Constructed Wetlands for Treating Water Containing Metals: An ARCO/Montana Tech Collaborative Project

William J. Drury, Ph.D., P.E.
Department of Environmental Engineering

The Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA, also known as Superfund) requires the remediation and restoration of improperly disposed historical wastes that present a significant human health hazard or have greatly degraded the environment. According to CERCLA regulations, 100 years of mine waste disposal practices in the Butte and Anaconda area have created a significant problem. The federal government and the state of Montana resolved to remedy the problems created by these mine wastes. Atlantic Richfield Company (ARCO) is a principal party responsible for the remediation of these problems.

ARCO is responsible for treating water containing regulated metals (copper, cadmium, zinc, and iron) at elevated concentrations. The company has been actively seeking to identify a water-treatment technology that would be inexpensive to operate over a long period. Current technologies, such as metal precipitation, require continual supplies of the pH adjustment chemical (usually lime) as well as a repository for disposal of the chemical sludge removed from the treated water. Water treatment to remedy these problems is likely to be required for many decades. Thus, at ARCO’s request, Montana Tech began its investigation of wetland technologies as an alternative method of treating the metal-laden water.

The ARCO-Montana Tech collaborative effort began in 1993 with laboratory and greenhouse experiments. These initial experiments showed that wetland technologies could treat the water effectively. In 1995-1996, ARCO built three test sites for field-scale wetlands. The first one, Wetlands Demonstration Project 1 (WDP1), was built at the corner of Kaw Avenue and George Street in Butte. Two other test sites, the Butte Reduction Works Demonstration Site (BRW) and the Colorado Tailings Demonstration Site (CT), were built south of Centennial Avenue near the Butte Metro Sewage Treatment Plant. Montana Tech then conducted field studies of these wetlands.

Wetlands Demonstration Project 1

The constructed wetlands tested at WDP1 (See Figure 1) work through the degradation of organic matter, which reduces sulfate to sulfide. Sulfide is an excellent precipitant of copper, cadmium, and zinc. At low concentrations, it forms metal sulfide solids that can be easily removed from the water, leaving a low metal concentration in the water. Metal sulfide solids are relatively dense and form less sludge than the solids produced by lime precipitation. Therefore, the sludge disposal costs for wetlands should be significantly less than for conventional lime precipitation. The organic matter required for sulfate reduction comes from placing a biodegradable organic solid such as compost in the wetland during construction. Cattails were planted in some of the wetlands; natural decay of the plants provided additional organic matter.

Three schemes of subsurface flow (SSF) wetlands for sulfide generation were tested. The first scheme consisted of two SSF wetlands built with substrates of gravel and planted with cattails. A second scheme was built with a substrate of 80% gravel and 20% Eko-compost (an organic material from wood wastes and sewage sludge). This scheme also was planted with cattails. The third scheme, with a substrate of 50% gravel and 50% Eko-compost, had no plants and was covered with polystyrene with a black tarp for insulation in winter.

Conversion of sulfate to sulfide can only occur in an anaerobic environment; therefore, the water must flow underground in SSF wetlands. Effluent water from the SSF wetlands flowed into surface flow (SF) wetlands for oxidation of iron and sulfide.

Gravel Subsurface Flow Wetlands: Cells One and Four

The larger of the two wetlands in the first scheme treated water effectively in summer, but the smaller wetland usually did not. It was apparently too small and did not retain the water long enough to produce enough sulfide to precipitate zinc.

The effectiveness of the wetlands in winter was a problem. The wetland that worked well in summer removed less zinc in the winter. Because of weather conditions, the process worked better in the winter of 1996-1997 than in winter 1997-1998, when frost penetrated 75% of its depth. With only 25% of the wetland volume free flowing, the water was not held for a long enough period to be treated. Winter temperatures also slowed the sulfide production rate.

Compost Subsurface Flow Wetlands:
Cells Two And Three

Deeper wetlands were not affected as much in winter because a smaller percentage of their volumes were penetrated by frost, and the free-flowing volume remained warmer because it was heated by the unfrozen subsurface soils.

While compost increased the sulfide production rate, it had a negative effect on the permeability of the substrates. Water must be able to pass through SSF wetlands rapidly; otherwise, the wetlands would have to be impracticably large to keep the water underground.

Compost particles are small and therefore make the permeability of the substrate small. The wetland with 50% compost, despite its ability to produce water of good chemical quality, was unable to show a measurable flow rate after 2.5 years of operation. The permeability of the wetland containing 20% compost was about one tenth that of the wetlands containing no compost.

All SSF wetlands were subject to clogging of inlet zones. This clogging may be caused by filtration of iron colloids that did not settle in the sedimentation pond preceding the SSF wetlands.

Surface Flow Wetlands: Cells Six and Seven

Iron was removed from the water effectively in the SF wetlands. Because iron precipitates best in an oxygen-filled environment, the iron that dissolved in the SSF wetlands quickly precipitated in the aerobic SF wetlands that followed the SSF wetlands.

Conclusions

Data indicate that SSF wetland technology may be feasible. Greater depths of gravel substrate wetlands, however, should be tested. Further research is planned by Montana Tech to develop a way of maintaining the permeability level at the inlet zone as well as in the rest of the wetland.

Butte Reduction Works And Colorado Tailings
Demonstration Wetlands

The BRW and the CT wetlands consisted of three ponds open to the atmosphere and separated with permeable treatment walls consisting of cobblestone and Eko-compost. The BRW wetlands had no chemical addition and did not create conditions that precipitate metals from water. Therefore, they did not produce water of acceptable quality. The CT wetlands had lime added to the influent water to raise the pH and improve metal precipitation. The CT wetlands did produce water of acceptable chemical quality as long as the lime-addition system worked properly. Because of construction activity, electric power could not be supplied to the CT wetlands early on. Therefore, a water-driven liming system, which experienced frequent failures, was used. In April of 1998, an electrically driven liming system was installed and worked successfully.

Field-scale experiments—the Wetlands Demonstration Project 1, the Butte Reduction Works Demonstration Site, and the Colorado Tailings Demonstration Site—investigated the use of constructed wetlands for treatment of water containing heavy-metal concentrations. Some wetland schemes produced promising results, especially the CT wetlands. Problems included a lack of wetland permeability, effects of winter temperatures, and the malfunctions of the water-driven liming system. ARCO continues to explore the results of these experiments.