CRL Energy Mine Drainage Framework

Appendix F. Reducing impacts – treatment techniques

F.1 AMD treatment (coal PAF regions)

F.1.1 Selection between active and passive treatment

A number of factors will influence the decision as to whether to use active or passive treatment. For example, where mine drainage exceeds specific thresholds and a large amount of neutralising material is required to ensure appropriate treatment, then a large passive system would be required, which can be prone to failure (see below), so active treatment is likely to be a better choice. Particularly relevant in New Zealand are variables such as flow rate and acid load. On the West Coast, for example, where the majority of AMD sites are located, annual rainfall is high (up to 6 m on the plateaus north of Westport), and AMD sites are often located in very isolated areas with steep topography. This can result in AMD with very high flow rates and acid loads that can be markedly influenced by rainfall events. This is very different to the south-eastern USA (West Virginia) where much of the early development of AMD treatment occurred. There, topography plays a minor role in the selection of treatment systems and the dominant variables are climatic – the very cold winters and hot summers. Therefore, for New Zealand AMD sites, base and peak flow rates and acid loads need to be carefully considered along with the degree of isolation and access to power when making a choice between active and passive treatment. Further details of how each parameter influences the choice of active or passive treatment is provided below.

Acidity and pH

Acidity is comprised of proton acidity (pH) and mineral acidity, which is from dissolved metal species which produce more proton acidity upon hydrolysis (Rose and Cravotta 1998). To treat a low pH and/or high acidity AMD with passive remediation, a very large system is necessary to achieve a long enough residence time for neutralisation and a very large amount of neutralising material is necessary to maintain system longevity. Very large systems can be prone to short-circuiting and failure as preferential pathways may develop over time. Rather than constructing one large system, it would be better to split the flow among several parallel systems. Waters et al. (2003) document that the most successful passive treatment systems treat AMD with pH > 2 and acidity < 800 mg/L (as CaCO₃). An acidity of 800 mg/L equates to a pH of 2, an Fe concentration of 50 mg/L and Al concentration of 30 mg/L. The relationship between acidity and pH is exponential, with much greater increases in acidity below pH 2. As such, AMD with extremely low pH (<2) and/or high acidity (>800 mg/L) is better treated with active systems than passive systems.

Flow rate

High flow rates are difficult to treat with passive systems because the systems need to be very large to achieve the necessary residence time and very large systems can be prone to short circuiting and failure. However, flow rates should be carefully considered along with the practicabilities of treatment. Significant flow rate variations during storm events are common at AMD sites in New Zealand, and treatment systems (active or passive) may not be able to treat the AMD 100% of the time, but rather might be designed for base flow conditions and allow high flow events to bypass the treatment system. Before this decision can be made, quantification of acid loads during base flow versus total flow (including storm events) should be completed and compared with the treatment requirements for the site. In general, active treatment systems can accomodate some variability in flow rates by changing chemical dosing rates to match flow rates, whereas, passive treatment systems can only manage variability if an equilisation header pond is constructed to dampen peak flows.

Acid load (acidity x flow rate)

A high acid load (generated by high acidity and/or high flow) consumes the neutralising material faster, and in a passive system this can limit life expectancy. In addition, high flow rates require construction of very large systems that can be prone to short circuiting and failure. Therefore, for a situation with very high acid load, it is recommended that consideration be given to using an active treatment system, although a passive system can perhaps be used if a shorter life expectancy (< 25 years) is acceptable. The choice will often be based on a cost–benefit assessment of active versus passive treatment. Alternatively, if sufficient space is available at the site, the AMD can be split into several smaller flows with lower acid loads, and multiple passive treatment systems can be constructed. This is essentially using a process principle to enable use of passive treatment in a high acid load situation. High acid loads also can result in the generation of much greater volumes of precipitates during neutralisation and the longevity of passive systems can be compromised if permeability drops as precipitates accumulate in system pore spaces (Waters et al. 2003).

As discussed above, significant flow-rate fluctuations are common at AMD sites in New Zealand , and systems (active or passive) must be designed with this in mind. But research is lacking on the effect of flow rate fluctuations on acid load in New Zealand; it is possible that correlations may be site specific and even variable for a given site. Research is also lacking on ecosystem effects resulting from a sudden pulse of high-acid-load water if treatment systems are not designed to accommodate storm events (where acid loads are increased). It is recommended that these issues be addressed prior to deciding on the level of treatment required.

Operational mine site with power

In general, active treatment systems are more commonly used at operational mine sites, whereas passive systems are typically used at closed and abandoned mines. Operational mine sites typically have limited space for treatment systems and a drainage chemistry and flow rate that can change as mining proceeds. These factors are addressed more easily with active treatment systems than with passive systems. However, if sufficient space is available, and chemistry and flow rates are not expected to change significantly with time, passive treatment can be a suitable solution at active mine sites. Further, passive treatment may be used to complement active treatment.

Power is a critical factor for active treatment systems. Pumps are often used to convey the water to the treatment system and between various components of the system; power is usually needed to meter additives to the water such as neutralising chemicals, flocculants, and coagulants, and power is necessary for mixing and oxidation of the water. If no power is available at the site, a semi-passive system, such as the Aquafix system, can be used, which relies on a paddlewheel in the AMD stream to operate a hopper that dispenses neutralising chemicals into the AMD (Skousen and Jenkins 1993). Otherwise, lack of power limits AMD treatment to passive systems unless it is warranted to bring power to the site.

F.2 AMD active treatment systems

F.2.1 Selection of neutralising chemicals – Dosing-with-alkali stage

The most commonly used chemicals for raising the pH of acidic mine drainages are soda ash (sodium carbonate; Na₂CO₃), hydrated lime (calcium hydroxide; Ca(OH)₂), quicklime (calcium oxide; CaO), caustic soda (sodium hydroxide; NaOH), and ammonia (NH₃). Magnesium oxide or hydroxide (MgO or Mg(OH)₂) and limestone (CaCO₃) are occasionally used. A brief description of these chemicals and their use is provided below.

Selection of an appropriate chemical is dependent on the flow rate of the AMD, and the concentration of dissolved Fe (see main text). Other common metal ions in AMD, such as Al, and to a lesser extent Zn, and Ni, are removed along with Fe, and are not a factor in deciding among the neutralising chemicals. Other factors will also influence chemical selection. These include: chemical cost, neutralising efficiency, maximum pH attainable, dispensing mechanism required, mixing mechanism required, health and safety issues, sludge settling rates and therefore requirement for flocculants or coagulants, and resulting sludge volume and density (Skousen et al. 2000; Waters et al. 2003; Means 2006). Chemical cost in particular can be significant, as over the long term the largest single cost component in most systems is the neutralising chemical (Waters et al. 2003). It is recommended that bench scale tests are conducted on various chemicals before final selection (Younger et al. 2002) and also a sequential titration acidity analysis as described in Hilton (2004) and in section F.3.1. A comparison of the characteristics of the common chemicals is provided in Table F1.

If Mn is present at high concentrations in mine drainage, there needs to be additional consideration of what chemicals to use. Manganese is a difficult metal to remove from solution as it exhibits high solubility over a broad pH range (4.5-8) and the chemical oxidation of Mn is kinetically slow (Bamforth et al. 2006). The most effective way to remove Mn from water is to raise the pH above 9 and allow Mn²⁺ to oxidise to Mn³⁺ or Mn⁴⁺ and form insoluble Mn oxides or Mn carbonates (Evangelou 1998). Where it is necessary to lower the concentrations of Mn, although Ca-based chemicals can be used, the most commonly-used chemicals for Mn removal are sodium hydroxide (NaOH) and ammonia (NH₃) (Skousen 1988; Skousen et al. 1990, 2000).

Table F1. Characteristic of chemicals used to neutralise AMD in active treatment systems (Skousen et al. 2000; Waters et al. 2003; Means 2006).

Chemical	Conversion factor¹	Neutralisation efficiency (%)²	Cost of chemical per tonne of acid neutralised³	Dispensing mechanism	Key benefits	Key limitations	Risk of failure
Soda ash or sodium carbonate (Na₂CO₃)	1.06	95–100 (powder) 60 (briquettes)	$830 – $870 (powder)	Briquettes placed in wooden box or large drum/reactor in AMD stream.	High efficiency in powder form, most metals precipitate, low sludge volumes.	Health and safety issues, poor sludge settling rates, potential sodium toxicity.	Potential for reduced treatment effectiveness when using briquettes if acidity loading rates increase significantly (best as an interim treatment or only for low flow, low acidity AMD).
Hydrated lime or calcium hydroxide (Ca(OH)₂)	0.74	90–95	$330 – $350	Silo or hopper with mechanical feed screw to dispense powder. Batching tank to mix powder with water. Can use aqueous slurry. Mixing suggested.	High efficiency, most metals precipitate, low cost, widely available.	Health and safety issues, reagent saturation can lower efficiency.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop. Poor maintenance can result in plugged dispensing mechanism and complete failure.
Quicklime or calcium oxide (CaO)	0.56	90	$233	Silo or hopper with mechanical feed screw to dispense powder or water wheel feeder with one-tonne storage bin (no power). Batching tank to mix powder with water. Mixing suggested.	High efficiency, most metals precipitate, very low cost, widely available.	Health and safety issues, reagent saturation can lower efficiency, possible armouring of pebbles.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop. Poor maintenance can result in plugged dispensing mechanism and complete failure. Must be watertight or will hydrate and form calcium hydroxide and plug dispensing mechanism.
Ammonia (NH₃ or NH₄OH)	0.34	100	$2,200	Compressed and stored as liquid in tank, gas injected near bottom of pond or water inlet. No mixing required.	Very high efficiency, most metals precipitate, low sludge volumes.	Health and safety issues, poor sludge settling rates, can be toxic to aquatic life, high cost.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop.
Caustic soda or sodium hydroxide (NaOH)	1063 (50% liquid)	100	$1,050	Stored as a liquid in tank, dispense through metering pump or valve and feeder hose near top of pond or water inlet. No mixing required.	Very high efficiency, most metals precipitate, low sludge volumes.	Health and safety issues, poor sludge settling rates, potential sodium toxicity, highest cost of all chemicals, low freezing point.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop. If insufficient antifreeze added, can freeze in winter resulting in complete failure.
Magnesium oxide or hydroxide (MgO or Mg(OH)₂)	0.40 or 0.58	90 - 95	Not common in NZ	Silo or hopper with mechanical feed screw to dispense powder. Batching tank to mix powder with water. Mixing suggested.	Very high efficiency, most metals precipitate, low sludge volumes, low cost.	Some health and safety issues, not widely available, lower reaction rate than calcium hydroxide.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop.
Limestone (CaCO₃)	1	approx 90	$111	Silo or hopper with mechanical feed screw to dispense powder. Batching tank to mix powder with water. Mixing suggested.	Safe to use, lowest cost of all chemicals, readily available, cannot overtreat.	Low efficiency, not all metals removed (ineffective for Mn), armouring.	If acidity loading rates increase beyond system capacity to neutralise and settle hydroxides, treatment effectiveness will drop.

¹ Conversion factor is the mass of chemical needed to neutralise the acidity relative to limestone. The conversion factor is used along with the neutralisation efficiency to calculate the tonnes of chemical needed to neutralise each tonne of acidity per year. For sodium hydroxide, the conversion factor gives litres of chemical needed per tonne of acid.

² Neutralisation efficiency estimates the effectiveness of the chemical in neutralising acidity relative to sodium hydroxide and is used along with the conversion factor to calculate the tonnes of chemical needed to neutralise each tonne of acidity per year. For example, if 46 tonnes of acid needs neutralisation per year, 38 tonnes of hydrated lime would be needed [46(0.74)/0.90].

³ Cost of chemical is per tonne of acid neutralised in 2010 NZD. This is an effective way to compare costs of treatment with different chemicals. Cost is calculated using purchase price, conversion factor and neutralisation efficiency. For example, if calcium oxide costs $375 per tonne, to neutralise one tonne of acidity would cost $233 [(375)(0.56)/0.90].

Factors to consider in selecting a calcium or sodium compound for an AMD treatment system are shown in Table F2.

Table F2. Factors that may influence the selection of a calcium or sodium compound for an AMD treatment system (from Skousen 1988).

Factor	Calcium	Sodium
Solubility	slow	fast
Application	requires mixing	diffuses well
Hardness	high	low
Gypsum formation	yes	no
High TSS or clay particles	helps settle clay	disperses clay particles and keeps clay in suspension
Chemical cost	lower	higher
Installation and maintenance costs	high	low

Soda ash or sodium carbonate (Na₂CO₃)

Soda ash is generally only used in small-flow cases with low amounts of Fe and where Mn is not a problem (Skousen et al. 1990). It can also be used as a good interim solution until a more permanent system is constructed (Means 2006). Selection of this chemical is sometimes based more on convenience than cost-effectiveness. The chemical is delivered in briquette form and placed in a box, barrel or drum. Water flows through this dispensing mechanism dissolving a small amount of the chemical. Gravity keeps the briquettes feeding into the water for constant treatment.

Figure F1. Treatment of AMD using sodium carbonate.

Hydrated lime or calcium hydroxide (Ca(OH)₂)

Hydrated lime is the most common chemical used for neutralising AMD. Although the capital cost for a system using hydrated lime is high, low chemical cost makes this a very economical treatment method over the long term. It is generally delivered in powdery form, which is hydrophobic, so extensive mixing is recommended for high flow rates to make it soluble in water (Skousen et al. 1990). It is usually contained in large hoppers and dispensed into the AMD using a mechanical screw feeder (Figures F2 and F3) (Means 2006).


Figure F2. Treatment of AMD using calcium hydroxide. Hoppers dispense chemical directly into AMD.	Figure F3. Treatment of AMD using calcium hydroxide in large water treatment plants.

Quicklime or calcium oxide (CaO)

Quicklime is the non-hydrated form of hydrated lime. Quicklime is the least expensive chemical used in treatment of AMD, but it must be stored in a water-tight container to avoid premature hydration and formation of hydrated lime (Means 2006). If it does form hydrated lime, it can bridge and eventually plug the dispensing mechanism. It is commonly dispensed from a hopper through a waterwheel-powered auger into the AMD (Figure F4), requiring less operation and maintenance than a conventional system, and useful for areas without power (Skousen and Jenkins 1993). At high flow rates, use of a water-driven mixer is recommended (Means 2006).

Figure F4. Treatment of AMD using calcium oxide. Chemical is dispensed from hopper (left) by a paddlewheel (right) driven by the AMD flow.

Caustic soda or sodium hydroxide (NaOH)

Caustic soda is the most expensive chemical used in AMD treatment, is hazardous to handle, and freezes more readily than lime; however, it is 100% efficient, can raise the pH very high and therefore is able to remove Mn, and a treatment system using caustic soda has low capital costs (Means 2006). It can be delivered to the site in pellet form or more commonly as an aqueous solution (20–50%) (Figure F5). It is often stored in large tanks and dispensed either through a valve and feeder hose or with a metering pump into the top of the AMD (it is denser than water). Along with ammonia, no mixing is required; however, the resulting sludge is loose and gelatinous (Skousen et al. 2000). Sodium hydroxide shows a nearly linear relationship with pH (to 12), unlike ammonia which shows a logarithmic curve, with only small changes in pH occurring above 9.2 with the addition of more chemical (Skousen et al. 1990). It is often not economical to use Ca-based chemicals for Mn removal due to the very long residence times needed to attain a high pH (slower dissolution rates than NaOH and NH₃) and the large amounts of unreacted chemical that are produced when raising the pH above 7 (Skousen 1988; Skousen et al. 2000; Means 2006).

Figure F5. Treatment of AMD using sodium hydroxide. Chemical is dispensed by gravity from tank into the AMD.

Ammonia (NH₃)

After caustic soda, ammonia is the next most expensive chemical used in AMD treatment and is very hazardous to handle. However, like caustic soda it is 100 % efficient, can raise the pH to 9.2 and therefore is able to remove Mn, and a treatment system using ammonia has low capital costs (Waters et al. 2003). It is delivered as a compressed gas (anhydrous ammonia), stored as a liquid in large tanks (Figure F6, left) and bubbled into the water near the bottom of the water column where it returns to the gaseous state (Figure F6, right) (Faulkner and Skousen 1991). As it is a compressed gas, it does not have to be pumped or gravity fed. No mixing is required, but the resulting sludge is loose and gelatinous (Skousen et al. 2000). If ammonia is used, a pH-driven monitoring system is recommended, because over application can lead to toxicity of the treated water. Ammonia reacts with sulphuric acid to form ammonium sulphate, consuming hydrogen ions (raising the pH), and forming hydroxyl ions which react with Fe and Al to form hydroxides (Faulkner and Skousen 1991). It is one of the fastest ways to raise the pH.

Figure F6. Treatment of AMD using ammonia. Chemical is dispensed from tank (left) under pressure through piping into the AMD and precipitates form immediately (right).

Magnesium oxide or hydroxide (MgO or Mg(OH)₂)

Although magnesium oxide has a relatively low cost (but slightly more expensive than calcium hydroxide) and is highly efficient, it needs mixing, has a low reaction rate, and most importantly, is not generally widely available (Waters et al. 2003). It is dispensed in a similar manner to and behaves like calcium hydroxide (Skousen et al. 1996).

Limestone (CaCO₃)

Limestone has the lowest material cost and is the safest and easiest to handle of the chemicals commonly used for AMD treatment. Unfortunately, its successful application is limited due to its low solubility and tendency to develop an external coating, or armour, of ferric hydroxide when added to AMD (Phipps et al. 1996). In addition, limestone has the lowest treatment efficiency of the common chemicals used for AMD treatment. Limestone is generally only applicable when Fe concentrations are below 5 mg/L and total acidity is less than 50 mg/L (Skousen et al. 1990).

If limestone is ground to a powder it can be dispensed into an AMD stream by a waterwheel-powered doser and is effective in raising the pH and lowering the concentrations of dissolved metals (Mills 1996; Zurbuch 1996). In Mills (1996) waterwheel-powered dry powder limestone dosers raised the pH to neutral and treated water with Fe concentrations of about 7 mg/L and Al concentrations of about 8 mg/L. Zurbuch (1996) describes water-powered rotary drums which autogenously grind limestone aggregate inside them into a slurry which is released into a river. The limestone drops into a small hopper where it enters the drum via an inverted screw feed which is part of the drum axle. Small holes in each drum vane permit water to enter the drum and CaCO₃ slurry to exit. A back-up limestone powder doser was constructed adjacent to the drum station. In New Zealand powdered limestone (< 0.1 mm in size) is currently being used to treat AMD at the Stockton Mine (Weber et al. 2007). Powdered limestone is released from a hopper via an electrically-driven screw feed into the Mangatini Stream (Figure F7). The rate of the screw feed is regulated by a pH probe placed downstream of the hopper. The main disadvantage of simply dispensing a neutralising chemical into an AMD stream is that metals precipitate and settle out of solution along the watercourse resulting in an increase in TSS in the stream and large amounts of Fe and Al floc (Weber et al. 2007).

Figure F7. Treatment of AMD using powdered limestone. Chemical is dispensed from hopper by screw feeder directly into AMD. Stockton Mine, New Zealand.

Rarely used chemicals

Other chemicals occaisionally used to neutralise AMD include kiln dust, fly ash, fluidised bed combustion ash, calcium peroxide, potassium hydroxide, barium carbonate, hydroxyapatite and seawater-neutralised red mud (Waters et al. 2003).

F.2.2 Oxidation

Examples of different oxidation systems are shown in Figures F8 and F9.

Figure F8. Oxidation using rotating blades (left) and trickle filter (right).

Figure F9. Oxidation using in-line venture aeration (top) and cascade aeration (bottom).

F.2.3 Sedimentation

The chemical composition of AMD sludge is generally hydrated Fe²⁺ or Fe³⁺ oxyhydroxides, CaSO₄, Al(OH)₃∙6H₂O, CaCO₃, and Ca(HCO₃)₂ with trace amounts of Si, PO₄, Mn, Cu, and Zn (Ackman, 1982). Leaching of trace elements from the sludge may influence disposal in a landfill or on-site and leach tests should be performed.

Standard leaching tests, both of which should be used, include the toxicity characteristic leaching procedure (TCLP; US Environmental Protection Agency test method 1311; EPA 1995) and the synthetic precipitation leaching procedure (SPLP; US Environmental Protection Agency test method 1312; EPA 1995). Each procedure produces a leachate from the sample which is then analysed for the chemicals of concern in the AMD. The TCLP assumes worse case scenario in which the test material is subjected to an acidic extraction buffer, whereas the SPLP uses an extraction buffer more similar to rainfall.

If significant leaching of trace elements occurs, stabilisation of the sludge may be required. Stabilisation of sludge involves incorporating an additive (often cement) to prevent leachate forming. See Brown et al. (1994a, b, c) for information on volume, stability, and composition of sludge generated from active treatment of AMD.

Examples of different sedimentation techniques are shown in Figures F10-F12, while photos of sludge dewatering techniques are shown in Figures F13-F14.

Figure F12. Equipment used to add coagulants during AMD treatment to assist sedimentation of the precipitates formed.

Figure F13. Dewatering of sedimentation pond and drying of precipitates.

Figure F14. Pumping of precipitate slurry using vacum trucks following AMD treatment.

F.3 AMD passive treatment systems

Ten passive remediation systems are described in this section. The first six utilise the oxidising strategy (open limestone channel, diversion well, limestone leaching bed, slag leaching bed, aerobic wetland, and dosing with limestone sand) and the next four utilise the reducing strategy (anoxic limestone drain, anaerobic wetland, vertical flow wetland, and sulphate reducing bioreactor). Each system is listed in Table F3, providing a general description, design factors, indicative cost, key benefits, limitations, risk of failure, and references. Full-scale systems can be designed using the computer program AMDTreat (Means et al. 2003). Figures F15 and F16 provide information on potential construction costs, lifetime treatment costs and overall effectiveness of passive treatment systems based on literature studies.

Table F3. Passive treatment systems

System type	General description	Design factors	Key benefits	Limitations	Construction cost¹	Risk of failure	References
Oxidising strategies
Open Limestone Channels (OLC) Open Limestone Drains (OLD)	Channel, ditch, or streambed lined with limestone cobbles. Dissolution of limestone and neutralisation occurs as AMD flows down the channel.	Rock sizes should be > 10 cm (15-30 cm). Slope should be >20%, if possible. Base residence time on acid load, limestone volume on alkalinity generation rate.	Low cost. Simple. High level of reliability.	Armouring with hydroxides. Can construct at >20% gradient to minimise but then must be long channel to achieve residence time.	$135,000	If Fe loading rates too high, armouring of limestone will reduce effectiveness and system may fail.	Ziemkiewicz et al. 1994, 1997; Ziemkiewicz and Brant 1996; Cravotta and Trahan 1999; Trumm et al. 2005, 2008
Diversion Wells (DW)	Round chamber filled with crushed limestone aggregate. AMD flows into chamber outpipe near bottom of well creating turbulence which abraids particles preventing armouring. Water flows upward and out of chamber.	Design based on trial and error. Well dimensions typically 1.5 m in diameter, 2 m deep, half-filled with 1-2 cm diameter limestone gravel (low hardness). Residence time about 15 min.	Low cost. High limestone efficiency.	Requires refilling with limestone chips about every 2-4 weeks. Requires elevation change and constant flow rate. Precipitates not captured. Pockets of air in piping can reduce flow rate.	NA	If limestone gravel too coarse it may not be churned enough to prevent armouring by Fe oxides and system may fail.	Sverdrup 1983; Arnold 1991; Faulkner and Skousen 1995; Skousen et al. 1998
Limestone Leaching Beds (LLB)	Rectangular chamber filled with limestone cobbles. AMD flows horizontally or vertically through cobbles dissolving limestone which neutralises acidity.	Limestone mostly between 38 and 90 mm size. 15 h residence time in unit.	Simple. High level of reliability.	Armouring with hydroxides if Fe concentrations too high.	$152,000	If Fe loading rates too high, armouring of limestone will reduce effectiveness and system may fail.	Black et al. 1999; Danehy et al. 2002; Hilton et al. 2003; Watzlaf et al. 2003
Slag Leaching Beds (SLB)	Rectangular chamber filled with steel slag fines. AMD flows horizontally or vertically through slag dissolving CaO which neutralises acidity.	Steel slag mostly less than 3 mm. Residence time 1-10 h depending on acidity.	High pH. Generates more alkalinity than limestone (up to 2,000 mg/L). Stable silicate sludge produced.	Armouring with hydroxides. Slow dissolution. Leaching of metals from slag.	NA	If Fe loading rates too high, armouring of slag may reduce effectiveness and system may fail.	Ziemkiewicz and Skousen 1998; Simmons et al. 2002; Trumm et al. 2009
Aerobic Wetland	Large surface area pond with emergent wetland species plants. With or without limestone. AMD flows horizontally through pond and over substrate. Oxidation reactions precipitate oxides and hydroxides.	Design based on removal rates: 10-20 g Fe/m²/d, 0.5-1 g Mn/m²/d, 3.33 g H₂SO₄/m²/d.	General precipitate storage. Low cost.	Best for pH > 5.5. Removes selected metals.	$25,349	If pH is too low, plants may not survive, oxidation rates may lower, and system may fail.	Skousen et al. 1992; Hedin et al. 1994; MEND 1999; Skousen et al. 2000; Batty and Younger 2002
Dosing With Limestone Sand	Large stockpile of crushed limestone placed on edge of AMD. Water washes limestone into stream and acidity neutralised.	Limestone very small grain size. Stockpile placed partly in stream.	Very simple, armouring prevented due to small grain size.	Requires restocking pile.	NA	If limestone sand is too coarse, it may armour with Fe oxides before it can dissolve and treat the water.	Mills 1996; Zurbuch 1996; Skousen et al. 1998; Watzlaf et al. 2003
Reducing Strategies
Anoxic Limestone Drains (ALD)	Buried limestone-filled drain. AMD flows horizontally through limestone in an anoxic environment. Alkalinity added by limestone dissolution.	Limestone 6-15 cm in diameter. 14 h residence time. Need to exclude oxygen.	Low cost. Simple.	Best for low Al, low DO, low Fe³⁺ or armouring occurs. Maintenance difficult.	$128,000	If DO levels are too high and/or Fe3+ concentrations too high, armouring with Fe oxides can cause system to fail. Too high Al concentrations can plug pore spaces.	Turner and McCoy 1990; Skousen 1991; Hedin and Watzlaf 1994
Anaerobic Wetlands	Large pond with a layer of organic substrate - typically spent mushroom compost with about 10% CaCO₃. AMD flows horizontally within the substrate. Emergent vegetation helps stabilise substrate and provide organic material. Sulphate reduction removes suphate and metals.	Hydraulic conductivity of substrate 10³-10⁴ cm/s. Design based on removal rates: 3.5 g acidity/m²/d; 2.9 g H₂SO₄/m²/d; sulphate reduction rate 300 mmoles/m³/d.	Relatively stable sulphide sludge storage.	Requires long residence time.	$57,938	Without sufficient organic matter, anaerobic conditions may not occur and Fe oxides can plug pore spaces causing system failure.	Hedin et al. 1994; Skousen et al. 1992, 2000
Vertical Flow Wetlands (VFW); Successive Alkalinity Producing System (SAPS); Reducing and Alkalinity Producing System (RAPS)	A rectangular unit with limestone at the base covered by spent mushroom compost and free water. AMD flows vertically down through the unit. Sulphate reduction removes sulphate and metals in the compost, alkalinity generated in limestone.	15-30 cm organic matter; 6-15 cm diameter limestone. Design based on 15 h residence time in limestone layer; 35-40 g acidity/m²/d removal rate.	Small area required. Good for high Fe, low Al.	High capital costs. Armouring and plugging with hydroxides.	$171,000	If compost thickness or composition inadequate to establish reducing conditions, metals will oxide, limestone will become armoured and system can fail. Inadequate flushing may not remove accumulated metals and lead to plugging of system.	Kepler and McCleary 1994; Zipper and Jage 2001; Watzlaf et al. 2004; Rose 2006
Sulphate Reducing Bioreactor (SRB)	A rectangular unit filled with a mixture of organic substrates such as hay, lucerne, sawdust, paper, or woodchips, crushed limestone, and compost or manure. AMD flows vertically through unit. Sulphate reduction removes sulphate and metals.	Design based on removal rates: 0.3 mol metals/m³ of substrate/d or 0.3 mol sulphate/m³/d.	Small area required. Good for high Fe, low Al.	High capital costs. Reduced permeability with time. Potential for armouring and plugging with hydroxides.	$274,000	Without sufficient organic matter, anaerobic conditions may not occur and Fe oxides can plug pore spaces causing system failure.	Gusek 2002, 2004; Neculita et al. 2007

¹Cost to treat hypothetical AMD (acidity 150 mg/L, Fe²⁺ 10 mg/L, Fe³⁺ 1 mg/L, Al 12 mg/L, pH 2.9, DO 1 mg/L, flow rate 10 L/s) determined using AMDTreat computer program (Table F4).


Figure F15. Comparision of potential construction costs for AMD passive treatment systems. A, Average cost in evaluation of 137 treatment units in the Eastern USA (Ziemkiewicz et al. 2003a, 2003b). B, Range in costs for treatment systems in Australia (Waters et al. 2003). Aew, aerobic wetland; Anw, anaerobic wetland; OLC, open limestone channel; ALD, anoxic limestone drain; RAPS, reducing and alkalinity producing system; SLB, slag leaching bed; LLB, limestone leaching bed; DW, diversion well; Bio, bioreactor.

Figure F16. Comparision of the cost-effectiveness of AMD passive treatment systems (Ziemkiewicz et al. 2003a, 2003b). A, Cost to treat each kg of acidity per year over the system lifespan (usually 20 years) using the following equation: Acid removal cost ($/kg/y) = [Construction Cost/(kg acid treated per year)(system lifespan)]. B, %age of installed systems which had positive treatment measured as net acid reduction in the AMD. Aew, aerobic wetland; Anw, anaerobic wetland; OLC, open limestone channel; ALD, anoxic limestone drain; RAPS, reducing and alkalinity producing system; SLB, slag leaching bed; LLB, limestone leaching bed.

Oxidising systems

Open limestone channel and open limestone drain (OLC, OLD)

Open limestone channels (OLCs) and open limestone drains (OLDs) are systems in which limestone is placed along the sides and bottom of culverts, ditches or stream channels (Ziemkiewicz et al. 1994; Cravotta and Trahan 1999) (Figure F17). As AMD flows down the channel, dissolution of the limestone neutralises the acidity and raises the pH.

Iron and Al are removed in OLCs by formation of metal hydroxide, oxyhydroxide or hydroxysulphate precipitates as the pH is raised (Nordstrom and Alpers 1999; Skousen et al. 2000). Iron precipitates above a pH of about 2.5 and Al precipitates above a pH of about 5 (Nordstrom and McCleskey 2006). As the AMD flows down the channel and is neutralised, the precipitates typically accumulate in the channel; however, due to the gradient in pH, the Al and Fe precipitates are spatially separated and little Fe-Al coprecipitation occurs (Bigham and Nordstrom 2000). Zinc and Mn are removed in these systems by adsorption onto reactive Fe hydroxide surfaces rather than formation of oxides (Stumm and Morgan 1996; Bostick et al. 2001).

One of the key benefits of an OLC is its simplicity and low cost. In an analysis of long-term performance and cost-benefit of passive treatment systems, Ziemkiewicz et al. (2003a, b) found that OLCs were one of three systems that provided a high level of reliability (success versus failure), high acid load removal, and low treatment cost.

A key limitation, however, is the tendency for the limestone to become armoured with precipitates which can reduce the dissolution rate and therefore the effectiveness of the system (Ziemkiewicz et al. 1997; Trumm et al. 2008). Selfridge et al. (2003) found that only a relatively thin coating of less than 1 μm is needed to reduce dissolution rates for iron-coated limestone by 20%. Eventually dissolution rates in OLCs can drop by 50-80% (Skousen et al. 2000). In addition, if the gradient is relatively low, precipitates can accumulate in pore spaces and this limits the reactive surface area of the limestone. Cravotta and Trahan (1999) recommend that velocities through the system are kept high enough (>0.1 m/min) so that the solids are kept in suspension for subsequent settlement in a lagoon, wetland, or settling pond. To reduce the amount of armouring of limestone and build-up of precipitates within pore spaces, Ziemkiewicz et al. (1997) recommend that OLCs be constructed at gradients of greater than 20%.

Even if OLCs are constructed at low gradients or used to treat AMD with very high Fe concentrations, recent work suggests that armouring on limestone can be removed through agitation. Santomartino and Webb (2007) found in a laboratory experiment that Fe hydroxide coatings on limestone consist of three distinct layers, with a 60-µm-wide void between the limestone substrate and the first layer. They suggest that mechanical agitation can dislodge the coatings and restore effectiveness of the treatment system.

Operation and maintenance, therefore, consists of periodically moving and breaking the rocks with a dozer to loosen and remove precipitates if they have accumulated in the channel.

Open limestine channels are sized according to residence time necessary to neutralise acidity. A model was developed by Ziemkiewicz et al. (1994) to estimate limestone volumes and channel dimensions for achieving neutralisation of AMD based on first-order kinetics. The required residence time to achieve target acidity from the initial acidity is first solved. Then the amount of limestone required is based on flow rate and measured acidity of the AMD. Wetted channel dimensions are computed to obtain sufficient residence time of the water in the limestone channel at specified water velocities and channel slopes (using the Manning equation).

In their initial publication (1994), Ziemkiewicz et al. recommended that channels be built five times larger than the model suggests to account for reduced dissolution rates as the limestone became armoured. However, in a later publication (Ziemkiewicz et al. 1997), they found that OLCs did not need to be built this big. Santomartino and Webb (2007) provide a formula to predict when an OLC (or open limestone drain) may need agitation to dislodge precipitates armouring limestone rocks. The formula incorporates reactive surface area, Fe concentration, fraction of Fe retained in the system, and flow rate.

Figure F17. Open limestone channel passive treatment system.

Diversion well

Diversion wells (DWs) are partially-buried large-diameter vertical chambers constructed of concrete or metal and half-filled with limestone chips ( Arnold 1991). Acid mine drainage flows through a PVC pipe that enters vertically down the centre of the well and ends shortly above the bottom (Figure F18). The hydraulic force of the pipe flow causes the limestone to turbulently mix and abrade into fine particles. The water flows upward and overflows the well through a notch or a hole, where it is diverted back into the stream. The limestone is replenished on a regular basis (typically 2-4 weeks); therefore, this is considered a semi-passive treatment system. Diversion wells were initially developed for treatment of acidity caused by acid rain in Norway and Sweden (Cementa 1983).

To ensure a velocity high enough to fluidise the limestone bed, there must be sufficient head drop above the DW from a dam or other AMD collection device (typically 2.5 m or more). Work by Arnold (1991) shows that a DW with a diameter of 1.5-1.8 m and height of 2-2.4 m, filled with limestone 1-2 cm in diameter, can treat a flow rate of 1500 L/s. The PVC piping should be 20-30 cm in diameter. Waters et al. (2003) suggest DWs can be constructed with diameters between 1 and 3 m and heights below 5 m.

Acid mine drainage is treated in a DW through dissolution of the limestone, which raises the pH and causes Fe and Al hydroxides to precipitate. Turbulence within the well abrades any armouring that forms on the limestone chips, thereby continually exposing fresh limestone surfaces for dissolution. Water flowing out the well carries Fe and Al floc as well as some limestone fines. Settling ponds can be constructed following the DW to collect the precipitates (Skousen et al. 1998). Sometimes only a portion of the AMD needs to flow through the DW, and the treated AMD can be mixed with untreated AMD in the settling ponds to complete the treatment.

The key benefits of a DW are its simplicity and the high efficiency of limestone dissolution due to the lack of armouring with Fe and Al hydroxides.

The major limitation is that the limestone must be replenished on a regular basis. In addition, because the residence time within the well is relatively short (about 15 min), alkalinity generation is limited and a pH rise of only 1 to 2 units can be expected (Arnold 1991). To compensate, several DWs can be constructed in series to treat a low-pH AMD. Other design considerations include increasing efficiency by installing a metal collar at the point of AMD discharge into the well with lateral holes so that the water is directed sideways rather than down against the bottom, and flaring the top of the well to cause reduced water velocity, which allows undissolved particles to sink back into the fluidised bed (Arnold 1991).

Wratten and Schwartz (1996) developed a pulsed-bed limestone-water contactor that accelerates limestone dissolution rates through use of a carbon dioxide pretreatment step. Mineral acidities in excess of 1000 mg/L were neutralised and unusually high levels of alkalinity were achieved during treatment. However, their design adds additional complexity to the system and requires power.

Operation and maintenance consists of regularly adding limestone chips to the DW and clearing piping of accumulated floating debris.

Diversion wells are sized according to flow rate and available head drop. Although some workers have gone to considerable lengths to calculate the exact consumption rate and particle size distribution under various conditions (Sverdrup 1983; Fraser et al. 1985), Arnold (1991) suggests that local variation will in most cases negate the accuracy of such calculations and that a trial-and-error process is necessary to achieve optimum results.

Figure F18. Diversion well passive treatment system, side (left) and top (right) views.

Limestone leaching bed

Limestone leaching beds (LLBs) are rectangular chambers or beds filled with limestone cobbles ranging in size from 5 to 10 cm in diameter (Black et al. 1999; Watzlaf et al. 2003) (Figure F19). Acid mine drainage typically flows through the bed horizontally and is neutralised through dissolution of the limestone. To ensure adequate residence time in the LLB, the water level is kept below the surface of the limestone.

Aluminium and Fe are removed in LLBs during neutralisation of AMD by formation of metal hydroxide, oxyhydroxide or hydroxysulphate precipitates as the pH is raised (Nordstrom and Alpers 1999; Skousen et al. 2000). Zinc and Mn are likely removed in these systems by adsorption onto reactive Fe hydroxide surfaces (Stumm and Morgan 1981; Bostick et al. 2001). Watzlaf et al. (2003) recommend using LLBs only if Fe is not present, and suggest that they primarily be used to remove Mn through biological Mn oxidation. Rose et al. (2003) used an LLB inoculated with Mn-oxidising bacteria to remove Mn from an AMD with a pH of 6.5. Hilton et al. (2003) used an LLB to successfully treat AMD containing low Fe (1.65 mg/L) but high Al (56.0 mg/L).

The key benefits of an LLB are its simplicity and its high treatment efficiency. In an analysis of long-term performance and cost–benefit of passive treatment systems, Ziemkiewicz et al. (2003a, b) found that LLBs were one of three systems that provided a high level of reliability (success versus failure), high acid load removal, and low treatment cost.

A key limitation, however, is the tendency for the limestone to become armoured with Fe precipitates, which can reduce the dissolution rate and therefore the effectiveness of the system (Ziemkiewicz et al. 1997; Trumm et al. 2008). Selfridge et al. (2003) found that only a relatively thin coating of less than 1 μm is needed to reduce dissolution rates for iron-coated limestone by 20%. Therefore LLBs are best reserved for AMD with low Fe concentrations.

A common problem with leaching beds used for treating AMD is the clogging of passages with precipitates that decrease permeability and, hence, residence time (Faulkner and Skousen 1994; Watzlaf and Hyman 1995). Often flow rates through the open passages in the bed increase, resulting in enlargement of these preferential passageways and complete plugging elsewhere. To help overcome this problem, researchers in the USA have modified an existing system (an LLB) to have a horizontal flow with vertical flush capability (Hilton et al. 2003). During normal operation, preferential horizontal pathways start to become established as precipitates build up. During rapid vertical flushing, the precipitates are dislodged and removed from the system, thereby restoring permeability throughout the bed and eliminating these preferential pathways. We recommended that this horizontal flow - vertical flush concept is used in the construction of LLBs.

Operation and maintenance consists of regularly flushing the system to remove built-up precipitates and capture them in a holding pond. Piping and valves may need cleaning or replacement with time.

Limestone leaching beds are sized according to residence time necessary to neutralise acidity. Residence times are typically 12-17 h (Black et al. 1999; Ziemkiewicz et al. 2003b).

Figure F19. Limestone leaching bed passive treatment system.

Slag leaching bed

Slag leaching beds (SLBs) are rectangular chambers or beds filled with steel slag generated from the arc furnace steel-making process (Figure 20). AMD typically flows through the bed horizontally and is neutralised through dissolution of the slag. The slag from New Zealand Steel Limited in the North Island contains approximately 15% CaO and ranges in size from 20 to 60 mm.

As with CaO active treatment systems, neutralisation by steel slag is relatively fast and the pH can be raised substantially, possibly in excess of 10 (Ziemkiewicz and Skousen 1998). Iron and Al are removed in SLBs by formation of metal hydroxide, oxyhydroxide or hydroxysulphate precipitates as the pH is raised. Manganese is typically removed in these systems through precipitation of oxides, hydroxides, or carbonates at high pH, and is the primary target metal of SLBs.

The key benefits of an SLB are its simplicity and its ability to raise the pH relatively high to remove metals such as Mn, due to a high neutralisation potential (Simmons et al. 2002). If the source of the slag is local it can prove to be as economical as limestone, otherwise it is typically more expensive than limestone.

Limitations of slag include cost, the potential to leach other elements, and the potential of armouring if Fe concentrations are significant. If the source of the slag is far from the site, it can be cost prohibitive. Slag is a by-product of the steel-making process, and as such, contains impurities from iron ore in the production of steel. There is the potential for these impurities to be leached from the slag as the CaO dissolves (Simmons et al. 2002). It is prudent, therefore, to conduct trials of the slag to determine the stability of these impurities in the slag prior to use in full-scale construction. As with limestone-based systems, armouring can be a significant problem which can reduce dissolution rates, and therefore, neutralising capability. Selfridge et al. (2003) found that only a relatively thin coating of less than 1 μm is needed to reduce dissolution rates for Fe-coated limestone by 20%. The upper limits for Fe concentration in SLBs are unknown.

Operation and maintenance consists of regularly flushing the system to remove built-up precipitates and capture them in a holding pond. Piping and valves may need cleaning or replacement with time.

Slag leaching beds are sized according to the residence time necessary to neutralise acidity. Residence times are typically between 1 and 10 h and are determined largely through trial and error (Trumm et al. 2009).

Figure F20. Slag leaching bed passive treatment system.

Aerobic wetland

Aerobic wetlands are the earliest passive treatment system used to treat AMD. Huntsman et al. (1978) and Wieder and Lang (1982) first noted treatment of AMD flowing through naturally occurring Sphagnum bogs in Ohio and West Virginia, USA. Since then, over a thousand wetlands have been constructed to treat AMD (Skousen et al. 2000) (Figure F21). Aerobic wetlands consist of Typha, Juncus and Scirpus sp. and other wetland vegetation planted in shallow (<30 cm), relatively impermeable sediments comprised of soil, clay or mine spoil (Skousen et al. 2000; Ziemkiewicz et al. 2003b). The plants encourage more uniform flow, help stabilise the substrate, help maintain microbial populations, add organic matter, and provide aesthetic qualities to the wetlands.

Water flows horizontally through the wetland with depths varying between 15 and 45 cm. Because of their extensive water surface area and slow flow, aerobic wetlands promote metal oxidation and hydrolysis, thereby causing precipitation and physical retention of Fe, Al and Mn hydroxides. Variations in water depth within the wetland cell may be beneficial for performance and longevity. Shallow water zones enhance oxygenation and oxidising reactions and precipitation, while deeper water zones provide storage areas for precipitates. Aerobic wetlands perform best when water is net alkaline (Skousen et al. 2000; Watzlaf et al. 2003; Ziemkiewicz et al. 2003a); however, they can be used for low pH water but need to be a larger size due to lower removal rates, and plants are seldom very vigorous at low pH (PIRAMID 2003).

Metals are removed in aerobic wetlands by sedimentation of suspended flocs, filtration of flocs by stems of plants, adsorption of aqueous metal species, precipitation of hydroxides on plant stems and the wetland sediment surface, and direct plant uptake of iron and other metals (which are retained primarily in the plant roots; PIRAMID 2003). Metal oxidation, hydrolysis, and precipitation are considered the most important metal removal mechanisms and it is best to have net alkaline water to neutralise acidity from hydrolysis (Skousen et al. 2000). Aeration prior to the wetland, via riffles and falls, increases the efficiency of the oxidation process and the microbial mass in the wetland is necessary for microbial Fe oxidation. Although plant uptake is generally considered less important, Batty and Younger (2002) have shown that it is crucial in removing the last few mg/L of Fe.

The key benefits of an aerobic wetland are its simplicity, low cost, aesthetic qualities, and ability to serve as general precipitate storage.

The most significant limitation of aerobic wetlands is their poor performance with low pH water. As such, they have not been included in the flow charts for selecting AMD passive treatment systems, but they are commonly used as polishing systems following other passive treatment systems. Ziemkiewicz et al. (2003a) found that wetlands treating water with pH > 6 showed optimal performance.

There is very little required operation and maintenance for aerobic wetlands. Inflow and outflow devices/channels may need to be regularly cleared of debris.

Aerobic wetlands are typically sized based on metal loading rates. For example, Skousen et al. (1992) recommend a sizing criteria of 20 g Fe/m²/d, Brodie et al. (1993) recommend 21 g Fe/m²/d and Hedin et al. (1994) recommend 10-20 g Fe/m²/d. PIRAMID (2003) provide examples of successful treatment of drainage containing over 50 mg/L Fe. A review of aerobic wetland performance and design criteria is provided in MEND (1999). When constructing aerobic wetlands, Watzlaf et al. (2003) suggest (1) precede the wetland with a pond 1.5-2 m deep with 8-24 h retention time or have a deep section at the start of the wetland, (2) plant reed rhizomes (or other wetland species) in soil or spread seeds, (3) design with an average depth of 10-50 cm with some shallow and deep marsh areas and a few 1-2 m deep areas, and (4) design with 1-m freeboard and size to hold 20-25 years’ accumulation of Fe. Construction details are also provided in Hammer (1992) and PIRAMID (2003). Plant selection and establishment details are provided in Skousen et al. (1992).

Figure F21. Aerobic wetland passive treatment system.

Dosing with limestone sand

Dosing with limestone sand is perhaps the simplest of passive treatment systems. With this technique, sand-sized limestone is dumped into and adjacent to AMD streams (Figure 22), the sand is picked up by the stream flow, dissolves, and thereby neutralises the acidity (Mills 1996; Zurbuch 1996; Skousen et al. 1998; Watzlaf et al. 2003).

There is no system as such through which water flows. Instead the neutralising material (limestone) is directly added to the AMD, similar to how an active treatment system works. During base flow conditions, only a small amount of the limestone is swept into the AMD stream adding some alkalinity; during storm flow conditions, much greater amounts of limestone are carried by the stream and transported throughout the watershed where the limestone is incorporated in the stream sediments and dissolved to add alkalinity over time (Watzlaf et al. 2003).

Aluminium and Fe are removed with this technique by formation of metal hydroxide, oxyhydroxide or hydroxysulphate precipitates as limestone dissolves and the pH is raised. Zinc and Mn are likely removed by adsorption onto reactive Fe hydroxide surfaces. Coating with Fe-hydroxides can occur but agitation and scouring of the limestone keeps fresh surfaces available (Skousen et al. 1998).

The key benefits of dosing with limestone sand are its simplicity, low cost, and the high efficiency of limestone dissolution due to minimal armouring with Fe and Al hydroxides.

There are two main limitations to this technique. The limestone pile needs replenishment on a regular basis, effectively making this a semi-active system. At 41 sites in the Middle Fork River in West Virginia treated with limestone sand dosing, workers found that limestone needed to be replenished three times per year to maintain a pH above 6 (Zurbuch 1996). Another limitation to this technique is that the length of stream in which the limestone is carried before dissolving is sacrificed for downstream improvement in water quality (Watzlaf et al. 2003).

Operation and maintenance revolves around ensuring that the correct amount of limestone is being carried away by the AMD stream. This requires having the pile in a location where the stream can continuously erode the limestone at an appropriate rate. It requires regular replenishment.

The only design criterion for this system is to choose an appropriate grain size for the limestone. If too large a grain size is chosen, Fe-hydroxides can armour the limestone before it is dissolved into the AMD. Too small a grain size may result in much of the limestone being blown away from the site as dust. Higher gradient streams, and hence, those with higher turbulence, can be treated with larger-grain-sized limestone. Optimal grain size can be chosen through trial and error. Zurbuch (1996) found that grain sizes between 0.075 and 2.36 mm work well for the Middle Fork River in West Virginia.

Figure F22. Dosing with limestone sand passive treatment system.

Reducing systems

Anoxic limestone drain

Anoxic limestone drains (ALDs) are buried trenches filled with limestone gravel (Hedin and Watzlaf 1994) (Figure 23). Water flows through the system under anoxic (low DO) conditions, dissolving the limestone thereby neutralising the acidity. ALDs are only effective when the majority of Fe is in the Fe²⁺ form and very little DO is present. The common problem that plagues most oxidising passive treatment systems, armouring of limestone with Fe oxides, can be avoided with ALDs. Ferrous iron requires oxygen to oxidise to Fe³⁺ and Fe³⁺ readily precipitates without the need for oxygen at pH around 3 (resulting in armouring). Therefore, if the majority of Fe is Fe²⁺ and no DO is present, the pH can be raised substantially without Fe oxides precipitating since Fe²⁺ will not oxidise to Fe³⁺ without oxygen. It is possible to form Fe²⁺ precipitates, but only at higher pHs than can be achieved through limestone dissolution (Stumm and Morgan 1996).

In an ALD, the limestone is buried because conditions within the trench must be anoxic, so that all dissolved iron remains as Fe²⁺, rather than converting to Fe³⁺, which would quickly lead to hydrolysis and armouring of the limestone with Fe precipitates, leading to a reduction in limestone dissolution rate and, ultimately, clogging of the pore spaces of the ALD. ALDs were developed by the Tennessee Valley Authority who found that AMD seeping through a coal refuse dam was being treated passively by high calcium carbonate limestone in an old haul road buried under a dam (Turner and McCoy 1990; Brodie et al. 1993).

Once water exits the ALD, sufficient area in the form of a ditch, settling pond, or wetland must be provided for metal oxidation, hydrolysis and precipitation to occur (Skousen 1991). Dissolved oxygen levels will increase, Fe²⁺ will oxidise to Fe³⁺ and Fe oxides will precipitate. Zinc and Mn are likely removed by adsorption onto reactive Fe hydroxide surfaces.

Aluminium, however, will precipitate in the ALD and, if the concentration is too high, will tend to reduce the permeability of the drain with time, potentially leading to eventual failure. Skousen et al. (2000) recommend a maximum Al concentration of 25 mg/L, but Watzlaf et al. (2000) found ALDs could fail with Al concentrations above 21 mg/L, and Watzlaf et al. (2003) recommend that Al concentrations be below 1 mg/L as an extra precaution against failure. We recommend ALDs can be used with Al concentrations up to 25 mg/L but extra permeability should be built into the ALD at the higher values.

The key benefit of an ALD is its high efficiency of limestone dissolution due to minimal armouring with Fe hydroxides. Other benefits of an ALD are its simplicity and low cost. In an analysis of long-term performance and cost–benefit of passive treatment systems, Ziemkiewicz et al. (2003a, b) found that ALDs were one of three systems that provided a high level of reliability (success versus failure), high acid load removal, and low treatment cost and they were the most consistently efficient passive treatment system in terms of cost per tonne of acid removed (Ziemkiewicz et al. 2003b).

The main limitations to this technique are its strict water chemistry limits. Dissolved oxygen levels should be below 2 mg/L and Eh at 0 or less (Skousen 1991). Some argue for DO to be less than 1 mg/L (PIRAMID 2003). Ferric iron should comprise at most 10% of the total Fe, but some workers recommend a concentration limit of 1 mg/L for Fe³⁺ (PIRAMID 2003; Watzlaf et al. 2003). And Al levels should be at most no greater than 25 mg/L (Skousen et al. 2000). In addition to these, Nairn et al. (1991) and Robbins et al. (1999) show that high levels of sulphate (> 1500 mg/L) can result in precipitation of gypsum in the drain.

There is very little required operation and maintenance for ALDs. The most important is to ensure that air is continuously excluded from the system. Inflow and outflow devices/channels may need to be regularly cleared of debris.

Design criteria are largely based on residence time. The theoretical maximum level of alkalinity that can be generated is 300 mg/L, mostly in the form of bicarbonate (HCO₃^-; Skousen 1991). Hedin et al. (1994) recommend a residence time of 14 h, and in a review of long-term performance of ALDs, Watzlaf et al. (2000) found maximum levels of alkalinity were reached after 15 h retention. Limestone particle size is another important design criterion. Too small a size and the drain may plug and too large reduces limestone reactive surface area substantially. Skousen (1991) recommends sizes between 4 and 10 cm, and PIRAMID (2003) recommend sizes between 5 and 7.5 cm. Some 8-25 cm rocks should be used as well, especially when Al is present (Faulkner and Skousen 1996). Construction details (such as recommended width and depth) and design equations are provided in Skousen (1991), Anonymous (2001) and Watzlaf et al. (2003).

Figure F23. Constructing an anoxic limestone drain passive treatment system.

Anaerobic wetland

Anaerobic wetlands (sometimes referred to as compost wetlands) are similar in design to aerobic wetlands, except the organic substrate thickness is much greater (> 30 cm) and the thickness of free-standing water is much less (0-8 cm; Hedin et al. 1994; Skousen et al. 2000; PIRAMID 2003; Waters et al. 2003; O'Sullivan 2005). The result of this design is that water flows horizontally through the substrate material in a reducing environment rather than over the substrate as it does in aerobic wetlands. Some wetlands have limestone mixed into the lower layers of the substrate. Emergent vegetation helps stabilise the substrate and helps maintain microbial populations, adds organic matter, and provides aesthetic qualities to the wetlands.

Anaerobic wetlands perform better with AMD containing low DO and pH above 3.5 (Waters et al. 2003), but they can tolerate high levels of DO and lower pH (Skousen et al. 2000).

Metals are removed in anaerobic wetlands by a combination of mechanisms (Skousen et al. 1992, 2000). In the aerobic surface layers, some metal oxidation and hydrolysis occurs, especially if the concentration of Fe³⁺ is significant. Aluminium will form Al hydroxides on the surface of the substrate material. In the upper layers of the substrate, metal exchange and complexation reactions can occur (but only initially until sorption sites filled).

However, the dominant mechanism is chemical and microbial reduction, which precipitates metals and neutralises acidity. Microbial sulphate reduction occurs in the presence of an oxidisable carbon source (the organic substrate), which results in the formation of bicarbonate and hydrogen sulphide (H₂S). The H₂S then combines with metals to form mono- and di-sulphides (such as FeS and FeS₂). Compared with diffuse amorphous Fe oxyhydroxides, Fe mono- and di-sulphides are compact and dense precipitates, which would therefore hinder permeability to a lesser extent (Skousen et al. 2000). The removal of metals from solution results in a lowering of the mineral acidity and excess bicarbonate produced through sulphate reduction will consume proton acidity, thereby raising the pH (Hedin and Nairn 1992). Dissolution of carbonates within the substrate material will also generate alkalinity (Skousen et al. 1992; Waters et al. 2003). Under the reducing environment within the substrate, the limestone will not become armoured with Fe oxyhydroxides.

The key benefits of an anaerobic wetland are its simplicity, low cost, aesthetic qualities, and ability to serve as general precipitate storage. Another main benefit is the high efficiency of limestone dissolution due to minimal armouring in the anaerobic environment. In addition to this, anaerobic wetlands are useful in locations where there is a minimal elevation drop between the AMD and the treatment system location.

Limitations of anaerobic wetlands include potential metal overloading and excessive acidity in the AMD. Metal overloading can result in exhaustion of sorption sites on the organic material and a lowering of the permeability of the organic material due to sulphide precipitation (Waters et al. 2003). High acid loads (> 300 mg/L) can overwhelm pH-sensitive microbial activities (Skousen et al. 2000). Mono-sulphides within the substrate material can oxidise and release acidity if exposed to air (Trumm 2007).

There is very little required operation and maintenance for anaerobic wetlands. Organic matter may need to be added to the wetland if performance drops (Waters et al. 2003). Inflow and outflow devices/channels may need to be regularly cleared of debris.

Anaerobic wetlands are typically sized based on acidity loading rates. For example, Hedin et al. (1994) recommend a sizing criterion of 3.5-7 g acidity/m²/d and MEND (1999) recommend a loading rate of 1 kg acidity per 200-500 m²/d. A review of anaerobic wetland performance and design criteria is provided in MEND (1999). Material selected for the substrate must provide sufficient dissolved organic carbon to support sulphate-reducing-bacteria activity. Typical material includes woodchips, sawdust, peat moss, compose, manure, leaf litter or straw/hay (Skousen et al. 2000; Waters et al. 2003). Skousen et al. (2000) suggest that selection and enrichment of naturally occurring microbial populations help long-term performance. Plant selection and establishment details are provided in Skousen et al. (1992) and Skousen et al. (2000). PIRAMID (2003) recommend following an anaerobic wetland with an aerobic wetland to increase the DO and remove any remaining Fe and Al.

Reducing and alkalinity producing system / Bioreactor

Reducing and alkalinity producing systems (RAPS), also known as vertical flow wetlands (VFWs) (Figre 24) and successive alkalinity producing systems (SAPS), are a combination of an ALD and an organic substrate (Kepler and McCleary 1994; Zipper and Jage 2001). Sulphate reduction and metal sulphide precipitation can occur in the compost material while the underlying limestone adds alkalinity to the AMD.

Water flows vertically downward through the system. Typically, there is 1 m of standing water over 13-30 cm of organic compost, underlain by 0.5-1 m of limestone (Kepler and McCleary 1994). The water first encounters the top of the compost layer where aerobic bacteria consume the available oxygen, then within the compost layer, where DO levels are very low, sulphate-reducing and Fe-reducing bacteria operate, then when the water reaches the limestone layer, dissolution of limestone can occur without significant armouring of Fe oxides since Fe³⁺ will have been reduced to Fe²⁺ (Skousen et al. 2000). The water leaves the system via a network of drainage pipes at the base of the limestone layer, and the outlet consists of a riser to maintain head within the system (Danehy et al. 2002).

In RAPS, reducing conditions in the compost layer reduce Fe³⁺ to Fe²⁺ and reduce sulphate to hydrogen sulphide, generating bicarbonate alkalinity (Hedin et al. 1994; Jage et al. 2001). The Fe²⁺, hydrogen sulphide and bicarbonate react to form Fe mono-sulphides. Laboratory studies demonstrate that under reducing conditions the pathway to pyrite involves the formation of mackinawite (FeS), followed by reaction with sulphide to form greigite (Fe₃S₄), which then reacts with sulphide to form pyrite (FeS₂; Hunger and Benning 2007). Zinc removal in RAPS is likely through formation of Zn sulphide or as impurities in FeS (likely to be sphalerite). Manganese does not precipitate as a sulphide at circum-neutral pH, therefore it is not removed in these systems (Kepler and McCleary 1994; Trumm 2007).

The main benefit of RAPS is that armouring of limestone with Fe oxyhydroxides is prevented. Any Fe³⁺ in the AMD is reduced to Fe²⁺ and precipitated as sulphides before the AMD comes in contact with the limestone. Therefore, dissolution of limestone can occur unhindered with armouring in a reducing environment, similar to that of the ALD. Another benefit of RAPS is that they can be constructed in relatively small areas.

In a review of the long-term performance of 40 RAPS, Rose (2006) found four main problemscan occur.

AMD with high Fe concentrations can result in precipitation of Fe hydroxides on top of the compost layer which can result in a reduction in the permeability of the compost, reducing flow rates through the compost and plugging the system. Solutions can include installing settling ponds before the RAPS to remove Fe, or Fe hydroxides can be removed from the top of the compost periodically.
AMD with high concentrations of Al (>20 mg/L) can result in plugging of the limestone pore spaces with Al precipitate and/or decreased limestone dissolution rates due to coating of the limestone with Al precipitates. Aluminium does not form sulphides, but rather precipitates as a hydroxide at a pH of about 5 (Nordstrom and McCleskey 2006). Although not to the same extent as Fe hydroxides, Al hydroxides can armour limestone to some extent and potentially reduce dissolution rates (Hammarstrom et al. 2003). To reduce problems with Al, RAPS can be constructed with automatic flushing systems or be flushed manually more frequently (Kepler and McCleary 1997; Watzlaf et al. 2002; Weaver et al. 2004).
Preferential pathways can form resulting in short circuiting of the water through the compost and limestone and into the underdrain without the required residence time for treatment. This can be prevented through proper design and construction.
If RAPS are under-designed for the flow rate and acidity levels, inadequate treatment can result. Systems are typically designed for at least 15 h residence time in the limestone layer; however, there is also an acidity treatment limit of 35-40 g/m²/d (Rose 2004).

Operation and maintenance for RAPS consist of regularly flushing the systems to remove built-up precipitates and capture them in a holding pond. Piping and valves may need cleaning or replacement with time.

Sizing for RAPS is based both on residence time and acidity removal rates. Residence time in the limestone layer should be at least 15 h (Zipper and Jage 2001; Watzlaf et al. 2004). Jage et al. (2001) found that alkalinity generation in a RAPS is at first rapid but decreases over time, with the response of alkalinity generation to residence time logarithmic. Design recommendations for acidity removal rates range from 25-30 g/m²/day (Watzlaf et al. 2004) to 35-40 g/m²/day (Rose 2004).

Typical systems are designed with 13-30 cm compost underlain by 0.5-1 m limestone (Kepler and McCleary 1994), although Watzlaf et al. (2003) suggest compost thicknesses between 15 and 60 cm and Demchak et al. (2001) suggest at least 50 cm of compost. Thicker layers are important for highly oxidised AMD. Limestone grain size should be between 6 and 15 cm. Details and guidelines on designing systems can be found in Kepler and McCleary (1994), Skousen et al. (1998), and Skovran and Clouser (1998).

Figure F24. Vertical-flow wetland passive treatment system.

Sulphate-reducing bioreactor

Bioreactors are similar to VFWs except that the neutralising material (usually limestone) is homogeneously mixed with the organic material and there are no plants growing in the substrate (Gusek 2002, 2004; Neculita et al. 2007). Bioreactors are relatively new to the arsenal of passive treatment for AMD.

As with VFWs, water flows vertically through the bioreactor (usually down, but sometimes up) and AMD is treated through sulphate reduction reactions and dissolution of limestone (Figure F25). The water leaves the system via a network of drainage pipes at the base of the system, and the outlet consists of a riser to maintain head within the system.

Metal removal is dominantly through reduction as metal sulphides through the activity of sulphate-reducing bacteria (Nordwick et al. 2006). Microbial sulphate reduction occurs in the presence of an oxidisable carbon source (organic material in the bioreactor), which results in the formation of bicarbonate and hydrogen sulphide (H₂S). The H₂S then combines with metals to form mono- and di-sulphides (such as FeS and FeS₂). Compared with diffuse amorphous Fe oxyhydroxides, Fe mono- and di-sulphides are compact and dense precipitates, which would therefore hinder permeability to a lesser extent (Skousen et al. 2000). The removal of metals from solution results in a lowering of the mineral acidity, and excess bicarbonate produced through sulphate reduction will consume proton acidity, thereby raising the pH (Hedin and Nairn 1992). Dissolution of carbonates within the substrate material will also generate alkalinity. Bioreactors, therefore, are good in situations where multiple metals (such as Cu, Cd, Zn) are present. Aluminium does not precipitate as a sulphide but rather as an Al hydroxysulphate (Thomas and Romanek 2002).

Other metal removal processes that can operate include ion exchange of metals by an organic-rich substrate, precipitation of metal hydroxides, adsorption of metals by precipitated Fe hydroxides, precipitation of carbonates, and precipitation of silicates (Gusek 2002, 2004; Nordwick et al. 2006).

Gusek (2002) provides a list of the benefits of bioreactors over other systems. They can treat low pH water containing multiple metals, they are resilient to metal loading, and they are less prone to plugging by Al since Al hydroxysulphate is more compact that Al hydroxide.

The main limitation of bioreactors is the potential for short circuiting through the unit (Gusek 2004). Careful selection of material and complete mixing can help to prevent this problem. Also, constructing multiple smaller systems is better than one large system.

Very little operation and maintenance is required for bioreactors. Piping and valves may need cleaning or replacement with time. Inflow and outflow devices/channels may need to be regularly cleared of debris.

Wildeman et al. (2006) recommend a design criterion of 0.3 moles of metal removal/m³ of substrate/day for bioreactors with a mixture of organic materials and crushed limestone. Gusek (2002, 2004) provides details on design and materials for bioreactors. Organic material can include wood chips, hay and straw (spoiled), sawdust, cardboard, rice hulls, soy bean hulls, yard waste, waste alcohols including antifreeze, mushroom compost, waste dairy products, animal manure, and sugar cane processing residue (bagasse). McCauley et al. (2009) used mussel shells for the neutralising material in laboratory trials of bioreactors treating low pH AMD containing multiple metals.

Figure F25. Sulphate-reducing bioreactor passive treatment system.

F.4 AMDTreat

AMDTreat is a free Windows-based computer program designed to estimate the capital and annual costs to treat mine discharges (Means et al. 2003). It was developed cooperatively by the Pennsylvania Department of Environmental Protection (http://www.dep.state.pa.us/), the West Virginia Department of Environmental Protection (http://www.dep.state.wv.us/), and the U.S. Office of Surface Mining Reclamation and Enforcement (http://www.osmre.gov/). It can be downloaded from the AMDTreat Home Page (http://amdtreat.osmre.gov/). Version 4.1C was released on 4 February 2008.

AMDTreat uses a three-step approach to estimate treatment costs:

Users enter water quality and quantity data.
Users ‘build’ an active and/or passive treatment system by selecting the applicable treatment components from the software menu.
Users customise each treatment system to site-specific conditions by controlling the size, quantity, and unit cost of treatment components.

The water quality and quantity data to be entered include: acidity, alkalinity, total Fe, ferric Fe, ferrous Fe, filtered Fe, total Al, filtered Al, total Mn, filtered Mn, sulphate, pH, specific conductivity, total dissolved solids, DO, typical flow, design flow. Not all of the parameters are used in the calculations (the program only requires acidity, alkalinity, total Fe, ferrous Fe, total Al, total Mn, sulphate, pH, typical flow, and design flow). If acidity is not known, an acidity calculator can be used to estimate the value.

The passive treatment systems that can be constructed using the program include vertical flow ponds, anoxic limestone drains, anaerobic wetlands, aerobic wetlands, manganese removal beds, oxic limestone channels, limestone beds, and bioreactors. Active treatment systems that can be constructed include caustic soda, hydrated lime, pebble quicklime, ammonia, oxidation chemicals, and soda ash. The program does not help the user to choose the most appropriate solution for a given water chemistry and flow rate. It is up to the user to identify which solutions may or may not work for the given site. The program then can be used to compare relative costs between different types of systems (Table F4).

To help select among appropriate active treatment systems, Means (2006) provides specific details for each chemical, including: system description, typical set-up, chemical facts, benefits, limitations, maintenance requirements, and economics.

The model combines costs from these treatment methods with costs of ancillary treatment components, such as settling ponds, roads, land access, ditching, and engineering costs to calculate a site-specific capital cost. Similarly, AMDTreat calculates annual costs by taking into account user-provided information regarding sampling, labour, maintenance, pumping, chemical consumption, and sludge removal. Capital and annual costs can be used in conjunction with AMDTreat’s financial forecasting utility to evaluate the economics of long-term treatment. Over 400 user-modifiable variables are provided for modelling costs.

AMDTreat also contains several scientific tools to help select and plan treatment systems. These tools include an acidity calculator, a sulphate reduction calculator, a mass balance calculator, and a passive treatment alkalinity calculator.

Table F4. Hypothetical AMD chemistry used for comparison of passive treatment system construction costs using the program AMDTreat. Selected chemistry can be treated by any passive treatment system. For construction costs see Table F3 and Figure 21 in main text.

Acidity	150	mg/L
Flow rate	10	L/s
Al	12	mg/L
Fe total	11	mg/L
Ferrous	10	mg/L
Ferric	1	mg/L
pH	2.9
DO	1	mg/L
Acid load	52.1	tonnes/year

F.5 Acidity and treatment

The acidity of a solution, not the pH, is probably the best single indicator of the severity of AMD (Rose and Cravotta 1998). The hydrolysis of metals in AMD, as the pH is raised, produces hydrogen ion acidity as shown in the following reactions (Stumm and Morgan 1996):

Fe²⁺ + 0.25 O₂ + H⁺ = Fe³⁺ + 0.5 H₂O

Fe³⁺ + 3 H₂O = Fe(OH)₃(s) + 3 H⁺

Al³⁺ + 3 H₂O = Al(OH)₃(s) + 3 H⁺

Mn²⁺ + 0.5 O₂ + H₂O = MnO₂(s) + 2 H⁺

Therefore, acidity in AMD is comprised of mineral acidity (the hydroxide ion demand by cations of Fe, Al, Mn and others) and hydrogen ion acidity (measured as pH units). Acidity can either be measured in a laboratory through the hot acidity procedure in which hydrogen peroxide and heating are used to oxidise and hydrolyse metals, followed by titration with base to a pH of 8.2 or 8.3 (USEPA 1979; APHA 1980; ASTM 1994), or it can be calculated using the concentrations of dissolved metals and pH according to the following formula (Hedin et al. 1994):

Acidity = 50 [2Fe²⁺/55.85 + 3Fe³⁺/55.85 + 3Al³⁺/26.98 + 2Mn²⁺/54.94 + 1000(10^-pH)]

(Fe²⁺, Fe³⁺, Al³⁺, Mn²⁺ are dissolved concentrations in mg/L)

Although Fe²⁺, Fe³⁺, Al³⁺, Mn²⁺ and H⁺ are the major components of acidity in coal-mine drainage (Ott 1986), other dissolved species that precipitate as hydroxides or oxides or change form during the acidity titration, including Mg²⁺, H₂CO₃, or H₂S, can contribute to acidity (Payne and Yeates 1970). Therefore, if possible, the laboratory acidity measurement is preferable to the calculation method.

Acidity is commonly expressed as milligrams of CaCO₃ per litre of solution (mg/L as CaCO₃) on the basis of the following stoichiometric relation:

2 H⁺ + CaCO₃ = Ca²⁺ + CO₂ + H₂O

F.5.1 Sequential titration methodology

Recent work has shown that often there is poor agreement between calculated acidity and laboratory-measured acidity and that sometimes neither method adequately predicts the amount of neutralising material necessary to neutralise the AMD and lower the concentrations of dissolved metals to acceptable levels (Means and Hilton 2004). Hilton (2004) suggests that multiple titrations are conducted in the field to determine the amount of neutralising material needed to treat the AMD to a specified pH level or specified concentrations of dissolved metals. This methodology is known as sequential titration. The procedure lends itself to field applications for instantaneous results, or for laboratory analysis where more detailed analyses are required.

In the methodology, raw AMD water is repeatedly titrated with NaOH. Each sequential titration is conducted with an increasing amount of NaOH and each titration is analysed for dissolved metals and pH. Once the goals for water quality treatment are determined, these titration data can be used to design an appropriately sized treatment system. See Trumm and Cavanagh (2006) for an example of the titration conducted at a New Zealand AMD site.

F.6 Small-scale trials for AMD passive treatment

For sites with existing AMD where passive treatment is being considered, data can be collected from the AMD over a period of time to add confidence to the selection of potential treatment techniques. Once potential treatment solutions have been identified through the use of the flow chart presented in the main text, Trumm (2007) recommends that small-scale trials be constructed on site to test the effectiveness of the various options before investing in full-scale system construction (see Trumm et al. 2006, 2008, 2009; McCauley et al. 2008 for examples of small-scale trials in New Zealand). Even if only one option is indicated through the use of the flow chart, field trials should still be conducted because unknown factors can influence the effectiveness of treatment systems. The choice of the full-scale system should be based on the results of the field trials and a review of the cost, effectiveness, limitations and risk of failure for each option. The following sections detail methodologies for small-scale trials.

F.6.1 Collect data and determine remedial action objectives

If possible, AMD water chemistry and flow rate should be measured monthly for at least 12 months. Water chemistry parameters should include the field measurement of pH, conductivity, temperature, dissolved oxygen, and ferrous Fe concentration, and laboratory analyses for total acidity, Fe and Al concentrations (dissolved and total), and sulphate. Any additional metals or metalloids should be added to the list of analytes that are at concentrations of concern. A sequential titration (Hilton 2004) should be conducted at least once (see section above).

Flow rate should be determined by either manual measurement on the same day as water chemistry is measured or preferably by data logger on a continuous basis. If possible, additional water quality and quantity data should be collected during high and low flow events to enable determination of flow variation on water chemistry.

Using the acceptable level of impact (from stakeholders), analyse the data to determine the levels by which metals should be reduced and the required increase in pH. Determine the acid load (kg/day) above acceptable limits and validate the calculations with the treatment level determined from the titration results (NaOH equivalent to lower concentration of metals and raise pH to acceptable levels).

F.6.2 Design and construct pilot trials

Using water chemistry data (Fe and Al concentrations, DO) and available land area, identify potential passive treatment solutions with the flow chart in the main text (there is often more than one). Further verify appropriate solutions by reviewing descriptions, examples, effectiveness, limitations, risk of failure, and relative cost of systems in the section AMD Passive Treatment Systems later in this appendix.

For each potential solution being considered for the site, design and construct a pilot trial to be operated at the site or to be operated in a controlled laboratory situation. If the trials are to be operated in a laboratory, AMD water from the site should be transported to the laboratory and sent through the systems. Three factors need to be considered when designing and constructing the systems:

Size and dimensions of the trials
Treatment media specifications for each system
Amount of AMD to be treated for each system.

The budget and availability of materials are typically used to determine the size and dimensions of the small-scale systems to be constructed. Typical sizes range from 0.5 m³ (Trumm et al. 2008) to 1 m³ (McCauley et al. 2008; Trumm et al. 2009), and containers and piping are usually plastic. For treatment media type, particle size, and volume, refer to system details in the section AMD Passive Treatment Systems later in this appendix. If particle sizes in the trials are much smaller than those recommended for full-scale systems, grain size surface area should be considered when later analysing the data for system performance and scaling up to full size. To determine the amount of AMD that can be treated by each system, the computer program AMDTreat can be used (see previous section above) and formulas for designing each system should be consulted (see section AMD Passive Treatment Systems below). Piping and valves are used to convey only the amount of AMD needed to pass through each system.

F.6.3 Operate pilot trials and collect data

During operation of the pilot trials, data should be collected on a regular basis according to the duration of the trials to monitor system effectiveness with time and variations in inlet water chemistry (weekly, fortnightly, or monthly). Typical trial durations are 3–6 months. Data should include the field parameters: pH, conductivity, temperature, DO, ferrous Fe concentration and flow rate, and samples should be collected for laboratory analysis of Fe and Al (dissolved and total concentrations), sulphate, sulphides (where applicable), acidity, and any other metals or metalloids of concern.

Ecotoxicity experiments should be conducted using treated water to verify that treatment will enable restoration of the aquatic ecosystem (for examples of ecotoxicity experiments see Trumm et al. 2003; Trumm and Gordon 2004).

To determine system limits and effectiveness, flow rates (and therefore, residence times) should be varied during the duration of the trial.

For systems with flushing mechanisms, such as vertical flow wetlands and leaching beds, the pilot trials should be flushed at least once to test flushing effectiveness at removing accumulated contaminants (for piping design and sampling during flushing see Danehy et al. 2002; Watzlaf et al. 2002; Weaver et al. 2004; Trumm et al. 2008).

F.6.4 Analyse data and select full-scale system

Once operation of the systems has ceased, the system should be deconstructed (autopsy) and additional data should be collected. This includes photographing treatment media, analysing armouring on treatment media using SEM or other applicable methods (see Hammarstrom et al. 2003), and analysing treatment media for contaminants of concern removed during the trial.

Data analysis from the trials should include the following:

Calculate residence times in the systems;
Calculate contaminant removal rates and compare to residence times and duration of system operation;
Compare contaminant removal rates per unit surface area of treatment media to performance predicted by models (such as AMDTreat and those specified in the section AMD Passive Treatment Systems earlier in this appendix);
Determine system lifespan, if possible, by comparing treatment media before and after the pilot trial and analysing applicable analytical results to determine consumption of neutralising material in kg/day;
Analyse data collected during flushing, and from system deconstruction to determine effectiveness of flushing mechanisms in removing accumulated contaminants;
Analyse system deconstruction data to determine where contaminants accumulated; and
Determine if treatment was effective in reducing toxicity of untreated water.

For examples of data analysis and interpretation of pilot trials see Trumm et al. (2006, 2008, 2009) and McCauley et al. (2008).

The results of pilot trials or laboratory experiments are used indicate the effectiveness of a given system in meeting the treatment goals. Selection of a full-scale treatement system is based on the effectiveness of that system, in addition to ease of implementation and cost. Ease of implementation considers the practicability of implementing the different treatment systems and can include consideration of the site-specific constraints (such as space availability, terrain, etc.), as well as the experience of personnel involved in the construction of different systems. Cost should be an order-of-magnitude cost estimate to allow comparison of the alternatives from a cost standpoint. These costs should approximate the total implementation cost, which would include design, construction, operation, and engineering services during construction. These estimates are developed using quotations by vendors and estimating values.

F.7 Arsenic treatment - hard rock gold mines

F.7.1 Active treatment systems

There are many active treatment technologies that can be used to remove As from water. The following discussion is largely based on USEPA (2002). A brief description is provided below, with further details, including benefits and limitations of the technology provided in Table F5.

Oxidation

Oxidation of As³⁺ to As⁵⁺ improves the performance of As removal technologies. It often forms a pretreatment step or may be an intrinsic part of another technology, although it is not typically used alone as an As treatment. Technologies that are typically considered as As removal processes are:

Coagulation/precipitation

This process uses chemicals to transform dissolved contaminants into an insoluble solid, either through direct precipitation or adsorbtion onto another species that is precipitated (co-precipitation). The precipitated solid is then removed from the liquid phase by clarification or filtration. Conventional chemicals used include Al and Fe hydroxides, ferric salts (e.g. ferric sulphate, ferric chloride) ferrihydrite, lime and limestone. The effectiveness of this technology is less likely to be reduced by characteristics and contaminants other than As, compared with other water treatment technologies. It is also capable of treating water characteristics or contaminants other than As, such as hardness or heavy metals. Systems using this technology generally require skilled operators; therefore, precipitation/coprecipitation is more cost effective at a large scale where labour costs can be spread over a larger amount of treated water produced. Chemical costs may also influence costs.

This technique is the most widely used active treatment technology in the US for As removal from a variety of As-contaminated waters (USEPA 2002). This process is also used in New Zealand, for the treatment of As-contaminated water at OceanaGold’s Globe Progess Mine at Reefton (below).

Case study: Active treatment at Globe Progress Mine, Reefton

OceanaGold’s Globe Progress Mine at Reefton uses a coagulation/precipitation process and treats water with As concentrations up to 2.8 mg/L. A schematic diagram of the treatment plant is shown in Figure F26. In the plant all water potentially containing As (mine water, tailings return water, waste-rock seepage) and plant sump return water is initially fed into an oxidation tank to which Fe chloride is added in small amounts. Pilot trials were used to determine the amount of Fe chloride to be added. This tank is aerated by an agitator and air injection. The water then enters the second tank (precipitation tank) where lime is added with mixing, to adjust the pH to the target of 7.0. Flocculent overflow from the clarifier and raw water are added to the tank. Flocculent is added to the water overflow from the precipitation tank, which then enters the clarifier. The solids settle out and are pumped from the bottom into a tailings dam. The clean water is pumped into a discharge lagoon that overflows to the nearby stream. The stream is monitored regularly for As and is required to remain below a trigger level set by the regional council (maximum As 3.3 mg/L, 90^th percentile As 0.25 mg/L).

Figure F26. Schematic diagram of the water treatment plant at OceanaGold’s Globe Progress Mine at Reefton.

Adsorption

Adsorption techniques are typically based on contaminated water passing though a fixed bed of media. Dissolved contaminants concentrate at the surface of the media (sorbent), thereby reducing their concentration in the water. As contaminated water is passed through the column, contaminants are adsorbed. When adsorption sites become filled, the column must be regenerated or disposed of and replaced with new media. Competition for adsorption sites could reduce the effectiveness of adsorption because other constituents may be preferentially adsorbed, resulting in a need for more frequent bed regeneration or replacement. Phosphates and silicates are reported to compete with As⁵⁺, while sulphate and chloride will compete more favourably with As³⁺.

Common adsorbents are activated alumina, activated carbon, Fe-based adsorption media (including granular ferric hydroxide, Fe-oxide-coated sand and Fe ferrihydrite). Of these, Fe-rich absorbents may be particularly useful in New Zealand as they are reasonably cheap and abundant, and often available naturally (Fe-Mn ore and hydrated Fe oxides from Fe-rich waters) or as a by-product from mining and processing such as AMD sludge (amorphous Fe oxide).

Ion exchange

Ion exchange is a physical/chemical process in which ions held electrostatically on the surface of a solid are exchanged for ions of similar charge in a solution. This process typically uses a resin made from synthetic organic materials, inorganic materials, or natural polymeric materials that contain ionic functional groups to which exchangeable ions are attached. Specifically, ions are removed from the aqueous phase by the exchange of cations or anions between the contaminants and the exchange medium. Ion exchange resins must be periodically regenerated to remove the adsorbed contaminants and replenish the exchanged ions.

Membrane filtration

In this technique, As is removed by passing contaminated water through a semi-permeable membrane or barrier. There are four types of membrane processes, categorised by the size of the particles that pass through the membranes: reverse osmosis (RO), nanofiltration (NF), ultrafiltration (UF), and microfiltration (MF). All four are pressure-driven and the force required to drive fluids across the membranes depends on the pore size; NF and RO require a relatively high pressure (50–150 pounds per square inch [psi]), while MF and UF require a relatively low pressure (5–100 psi). The low-pressure processes primarily remove contaminants through physical sieving and the high-pressure processes primarily remove contaminants through chemical diffusion across the permeable membrane. These processes are often expensive and membranes are prone to fouling due to the small pore sizes, and it is more likely to be used as a finishing step for As removal.

Biological

Biological treatment for As removal is an emerging technique that uses biological activity to promote precipitation of As from water. Arsenic removal may occur either as a result of biological activity, typically using microbes, creating ambient conditions that result in As precipitating or by transforming As species into those that are more amenable to precipitation. The microbes may be suspended in the water or attached to a submerged solid substrate. Iron or hydrogen sulphide may also be added.

F.7.2 Passive treatment systems

There are few examples of passive treatment of mining water contaminated with As in the international literature. Treatment systems typically use Fe-rich adsorbants, and Fe oxyhydroxides as these are naturally formed. For example, The Wheal-Jane constructed wetland can remove As (initial As <2 mg/L) by precipitation of amorphous Fe oxyhydroxides in acidic conditions (pH 3-4) (Swash and Monhemius 2005). This system consists of a pH-conditioning cell, 5 aerobic cells, 1 anaerobic cell and 9 rock filtration cells in which As is removed in the aerobic cells and rock filter (Swash and Monhemius 2005). Wildeman et al. (2006) describe benchtop passive treatment of As (initial concentration 129.8 mg/L) and cyanide-rich water at a gold mine site in Brazil by zerovalent Fe.

More recently, passive treatment of As using AMD sludge (which contains Fe oxyhydroxides) has been investigated in New Zealand and has provided promising results. This research is described below.

Case study: Passive treatment of As-contaminated water at Waiuta, West Coast, using AMD sludge

The efficiency of AMD sludge at removing As from surface water was investigated in laboratory and field experiments. AMD sludge was obtained from active treatment of AMD at a working coal mine (Stockton, ST) and natural precipitation from untreated AMD at the abandoned Blackball Coal Mine (BB) on the West Coast, New Zealand. The sludge contained Fe oxide at concentrations of 13 wt% (Stockton sludge) and 74 wt% (Blackball sludge) respectively. Water used for laboratory trials was obtained from a contaminated gold mine site at Waiuta, West Coast, New Zealand (Haffert and Craw 2008a, b). Laboratory batch and column experiments were initially performed to determine the ability of AMD precipitate to remove As from mine drainage. Subsequently, a field trial was established at the Waiuta site. Further details are available in Rait et al. (in press).

Laboratory trials

Batch and column experiments were performed in the laboratory to determine the ability of acid mine drainage (AMD) precipitate to remove As from mine drainage. Water used in the laboratory experiments contained up to 99 mg/L As. In the initial batch experiments rapid reduction in As concentrations occurred within 15 h, with BB sludge reducing As levels below that for ST sludge. After 48 h at a ratio of 10 g sludge to 1 L water As concentrations were lowered from 99 to 0.016 mg/L (BB) and 0.55 mg/L (ST). For a higher ratio of 50 g sludge to 1 L water As concentrations after 48 h were lowered to 0.0017 mg/L (BB) and 0.008 mg/L (ST). A column leaching experiment was then conducted to determine long-term adsorption potential. Water (at average concentrations of 40 mg/L As) was passed through columns, at an average flow rate of 1.8 L/day, with sand coated in the powdered AMD sludge for 10 days at a constant contact residence time of 1 day. Two columns were packed at a ratio of 50 g sludge (ST and BB) to 1 L water and one at a ratio of 10 g sludge (BB) to 1 L water. All columns lowered As concentrations in the effluent to <0.01 mg/L for the first 2 days. However, after 2 days the columns with ST sludge and BB sludge at a ratio of 10:1 showed a steady increase in effluent As concentrations, suggesting that adsorption sites were being exhausted. The column with the BB sludge at a ratio of 50:1 continued to have effluent As concentrations of <0.01 mg/L for 9 days, only increasing to 0.04 mg/L on day 10, suggesting many more available adsorption sites on the naturally-precipitated AMD sludge. A toxicity characteristic leaching procedure test on the BB sludge/sand mixture packed at the 50:1 ratio showed <0.021 mg/kg As, indicating relatively good stability of the adsorbed As. These results suggest that As can be treated with AMD sludge if the ratio of sludge to water and contact residence time are optimised.

Field trials

Based on the above laboratory trials, a small-scale field trial was established at the Waiuta site. At this location, water from a wetland flows through a manmade dam and emerges on the other side at a flow rate of about 2 L/s and As concentrations about 2.4 mg/L. The dam acts as a natural As removal system reducing concentrations from 52 (wetland) to 2.4 mg/L (Haffert and Craw 2008a). The field trial was constructed using 1000-L plastic tubs (filled with sand/AMD mixtures), PVC piping and valves. Three tubs were set up: one with ST sludge and two with BB sludge (BB-1andBB-2) at loading ratios of 139, 99 and 99 g sludge/L water, respectively. A flow rate of 0.006 L/s was used as the primary flow rate, which provided the optimal residence time of 20 h. Water inflow and outflow were sampled on a daily basis for the first 3 days, and weekly thereafter for 11 weeks. The influence of residence time on As removal was examined by varying the inflow rate, which is inversely proportional to residence time, for BB-2 to assess As removal at 1x, 2x, 3x, and 4x the optimal residence time. Longer residence times (i.e. closer to 20 h) were more effective in lowering As concentrations. At present the trials have run for 11 weeks. One outcome yet to be addressed is how long the sludge can remain effective. Fortnightly sampling will be continued over the next year to determine this. Analysis of results will provide information on the effectiveness of the treatment system and allow for later design of a full-scale remediation system.

Table F5. Summary of arsenic removal technologies

Technology	Techniques	Chemicals and benefits	Limitations	References
Oxidation	Air oxidation, chemical oxidation, photo catalytic, electrochemical preoxidation	Mainly used in conjunction with other technologies as oxidises As(III) to As(V) for removal. By stirring or cascade in ponds. Relatively simple, low-cost but slow process. Also oxidises other inorganic and organic constituents in water. Simple, rapid process. Minimum residual mass. Chemicals include calcium hypochlorite, ozone, bleach, Mn oxide, permanganate, chlorine. Fe(VI) electrochemical oxidation of As(III). Can be used on a wide range of water contaminants.	Mainly removes As(V) and accelerates the oxidation process. Efficient control of the pH and oxidation step is needed. Can be costly. Interfering reductants can inhibit. As sludge needs to be disposed.	Buisson et al. (1979), Hug et al. (2001), Arienzo et al. (2002), Lee and Choi (2002), Ghurye and Clifford (2004), Dutta et al. (2005), Light and Yu (2005), Moore (2005), Pokhrel et al. (2005), Swash and Monhemius (2005), Ferguson and Hering (2006), Yuan et al. (2006), Zhang and Itoh (2006), Habuda-Stanic et al. (2007)
Coagulation/ precipitation	Al coagulation, Fe coagulation, ferric ions plus calcite, lime and limestone	Low capital cost. Simple process. Al and Fe hydroxide. Effective over a wider range of pH. Common chemicals are available, ferric sulphate, ferric chloride, ferrihydrite. Medium to high removal of As(III). Fe₂(SO₄)₃.5H₂O with calcite to enhance removal. Most common chemicals are available commercially. Granulated lime, calcium carbonate, limestone.	Produces sludge. Disposal of sludge required. Anaerobic bacteria can reduce ferrihydrite to mobilise As. Competing species. Sedimentation and filtration needed	Buisson et al. (1979), Meng et al. (2001), Chen et al. (2002), Brown et al. (2003), Wickramasinghe et al. (2004), Dousova et al. (2005), Tadanier et al. (2005), Davis et al. (2006), Parks and Edwards (2006), Song et al. (2006), Violante et al. (2006), Yuan et al. (2006), Lee et al. (2007), Masue et al. (2007)
Adsorption	Activated alumina (AA), activated carbon, Fe-coated sand, Fe oxides, ferrihydrite, zero valent iron, Fe-Mn oxides,	Relatively well known and commercially available. Manganese amended AA higher removal efficiencies Granular Activated Carbon-Fe (GAC) Expected to be cheap. Ferralite used to make simple household filters. Remove both As(III) and As(V). Natural red earth-natural Fe-coated sand. Low capital costs.	Arsenic-rich sludge disposal. Phosphates and silicates inhibit removal. Coating of humic acid on surface of natural Fe-Mn nodules retard As(III) adsorption. Mn may leach from GAC.	Viraraghavan et al. (1999), Altundogan et al. (2000), Ramaswami et al. (2001), Su and Puls (2001), Altundogan et al. (2002a, 2002b), Chakravarty et al. (2002), Manning et al. (2002), Petrusevski et al. (2002), Selvin et al. (2002), Wang et al. (2002), Bhattacharyya et al. (2003), Kunzru and Chaudhuri (2005), Cheng et al. (2005), Dutta et al. (2005), Kanel et al. (2005, 2006), Leupin and Hug (2005), Pokhrel et al. (2005), Chen et al. (2006), Ferguson and Hering (2006), Gillman (2006), Kumpiene et al. (2006), Makris et al. (2006), Nguyen et al. (2006), Ramakrishna et al. (2006), Sun et al. (2006), Wildeman et al. (2006), Zhang and Itoh (2006), Kundu and Gupta (2007), Guo et al. (2007a, 2007b), Mondal et al. (2007), Zhang et al. (2007)
Membranes, reverse osmosis, exchange resin	Nanofiltration, ultrafiltration, reverse osmosis, electrodialysis, compost-based permeable reactive barrier, on exchange membrane bioreactor	Nanofilitration is well-defined and high-removal efficiency. Pre-oxidation required. Micellar enhanced ultrafiltration with high molecular weight cut-off membranes.	High-cost medium. Not very economic if high TDS. Disposal of residuals. Toxic wastewater produced. Requires high-tech operation and maintenance. May be retarded by other trace metals	Ning (2002), Wang et al. (2002), Clifford et al. (2003), Moore (2005), Köber et al. (2005), Nguyen et al. (2006), Oehmen et al. (2006) Ramakrishna et al. (2006), Beolchini et al. (2007), Habuda-Stanic et al. (2007)
Biological	Sulphate-reducing bacteria, plant accumulation, oxidation, constructed wetlands	Passive. Low concentrations. Aided by phosphates. Aerobic oxidation and adsorption on Fe oxyhydroxides.	Can release As. Disposal of solid phase containing As. Further investigations required to obtain higher removal efficiencies. Large area needed.	Gammons and Fradsen (2001), Lehimas et al. (2001), Katsoyuannis et al. (2002), Jong and Parry (2005), Swash and Monhemius (2005), Zouboulis and Katsoyiannis (2005), Fayiga and Ma (2006), Anderson and Walsh (2007)

F.8 References

Altundogan HS, Altundogan S, Tumen F, Bildik M 2000. Arsenic adsorption from aqueous solutions by adsorption on red mud. Waste Management 20: 761-767.

Altundogan HS, Altundogan S, Tumen F, Bildik M 2002a. Arsenic adsorption from aqueous solutions by activated red mud. Waste Management 22: 357-363.

American Public Health Association (APHA) 1980. Standard methods for the examination of water and wastewater (15th). Washington, DC: American Public Health Association. Pp. 2-30 to 2-34.

American Society for Testing Methods (ASTM) 1994. Standard test methods for acidity or alkalinity of water. Annual Book of ASTM Standards, v. 11.01, Method D1067-92. Pp. 257-263.

Anderson L, Walsh MM 2007. Arsenic uptake by common marsh fern Thelypteris palustris and its potential for phytoremediation. Science of the Total Environment 379: 263-265.

Anonymous 2001. The science of acid mine drainage and passive treatment. Pittsburgh, PA: Pennsylvania Department of Environmental Protection. Available at: http://www.depweb.state.pa.us/abandonedminerec/cwp/view.htm?a=1466&q=457768.

Arienzo M, Adamo P, Chiarenzelli J, Bianco MR, De Martino A 2002. Retention of arsenic on hydrous ferric oxides generated by electrochemical peroxidation. Chemosphere 48: 1009-1018.

Arnold DE 1991. Diversion wells - A low-cost approach to treatment of acid mine drainage. Proceedings, Twelfth West Virginia Surface Mine Drainage Task Force Symposium, 3-4 April 1991, Morgantown, WV .

Bamforth SM, Manning DAC, Singleton I, Younger PL, Johnson KL 2006. Manganese removal from mine waters - investigating the occurrence and importance of manganese carbonates. Applied Geochemistry 21: 1274-1287.

Batty LC, Younger PL 2002. Critical role of macrophytes in achieving low iron concentrations in mine water treatment wetlands. Environmental Science & Technology 36: 3997-4002.

Beolchini F, Pagnanelli F, De Michelis I, Veglio F 2007. Treatment of concentrated arsenic(V) solutions by micellar enhanced ultrafiltration with high molecular weight cut-off membrane. Journal of Hazardous Materials 148: 116-121.

Bhattacharyya R, Jana J, Nath B, Sahu SJ, Chatterjee D, Jacks G 2003. Groundwater As mobilization in the Bengal Delta Plain, the use of ferralite as a possible remedial measure - A case study. Applied Geochemistry 18: 1435-1451.

Bigham JM, Nordstrom DK 2000. Iron and aluminum hydroxysulfates from acid sulfate waters. In: Alpers CN, Jambor JL, Nordstrom DK eds Sulfate minerals. Reviews in Mineralogy and Geochemistry 40. Washington, DC: Mineralogical Society of America. Pp. 351–403.

Black C, Ziemkiewicz P, Skousen J 1999. Construction of a limestone leach bed and preliminary water quality results in Beaver Creek. Morgantown, WV. Proceedings, 20th West VirginiaSurface Mine Drainage Task Force Symposium.

Bostick BD, Hansel CM, LaForce MJ, Fendorf S 2001. Seasonal fluctuations in zinc speciation within a contaminated wetland. Science Technology 35: 3823-3829.

Brodie GA, Britt CR, Tomaszewski TM, Taylor HN 1993. Anoxic limestone drains to enhance performance of aerobic acid drainage treatment wetlands: experiences of the Tennessee Valley Authority. In: Moshiri GAed. Constructed wetlands for water quality improvement. Boca Raton, FL: Lewis. Pp. 129-145.

Brown CJ, Eaton RA, Cragg SM, Goulletquer P, Nicolaidou A, Bebianno MJ, Icely J, Daniel G, Nilsson T, Pitman AJ, Sawyer GS 2003. Assessment of effects of chromated copper arsenate (CCA)-treated timber on nontarget epibiota by investigation of fouling community development at seven European sites. Archives of Environmental Contamination and Toxicology 45: 37-47.

Brown H, Skousen J, Renton J 1994a. Floc generation by chemical neutralization of acid mine drainage. Green Lands 24(1): 44-51. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd edn. West Virginia University, Morgantown, WV, USA. Pp. 217-224.

Brown H, Skousen J, Renton J 1994b. Stability of flocs produced by chemical neutralization of acid mine drainage. Green Lands 24(3): 34-39. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University, Morgantown, WV, USA. Pp. 231-236.

Brown H, Skousen J, Renton J 1994c. Volume and composition of flocs from chemical neutralization of acid mine drainage. Green Lands 24(2): 30-35. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd edn. West Virginia University, Morgantown, WV, USA. Pp. 225-230.

Buisson DH, Rothbaun HP, Shannon WT 1979. Removal of arsenic from geothermal discharge waters after absorption ion iron floc and subsequent recovery of the floc using dissolved air flotation. Geothermics 8: 97-110.

Cementa M 1983. Liming running waters. Malmo, Sweden: Cementa AB. 16 p.

Chakravarty S, Dureja V, Bhattacharyya G, Maity S, Bhattacharjee S 2002. Removal of arsenic from groundwater using low cost ferruginous manganese ore. Water Research 36: 625-632.

Chen ASC, Fields KA, Sorg TJ, Wang L 2002. Field evaluation of As removal by conventional plants. American Water Works Association Journal Pro Quest Science Journal 94(9): 64.

Chen Z, Kim KW, Zhu YG, McLaren R, Liu F, He JZ 2006. Adsorption (As^III,V) and oxidation (As^III) of arsenic by pedogenic Fe-Mn nodules. Geoderma 136: 566-572.

Cheng Z, Van Geen A, Louis R, Nikolaidis N, Bailey R 2005. Removal of methylated arsenic in groundwater with iron filings. Environmental Science and Technology 39: 7662-7666.

Clifford DA, Ghurye GL, Tripp AR 2003. As removal using ion exchange with spent brine recycling. American Water Works Association Journal Pro Quest Science Journal 95(6): 119.

Cravotta CA, Trahan MK 1999. Limestone drains to increase pH and remove dissolved metals from acidic mine drainage. Applied Geochemistry 14: 581-606.

Danehy TP, Hilton T, Watzlaf GR, Johnson F, Busler SL, Denholm CF, Dunn MH 2002. Vertical flow pond piping system design considerations. Proceedings, 2002 National Meeting of the American Society for Surface Mining and Reclamation, Lexington, Kentucky, 9-13 June 2002. Pp. 916-934.

Davis AD, Webb CJ, Dixon DJ, Sorenson JL, Dawadi S 2006. Arsenic removal from drinking water by limestone-based material. St Louis, MO: SME Annual Meeting. Pp. 1-4.

Demchak J, Morrow T, Skousen J 2001. Treatment of acid mine drainage by four vertical flow wetlands in Pennsylvania . Geochemistry: Exploration, Environment, Analysis, 1: 71-80.

Dousova B, Kolousek D, Kovanda F, Machovic V, Novotna M 2005. Removal of As(V) species from extremely contaminated mining water. Applied Clay Science 28(1-4 Special issue): 31-42.

Dutta PK, Pehkonen SO, Sharma VK, Ray AK 2005. Photocatalytic oxidation of arsenic (III): Evidence of hydroxyl radicals. Environmental Science and Technology 39: 1827-1834.

Evangelou VP 1998. Environmental soil and water chemistry. New York: John Wiley. 564 p.

Faulkner BB, Skousen, J 1991. Using ammonia to treat mine waters. Green Lands 21(1): 33-38. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 181-186.

Faulkner BB, Skousen, J 1996. Treatment of acid mine drainage by passive treatment systems. In: Skousen JG, Ziemkiewicz PF compilers Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 267-274.

Fayiga AO, Ma LQ 2006. Using phosphate rock to immobilize metals in soil and increase arsenic uptake by hyperaccumulator Pteris vittata. Science of the Total Environment 359: 17-25.

Ferguson MA, Hering JG 2006. TiO₂-photocatalyzed As(iii) oxidation in a fixed-bed, flow-through reactor. Environmental Science and Technology 40: 4261-4267.

Fraser JE, Britt DL, Kinsman JD, DePinto J, Rodgers P, Sverdrup HU, Warfvinge P 1985. APMP Guidance Manual - Volume II: Liming materials and methods. U.S.Fish and Wildlife Service, Eastern Energy and Land Use Team, Biological Report 80(40.25). 197 p.

Gammons CH, Fradsen AK 2001. Fate and transport of metals in H₂S-rich waters at a treatment wetland. Geochemical Transactions 2: 1-15.

Ghurye G, Clifford D 2004. As(III) oxidation using chemical and solid-phase oxidants. American Water Works Association Journal Pro Quest Science Journal 96(1).

Gillman GP 2006. A simple technology for arsenic removal from drinking water using hydrotalcite. Science of the Total Environment 366: 926-931.

Gusek JJ 2002. Sulfate-reducing bioreactor design and operating issues: is this the passive treatment technology for your mine drainage? In: Proceedings, the 2002 National Association of Abandoned Mine Land Programs Annual Conference. Park City, UT.

Gusek JJ 2004. Scaling up design challenges for large scale sulfate reducing bioreactors. In: Proceedings, National Meeting of the American Society of Mining and Reclamation and the 25^th West VirginiaSurface Mine Drainage Task Force. Morgantown, WV. Pp. 752-765.

Habuda Stanic M, Kules M, Kalajdzic B, Romic Z 2007. Quality of groundwater in eastern Croatia. The problem of arsenic pollution. Desalination 210: 157-162.

Haffert L, Craw D 2008a. Mineralogical controls on environmental mobility of arsenic from historic mine processing residues, New Zealand. Applied Geochemistry 23: 1467-1483.

Haffert L, Craw D 2008b. Processes of attenuation of dissolved arsenic downstream from historic gold mine sites, New Zealand. The Science of the Total Environment 405: 286-300.

Hammarstrom JM, Sibrellb PL, Belkina HE 2003. Characterization of limestone reacted with acid-mine drainage in a pulsed limestone bed treatment system at the Friendship Hill National Historical Site, Pennsylvania, USA. Applied Geochemistry 18: 1705–1721.

Hammer DA 1992. Creating freshwater wetlands. Boca Raton, FL: Lewis. 298 p.

Hedin RS, Nairn RW 1992. Passive treatment of coal mine drainage. Course notes for workshop. Pittsburgh, PA: U.S.Bureau of Mines.

Hedin RS, Watzlaf GR 1994. The effects of anoxic limestone drains on mine water chemistry. Third International Conference on the Abatement of Acidic Drainage, Pittsburgh, PA.Pp. 185-194.

Hedin RS, Nairn RW, Kleinmann RLP 1994. Passive treatment of coal mine drainage. US Bureau of Mines Information Circular Number 9389. 35 p.

Hilton T 2004. Draft report, AMDTreat/titration evaluation, Kanes Creek of Decker Creek, sites KCS 1, KCS 3-1,2,3, KCS 3-4, and KCS 3-6. Unpublished 12 October 2004 report by WOPEC (Working on Peoples Environmental Concerns).

Hilton T, Dunn M, Danehy T, Denholm C, Busler S 2003. Harbison Walker - A hybrid passive treatment system. Presented at the 24^th Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV. Also available at: http://wvmdtaskforce.com/proceedings/03/Hilton03.pdf.

Hug SJ, Canonica L, Wegelin M, Gechter D, Von Gunten U 2001. Solar oxidation and removal of arsenic at circumneutral pH in iron containing waters. Environmental Science and Technology 35: 2114-2121.

Hunger S, Benning L 2007. Greigite: a true intermediate on the polysulfide pathway to pyrite. Geochemical Transactions 8: No. 1. doi:10.1186/1467-4866-8-1.

Huntsman BE, Solch JB, Porter MD 1978. Utilization of a Sphagnum species dominated bog for coal acid mine drainage abatement. In: Abstracts, 91st annual meeting Geological Society of America, Ottawa, Ontario, Canada.

Jage CR, Zipper CE, Noble R 2001. Factors affecting alkalinity generation by successive alkalinity-producing systems: regression analysis. Journal of Environmental Quality 30: 1015-1022.

Jong T, Parry DL 2005. Evaluation of the stability of arsenic immobilized by microbial sulfate reduction using TCLP extractions and long-term leaching techniques. Chemosphere 60: 254-265.

Kanel SR, Manning B, Charlet L, Choi H 2005. Removal of arsenic(III) from groundwater by nanoscale zero-valent iron. Environmental Science and Technology 39: 1291-1298.

Kanel SR, Choi H, Kim JY, Vigneswaran S, Shim WG 2006. Removal of arsenic(III) from groundwater using low-cost industrial by-products - Blast furnace slag. Water Quality Research Journal of Canada41: 130-139.

Katsoyiannis I, Zouboulis A, Althoff H, Bartel H 2002. As(III) removal from groundwaters using fixed-bed upflow bioreactors. Chemosphere 47: 325-332.

Kepler DA, McCleary EC 1994. Successive alkalinity producing systems (VFW) for the treatment of acidic mine drainage. Proceedings of the International Land Reclamation and Mine Drainage Conference, Pittsburgh, PA.USBM SP 06A-94. Pp. 195–204.

Kepler DA, McCleary EC 1997. Passive aluminum treatment successes. Presented at the 18^th Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Köber R, Daus B, Ebert M, Mattusch J, Welter E, Dahmke A 2005. Compost-based permeable reactive barriers for the source treatment of arsenic contaminations in aquifers: Column studies and solid-phase investigations. Environmental Science and Technology 39: 7650-7655.

Kumpiene J, Ore S, Renella G, Mench M, Lagerkvist A, Maurice C 2006. Assessment of zerovalent iron for stabilization of chromium, copper, and arsenic in soil. Environmental Pollution 144: 62-69.

Kundu S, Gupta AK 2007. Adsorption characteristics of As(III) from aqueous solution on iron oxide coated cement (IOCC). Journal of Hazardous Materials 142: 97-104.

Kunzru S, Chaudhuri M 2005. Manganese amended activated alumina for adsorption/oxidation of arsenic. Journal of Environmental Engineering 131: 1350-1353.

Lee H, Choi W 2002. Photocatalytic oxidation of arsenite in TiO₂ suspension: kinetics and mechanisms. Environmental Science and Technology 36: 3872-3878.

Lee M, Paik IS, Kim I, Kang H, Lee S 2007. Remediation of heavy metal contaminated groundwater originated from abandoned mine using lime and calcium carbonate. Journal of Hazardous Materials 144: 208-214.

Lehimas GFD, Chapman JI, Bourgine FP 2001. Arsenic removal from groundwater in conjunction with biological-iron removal. Journal of the Chartered Institution of Water and Environmental Management 15: 190-192.

Leupin OX, Hug SJ 2005. Oxidation and removal of arsenic (III) from aerated groundwater by filtration through sand and zero-valent iron. Water Research 39: 1729-1740.

Light S, Yu X 2005. Electrochemical alkaline Fe(VI) water purification and remediation. Environmental Science and Technology 39: 8071-8076.

Makris KC, Sarkar D, Datta R 2006. Evaluating a drinking-water waste by-product as a novel sorbent for arsenic. Chemosphere 64: 730-741.

Manning BA, Hunt ML, Amrhein C, Yarmoff JA 2002. Arsenic(III) and arsenic(V) reactions with zerovalent iron corrosion products. Environmental Science and Technology 36: 5455-5461.

Masue Y, Loeppert RH, Kramer TA 2007. Arsenate and arsenite adsorption and desorption behavor on coprecipitated aluminum: iron hydroxides. Environmental Science and Technology 41: 837-842.

McCauley CA, O’Sullivan AD, Weber PA, Trumm DA 2008. Performance of mesocosm-scale sulfate reducing bioreactors for treating acid mine drainage in New Zealand. Proceedings: 2008 Meeting of the American Society of Mining and Reclamation. Richmond, Virginia, 14-19 June 2008. Pp. 662-698.

McCauley CA, O’Sullivan AD, Milke MW, Weber PA, Trumm DA 2009. Sulfate and metal removal in bioreactors treating acid mine drainage dominated with iron and aluminum. Water Research 43: 961-970.

Means B 2006. AMDTreat 101 The basics, Introduction to AMDTreat chemical treatment modules. Pennsylvania Statewide Conference on Abandoned Mine Reclamation, 24-26 August 2006. Available at: http://2006.treatminewater.com/B/materials/activetreatment.pdf.

Means B, Hilton T 2004. Comparison of three methods to measure acidity of coal-mine drainage. Presented at the National Meeting of the American Society of Mining and Reclamation and the 25^th Annual West Virginia Surface Mine Drainage Task Force Symposium, 19-24 April.

Means B, McKenzie B, Hilton T 2003. A computer-based model for estimating mine drainage treatment costs. Presented at the 24^th Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV. Also available at: http://wvmdtaskforce.com/proceedings/03/Means03.pdf. Computer program available for free at: http://amdtreat.osmre.gov/

Meng X, Korfiatis GP, Christodoulatos C, Bang S 2001. Treatment of arsenic in Bangladesh well water using a household co-precipitation and filtration system. Water Research 35: 2805-2810.

Mills JE. 1996. The North Branch of the Potomac River: Results of two years of lime dosing. Presented at the Seventeenth Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV, 2-3 April. Also available at: http://wvmdtaskforce.com/proceedings/96/96MIL/96MIL.HTM.

Mine Environment Neutral Drainage (MEND) Program 1999. Review of Passive Systems for Treatment of Acid Mine Drainage. MEND Report 3.14.1.

Mondal P, Balomajumder C, Mohanty B 2007. A laboratory study for the treatment of arsenic, iron, and manganese bearing ground water using Fe³⁺</sup> impregnated activated carbon: Effects of shaking time, pH and temperature. Journal of Hazardous Materials 144: 420-426.

Moore K 2005. Treatment of arsenic contaminated groundwater using oxidation and membrane filtration. University of Waterloo, 161 p.

Nairn RW, Hedin RS, Watzlaf GR 1991. A preliminary review of the use of anoxic limestone drains in the passive treatment of acid mine drainage. Presented at the Twelfth Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Neculita CM, Zagury G, Bussiere B 2007. Passive treatment of acid mine drainage in bioreactors using sulfate-reducing bacteria: critical review and research needs. Journal of Environmental Quality 36: 1-16.

Nguyen TV, Vigneswaran S, Ngo HH, Pokhrel D, Viraraghavan T 2006. Iron-coated sponge as effective media to remove arsenic from drinking water. Water Quality Research Journal of Canada 41: 164-170.

Ning RY 2002. Arsenic removal by reverse osmosis. Desalination 143: 237-241.

Nordstrom DK, Alpers CN 1999. Geochemistry of acid mine waters. In: Plumlee G, Logsdon MJ ed. Reviews in Economic Geology Volume 6A: Environmental Geochemistry of Mineral Deposits, Part A, Processes, Techniques and Health Issues. Littleton, CO: Society of Economic Geologists. Pp. 133-160.

Nordstrom DK, McCleskey RB 2006. Mineral solubility and weathering rate constraints on metal concentrations in groundwaters of mineralized areas near Questa, New Mexico. In: Barnhisel, RIed. Seventh International Conference on Acid Rock Drainage. 26-30 March, St. Louis MO. Lexington, KY: American Society of Mining and Reclamation (ASMR).

Nordwick S, Zaluski M, Park B, Bless D. 2006. Advances in development of bioreactors applicable to the treatment of ARD. Presented at the 7th international cobnference on acid rock drainage (ICARD), March 26-30 2006, St. Louis MO. RI Barnhisel (ed) published by the American Society of mining and reclamation.

Oehmen A, Viegas R, Velizarov S, Reis MAM, Crespo JG 2006. Removal of heavy metals from drinking water supplies through the ion exchange membrane bioreactor. Desalination 199: 405-407.

O’Sullivan A 2005. Passive treatment technologies for managing metal mine wastes: lessons learnt from global applications. In: Moore TA, Black A, Centeno JA, Harding JS, Trumm DA ed. Metal contaminants in New Zealand, sources, treatments, and effects on ecology and human health. Christchurch: Resolutionz Press. Pp. 279–300.

Ott AN 1986. Estimating iron and aluminum content of acid mine discharge from a north-central Pennsylvania coal field by use of acidity titration curves. U.S.Geological Survey Water Resources Inv. Report 84-4335, 25 p.

Parks JL, Edwards M 2006. Precipitative removal of As, Ba, B, Cr, Sr, and V using sodium carbonate. Journal of Environmental Engineering 132: 489-496.

Payne DA, Yeates TE 1970. The effects of magnesium on acidity determinations. Pittsburgh, PA: 3rd Symposium on Coal Mine Drainage Research, National Coal Association/Bituminous Coal Research, Inc. Pp. 200-223.

Petrusevski B, Sharma SK, Kruis F, Omeruglu P, Schippers JC 2002. Family filter with iron-coated sand: Solution for arsenic removal in rural areas. Water Science and Technology: Water Supply 2: 127-133.

Phipps TT, Fletcher JJ, Skousen JG 1996. Costs for chemical treatment of AMD. In: Skousen JG, Ziemkiewicz PF compilers Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 187-212.

PIRAMID Consortium 2003. Engineering guidelines for the passive remediation of acidic and/or metalliferous mine drainage and similar wastewaters. European Commission 5th Framework RTD Project no. EVK1-CT-1999-000021 ‘Passive in-situ remediation of acidic mine / industrial drainage’ (PIRAMID). Newcastle Upon Tyne, UK: University of Newcastle upon Tyne. 166 p.

Pokhrel D, Viraraghavan T, Braul L 2005. Evaluation of treatment systems for the removal of arsenic from groundwater. Practice Periodical of Hazardous Toxic and Radioactive Waste Management 9: 152-157.

Rait R, TrummD, Pope J, Craw D, Newman N, MacKenzie H In press. Adsorption of arsenic by iron rich precipitates from two coal mine drainage sites on the West Coast of New Zealand. New Zealand Journal of Geology and Geophysics.

Ramakrishna DM, Viraraghaven T, Jin YC 2006. Iron oxide coated sand for arsenic removal: Investigation of coating parameters using factorial design approach. Practice Periodical of Hazardous Toxic and Radioactive Waste Management 10: 198-206.

Ramaswami A, Tawachsupa S, Isleyen M 2001. Batch-mixed iron treatment of high arsenic waters. Water Research 35: 4474-4479.

Robbins EI, Cravotta III CA, Savela CE, Nord Jr GL 1999. Hydrobiogeochemical interactions in ‘anoxic’ limestone drains for neutralization of acidic mine drainage. Fuel 78: 259–278.

Rose AW 2004. Vertical flow systems - Effects of time and acidity relations. Proceedings, American Society of Mining and Reclamation, Lexingon, KY, 9-13 June. Pp. 776-797.

Rose AW 2006. Long-term performance of vertical flow ponds - An update. Proceedings, 7th International Conference on Acid Rock Drainage (ICARD), 26-30 March 2006, St. Louis, MO.The American Society of Mining and Reclamation (ASMR). Pp. 2142-2158.

Rose AW, Cravotta III CA 1998. Geochemistry of coal mine drainage. In: Brady KBC, Smith MW, Schueck J ed. Coal mine drainage prediction and pollution prevention in Pennsylvania. Harrisburg, PA: Pennsylvania Department of Environmental Protection. Pp. 1-1 to 1-22.

Rose AW, Means B, Shah PJ 2003. Methods for passive removal of manganese from acid mine drainage. Presented at the annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Santomartino S, Webb JA 2007. Estimating the longevity of limestone drains in treating acid mine drainage containing high concentrations of iron. Applied Geochemistry 22: 2344-2361.

Selfridge JA, McDonald LM, Skousen JG, Ziemkiewicz PF 2003. Effect of iron coatings on limestone surface area and dissolution rates. Presented at the 24^th Annual West VirginiaSurface Mine Drainage Task Force Symposium. Also available at: http://wvmdtaskforce.com/proceedings/03/Selfridge.pdf.

Selvin N, Upton J, Simms J, Barnes J 2002. Arsenic treatment technology for groundwaters. Water Science and Technology: Water Supply 2: 11-16.

Simmons J, Ziemkiewicz P, Black DC 2002. Use of steel slag leach beds for the treatment of acid mine drainage. Mine Water and the Environment 21: 91-99.

Skousen J 1988. Chemicals for treating acid mine drainage. Green Lands 18(2): 36-40. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 157-162.

Skousen J 1991 Anoxic limestone drains for acid mine drainage treatment. Green Lands 21(4): 30-35. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 261-266.

Skousen J, Jenkins M 1993. Acid mine drainage treatment with the aquafix system. Green Lands 23(4): 36-38. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 213-216.

Skousen J, Politan K, Hilton T, Meek A. 1990. Acid mine drainage treatment systems: chemicals and costs. Green Lands 20(4): 31-37. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 163-172.

Skousen J, Sexstone A, Garbutt K, Sencindiver J 1992. Wetlands for treating acid mine drainage. Green Lands 22(4): 31-39. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 249-260.

Skousen J, Hilton T, Faulkner B 1996. Overview of acid mine drainage treatment with chemicals. Green Lands 26(3): 40-49. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers. Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 237-247.

Skousen J, Rose A, Geidel G, Foreman J, Evans R, Hellier W 1998. A handbook of technologies for avoidance and remediation of acid mine drainage. Acid drainage technology initiative, published by the National Mine Land Reclamation Center at West Virginia University.

Skousen J, Sextone A, Ziemkiewicz PF 2000. Acid mine drainage control and treatment. In: Barnhisel RI, Darmody RG, Daniels, WL ed. Reclamation of drastically disturbed lands. Monograph Number 41. American Society of Agronomy, Madison, WI, USA. Pp. 131-168.

Skovran GA, Clouser CR 1998. Design considerations and construction techniques for successive alkalinity producing systems. Proceedings, 1998 National Meeting of the American Society for Mining and Reclamation, 16-21 May, St Louis, MO, USA.

Song S, Lopez-Valdivieso A, Hernandez-Campos DJ, Peng C, Monroy-Fernandez MG, Razo-Soto I 2006. Arsenic removal from high-arsenic water by enhanced coagulation with ferric ions and coarse calcite. Water Research 40: 364-372.

Stumm W, Morgan JJ 1996. Aquatic chemistry: Chemical equilibria and rates in natural waters. 3rd ed. Wiley-Interscience. 470 p.

Su C, Puls RW 2001. Arsenate and arsenite removal by zerovalent iron : effects of phosphate, silicate, carbonate, borate, sulfate, chromate, molybdate, and nitrate, relative to chloride. Environmental Science and Technology 35: 4562-4568.

Sverdrup HU 1983. A short description of the diversion wells located at Piggaboda, Sweden. Translation from the symposium proceedings ‘Kalkning av Rinnande Vatten’, 18 September 1983, Alvesta, Sweden. 11 p.

Swash PM, Monhemius AJ 2005. Characteristics and stabilities of residues from the Wheal Jane constructed wetlands. Science of the Total Environment 338(1-2: Special issue): 95-105.

Tadanier CJ, Schreiber ME, Roller JW 2005. Arsenic mobilization through microbially mediated deflocculation of ferrihydrite. Environmental Science and Technology 39: 3061-3068.

Thomas RC, Romanek CS 2002.Acid rock drainage in a vertical flow wetland II: Metal removal, in Proceedings of the 19th Annual National Conference, American Society of Surface Mining and Reclamation, Lexington, KY, p. 752 - 775.

Trumm D 2007. Selection of passive AMD treatment systems – flow charts for New Zealand conditions. In: The Australasian Institute of Mining and Metallurgy New Zealand Branch 40th Annual Conference, August 2007, Christchurch, New Zealand. Pp 23-27.

Trumm D, Cavanagh JE 2006. Investigation of remediation of acid-mine-impacted waters at Cannel Creek. Landcare Research Contract Report LC0506/169, CRL Report 06-41101. Prepared for West Coast Regional Council. Available at: http://www.wcrc.govt.nz/Resources/Documents/Publications/Environmental%20Management/BellvueMineRemediation.pdf.

Trumm DA, Gordon K 2004. Solutions to acid mine drainage at Blackball and Sullivan Mines. In: The Australasian Institute of Mining and Metallurgy New Zealand Branch 37th Annual Conference, August 2004, Nelson, New Zealand. Pp 103-109.

Trumm D, Black A, Cavanagh J, Harding J, de Joux A, Moore TA, O’Halloran K 2003. Developing assessment methods and remediation protocols for New Zealand sites impacted by Acid Mine Drainage (AMD). Sixth International Conference on Acid Rock Drainage, 12-18 July, Cairns, Queensland, Australia. Pp. 223-232.

Trumm DA, Black A, Gordon K, Cavanagh J, deJoux A 2005. Acid mine drainage assessment and remediation at an abandoned West Coast coal mine. In: Moore TA, Black A, Centeno JA, Harding JS, Trumm DA ed. Metal contaminants in New Zealand, sources, treatments, and effects on ecology and human health. Christchurch: Resolutionz Press. Pp. 317-342.

Trumm DA,bWatts M, Gunn P 2006. AMD treatment in New Zealand– use of small-scale passive systems. Proceedings, 7th International Conference on Acid Rock Drainage (ICARD), 26-30 March 2006, St. Louis, MO.The American Society of Mining and Reclamation (ASMR). Pp. 2142-2158.

Trumm D, Watts M, Pope J, Lindsay P 2008.Using pilot trials to test geochemical treatment of acid mine drainage on Stockton Plateau. New ZealandJournal of Geology and Geophysics 51: 175-186.

Trumm D, Pope J, Newman N 2009. Passive treatment of neutral mine drainage at a metal mine in New Zealand using an oxidising system and slag leaching bed. Eighth International Conference on Acid Rock Drainage, 23-26 June 2009, Skelleftea, Sweden.

Turner D, McCoy D 1990. Anoxic alkaline drain treatment system, a low cost acid mine drainage treatment alternative. In: Proceedings of the National Symposium on Mining. Lexington, KY: University of Kentucky. Pp. 73-75.

U.S.Environmental Protection Agency (USEPA) 1979. Methods for chemical analysis of water and wastes. Cincinnati, OH: U.S.Environmental Protection Agency.

U.S.Environmental Protection Agency (USEPA) 1995. US Environmental Protection Agency (USEPA) 1995. Test methods for evaluating solid waste. SW-846. 3rd ed. through update llB.

U.S.Environmental Protection Agency (USEPA) 2002. Arsenic treatment technologies for soil, water and waste. US EPA EPA-542-R-02-004. Cincinnati, OH.

Violante A, Ricciardella M, Del Gaudio S, Pigna M 2006. Coprecipitation of arsenate with metal oxides: Nature, mineralogy, and reactivity of aluminum precipitates. Environmental Science and Technology 40: 4961-4967.

Viraraghavan T, Subramanian KS, Aruldoss JA 1999. Arsenic in drinking water - Problems and solutions. Water Science and Technology 40: 69-76.

Vithanage M, Senevirathna W, Chandrajith R, Weerasooriya R 2007. Arsenic binding mechanisms on natural red earth: A potential substrate for pollution control. Science of the Total Environment 379: 244-248.

Wang L, Chen ASC, Sorg TJ, Fields KA 2002. Field evaluation of AS removal by IX and AA. American Water Works Association Journal Pro Quest Science Journal 94(4): 161.

Waters JC, Santomartino S, Cramer M, Murphy N, Taylor JR 2003. Acid rock drainage treatment technologies - Identifying appropriate solutions. Proceedings, Sixth International Conference on Acid Rock Drainage (ICARD), 12-18 July 2003. Cairns, Australia. Pp. 831-843.

Watzlaf GR, Hyman DM 1995. Limitations of passive systems for the treatment of mine drainage. In: Proceedings, Seventeenth Annual Conference of the National Association of Abandoned Mine Lands, October 1995, French Lick, IN.

Watzlaf GR, Schroeder KT, Kairies CL 2000. Longterm performance of anoxic limestone drains. Mine Water and the Environment 19: 98-110.

Watzlaf GR, Kairies C, Schroeder KT, Danehy T, Beam R 2002. Quantitative results from the flushing of four reducing and alkalinity-producing systems. Presented at the Annual West Virginia Surface Mine Drainage Task Force Symposium, 16-17 April, Morgantown, WV.

Watzlaf GR, Schroeder KT, Kleinmann RLP, Kairies CL, Nairn RW 2003. The passive treatment of coal mine drainage. US Department of Energy.

Watzlaf GRm Schroeder KT, Kleinman RKLP, Caries CL, Nairn rW. 2004. The Passive Treatment of Coal Mine Drainage. US Department of Energy DOE/NETL- 2004/1202, Available from: <ftp://ftp.netl.doe.gov/pub/Watzlaf/NETL-1202.pdf>, accessed October 2004

Weaver KR, Lagnese KM, Hedin RS 2004. Technology and design advances in passive treatment system flushing. Presented at the Twenty-fifth Annual West Virginia Surface Mine Drainage Task Force Symposium, 18-24 April, Morgantown, WV.

Weber PA, Hughes JB, Lindsay P 2007. Stockton Mine acid rock drainage remediation - Part 2 - Treatment. In: The Australasian Institute of Mining and Metallurgy New Zealand Branch 40th Annual Conference, August 2007, Christchurch, New Zealand. Pp. 41-48.

Wickramasinghe SR, Han B, Zimbron J, Shen Z, Karim MN 2004. Arsenic removal by coagulation and filtration: Comparison of groundwaters from the United States and Bangladesh. Desalination 169: 231-244.

Wieder RK, Lang GE 1982. Modification of acid mine drainage in a freshwater wetland. In: Proceedings Symposium Wetlands Unglaciated Appalachian Region, October 1982, West Virginia University, Morgantown, WV, USA. Pp. 43-53.

Wildeman TR, Pinto AP, Tondo LA, Alves LA 2006. Passive treatment of a cyamide and arsenic laden process water at the RPM gold mine, Minas Gerais, Brazil. 7th International Conference on Acid Rock Drainage (ICARD), St Louis, MO.Pp. 2377-2384.

Wratten BJ, Schwartz MF 1996. Carbon dioxide pretreatment of AMD for limestone diversion wells. Presented at the Seventeenth Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Younger PL, Banwart SA, Hedin RS 2002. Mine water: Hydrology, pollution, remediation. Kluwer.

Yuan T, Hu JY, Luo QF, Cheng JP, Ong SL 2006. Stopgap coagulation technology for arsenic removal from rural household drinking water. Water Science and Technology: Water Supply 6: 273-279.

Zhang FS, Itoh H 2006. Photocatalytic oxidation and removal of arsenite from water using slag-iron oxide-TiO₂ adsorbent. Chemosphere 65: 125-131.

Zhang GS, Qu JH, Liu HJ, Liu RP, Li GT 2007. Removal mechanism of As(III) by a novel Fe-Mn binary oxide adsorbent: Oxidation and sorption. Environmental Science and Technology 41: 4613-4619.

Ziemkiewicz PF, Brant DL 1996. The Casselman River restoration project. Presented at the Eighteenth Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Ziemkiewicz PF, Skousen JG 1998. The use of steel slag in acid mine drainage treatment and control. Proceedings, 19^th Annual West Virginia Surface Mine Drainage Task Force Symposium, Morgantown, WV.

Ziemkiewicz PF, Skousen JG, Lovett R.J. 1994. Open limestone channels for treating AMD: a new look at an old idea. Green Lands 24(4): 31-38. Also published in 1996 In: Skousen JG, Ziemkiewicz PF compilers Acid mine drainage control and treatment. 2nd ed. West Virginia University. Pp. 275-280.

Ziemkiewicz PF, Skousen JG, Brant DL, Sterner PL, Lovett RJ 1997. Acid mine drainage treatment with armoured limestone in open limestone channels. Journal of Environmental Quality 26: 1017-1024.

Ziemkiewicz P, Skousen J, Simmons JS 2003a. Long-term performance and cost benefit analysis of passive acid mine drainage treatment systems installed in the Appalachian Coal Region of the Eastern United States. Proceedings, Sixth International Conference on Acid Rock Drainage (ICARD), 12-18 July 2003, Cairns, Australia. Pp. 855-862.

Ziemkiewicz P, Skousen J, Simmons JS 2003b. Long-term performance of passive acid mine draiange treatment systems. Mine Water and the Environment 22: 118–129.

Zipper C, Jage C 2001. Passive treatment of acid-mine drainage with vertical-flow systems. Virginia Polytechnic Institute and State University, Virginia Cooperative Extension, Publication Number 460-133. 16 p.

Zouboulis AI, Katsoyiannis IA 2005. Recent advances in the bioremediation of arsenic-contaminated groundwaters. Environment International 31: 213-219.

Zurbuch PE1996. Early results from calcium carbonate neutralization of two West Virginia rivers acidified by mine drainage. Presented at the Seventeenth Annual West Virginia Surface Mine Drainage Task Force Symposium, 2-3 April, Morgantown, WV. Also available at: http://wvmdtaskforce.com/proceedings/96/96ZUR/96ZUR.HTM

◄ previous page │ next page►


	HOME	SERVICES	DOWNLOADS	ABOUT US	NEWS	CONTACT US