PRELIMINARY REVIEW:

Wetland treatment concepts in EPA/A.D. LITTLE draft Engineering Evaluation/Cost Analysis (EE/CA) Report on the Elizabeth Mine

PROPOSED CONCEPTS:

The tailings at the Elizabeth Mine are discharging acidic and metal laden waters to the detriment of the local streams and the general environment. The proposed treatments are intended to remove metals to acceptable levels and to neutralize the acidity in the wastewater.

The tailing piles are designated as TP-1 and TP-3 in this review (TP-2 is combined with TP-1 for the proposed treatments). It is proposed to cap TP-1 to prevent future infiltration and to divert surface runoff and shallow groundwater. There will be a decreasing amount of seepage from TP-1 after the capping is complete; it is expected to reach a low, steady-state rate after a few years. The proposed treatment sequence for this TP-1 seepage includes a holding pond followed by Successive Alkalinity Producing Systems (SAPs), followed by a constructed aerobic wetland. These treatment components will be described and discussed below. For historic reasons, it is proposed to leave some or all of TP-3 as it now exists, without capping or removal of the acidic materials. Adjacent surface runoff will be diverted around TP-3, but the treatment system will have to deal long term with the full seepage volume caused by infiltration of precipitation and possibly some groundwater. The same treatment sequence described previously is also proposed for TP-3.

CONCEPT DESCRIPTIONS & DISCUSSION

GENERAL: The proposed treatment system is termed a "passive" system. It is a passive system in the sense that pumps and motors and chemical additions are not involved but this does not mean that maintenance is not required. These maintenance activities will include routine monitoring, very, very infrequent removal of the sediments in the holding ponds, very infrequent removal of sediments from the wetland, infrequent removal of sediments from the SAPs, and possibly at least an annual flushing of the SAPs. The frequency of sediment removal will depend on the flow rate and concentration of materials entering the treatment units. The flushing of the SAPs is required to flush out the accumulated precipitates which tend to clog the media in the SAP and thereby reduce the permeability of the filtering layers in the SAP. In a recent paper, Brenner (1) has estimated that the SAP might have to be flushed at six to eight month intervals. In addition to these cleaning activities the limestone and compost filtering media in the SAPs will have to be periodically removed and replaced. The EE/CA report estimates a 12 to 15 year replacement cycle for this media. These maintenance activities will be the responsibility of the State of Vermont.

It should be remembered that the metals removed in this system do not disappear, they are retained and accumulate in the system components. It is unlikely that they will reach toxic levels for the plants or wildlife, but the associated accumulation of sediments can in time interfere with system performance, so flushing, cleaning and eventually removals will be required. At TP-1, with an effective cap over the tailings, the quantity of seepage will be significantly reduced and this will serve to significantly increase the useful life of the system. At TP-3, because seepage is not controlled the maintenance requirements for this treatment system may persist for generations. The EE/CA presents cost estimates for maintenance based on a 30 year present worth basis. That may be adequate for the system at TP-1, but is likely to be inadequate for TP-3 where the need for continued maintenance may extend beyond 30 years.

HOLDING POND: This lined pond is intended to temporarily collect the seepage from the tailing piles. Most of the suspended solids entering with the seepage are intended to settle out in this pond. The pond is also intended to stabilize the flow rate so that water flows to the next components at a relatively steady rate. It is likely that such a condition can be achieved at TP-1 after the capping is complete but it is unclear how that will be possible at TP-3 with future infiltration and seepage uncontrolled. The pond would be six feet deep, providing adequate storage for long term accumulation of sediments. Winter ice will certainly form on this pond. Depending on the snow cover, typical ice depths in the Upper Valley range from 20 to 30 inches. In a very cold winter with little snow the ice depth could exceed three feet. This ice and other winter conditions are unlikely to interfere with the function of the holding pond.

The size of the holding pond has been determined with conventional criteria for runoff and seepage and appears reasonable.

SUCCESSIVE ALKALINITY PRODUCING SYSTEM (SAP): This component is also a lined pond, but designed for vertical flow through the media placed on the bottom of the pond. These SAPs are also called vertical flow ponds (VFP) in Pennsylvania because of this vertical flow path. The pond is lined to prevent loss of partially treated water to the groundwater. The bottom layer of media is typically limestone gravel, a three foot thick layer is proposed for this project. On top of the limestone is a layer of compost, a two foot thick layer is proposed. The design depth of water over the compost is three feet. Underdrain pipes in the limestone layer collect the treated water for discharge to the next treatment component. Some of the metals are removed, through a variety of reactions as the water passes down through the compost and some alkalinity is also added. As the water moves through the bottom layer the limestone is very gradually dissolved producing significantly more alkalinity in the water. This additional alkalinity is needed to neutralize the acidity in the water and raise the pH to acceptable levels. The seepage is already acidic coming out of the tailings, but the acidity can increase further as some forms of dissolved iron oxidize and precipitate.

Wetlands have been used for almost 30 years for treatment of acid mine drainage, but the need for additional alkalinity was recognized and some early systems actually added chemicals for this purpose, at great expense. The need for lower cost "passive" systems was recognized and the use of limestone as a source of alkalinity was recognized. In the early systems the limestone bed was exposed to the atmosphere and was "aerobic." As the water passed through the bed the iron and other metals would precipitate and form a coating directly on the limestone. This coating would serve to "armor" the limestone so it would no longer dissolve and provide a source of alkalinity. It was discovered that if the limestone were kept in an oxygen free ("anoxic") environment the alkalinity would be dissolved and the iron would not precipitate until the water were again in an "aerobic" environment. The first anoxic limestone system used crushed rock in a buried trench, the water passed through one sidewall of the trench and came out the other. This was called an anoxic limestone drain (ALD). The water from the ALD would then go to an aerobic wetland where the iron could precipitate and settle out, the pH would still be acceptable because of the previous dissolution of alkalinity from the limestone. The use of an ALD is discussed in the EE/CA report (page 3-7) for use at TP-1, but the drawings all show only SAPs.

Early efforts in Colorado experimented with the use of compost for the treatment of acid mine drainage from metal mines. Typically, spent mushroom compost has been used. This is compost made from animal manures that is used for mushroom cultivation. After several cycles of mushroom production it is necessary to replace the compost and it is this "spent" material that has been used. Mushroom cultivation is a major industry in Pennsylvania so many SAP systems in that State use this compost material. The source of the compost for this project is not specified in the EE/CA report.

The SAP concept combines the contributions from compost and limestone in the same system with vertical flow to obtain maximum possible water flow through the media. The EE/CA report cites two SAP systems in Pennsylvania as evidence of successful performance. It is necessary to remove and replace both the compost and the limestone in these systems and the EE/CA report estimates a frequency of 12 to 15 years based on the projects at Howe Bridge, PA and Oven Run, PA. It is unclear how these estimates were established since Kepler and McCleary (2) first proposed the SAP concept in 1994. A quick Internet search for systems in Pennsylvania did reveal numerous systems including one at Oven Run where construction started in July 1995. It is unclear therefore how the estimates of a 12 to 15 year life for the media were established. The SAP is a developing technology rather than a well established technology. The Internet search revealed a large number of systems in Pennsylvania using the concept. All systems used a combination of compost and limestone, but the proportions varied considerably and one system used a mixture of the two in a single layer. It can be concluded that the SAP can function as intended for partial metals removal and alkalinity production, but the long term reliability and maintenance requirements are less well established.

There are two SAPs, constructed in series, shown on the drawings for both TP-1 and TP-3. It would be prudent to add suitable piping to allow parallel operation also to facilitate maintenance operations. Winter conditions could have a significant effect on performance of these systems. A three foot ice layer would sit on top of, and possibly compress, the compost layer. The most significant impact would be in the spring if the SAP ice did not thaw in time to receive the snow melt seepage from the tailing piles.

The proposed SAPs have been sized using criteria derived from their use elsewhere. These criteria seem to be reasonable and the SAP appears to be capable of performing itÕs intended function. However, there needs to be additional consideration of the long term reliability issues and the maintenance requirements.

AEROBIC WETLAND: The wetland is the final treatment component in the proposed treatment system. It is termed a "polishing" step by the EE/CA report. However, the report also assumes that only 50 percent of the iron and manganese will be removed in the SAPs, so the wetlands still have a significant treatment job to do. These wetlands are similar in concept and components to a natural marsh with emergent vegetation (ie: cattails, etc). The wetland is also lined to prevent contamination of groundwater. A layer of soil on top of the liner supports the root structure for the emergent plants. The plants provide minimal direct treatment benefits. Metals are taken up by the plants but are returned to the sediments in the fall of the year. The plants are, however an essential component in the system. They provide: shade for algae control, temperature modulation, wind reduction, and the annual die back provides additional organic materials for metals retention. Cattails typically provide more organic material than many other emergent plant species.

The major metals removal pathways in these wetlands are chemical and biochemical. The near surface water in the wetland is exposed to the atmosphere and is aerobic (ie: contains dissolved oxygen). Dissolved iron can oxidize, precipitate and then settle out. The lower water depths, even in a two foot deep wetland, lack oxygen and are anoxic (ie: no/or very little dissolved oxygen). The conditions here replicate those in the organic layer in the SAP. As a result additional dissolved metals can be removed and retained in these anoxic sediments. These sediments accumulate with time and eventually must be removed. Experience with wastewater treatment wetlands suggest that average sediment depth increases about one millimeter per year (1 inch every 25 years). The EE/CA report estimates wetland clean-out at 12 to 15 years. That may be a reasonable estimate for TP-3 where seepage volume will not be reduced in time. At TP-1, a reasonable estimate might require cleanout at 40 to 50 years. Most of the accumulation will occur near the front end of the wetland, so cleanout may not be required for the entire unit.

Winter conditions may severely impact on the function of these wetlands. With only a two foot water depth and a low flow and relatively long detention time they may freeze to the bottom except in the mildest winters. If groundwater is a significant factor in seepage from the tailings the impacts could be serious. The ability of the wetland to function during the spring snow melt period would also be at question if the ice does not melt quickly. During the winter months, with a continuous ice cover on the wetland critical treatment functions will be lost. Low temperatures inhibit biological activity, but the most serious impact will be the lack of an aerobic zone in the water beneath the ice. This lack of oxygen will limit the oxidation reactions needed for iron removal. As a result, the systems may not achieve their water quality goals during these ice covered periods. This may not be a serious problem depending on the long term sources of the tailing pile seepage. Infiltration of precipitation will be minimal after TP-1 is capped. Infiltration of precipitation on TP-3 would also be minimal as long as freezing weather persists. However, if long term TP seepage comes primarily from groundwater sources there may be a continuous year-round discharge and the wetlands may not meet effluent goals during the ice covered period.

The area of the proposed wetlands has been determined in the EE/CA report with criteria derived from experience with acid mine drainage elsewhere and appears to be reasonable. A very large safety factor has also been adopted, that is a conservative approach at this stage, but may not be necessary for the final design. The wetlands are shown on the EE/CA drawings a single unit for each of the tailings piles. That is acceptable at this stage, but consideration should be given in the final design to parallel wetland units at each location to allow maintenance activities. Consideration might also be given to operating the wetlands with a three foot water depth in the winter months to compensate for ice formation. This would require an appropriate outlet device and adjustment in the late fall and then restoration of the two foot depth in late spring. These activities would add to the maintenance responsibilities by the State of Vermont.

PERFORMANCE EXPECTATIONS

It is not possible to conduct a detailed analysis with the data provided in the EE/CA draft report.

The treatment systems must produce effluents satisfying both federal and Vermont water quality standards. However, the Vermont Standards (VTWQS) are not presented in the report. Table 1-10 in the draft report presents what appear to be the federal requirements for the wetland effluent.

These are given in terms of micro-grams per liter (µg/L). These values are:

• Aluminum <= 87 µg/L
• Cobalt <= 3.00 µg/L
• Copper <= 5.00 µg/L
• Iron <= 1000.00 µg/L
• Manganese <= 80.00 µg/L
• Zinc <= 120 µg/L

There are other metals in the seepage; the set presented above are termed the "Primary Contaminants of Concern" (COC) in the EE/CA draft report. That appears to be a reasonable approach at this stage. If the proposed treatment system can remove the COCs to the desired levels it is likely the concentrations of the other metals will be acceptable also.

Having established the federal discharge requirements it is then difficult to determine the effectiveness of the various treatment components because the draft report does not give actual data on the concentrations in the tailing piles seepage. The report presents "Hazard Quotients" instead. There is one tabulation, in an Appendix, of 1998 data collected by the State of Vermont at various locations on the site. The most applicable set, for the TP seepage, seems to be:

• Aluminum 13,000 µg/L
• Cobalt 120 µg/L
• Copper 1,200 µg/L
• Iron 380,000 µg/L
• Manganese 3,700 µg/L
• Zinc 740 µg/L

The report also assumes a 50 percent removal of iron and manganese in the SAPs . It then bases the wetland design for TP-1 on an iron concentration of 231,000 µg/L, and a manganese concentration of 2000 µg/L, with a design flow of 7.5 gallons per minute. At TP-3, the wetland influent is assumed to contain 44,000 µg/L of iron and 2000 µg/L of manganese, with a design flow of 40 gallons per minute. Based on these values, the assumed tailing pile seepage would contain 462,000 µg/L of iron and 4000 µg/L of manganese at TP-1 and 88,000 µg/L of iron and 4000 µg/L of manganese at TP-3. These values are consistent with the Vermont data presented above. There are significant differences in flow rate and iron concentration at the two tailing piles. The resulting mass loadings (from the untreated seepage) on the two systems would be: TP-1 42 pounds per day of iron and 0.4 lb/d of manganese; TP-3 42 lb/d of iron and 2 lb/d of manganese.

Based on experience elsewhere the assumption of 50 percent removal in the SAP seems reasonable. The SAP data reviewed on the Internet from systems in Pennsylvania indicated almost complete aluminum removal for all of the systems listed. The SAP concept has the theoretical capability to do even better than 50 percent removal. The issues of concern are with the long term reliability and maintenance requirements. The aerobic wetlands proposed have also provided excellent service for metals removal in many locations, and for many years. These wetlands are more effective for iron removal than for manganese or aluminum (3,4). It is possible that the wetland at TP-3 may not always meet discharge requirements, depending on the input concentrations. As discussed previously, the winter performance under a continuous ice cover may also be poor for both wetlands. To avoid these complications it is suggested that the State of Vermont and the U.S. EPA agree on annual average discharge limitations for the combined wetland discharges to compensate for seasonal fluctuations.

CONCLUSIONS

On a conceptual level the proposed treatment sequence is believed to be capable of meeting discharge requirements, except for winter periods with a continuous ice cover on the wetland. Further refinements will be necessary during the final design.

There will be seasonal variations in wetland effluent quality. The best performance will be obtained during the warm summer months. Complications could be avoided in the State of Vermont and the U.S. EPA could adopt annual average water quality limits for the combined wetland effluent flow.

The long term reliability and maintenance requirements for the SAPs need further definition. Major concerns are the frequency of flushing, the source of water for flushing and its disposal, and the replacement frequency for the compost and limestone media.

The EE/CA report would be more useful if it presented actual seepage concentrations and the Vermont Water Quality Standards.

It would be prudent to design the wetland as two parallel cells and to provide for water level adjustment. A winter water depth of three feet would compensate for some of the expected ice formation.

It would be prudent to design the SAPs with the capability to operate with parallel cells so maintenance activities will not be disruptive.

Maintaining all of TP-3 in itÕs existing condition increases the risk of long term water quality problems and long term maintenance problems. Removal of some of the more acidic materials should allow a reduction in the size of the treatment system and significantly decrease the risk of long term problems.

Selection of the most cost effective alternative is not possible at this time. The costs presented for the various options are all within the typical margin of error for estimates of this type. A "rule-of-thumb" for passive projects of this type is to spend whatever it takes on capital and construction costs to minimize the operation and maintenance tasks and costs.

The cost analysis presents the present worth of the operation and maintenance costs for a 30 year period. That may be acceptable for the TP-1 treatment system, but is likely to be unacceptable for TP-3. If TP-3 is retained in itÕs present state the treatment maintenance needs are likely to persist for generations.

REFERENCES

1.Brenner, F.J., 2000. Use of Constructed Wetlands for Acid mine Drainage Abatement and Stream Restoration, in: Proceedings, Wetland Systems for Water Pollution Control, IAWQ 7th International conference, November 2000.

2. Kepler, D.A., E.C. McCleary, 1994. Successive alkalinity producing systems (SAPs) for the treatment of acidic mine drainage, in: Proceedings International Land Reclamation and Mine Drainage Conference, and Third International Conference on Abatement of Acid Drainage.

3. Reed, S.C., R.W. Crites, E.J. Middlebrooks,1995. Natural Systems for Waste Management and Treatment, McGraw Hill, New York, NY.

4. Kadlec. R. H., R.L. Knight 1996. Treatment Wetlands, Lewis Publishers, Boca Raton, FL.

Sherwood C. Reed, P.E.
Principal, E.E.C.

top