream. The reason for the positioning of this station is to receive an efficient mix of water before sampling. The objective of the measurements in this station was to certify that the approved terms concerning impact on Emån during remediation were fulfilled.

During dredging inside the closed off area, the content of suspended solids and PCB in water was measured. These measurements were to provide a basis for deciding when to remove the silt curtain.

At these stations turbidity was measured automatically and registered continuously while samples for analysis of suspended solids and PCB were collected automatically and combined to form daily (suspended solids) and weekly (PCB) average samples. The content of total solids (TS) in dewatered sediments was measured in two samples each shift by the contractor. The aim of the measurements was to certify the geotechnical stability of the landfill.

The aim of measuring the PCB remaining in sediments after dredging was to certify that the total content of PCB in the remaining sediments was sufficiently low. The results of these analyses determined whether additional dredging should take place. Measurement of PCB contamination on the silt curtain after completed dredging in the closed-off eastern area was used to determine whether the curtain also could be used, if necessary, to close off the western part of the lake.

Extended Monitoring
Besides the environmental monitoring required for management of the operation, additional monitoring was performed in order to gain experience for future remediation. For this purpose, measurements of turbidity, suspended solids and total organic content in Emån were made at three stations further downstream of Järnsjön. At two more stations, samples were collected so that analyses of PCB could be performed if desired. Moreover, at some of the stations established at the watercourse, monitoring was supplemented with measurements of TOC, phosphorus and nitrogen.

The extended monitoring of the watercourse also included a separate program for following up environmental effects. This program involved: i)studies of PCB in water at two stations upstream of Järnsjön, at the outlet form the lake and at two stations downstream. Sampling was carried out at intervals of one week; ii)studies of PCB in fish (perch) at two stations upstream of Järnsjön, in the lake and at one station downstream. iii) chemical and biological characterization of suspended solids in Emån which involved analysis of PCB, PAH, and PCDD/PCDF at two stations upstream of Järnsjön, at the outlet from the late and at two stations downstream. Sampling was carried out at intervals of 10 weeks; iv)fish-physiology studies focused on the liver's detoxification system, enzyme activity, liver and gonad size and histopathological studies of gills and liver.

Besides analyses of environmental monitoring of the watercourse as described above, the extended program also includes measurement of PCB content in air during removal and landfilling of contaminated sediments and monitoring of groundwater quality near the disposal area. Measurement of PCB content in the air involved 11 stations located between 5 and 1000 m from the disposal area and one reference station located 12 km away. These stations were in place throughout the remediation process until June 1995. Samples for determination of PCB, TOC, ph, conductivity and trace-element contents in groundwater involved 8 groundwater monitoring wells close to the disposal area and 6 drinking-water wells in the vicinity. The program was operated until 1997 after which measuring intervals for the long term were decided by the authorities.

Results from the Extended Environmental Program

PCBs in sediments, water, fish and air
Remediation resulted in lowered PCB concentrations in sediments, decreasing from 5 ug/g d.w. to 0.06 ug/g d.w. (28). The concentration of PCBs in the river downstream of Järnsjön during remediation was in the same range as that monitored before remedial actions were carried out. After remediation, the concentration of PCBs in the lake decreased from 8.6 ng/L in 1991 to 2.7 ng/L in 1996. Remediation resulted in a lowering of the PCB concentration in 1-yr old fish hatched in the summer after cleanup (Fig. 12). PCB concentration in fish from Järnsjön decreased from 34 ug/g extractable fat in 1991 to 16 ug/g in 1996.

Figure 12. Histogram showing the situation before and after remediation of Lake Järnsjön. Values for water, fish (1-yr perch), SPM (settling particular matter) and EROD are normalized against values for Aspo 1991. Aspo is located upstream Järnsjön and Gronskogssjon is located downstream.

During handling of the sediments at the landfill, PCB concentrations in air were found to be elevated. In the air above the landfill under construction, PCB concentration was 2.5 ng/m and thus within the range found in urban areas. After the landfill had been closed, the PCB concentrations in the air were similar to the background levels (29).

Dioxin-like effects
Settling particular matter and sediments were collected in Emån before, during and after dredging of Järnsjön. The samples were tested for dioxin-like effects using a sensitive bioassay based on EROD-induction (ethoxyresorufin-O-deethylase activity) in cultured chicken embryo livers. The dioxin-like activity of the diaromatic fraction in sediment collected after dredging had only about 1% of the activity of the pre-dredging sediment, showing that remediation of the lake was successful. Compared with the levels in the 1993 samples of settling particulate matter, the levels of diaromatic dioxin-like compounds in the 1996 samples were substantially lower (30).

Fish physiology
Caging juvenile rainbow trout in Järnsjön prior to remediation resulted in induction of liver EROD-Activity, skin lesions and fin erosion. During remediation, pronounced induction of liver EROD, bile duct proliferation and a relatively high incidence of necrotic hepatocytes were seen in the rainbow trout caged in the lake. In 1996, two years after the remedial activities, EROD activity was still slightly induced in caged rainbow trout, but the fish showed no sign of histopathological changes. In addition, liver EROD activity was induced in some sites downstream of Järnsjön. The results indicate a downstream transport of contaminants following the remedial actions. (31).

Health of perch (Perca fluviatilis) was studied with physiological and morphological methods before and after removal of the PCB contaminated sediment (32). Before the remedial activity, perch from the PCB-polluted Järnsjön were morphologically characterized by gill and liver damage with small biochemical/physiological effects. Two years after the removal of PCB the perch in Järnsjön appeared healthy.

References:

Swedish Environmental Protection Agency, 1995. We're well on the way. Site Remediation Action Plan.
Johansson, A. 1996. The River Emån Catchment Basin. The River Emån project. Kalmar County Administration
Swedish Meteorological and Hydrological Institute 1993. SVAR. Catchment Areas in Sweden. (In Swedish).
The River Ema Water Board 1996. Bentic Fauna in River Ema 1996. Vetlanda Municipality. (Troedsson, B.). Ekologgruppen i Landskrona AB (Holmstrom C.) (In Swedish).
Wiederholm, T. and Johnson R.K. 1997. Monitoring and Assessment of Lakes and Watercourses in Sweden. Temporal Reference Rivers. Department of Environmental Assessment, Swedish University for Agricultural Sciences.
The River Ema Water Board. 1996 Recipient Control Annual report 1996. Vetlanda municipality. (Troedsson, B.). (In Swedish).
Von Post, H. and Wik, O. 1992. Remediation of Lake Järnsjön in River Ema Investigation of contaminated sediments. Swedish Environmental Protection Agency, Report 3998. (In Swedish. English summary).
Larsson, P., Okla, L., Ryding, S-O. and Westoo, B. 1990. Contaminated sediment as a source of PCBs in river system. Can. J. Fish. Aquat. Sci 47, 746-754.
Larsson, P., Ryding, S-O and Westoo, B. 1988. Contaminated sediments as source of PCBs in the River Ema. Environmental research Institute Report B 896. (In Swedish).
Enell, M ,and von Post, H. 1988. Feasibility study for clean-up of PCB in Lake Järnsjön. Environmental Research Institute Report B 930. (In Swedish).
Gullbring, P. 1991. Cleanup of Lake Järnsjön in Emån-Main Study. Swedish Environmental Protection Agency Report 3999. (In Swedish. English summary).
Bjorklund, I. 1991. Scenarios for different PCB-loadings situations in Emån during and after cleanup of Lake Järnsjön. Swedish Environmental Protection Agency Report 4000. (In Swedish. English summary).
Rogbeck, J. 1992 Clean-up of Lake Järnsjön in Emån. Methods for sediment cleanup in industrially polluted water areas. Swedish Environmental Protection Agency Report 4001. (In Swedish. English summary).
Elander, P. 1992 the Lake Järnsjön Project, Disposal of PCB -contaminated Sediments. SwedGeo Enviro-geotechnical report. Linkoping 31 March 1992 (In Swedish.)
Larsson, P., Okla., L. and Bremle, G. 1992 Clean-up of Lake Järnsjön in the River Ema PCB in water and fish. Swedish Environmental Protection Agency, Report 4069. (In Swedish. English summary).
Forlin, L., Balk, L., and Norrgren, L. 1992 Cleanup of Lake Järnsjön in the River Ema. Physiological and morphological studies on perch and rainbow trout. Swedish Environmental Protection Agency, Report 4070. (In Swedish. English summary).
Brunstrom, B., Engwall, M., Hjelm, K., Norrgren, L., Broman, D., Ishaq, R., Naf, C. and Zebuhr, Y. 1992. Remediation of Lake Järnsjön on the river Ema. Extracts from sediment and seston-Toxic potency and PCB, PCDD/F and PAH concentrations. Swedish Environmental Protection Agency, Report 4074. (In Swedish. English summary).
Thuren, A. and Svensson, A. 1993. Remediation of Lake Järnsjön in the River Ema. Toxicity in porewater and sediments. Swedish Environmental Protection Agency, Report 4075. (In Swedish. English summary).
De Wit, C. 1992. Clean-up of Lake Järnsjön in the river Ema. Analysis of polychlorinated dibenzodioxins and dioxinlike substance. Swedish Environmental Protection Agency Report 4073. (In Swedish. English summary.)
Ekstrom, C. 1993. Clean-up of Lake Järnsjön n the River Ema. Investigation of bottomfauna 1992. Swedish Environmental Protection Agency, Report 4166. (In Swedish. English summary).
Ljungberg, M. 1992. Cleanup of Lake Järnsjön in the River Ema. Water quality and transport in the river Ema and tributary streams 1987-1991. Swedish Environmental Protection Agency, Report 4071 (In Swedish. English summary).
Troedsson, B. 1992 Cleanup of Lake Järnsjön in the river Ema. Physical-chemical water survey 1991. Swedish Environmental Protection Agency, Report 4068. (In Swedish. English summary).
Ljungberg, M. and Danielsson, J. 1991. Clean-up of Lake Järnsjön in the River Ema. Investigation of vegetation in Lake Järnsjön. Swedish Environmental Protection Agency, Report 4072. (In Swedish.)
Berg., H. and Kullberg, S. 1992. Clean-up of Lake Järnsjön in the River Ema. Current conditions in Lake Järnsjön. Swedish Environmental Protection Agency, Report 4164. (In Swedish. English summary.).
Nilsson, B. 1992. Clean-up of Lake Järnsjön in the River Ema. Assessment of erosion sensitivity and sedimentation properties of the bottom sediments in Lake Järnsjön. Swedish Environmental Protection Agency, Report 4115. (In Swedish. English summary).
Elander, P. and Hammar, T. 1998. The remediation of Lake Järnsjön, project implementation Ambio 26, 396-398.Helgee, A.,Troedsson, B., Elander, P. and Hammar, T. 1997. Cleanup of Lake Järnsjön in Emån. Monitoring programme. Swedish Environmental Protection Agency, Report 4746. (In Swedish. English summary.)
Bremle, G., Okla, L. and Larson, P. 1998. PCB in water of a lake after remediation of contaminated sediment. Ambio 27, -398-403.
Bremle, G. and Larsson P. 1998. PCB in Emån river ecosystem. Ambio 27, 384-392.Engwall, M., Naf, C. and Brunstrom, B. 1998. Biological and chemical determination of contaminant levels in settling particulate matter and sediments collected in a Swedish river system before, during and after dredging of PCB-contaminated lake sediments. Ambio 27, 403--410.
Blom, S., Norrgren, L. and Forlin, L. 1998. Sublethal effects in caged rainbow trout during remedial activities in Lake Järnsjön, Ambio 27, 411-418.
Forlin, F. and Norrgren, L. 1998 Physiological and morphological studies of feral perch before and after remediation of a PCB contaminated lake- Lake Järnsjön. Ambio 27, 418-424.
Clean-up of Lake Järnsjön in River Ema. 1995. Video in English.

Source: AMBIO: A Journal of the Human Environment. Vol. XXVII Nbr 5, August 1998.
Authors: Per Gullbring, Tommy Hammar, Anders Helgee, Bo Troedsson, Kjell Hansson and Fredrik Hansson.