March 22, 2002 Report # 6 - Review of Final EE/CA Report - Elizabeth Mine

Introduction

I attended the meeting on March 13, 2002, and have since reviewed the Final EE/CA Report documents presented to the CAG. My review focused on EPA's proposals for the "passive" treatment systems. The treatment systems and costs presented in the Final Report are the same as presented by the EPA at a meeting on January 23, 2002. My Report # 5 provided detailed comments on those January 23rd proposals and costs, and it can be found on the Elizabeth Mine Web site. In addition, to avoid repetition I have added that Report # 5 as an appendix to this report. This report will first present a summary and conclusions. At the end there are detailed technical comments on the report which may be of interest to only the EPA and AD Little. In addition, I have become concerned about misstatements of fact and unsupported assumptions which have been circulating recently in the community and a response to some of these is included also.

Summary & Conclusions

I have been involved as a Technical Advisor to the CAG since September 2001. I had no pre-conceived position or point-of-view when the work commenced and have tried to review the EPA materials in an objective, non-biased, professional manner.

When the work commenced, I was convinced, as a long term supporter of "natural systems" that it would be possible for the EPA to develop "passive" systems for the treatment of seepage from both TP3 and TP1/2. However, between September 2001 and November 2001 the EPA changed the proposed treatment systems three times. I had serious concerns with each iteration but still was confident a passive system could be feasible for TP3. In January 2002 the EPA presented another revision to the treatment systems, this time with a "Semi Passive" component for TP3 and a strong dependence on a unique algae growth in the wetland components. In my Report #5, I expressed serious concerns about the ability of these proposed systems to produce the desired water quality results. The EPA has now presented their Final EE/CA Report and the proposed treatment systems are the same as presented in January. I still have serious reservations regarding the capability of the systems, and the best the EPA can offer, for the TP3 treatment system is: "..... a reasonable possibility of success" (page 3-46 Final EE/CA Report). I don't believe a "reasonable possibility" of success is sufficient. When we consider the effort required and the amount of money the State of Vermont is being asked to spend on O&M a more positive level of assurance should be expected.

In view of these facts, I must reluctantly conclude that a reliable and cost effective passive treatment system is not possible for TP3, and it will be necessary to remove these tailing piles completely.

I believe the EPA and their expert consultants have done their very best to identify a passive treatment system for TP3. In the four iterations, so far, they have offered almost every combination of passive components available, but there is no positive assurance any of them will work, even after pilot testing. It should be noted that the EPA proposed treatment systems would be the same whether all of TP3 is preserved or up to 80 percent removed. A treatment system of some kind for TP3 is certainly possible, but it is likely to be much less passive, much more energy intensive and more mechanical with higher costs to build and much more intensive O&M and related costs.

I also still have serious concerns regarding the proposed treatment system for TP1/2, and believe a careful pilot study is absolutely necessary to prove the concept and then develop sizing criteria. This raises a related concern regarding the EPA timeline. Their proposed system includes emergent wetland vegetation and algae; these biotic components require time to become established. I am concerned that there is not now sufficient time to set-up, and run a proper pilot study over a complete annual cycle if the EPA expects to "begin the cleanup in 2004." A proper pilot study would be established on the site in late spring, several months allowed for the plants and biota to come to equilibrium, and then pollutant testing commenced in late summer and continuing until the following summer. It is not likely that such an effort could commence before the spring of 2003 and be completed before late 2004.

A recent Email circulation indicated: "...there is absolutely no reason for the $400,000 figure suddenly put out by the state." That statement is not correct! The $400,000 per year O&M cost was produced by the professional staff at AD Little. It represents the annual O&M costs for treatment if TP3 is preserved. I have reviewed these costs very carefully (See the Appendix at the end of this report) and conclude they are valid and were properly derived. Cost estimates at this stage of project development are typically expected to be within 20 to 30 percent plus or minus of the actual final costs. It has been my experience that the final project costs usually always tend toward the higher limit, so in this case the final costs could be as high as $500,000 per year (for a proposed system that I do not now believe will work).

In a recently circulated memo John Freitag has claimed that: if 80% of of TP3 is removed that treatment of the seepage from the remainder "could be accomplished by passive wetlands without the treatment cells." There is no basis for that claim! The EPA experts have proposed the same treatment process whether the entire site is preserved or is 80 percent is removed. The difference between the two cases is only a matter of size and cost of the treatment system.

In the same memo, John Frietag also claims that adoption of his proposal "would allow for 80 to 90 percent of the problems with the metals in the water to be solved." That is not true! A simple wetland, by itself, will not handle the complex mixture of metals present in the seepage from either TP3 or TP1/2.

Specific Technical Issues for EPA

The success of the proposed SAP's, SRB's, and wetlands depend significantly on steady state flow conditions. Assuming TP1 is capped, the major long term source seems to be groundwater and so near steady flow conditions may be possible. In the case of TP3 the major source seems to be surface runoff and precipitation. This means that in the coldest winter months there may be no flow except during winter thaw periods and at spring snow melt. The capability of the TP3 treatment components to respond to these unpredictable events is questioned. The "semi passive" alkalinity feeder, for example, is likely to clog up if not kept in motion, so when the next flow event occurs it will not operate properly.

If winter flow does occur at either TP3 or TP1/2 the water will have a low temperature and will be exposed to low ambient temperatures. The result will be icing in the proposed aeration channels and possibly continuous ice covers on any ponds and wetlands. Such an ice cover will drastically reduce the amount of atmospheric oxygen transfer to the under ice water. This in turn will affect the expected reactions. To overcome these winter problems, it might be possible to significantly enlarge the holding ponds for winter storage and only operate the treatment systems in the warm months of the year.

It is not clear why both an ALD and SAP's are required for TP1/2, since both devices are intended to provide alkalinity? It is also not clear why you propose SAP's for TP1/2 and SRB's for TP3? It would seem that O&M could be simpler if the same device were used at both locations. Since you must have SRB's at TP3 then you probably should use them at TP1/2 also.

If you put an ALD in TP1, when does limestone media have to be replaced?

The document says that SAP's are flushed every quarter. Where does the flushed material go?

Howe Bridge, PA is cited as the source for the assumed 15 year life for SAP's. I don't think that is sufficient justification. The SAP concept was first presented in the literature in 1994 and it is unlikely that any actual system is much older than seven or eight years.

The report describes the successful performance of a SRB in Wyoming under winter conditions. I believe an SRB can also function in our Vermont climate, but at a reduced rate. I am more concerned about an SRB at TP3 during the winter months with possibly significant periods of no flow. The bacteria in the SRB may not respond if allowed to dry out. The projected useful life of SRB's is also questioned since none have actually been in service that long.

The report postulates oxygenation of SRB effluent in a rock channel "during all seasons," those channel may fill with ice under sever winter conditions.

The wetland is expected to remove Both BOD (leaking from the SRB) and manganese. However, the oxygen requirements for BOD removal will supersede the oxygen requirements for manganese removal. As a result there may be little or no manganese removal until very low BOD concentrations (< 5 mg/L) are achieved.

It is clear that the writer of this section does not have any experience with wetland systems or an appreciation of the O&M requirements. The proposed system is supposed to contain algae covered rocks for manganese removal. In previous comments (see Appendix below) I had indicated that growing plants in the marsh would shade the algae and interfere with their function. In responding to that comment, the writer of the final report suggests that the plants could be routinely harvested during the growing season and the litter harvested in the fall of each year. Such activity would represent an intolerable and costly maintenance burden and cannot be seriously considered.

It will be impossible for such an algae system to function under the typical ice and snow cover in the winter months, and those conditions can persist for five months each year. If you have to have an algae component, it is obvious that their rock bed should remain unplanted so you get maximum sunlight directly on the algae. Therefore, such a system should have an initial typical vegetated marsh zone for BOD removal, then a zone with the algae rocks, then a final vegetated marsh zone for final polishing of suspended solids, etc.

Since EPA expects the TP1 treatment system to function year-round I would suggest your experts go back to the drawing boards and look for a more effective manganese removal process, and/or plan for winter storage since the algae are not going to do it for at least five months out of the year. There have been articles in the literature about successful use of rock bed subsurface flow wetlands for manganese removal, but oxygen levels were high and organic levels low and I donÕt think we can reproduce those conditions at the Elizabeth mine.

The statement on page 3-45 is quite surprising! If the best the EPA can promise is "a reasonable probability of success" for treatment at TP3 I think you should bite the bullet and recommend that all of TP3 be removed as has been recommended by all of your experts. It is unprofessional for the EPA to shift a decision on this matter to the State of Vermont on the basis of O&M costs.

Appendix

Report # 5 Review Comments: The EPA/ADL Presentation at CAG Meeting January 23, 2002

I was unable to attend the meeting, but I have reviewed the written material presented by EPA and have exchanged Emails with Ed Hathaway to request additional data and to clarify several issues. These data, consisting of detailed spreadsheets describing the development of the new O&M costs have been sent to me by ADL.

At the meeting, the EPA presented:
(1) The reports by their experts who had reviewed the EE/CA report and actually visited the site in November 2000,
(2) a "new" treatment sequence for TP3 if it is decided to preserve the entire site, or a portion of the site, and
(3) revised costs for the entire project, these included a large and rather startling increase in operation and maintenance (O&M) costs for the TP3 treatment system.

The acronyms for the various treatment system components in this project are becoming somewhat confusing and the number seems to grow with each iteration. At the end of this report, I've made a list of the acronyms, and some of the other critical technical terms, in current use with a brief description of each.

(1) Expert's Review

These individuals included five internationally recognized consultants with significant experience in acid mine drainage (AMD) work, an expert from the U.S. Geological Survey, and an engineer from a company with experience in constructing and managing corrective action at AMD sites. All of these people have reviewed the draft EE/CA report and inspected the site. Their comments can be found in the report "Elizabeth Mine Engineering Evaluation/Cost Analysis Comments and Technical Review Reports," presented at the January 23rd CAG meeting. The two Technical Advisors to the CAG had previously reviewed the draft EE/CA report and also offered comments.

It is the consensus professional recommendation of all of these qualified reviewers that the tailings piles at TP3 should be removed completely. The reason behind this consensus opinion is the presence of the highly acidic seepage containing high concentrations of metals. This seepage will be very difficult and expensive to treat. However, because of the local interest, options were presented for retention of TP3 and treatment of the seepage. There was no consensus by the experts on the proposed treatment (ALD followed by a SAP followed by an aerobic wetland in the draft EE/CA report). The majority believed the proposed system would not perform as intended without excessive operational and maintenance attention and costs. I shared those concerns.

As a result of these concerns the EPA then tentatively proposed SRB's instead of SAP's. Concerns were then expressed because of the high iron and aluminum concentrations in the TP3 seepage. A significant amount of these metals might be separated within the SRB, this in turn would result in clogging of the bed resulting in failure and/or requiring very frequent maintenance. To overcome these problems, a third iteration of the TP3 treatment sequence has now been proposed, and will be described and discussed below.

The fact that we are in the third iteration of a possible treatment sequence should not be taken as a criticism of EPA or ADL. Such changes are normal in the development of most engineering designs and it should be remembered we are still in the conceptual stage on this project. However, the fact that we are now at the third iteration is strong evidence of the difficulty of treating the acidic TP3 seepage. As noted in my comments below, I still have reservations regarding the ability of this "new" system to perform reliably on a year-round basis here in Vermont. It is possible that the system which is finally selected could be even more complex and more expensive.

Acid mine drainage has been successfully treated at a large number of locations in the U.S., with some systems being operated for over 20 years. These, however, are mostly coal mining operations and do not contain the very high concentrations of the types of metals present in the TP3 seepage. The treatment sequence now proposed for TP3 has not been used at any other site, to my knowledge. Some of the components have been in successful use for three to six years (an exception is conventional aerobic wetlands, these have been used for many decades), but there is no experience with the long term period required at TP3. The revised projections to 100 and 200 years, instead of the typical 30 years used at most engineering projects, are valid. I cautioned in my first review of the draft EE/CA report that O&M would be required for "generations," apparently the EPA now concurs. The rock particles in the TP3 tailings piles have been exposed to the environment for many years already and the seepage is still highly contaminated. These rocks will not be "clean" in the near future, contaminated seepage will continue until the acidic tailings piles are almost completely dissolved, or completely covered, or removed.

The "New" Treatment Process for TP3

This new treatment sequence has been developed to solve perceived problems with the earlier proposals. Both the SAP's and SRB's are in some ways similar to a filter bed. The water passes down through the media and biochemical transformations occur which then permit the settling and separation of metals in the unit or in subsequent components. In the case of TP3 seepage, the water has very high concentrations of aluminum and iron, and may also contain significant dissolved oxygen. These factors could allow the precipitation of aluminum and iron inside the SRB resulting in unacceptable clogging. The EPA consultants now believe it will be necessary to remove the iron and aluminum prior to the SRB, so we now have a SPAD (Semi Passive Alkalinity Doser), a mixing channel, and a settling pond ahead of the SRB unit. As the name implies the SPAD is semi-passive and uses a water wheel to drive the feed mechanism for adding an alkaline substance to the water. The increase in alkalinity allows the iron and aluminum precipitation reactions to begin; the mixing channel, probably lined with coarse rocks to induce turbulence, insures high oxygen levels in the water so the iron and aluminum can oxidize, precipitate, and settle out in the settling pond. This reduces the SRB clogging potential considerably, and the water then passes through the SRB for removal of copper, lead, zinc and cadmium. Effluent from the SRB then runs down an aeration channel to the wetland. The purpose of this rock lined aeration channel is to increase the oxygen content of the water before it enters the wetland. The final component is the aerobic wetlands. The wetlands are planned for BOD and manganese removal with a "combination of rock filters and algae mats." Details were not provided on the configuration of these features but it is indicated that manganese oxidation will occur by contact with algal mats of leptothrix discophora algae. In accordance with federal and Vermont standards the water at the end of the treatment system discharge pipe, prior to any mixing or dilution in a stream, would be the point of compliance monitoring. The cost estimates provide an allowance for sampling and testing.

It is indicated that all of these components are in use at ADM sites elsewhere. For example, a list is given of 19 sites where the SRB unit is in use. Scanning the list indicates that 17 of these systems are pilot scale, or even smaller "bench" scale units. The one full scale application, at a mine in Missouri, designed for lead and zinc removal, has been in operation for up to 5 years. We have a much more severe winter climate than Missouri so extrapolations are not possible. A pilot test of the entire proposed treatment process at the Elizabeth Mine would be absolutely essential prior to final design.

I am still concerned about the impact of our winter conditions on the proposed system, both on performance and on the ability to function at all. The design cannot assume that Global Warming will convert our climate to the equivalent of Pennsylvania's or Missouri, if the system is expected to run continuously throughout the winter it must contend with worst case conditions. Based on 42 years of record at the Lebanon, NH airport, the coldest winter in the Upper Valley occurred in 1970, the average January temperature was 18°F, in February 21°F, and in March 31°F. The extreme temperature in January was - 28°F. Under these conditions, with three continuous months of sub-freezing weather, ice will certainly form on all of the ponds, on the wetlands, and on the mixing and aeration channels. The latter are most susceptible to icing and a long period of sub freezing weather could result in creation of an ice glacier and subsequent overflow from the channels. It is also unclear how the algae in the wetland can perform their intended function at expected winter temperatures and under an ice or snow cover on the wetland since either ice or snow cover on these wetlands will occur every winter, and the algae require exposure to sunlight to function. It is true that a deep snow pack will act as an insulating barrier and retard freezing. The ground in my front yard will remain unfrozen if we get an early snowfall in November and the snow persists and accumulates all winter. That, however, is not the typical case. We more often have a cold December and no snow until January. The most reasonable solution for TP3 treatment might be to design this system for seasonal operations, say from early April to late November, with the seepage being stored in the holding pond during the colder months.

I also have concerns about the summer performance of the proposed wetland. It is not quite the same as the typical constructed cattail marsh used at many coal mine AMD sites. The wetland is to have rock cobbles on the bottom, presumably to provide the surfaces for the algae to attach and grow, and at least three species of plants will be planted. However, the algae require exposure to sunlight and if the plants grow and multiply they will shade the water surface, and without sunlight the algae will die. These aquatic plants also die back each fall, if the litter is not harvested, and removed, it may smother the algae. Apparently the algae mat for manganese removal is being proposed since the familiar cattail marsh is not too effective for removal of this metal. I have no personal experience with using algal mats, but am aware they have been demonstrated in Florida and California and achieved excellent removals of phosphorus and metals. Those systems incorporated frequent algae harvesting because if the algal cell is allowed to die and decompose there is a release of substances back into the water. Algae harvesting would not be feasible at Elizabeth Mine. Algae also require nutrients (ie: nitrogen, phosphorus, etc) to grow and function. Such nutrients are not normally present in the AMD waste flow, but may leak in sufficient quantities from the SRB component.

Another proposed wetland function is removal of BOD, because the SRB will leak organic material as BOD (Biochemical Oxygen Demand). The concentration may be significant for three to six months and stabilize at 10 mg/L according to the EPA handouts. A 10 mg/L BOD concentration is already below typical discharge standards in Vermont so it is unclear why the wetland is needed for this purpose. In my experience, a wetland system will discharge from 6 to 8 mg/L BOD as a "background" level due to the decomposition of natural organic materials present in the wetland.

The title "Aerobic Wetland" may also be a misnomer. The term "aerobic" implies the presence of oxygen. It has been my experience, that in constructed wetlands with water depths of one foot or more, only the water very near the surface has any oxygen, the bulk of the water in the wetland is essentially devoid of oxygen (or "anoxic"). If the water depth is limited to a few inches the entire depth may be "aerobic." In the winter time, after ice forms, even the surface layer of water in the deeper wetland is essentially devoid of oxygen, except near the front end where the water enters. The "new" treatment sequence proposes the use of an aeration channel ahead of the wetland in order to increase the dissolved oxygen in the water entering the wetland. It has been my experience, with wetland systems elsewhere that such oxygen will be rapidly utilized near the front end of the wetland and will not contribute to aerobic conditions in the remainder of the wetland.

In summary, I still have reservations about the ability of the "new" TP3 treatment system to provide all of the intended treatment responses and to even function effectively in the Vermont winter. I believe the most prudent choice may be to completely remove the tailings piles at TP3. However, if it is decided to preserve all or part of TP3 I also believe a functional treatment system can be developed after pilot testing. Such a system will be very complex and very expensive to operate as evidenced by the increasing cost estimates presented so far by the EPA.

Cost Estimates

The most recent cost estimates for operation and maintenance of a TP3 treatment system have risen dramatically. Sufficient details were not given at the January 23rd presentation so it seemed desirable to review the estimating data and procedures carefully to be absolutely sure the numbers were valid. I asked the EPA for such data and it was immediately provided by ADL. It consists of a massive (>5000K) multi paged spread sheet. This spread sheet contains a preliminary sizing of the various treatment components, made by the consultant engineers, estimates of the materials and labor and costs required for construction and estimates of costs for operation and maintenance (O&M). A one page summary of O&M costs was also provided and a hard copy of this is now on file at the Copperas Hill Coalition (Kathy Hardy). I have reviewed all of these these data and conclude that the projected costs are valid and realistic and were developed using standard procedures. It must also be remembered that we are still in the conceptual design stage and any cost estimate at this point is at best valid to within about 20 percent of the real costs. It is likely that as we get closer to the final design these costs will continue to change and are likely to get even higher. As an example of the estimating procedure I have presented below more detailed values from the ADL spread sheet to demonstrate where one of the summary costs presented by EPA came from.. The O & M (called PRSC costs by EPA & ADL) include two major components, the actual annual costs that occur every year and recurring costs that occur on a less frequent schedule. In the latter category, for example, would be replacement of the media in the SRB. A more realistic replacement schedule of 10 years has now been adopted in the ADL estimate. The original, somewhat optimistic schedules ranged from 15 to 25 years, but there was no actual experience to back up those estimates. I objected to that in my original review and believe the 10 year cycle is more realistic. Another major change from the draft EE/CA report is the assumed life cycle cost period. As with most conventional engineering designs the draft report assumed a 30 year period. In my review at the time, I indicated the O&M costs would continue for generations. EPA now agrees and has assumed a period of at least 100 years for costing purposes. This means that the recurring costs will be repeated 10 times over the 100 year period instead of just once with the 30 year assumption.

The actual estimated annual costs, assuming a non-hazardous sludge, for TP3, include:

The recurring costs occur on a ten year cycle, the present worth of these costs is determined and then distributed as an equivalent annual amount. These annualized amounts are:

When you add these two sets of costs, the total is $254,330 which is the same as the summary value for the complete preservation of TP3 with a non-hazardous sludge, presented by EPA at the November CAG meeting. The annual O&M costs for a hazardous sludge were $400,000. At this stage, I think it would be conservative and prudent to assume that the sludge requiring disposal would be hazardous. Hopefully, information would come from a pilot test to define sludge characteristics.

In summary, I believe these cost estimates are reliable for this stage of conceptual planning, and were derived using appropriate methods.

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

top