Bioremediation for Acid Mine Drainage and Heavy Metals Contamination

Acid mine drainage (AMD) is an extremely environmentally hazardous by-product of mining. Often from natural leaching of groundwater, excess precipitation, or flooding of a mine, water interacts with minerals and rock containing sulfur (commonly pyrite, or FeS₂) and creates sulfuric acid (H₂SO₄) and a dissolved iron ion through a series of oxidation reactions. This low pH environment allows metal ions, like iron, aluminum and manganese, to remain in solution. However, increasing pH, like through mixing with more neutral water, can cause precipitation of metal-oxides. When this liquid escapes the mine area, perhaps through leaching, unsealed portals/audits, overflow, or containment failure, the acidic conditions alone create unsustainable conditions for wildlife and plants. The metal precipitation additionally allows streams to be smothered by a thick layer, reducing sunlight penetration and killing aquatic life (Smith and Wildeman, 2003; Bless et al, 2008). This pollution affects source waters and all downstream until sufficiently remediated or diluted, this can cause entry into groundwater or leaching into soil and allow the negative effects to spread even further to harm all that use the stream or environment surrounding it (NewFields, 2016).

One method of remediating acid mine drainage includes the usage of bioreactors, powered by sulfate-reducing bacteria (SRB). Sulfate reducing bacteria is grown with some sort of organic material, like cow manure or peat. These bacteria, in anaerobic conditions, use the organic carbon (CH₂O) from these environments to reduce sulfate (SO₄^2-) to sulfide (H₂S), which reacts with the dissolved metal ions created by the reactions with pyrite and water to precipitate them. This decreases available concentrations of sulfate and dissolved metal ions from the original acid mine effluent, neutralizes pH from reduction of sulfate and removal of free hydronium (H₃O⁺) ions, and additionally produces the pH buffer bicarbonate (HCO₃^-), which increases the pH of the acid mine effluent (produces alkalinity) (Bless et al, 2008; Nordwick et al, 2008; Forms of Pollution, 2004). By exposing this effluent to these bacteria, cleaner water can be discharged from mines.

To examine this process, two case studies, that of Lilly and Orphan Boy Mine and Sure Thing Mine, are explored and compared to understand the efficiencies of bioreactors with sulfate-reducing bacteria and the evolution in their designs. Together, these two mines are both located within similar regions in mountainous Montana, with similar climate, geology, and therefore contaminants, but have extremely different methods of utilizing SRBs in bioreactors; as Lilly and Orphan Boy Mine remediation acted as a precursor to Sure Thing, we can easily compare the design ideology and lessons learned from one mine to the other.

This location receives heavy snow events, has an average temperature of 6 degrees Celsius and is at a high elevation of about 7000 feet (Doshi 2006). The geology of the site is characterized by granitic rock (Boulder Batholith) comprised of quartz monzonite with intrusions of aplite (NewFields 2016). Within the mine property and in range of the mine was a mineralized vein, containing galena, pyrite, sphalerite, arsenopyrite and tetrahydride. Run off drains into Telegraph Creek, which runs through the property, which includes four rock waste piles, the larger two immediately adjacent to the main mine shaft (NewFields 2016).

Table 1: Initial (pretreatment) acid mine drainage water dissolved metals characterization of averages over several months between September 1993 and August 1994, in mg/L, and pH (Bless et al, 2008; Nordwick et al, 2008).

The Lilly/Orphan Boy Mine remediation plan included a subsurface (in situ) bioreactor, meaning the mechanism for removal was implemented within the mineshaft itself; the concept was to create various plugs of organic substrate on which to grow SRB in the pathway of the flowing AMD to precipitate metals out and neutralize it. 

Thirty feet below the water level within the main shaft, platforms were built for the substrate, which consisted of a 7:2:1 ratio of cow manure, decomposed wood chips, and alfalfa straw, respectively, to be packed in. Two injection wells were drilled vertically downward from the surface to intersect the 90 degree arm of the shaft that extended in the direction of Telegraph Creek; here two more plugs of organic substrate were also packed to provide a total of three layers of substrate before exiting the shaft through the mine portal at the end of the arm.

It was installed in 1994 and until 2005 consistently monitored two different locations for water quality, one from a drilled monitoring well after the final substrate plug and one exiting the portal. This locations tested s mix of parameters to understand how well the bioreactor was removing the AMD pollutants from the water, like total and dissolved metals, sulfate and sulfide, dissolved oxygen (DO)/biochemical oxygen demand (BOD)/chemical oxygen demand (COD), alkalinity and pH, temperature, redox potential, and volatile fatty acids. Flow rate through the system varied seasonally due to spring snowmelt and increased run off, but often was between 1-8 gallons per minute. HRT could not be calculated due to lack of accurate measurements of total substrate volume in the system (Bless et al, 2008; Doshi et al 2006).

A summary of the treated effluent contaminant concentration is as follows: removal for aluminum, cadmium, copper and zinc were significant, but this design was inefficient for arsenic, manganese, and iron removal. Additionally, the effluent tended to increase in dissolved metals concentration from the last substrate layer located in the tunnel to the portal (Bless et al, 2008). Design goals for some of the parameters were met and differed strongly in correlation to season, or snowmelt. Target pH range was between 6-8 for the final effluent, which was achieved for both locations and time frames, except for May 2001 testing of the treated portal effluent, which was the same to the pretreatment portal value of 3.4.  Sulfate for this same location and time actually increased in comparison to original values at that location, from 213 mg/L to 394 mg/L. This correlates well to the decrease in pH, but also because the pretreatment values were sampled and averaged over a longer period of time, it is possible values this high were reached during events of snowmelt, but is not shown within the overall average.

Dissolved iron and arsenic did not meet any of the demonstration goals for any of the timeframes or sampling locations. Cadmium, aluminum and copper met all of the goals with exception to May 2001, treated portal effluent. Zinc only met the goal for the tunnel sampling locations, but not the portal location.

The full table is shown below.

In the first half of the 20th century, the Elliston District of Powell County, Montana was known as a producer of lead-zinc ores along with gold and silver. The extraction of these minerals exposed sulfide mineralization to the environment, leading to acid rock drainage from the mine adit (Bless et al, 2006). 

The Surething Mine consists of a near surface adit into a steep hillside with narrow horizontal single-layer workings. These workings were mostly exploratory, extending only a few hundred feet into the mountain. A pile of mine tailings and waste rock was dispositive just downhill from the Surething Mine adit. This adit historically discharged 2 gallons per minute of pH 2 acid mine drainage associated with groundwater from the workings. It is noted that the flow rate of discharge is heavily influenced by surface precipitation and snow melt. Although the discharge flows towards O’Keefe Creek, the discharge infiltrates into the subsurface roughly 120 feet from the adit well before reaching O’Keefe Creek (Nordwick and Bless, 2008).

Table 4: Typical Water Chemistry of Surething Mine Drainage

The technology’s multi-stage process at the Surething Mine involved sequential passage of acid mine drainage from the mine adit through three adjacent anaerobic reactors and an aerobic reactor. The entire system relies on gravity-driven flow, and was designed for 2 gpm of discharge. During operation, actual discharge reached peaks of 10 gpm during spring runoff (Nordwick and Bless, 2008). The configuration of the constructed treatment system can be found in the figure below.

Drainage water then flowed passively through a limestone reactor, which was constructed with limestone cobbles to add alkalinity to the water. The limestone reactor was sized for a 1.25-day residence time at 2 gpm (Nordwick and Bless, 2008). However, field measurements of the constructed reactor indicate that the actual residence time of the cell is 1.3 days. Flow in the limestone reactor was directed in the same manner as in the first SRB reactor. This cell was also covered with a 6-inch layer of alfalfa to provide thermal insulation.

The second SRB reactor was sized for a 1.25-day residence time at 2 gpm. However, field measurements of the constructed reactor indicate that the actual residence time of the cell is 2.02 days (Nordwick and Bless, 2008). Similar to the first SRB reactor, the second SRB reactor contained a 50%-50% mixture by volume of cow manure and walnut shells. This cell was designed to serve as the primary driver of sulfide-precipitating reactions to remove the contaminants from solution. With the exception of manganese, concentrations of all of the targeted metals exiting this cell were below state water quality standards.

The drainage water leaving the second SRB reactor was then aerated by routing the flow through 165 feet of 8-inch corrugated pipe which was snaked down the hillside. Three vertical air vents were placed along the length of the pipe to provide air to the line and provide sufficient amounts of oxygen for the manganese-oxidizing bacteria that the water would interact with in the next reactor. The original aerobic reactor was sized for a 10-day residence time at 2 gpm (Nordwick and Bless, 2008). The crushed limestone provided an environment for indigenous manganese-oxidizing bacteria to thrive and for subsequent removal of manganese as a precipitate. The limestone in this reactor was piled above the water level in order to avoid outlet plugging from the surrounding foliage.

Table 5: Final Effluent Contaminant Concentrations after all System Modifications Completed

Generally speaking, most of the heavy metals were effectively removed after passing through the first bacteria reactor, especially during the early years of operation. Notable exceptions include iron and aluminum, which show only partial removal of the contaminants after passing through the first SRB reactor during the later years of operation. However, these contaminants were still effectively removed after passing through the limestone and second bacteria reactor. Another notable exception to this general trend was manganese, where the bulk of the contaminant did not get removed until it passed through the aerobic limestone bed (Nordwick and Bless, 2008).

In both the first SRB reactor and the limestone reactor, the original feed distribution system failed due to plugging as precipitates built up on the bottom of the reactor. In order to fix this issue, the feed distribution was modified in order to allow for vertical distribution of the feed at three locations to depth of 6 feet near the front end of the reactor and to allow for easier cleanout of the system.

In addition to the aforementioned changes, an additional 250 gallon hold tank was connected to the top front end of the limestone reactor in order to promote even feed distribution and serve as a collection settler for additional precipitates carried over from the previous cell. Liner material was also added over the top of this reactor to provide some inhibition of oxidation in the reactor and help maintain the anaerobic conditions between the two SRB reactors.

It is noted that the second SRB reactor was not modified over the course of the demonstration. However, if the system were to continue operation, a vertical feed distribution system would need to be added to the cell in order to continue flow (Nordwick and Bless, 2008).

Analysis of the original aeration line indicated that the original design did not sufficiently oxidize the water. To remedy this, the length of the aeration pipe was increased to 300 feet and slits were made on top of the aeration line at two-foot intervals in order to allow for the placement of wooden weirs. These modifications helped to increase the retention time and provide turbulence to help mix the oxygen.

After multiple attempts to modify the original aerobic reactor to be able to remove enough manganese from the drainage water, a new aerobic reactor was eventually constructed on top of the existing aerobic reactor. New liner material was placed over the old reactor to create a shallower, 1-foot deep reactor. This pool was filled with limestone and featured vertical baffles spaced at 10-foot intervals in order to force water to the surface as it flowed along the reactor.

As demonstrated by the aforementioned case studies, the usage of SRB bioreactors are effective for passive acid mine drainage treatment, especially for remote, hard-to-access, or abandoned mines. In both cases, the organic nutrients applied to stimulate the sulfate-reducing bacteria only needed to be applied in the first year, as the systems were able to sustain themselves and resulted in virtually no operating costs. Additionally, because the physical features and surrounding geography of the mines were able to be used as part of the treatment processes, the capital investment for equipment was significantly lower than having to construct an active treatment system. Overall, the improvements made to Sure Thing mine from the lessons of Lilly and Orphan mine worked significantly better at removing heavy metals, especially for dissolved iron and aluminum with the addition of the limestone and second anaerobic reactor.

When considering improvements that could be made to the design of the organic substrate, one notable change was the replacement of wood chips and alfalfa straw with walnut shells. Walnut shells were chosen due to the high percentage of organic carbon, which can be up to 60%, and slow biodegradation rate. Walnut shells would serve to enhance the overall long-term bioreactor performance. Additionally, due to their shape, walnut shells increase the porosity of the mix and provide an internal structure that prevents settling of the material, thus preserving high permeability of the organic medium (Nordwick et al., 2003). 

Lilly and Orphan Boy mine acted as a case study for proving an in situ bioreactor within a mineshaft was viable, but with adjustments for locations of the substrate, for compact remote areas. Surething improved upon this further with additional reactors and limestone, but met other issues such as needing to configure the systems so that maintenance could easily be performed, since the precipitates could plug the pipe system and impede flow. Additionally, Surething’s design needed to add features that promote even flow distribution throughout the bioreactor, including perforated vertical pipes and hold tanks, and help improve bioreactor function.

Together, each of these proved that the application of sulfate reducing bacteria was possible in cold climate extremes. However, although these systems were very effective at removing the metals, exposure to acid mine drainage with high acidity and high metal concentrations may reduce the expected bioreactor lifetime and would need to be taken into account for future design evolutions. 

1. Bless, D. (2006, March). MWTP Demonstrates Integrated Passive Biological System for Treating Acid Rock Drainage. Technology News and Trends. https://clu-in.org/products/newsltrs/tnandt/view.cfm?issue=0306.cfm 

2. Bless, D., Park, B., Nordwick, S. et al. Operational Lessons Learned During Bioreactor Demonstrations for Acid Rock Drainage Treatment. Mine Water Environ 27, 241–250 (2008). https://doi.org/10.1007/s10230-008-0052-6 

3. Doshi, S. Bioremediation of Acid Mine Drainage Using Sulfate-Reducing Bacteria. (2006). https://clu-in.org/download/techdrct/S_Doshi-SRB.pdf 

4. Forms of Pollution. (2004). http://www.cotf.edu/ete/modules/waterq/wqchemistry.html

5. Hargrave, P., et al. Abandoned - Inactive Mines of the Blackfoot and Little Blackfoot River Drainages. (1998). http://mbmg.mtech.edu/pdf-open-files/MBMG368_Blackfoot.pdf 

6. Ness, I., Janin, A. and Stewart, K. Passive Treatment of Mine Impacted Water In Cold Climates: A review. Yukon Research Centre, Yukon College. https://casinomining.com/_resources/Passive_treatments_review_-_Cold_Climate_-_YRC2014.pdf

7. NewFields Companies, LLC. Expanded Engineering Evaluation & Cost Analysis: Lilly/Orphan Boy Mine Powell County, Montana. (2016). https://deq.mt.gov/Portals/112/Land/AbandonedMines/documents/ProjectDocuments/OrphanBoy/Final%20Lilly%20Orphan%20Boy%20EEECA.pdf?ver=2016-04-15-132002-563 

8. Nordwick, S. et al. Mine Waste Technology Program: In Situ Source Control Of Acid Generation Using Sulfate-Reducing Bacteria. (2008) https://drive.google.com/file/d/1uUweZq8gv-0MRyGRYMFQeOEvO9utZjoG/view 

9. Nordwick, S., & Bless, D. (2008, September). Integrated Passive Biological Treatment System: Mine Waste Technology Program Activity III, Project 16 - Final Report. US Environmental Protection Agency. https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NRMRL&dirEntryId=208530 

   Nordwick, S., Zaluski, M., Park, B., & Bless, D. (2003). Advances in the Development of Bioreactors Applicable to the Treatment of ARD. http://mwen.info/docs/imwa_2006/1410-Nordwick-MT-2.pdf