Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A fuzzy expert system for acid rock drainage site remediation Balcita, Judita Veronika 2001

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2001-0151.pdf [ 7.01MB ]
Metadata
JSON: 831-1.0089908.json
JSON-LD: 831-1.0089908-ld.json
RDF/XML (Pretty): 831-1.0089908-rdf.xml
RDF/JSON: 831-1.0089908-rdf.json
Turtle: 831-1.0089908-turtle.txt
N-Triples: 831-1.0089908-rdf-ntriples.txt
Original Record: 831-1.0089908-source.json
Full Text
831-1.0089908-fulltext.txt
Citation
831-1.0089908.ris

Full Text

A F U Z Z Y E X P E R T S Y S T E M F O R A C I D R O C K D R A I N A G E S I T E R E M E D I A T I O N by JUDITA VERONIKA BALCITA B.A.Sc., University of Toronto, Canada, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Mining and Mineral Process Engineering) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA March 2001 © Judita Veronika Balcita, 2001 UBC Special Collections - Thesis Authorisation Form Page 1 of 1 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department The University of B r i t i s h Columbia Vancouver, Canada http ://www.library.ubc. ca/ spcoll/thesauth.html 4/16/01 ABSTRACT This thesis describes the development of an expert system using fuzzy logic-based techniques to deal with the remediation of Acid Rock Drainage contaminated sites. Acid Rock Drainage (ARD) is one of the major environmental challenges faced today by the mining industry. The interdisciplinary nature of ARD, involving biology, chemistry, geology, and mining knowledge is not yet fully understood nor has a viable and permanent solution to the problem been found. Development of a fuzzy expert system for ARD is considered useful in producing a standardized, yet adaptable, approach to the problem which can be used to generate quick advice to the user in a field where expertise is lacking. A fuzzy expert system can deal with missing, inaccurate, or heuristic data. A fuzzy expert system actually thrives on such conditions. Fuzzy sets are defined as the degree of belief in a particular qualitative concept as a function of one or more quantitative values. Rules are developed from these fuzzy sets during interviews with a chosen expert in the field. From user input on site data and site characterization, association within these rules of the degrees of belief in different concepts can flow through the system to generate a decision or conclusion. The system includes a fuzzy controller, separate control modules for treatment options and interactive hypertext documents for user control. The controller and control modules work together to follow the decision-making process of experts in the field in choosing appropriate remediation options for a site of possible or existing ARD. The hypertext documents are set up as user help resources and to provide system output information. They can also serve as teaching i i aids on treatment of ARD or to obtain better understanding of the possibility of implementing a treatment system. Validation and verification of the system has been carried out by comparing the results obtained for certain specific case studies with those determined by an expert. iii TABLE of CONTENTS A B S T R A C T \ \ T A B L E O F C O N T E N T S IV L I S T O F T A B L E S V> L I S T O F F I G U R E S M j A C K N O W L E D G E M E N T MlV\ I N T R O D U C T I O N 1 A C I D R O C K D R A I N A G E - T H E P R O B L E M 3 M I T I G A T I O N O P T I O N S F O R A R D 5 INTRODUCTION 5 ACTIVE METHODS 5 PASSIVE TREATMENT OPTIONS 10 E X P E R T S Y S T E M S 15 WHAT ARE EXPERT SYSTEMS? 15 WHY IS AN EXPERT SYSTEM USEFUL FOR A R D REMEDIATION? 15 How ARE EXPERT SYSTEMS BUILT? 16 COMPONENTS OF THE A R D x EXPERT SYSTEM 16 S Y S T E M C O N F I G U R A T I O N 21 OVERALL APPROACH 21 K N O W L E D G E B A S E M O D U L E S 27 A R D x MAIN MODULE 2 7 WASTE ROCK AND TAILINGS 2 9 MINE WORKINGS AND OPEN PIT 3 9 DIVERSION SUB-MODULE 4 0 PASSIVE SUB-MODULE 41 C O S T M O D U L E 43 INTRODUCTION To COST MODULE 43 WATER COVER COSTS 43 DRY COVER COSTS 4 6 TREATMENT COSTS 48 DIVERSION COSTS 4 9 PASSIVE TREATMENT COSTS 5 0 COSTS CALCULATIONS AND ADJUSTMENTS 5 2 S Y S T E M V A L I D A T I O N A N D V E R I F I C A T I O N 54 SYSTEM VALIDATION 54 SYSTEM VERIFICATION... 55 C O N C L U S I O N S 63 iv R E C O M M E N D A T I O N S ; 64 R E F E R E N C E S 66 B I B L I O G R A P H Y 70 A P P E N D I X A - F A M M A P S 77 A P P E N D I X B - F U Z Z Y SETS 95 A P P E N D I X C - H Y P E R T E X T D O C U M E N T S 110 BACTERICIDES (BACTERHY.DOC) i l l DRY COVERS (COVERSHY.DOC) 114 WATER DIVERSION (DIVERSHY.DOC) 119 PASSIVE TREATMENT (PASSIVHY.DOC) 121 SULPHIDE REDUCTION (SULPHHY.DOC) 138 WATER TREATMENT (TREATHY.DOC) 140 WATER COVERS (WATERHY.DOC) 144 HELP DOCUMENT (HYPERARD.DOC) 147 A P P E N D I X D - INPUTS A N D O U T P U T S 164 RUN#1 AND RUN #2 165 RUN #3 AND RUN #4 168 v LIST of TABLES Table 1. Cost details for flooding of tailings 43 Table 2. Cost details for in-pit flooding of waste rock 44 Table 3. Cost details for sub-marine disposal of tailings 45 Table 4. Cost details for sub-marine disposal of waste rock. 46 Table 5. Cost details for tailings dry covers 47 Table 6. Cost details for waste rock dry covers 48 Table 7. Cost details for water treatment 48 Table 8. Cost details for water diversion 50 Table 9. Cost details for anoxic limestone drain 50 Table 10. Cost details for aerobic and anaerobic wetland 51 Table 11. Summary of Results; Run #1 vs Run #2 57 Table 12. Summary of Results; Run #3 vs Run #4 59 Table 13. Ranking of The Reclamation Options For Each Test Run 61 Table 14. Recommended Reclamation Options 62 vi LIST of FIGURES Figure 1. Flooded tailings in a constructed impoundment (Compiled from Placer Dome Inc., 1997) 7 Figure 2. Multiple Layer Cover. (Compiled from Lawrence, 1997) 8 Figure 3. Anoxic Limestone Drain (Compiled from Gordon & Robinson, 1995; Skousen, et al, 1995) 11 Figure 4. Aerobic Cell (Compiled from Gordon & Robinson, 1995; Skousen, et al, 1995) .12 Figure 5. Anaerobic cell (Compiled from Gordon & Robinson, 1995; Skousen, et al, 1995)14 Figure 6. Example of a rule within the knowledge base 18 Figure 7. Example of a Procedure within the cost module 19 Figure 8. Example of a F O R M used as a user interface 20 Figure 9. Basic A R D E X configuration flow-chart 21 Figure 10. Basic ARDx Flowchart 24 Figure 11. Example of a F A M map 25 Figure 12. Fuzzy set for "enviro.factors.high" 26 Figure 13. Water module flowchart 31 Figure 14. Covers module flowchart for waste rock.... 33 Figure 15. Covers module flowchart for tailings 34 Figure 16. Treatment module flow sheet 35 Figure 17. Fuzzy Sets for aluminum concentrations based on the fractional difference from the user input regulation value 36 Figure 18. Diversion module flowchart 41 Figure 19. Passive treatment module flowchart 42 v i i ACKNOWLEDGEMENT I would like to thank the following people for all their help and support with this project: John Meech for his guidance, patience, and assistance, Mory Ghomshei for his expertise, Marcello Veiga for his advice, Cy and Emerald Keyes Fellowship for their financial support. v i i i A Fuzzy Expert System on Acid Rock Drainage Site Remediation INTRODUCTION Acid Rock Drainage (ARD) is contaminated acidic drainage resulting from the spontaneous weathering and oxidation of pyrite and other sulfide minerals (Lawrence, 1994). Weathering increases the solubility of heavy metals, radionuclides and sulfates; and reduces pH. ARD will impact on watershed characteristics and will adversely affect a local ecosystem. ARD occurs when pyrite and other sulphide minerals are exposed to the surface environment by mining mineral resources. The problem is common in both coal and metal mines. Once exposed, waste rock generating acidic and metal-contaminated drainage may continue to do so for decades. Prediction and prevention are the primary methods used to deal with and control ARD. However, in active or abandoned mine sites where the problem already exists, or as a supplement to preventive measures in new mines; treatment of the contaminated drainage is necessary and costly (Filipek et al., 1996). Dealing with ARD should primarily concern the inhibition of reactions at the source. Any existing acidic drainage can then be controlled by directing contaminant migration toward a treatment pond. Any subsequent acidic drainage should be collected and treated prior to release to the environment. Problems in treating ARD include the fact that mine sites are usually in remote locations. Fluctuations in flow rates, large volumes of low intensity sludge, and the unknown stability of the chemical sludge are all factors that affect the long term treatment of ARD using chemical methods. The mining industry is faced with the challenge of finding an economic and sustainable solution to ARD. Expertise within the field of ARD is controversial and scarce. Fundamental knowledge is lacking, and new information is continually being sought and applied. Data to assess and deal 1 A Fuzzy Expert System on Acid Rock Drainage Site Remediation with ARD problems are often missing and so heuristics play an important role in decision-making. A fuzzy expert system is able to handle and manipulate missing and inaccurate, or heuristic data. (Bowen, 1995) The objectives of this thesis are to develop a fuzzy expert system providing recommendations on ARD mitigation. A logical, standardized adaptable approach to decision making when faced with ARD problems at mine sites will be implemented during the development of the system. The system will examine both technical and economic factors to decide on final recommendations for mitigation. The system will also provide an environment for teaching and training through hypertext documents and access to "help" information. The structure of the fuzzy expert system will allow for additions and updates of information to the developed system with relative ease. 2 A Fuzzy Expert System on Acid Rock Drainage Site Remediation ACID ROCK DRAINAGE - THE PROBLEM Pyrite is abundant in virtually all sulphide ore bodies and is often the main culprit contributing to the generation of ARD. The reactions taking place to cause highly-acidic metal-contaminated drainage can be studied through the oxidation of pyrite. When exposed pyrite comes into contact with oxygen and water, it oxidizes to produce ferrous iron, sulphate and hydrogen ions according to the following reaction: The reaction produces ferrous iron and sulphate to increase the total dissolved solids and ions leading to a decrease in pH. Trace metals such as copper, lead, and zinc that were bound up within the pyrite or other minerals are released as the pyrite is oxidized. Ferrous ions can be further oxidized to the ferric state: This normally slower reaction is facilitated by the presence of certain bacteria the most commonly known as Thiobacillus Ferrooxidans. Other bacteria include Thiobacillus Thiooxidans and Sulfolobus. All of these bacteria use energy from the above reaction to sustain their life processes. The presence of ferric ions can oxidize pyrite further via the following fast reaction: FeS2(s) + i 0 2 i g ) + H20=> Fe2+ + ISO2/ + 2H •+ 1. 2Fe2+ +^02(g)+2H+ => 2Fei+ + H20 2. FeS2 +14Fe*+SH20=* \5Fe2+ + 2S042" +16H + 3. 3 A Fuzzy Expert System on Acid Rock Drainage Site Remediation This reaction creates additional acidity, driving the pH down quickly and releasing more metals into the environment. As the drainage moves down the flow path and comes into contact with other rocks and minerals, further reactions occur, changing the drainage characteristics. These may include neutralization due to interaction with minerals such as calcium carbonate, and the acid/oxidative leaching of metals and iron hydrolysis. In a mining environment where ARD can contaminate the watershed and downstream environment, control and treatment must be implemented. Prevention and control is by far the best means to eliminate the problem. Unfortunately, technology has not yet found a fail-safe universal solution since many of the processes and reactions are inter-linked, not completely understood and are often site-specific. Many contaminated sites are a result of prior abandoned mining activity when knowledge of ARD was scarce or non-existent and preventative measures were not taken. This leaves behind an ongoing contamination problem. For mines that are in operation today and those that are in their preliminary planning stage, prevention should be examined initially as a means of abatement. Today's prevention methods are not foolproof although research is being done to find additional methods. So the mining community is faced with the problem of finding ways to clean up problems at old, and active mine sites. 4 A Fuzzy Expert System on Acid Rock Drainage Site Remediation MITIGATION OPTIONS FOR ARD Introduction Mitigation options for control and treatment of ARD are generally grouped into active and passive methods. Active methods refer to those requiring constant intervention or continuous operation. Within the expert system developed in this work, the active methods include sub-aqueous disposal, dry covers, chemical treatment and water diversion control. Passive method options include wetlands (aerobic and anaerobic), settling ponds, and anoxic limestone drains (ALD). It is the investigation of these mitigation options and their suitability for site specific application that forms the basis of this expert system - ARDx. A brief review of active and passive methods follows; while more detailed information may be found in Appendix C. Active Methods Sub-aqueous disposal (or water coverage) is one of the most promising methods for controlling ARD. A water cover limits the amount of oxygen that comes in contact with the waste due to the low solubility of oxygen in water and its low aqueous diffusivity. The solubility of oxygen in water is normally below 1 lmg/L, decreasing with increased temperature. At 10°C, the level is 5.8 mg/L. This low solubility leads to a low transport of oxygen to the waste. The majority of this dissolution takes place in the upper meter of the water cover (epi-limnion) and so without significant mixing, high O2 levels at depth are not possible. Diffusion of oxygen is four times lower in water than in air (Broughton, 1992). With the absence of oxygen, the rate of oxidation of waste material is reduced to a level such that ARD essentially does not occur. Sub-aqueous deposition can be achieved by: 5 A Fuzzy Expert System on Acid Rock Drainage Site Remediation • disposal of waste rock and/or tailings in an ocean or natural lake - Effective control of lake disposal requires an understanding of seasonal fluctuations in water levels, wave action, convection and seasonal lake overturn. • flooding of tailings in constructed impoundment - Disposal of waste in a constructed impoundment or lake involves maintenance to prevent initiation of oxidation. • flooding of disposed waste rock within an open pit - Flooding of an open pit containing waste rock requires re-handling and relocation of waste from its established location to an open pit that is flooded to prevent exposure to oxygen. • flooding of mine workings - Flooding of mine workings requires good understanding of local hydrology. The effectiveness of control requires the maintenance of a permanent water cover. Site considerations for choosing sub-aqueous deposition include knowledge of seasonal fluctuations in the water balance to ensure an adequate cover can be sustained. The flooding of tailings within a constructed impoundment is shown in Figure 1 below. 6 A Fuzzy Expert System on Acid Rock Drainage Site Remediation T A I L I N G S Figure 1. Flooded tailings in a constructed impoundment (Compiled from Placer Dome Inc., 1997) Dry covers are often used to control ARD. They work by limiting the amount of water, oxygen or both, coming into contact with the waste. Dry covers can be effective at restricting the amount of infiltration from rainfall which is beneficial as a large portion of seepage from a waste dump results directly from precipitation on the dump. Although wet or saturated covers inhibit oxygen entry to a far greater extent then do dry covers, they can not always be used because of hydrology factors and other site characteristics. Oxygen diffusion through a wet cover may be a magnitude lower or more than a dry cover. The effectiveness of a dry cover depends on grain size distribution, compaction, moisture content, permeability and long-term integrity. Various cover types are used by the mining industry as control techniques for ARD (Filipek, L., A. Kirk, W. Schafer, 1996). These include: • Simple soil covers - Simple soil covers consist of a single layer of a lower permeability soil ~ compacted till, clay, silt, or other material locally available. 7 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Multiple layer covers - Multiple soil covers consist of several layers with specific purposes. These covers are often used in high rainfall areas and are most effective at restricting oxygen from entering the waste dump. Complex clay covers - Complex clay covers are similar to multiple soil covers and also consist of several layers with specific purposes, one of which is the restriction of water flow through a low-permeability clay layer. Complex synthetic covers - Complex synthetic covers are also similar to multiple soil covers with the added feature of a synthetic material (liner) to provide a permanent and effective solution (life span of approximately 100 years). A Multiple layer cover is shown in Figure 2 below. Humic layer (moisture retention zone supporting growth) Coarse grained material (anti intrusion) Drainage layer Fine grained material (capillary retention) Non-capillary layer Acidic tailings Figure 2. Mul t ip le Layer Cover. (Compi led from Lawrence, 1997) 8 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Dry covers can be constructed using many different materials. Materials found on or near the site can often be used and may be the less expensive option. Cover materials include soil, clay, till topsoil, concrete, asphalt and synthetic materials. Compacted clay although low in permeability (10~9 to 10"nm/s) is subject to erosion, cracking and root penetration; and may not be readily available in all areas. Clay can be a good sealing material if it is well maintained. Compacted till (10"7 to 10~9m/s) and compacted topsoil (10"5 to 10"8m/s) have similar characteristics to compacted clay while being less permeable. Concrete and asphalt are both subject to cracking, frost and mechanical damage. Synthetic materials are highly impermeable albeit they require the necessary bedding and cover to protect from mechanical and root penetration (Lawrence, 1997). Migration control seeks to minimize the amount of water entering the waste site. Groundwater and surface water diversion and interception are methods of migration control. The objective is to reduce the amount of water entering the exposed area. Selecting a site with little or no ground water will avoid ground water infiltration and discharge. The site should also avoid a high runoff. Interception, collection, and isolation of water upstream of a waste dump reduces ground-water entrance. Surface water is diverted upstream with ditches and berms. Below the waste, water may require diversion into treatment areas. This is usually a short term control measure and can be applied to both proposed and existing operations. Collection and treatment of water are required in all cases where water discharge limits are not met by other means. Water treatment is a continuous process of adding energy, manpower and alkaline chemicals to decrease acidity and increase the pH of the effluent. 9 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Passive Treatment options Anoxic Limestone Drains use limestone to raise pH and add alkalinity to ARD waters. ALD's have been shown to generate up to 300mg/L of alkalinity when installed correctly and water quality parameters comply (Skousen, 1991). The drains must be constructed such that anoxic conditions prevail and therefore are buried in underground channels or trenches such that oxygen is unavailable. ALD's work according to the following simple dissolution reaction (Skousen, et al 1992): CaC03 + H+ = Ca+2 + HCO; 4 As acidic water passes through the drain, the limestone reacts causing the pH to increase. The reaction buffers against pH drops that can result from iron precipitation reactions. ALD's are not directly responsible for removal of dissolved metal ions. Water leaving the drain must be exposed to the atmosphere to allow for oxidation, hydrolysis and precipitation of metals. Under correct conditions, ALD's can generate alkalinity for many years. The effect of the drains is hindered by a number of factors that can lead to limestone armouring or plugging of the drain. Oxygen introduced into the system will result in limestone armouring. The dissolved oxygen content of water entering should be kept below 2 mg/1 for effective life of the system (Skousen, et al, 1992). A low concentration of ferric iron (Fe ) is necessary. As more oxygen is introduced, the ferrous iron (Fe ) concentration will decrease as it oxidizes to ferric iron. Ferric iron will precipitate in the drain coating the limestone. 10 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Furthermore, a high concentration of aluminum may also result in coating of the limestone. Aluminum precipitates at pH levels > 5 through hydrolysis and as such is unaffected by the availability of oxygen. Recently, aluminum has been shown to have an affinity for limestone and will precipitate within the drain. This leads to the clogging the drain which can inhibit water flow. The flow rate of water through a drain must be high enough to flush out precipitates, yet low enough so the residence time is sufficient for the limestone to be effective. Figure 3 shows a typical ALD. nine Water input — intercepted While Width = 2-9 f l Water is st i l l anoxic A - 2-4 f t fine soil / clay B - 20-40 ml plastic liner C - limestone bed / trench particles 1.5-4 inch d e U a - 5 f \ O ^ o ° < ? 0 * o " O ^ - Q ^ - ° v / t t d w t , \ „ ^ y r\ „ ^ V . J° ^ / treated Water i f * CaCD, • H* = C a " + HCD , $ * ^ / d ' s c h ^ 9 e length 100-2000 f t Figure 3. Anox i c Limestone Drain (Compi led from Gordon & Robinson, 1995; Skousen, et a l , 1995) Aerobic cells are built to encourage oxidation. They are used to collect water and remove metals in the form of oxides by providing residence time within the cell. In ARD waters, iron is primarily removed within an aerobic cell. Other metals removed are Al, Mg, As, and Hg. The shallow construction of the cell increases the surface area exposed to oxygen. Treatment occurs on the surface sometimes requiring large area depending on the input water 11 A Fuzzy Expert System on Acid Rock Drainage Site Remediation quality and the demands of the output quality. The shallow nature of an aerobic cells can result in undesired freezing during the winter months. To further increase the aerobic conditions, vegetation (reed beds) or green algae (release oxygen during photosynthesis) are added. Oxygen is introduced through the root structure of the vegetation facilitating metal precipitation. The vegetation also serves as a filter for suspended materials and the roots can hold the substrate together. This prevents channel formation and serves to increase mean residence time. Furthermore, vegetation adds aesthetic value to the treatment area. Figure 4 shows a typical aerobic cell. Figure 4. Aerobic Ce l l (Compiled from Gordon & Robinson, 1995; Skousen, et a l , 1995) Anaerobic cells work primarily to generate alkalinity by promoting reducing conditions. Treatment and water flow occur in the subsurface of the cell. Since the anaerobic cells are deeper than aerobic ones, they are able to function in sub-freezing climates. Sulfate-reducing bacteria (SRB's) such as desulfouibrio and desulfotomaculum play a large role by utilizing organic matter 12 A Fuzzy Expert System on Acid Rock Drainage Site Remediation as a carbon source. SRB's can generate up to 200 mg/L of sulfides (Faulkner and Skousen, 1995). By using sulfate as an electron acceptor for life processes, SRB's convert sulfate into hydrogen sulfide through the reaction (Robb and Robinson, 1995): so;2 +2CH20 = H2S+2HCO; 5. The process generates bicarbonate alkalinity. Metals such as iron, copper, sulfate, lead, zinc, mercury, cadmium, aluminum and uranium react with hydrogen sulfide to precipitate out as metal sulfides according to the example reaction: Zn+1 + H2S » ZnS + H+ 6 . Although this reaction appears to produce acidity, the sulfate-reducing reaction produces more alkalinity to produce a net increase in alkalinity for the system (Robb and Robinson, 1995). Anaerobic systems are constructed to facilitate growth of SRB's. An organic layer is present to promote anaerobic conditions. The layer consists of spent mushroom compost, manure, sawdust or other suitable materials available at the site. A floating vegetative cover can also be beneficial to the cell. The cover can decrease aerobic mixing, provide a supply of organic matter, and promote bacterial populations within the water column (Kalin and Smith, 1997). Many anaerobic cells also use a limestone layer below the organic layer to introduce further alkalinity. This would usually depend on the pH of the inlet water to keep it above 5 for bacteria to flourish; although the system does work at pH levels below 2.5. Providing optimum conditions to sustain a healthy bacteria population is key to a working anaerobic cell. Sugarcane 13 A Fuzzy Expert System on Acid Rock Drainage Site Remediation has been found to enhance the growth of SRB's by supplying acetate, formate, and lactate used by the bacteria as food ( Robb and Robinson, 1995). Adding a floating mat of vegetative cover to the anaerobic microbial cell brings many benefits. The technology has been implemented by Kalin in ARUM(acid reduction using microbiology) cells. The vegetation cover decreases mixing preventing oxygen from entering the cell. It is a constant source of new organic matter to replace the older material and sustain microbe populations. It provides an increased surface area for microbes, extending treatment above and below anaerobic sediments within the cell (Kalin, and Smith, 1997). Figure 5 shows a typical anaerobic cell. Mine Water input Sulphate Reduction 2CHgU + SD 2 = H gS + 2HCu 3 " H 2 S + In2* = ZnS + 2H + . Water <l-3 inches) Organic Matter (2-3 f t ) J ^ / L i ner d e p t h 2 " 5 ^ o l H T T T - ^ 3 = 7? t rea ted Water S O 0 • D o 2 " ' V ft r « fT- O # d ischarge y o _ v _ o f "° Linestone (1 f t ) , 0 o _ — Liner Figure 5. Anaerobic cel l (Compi led from Gordon & Robinson, 1995; Skousen, et a l , 1995) 14 A Fuzzy Expert System on Acid Rock Drainage Site Remediation E X P E R T S Y S T E M S What are Expert Systems? Expert systems are a form of artificial intelligence with the ability to transfer human intelligence into a computer program and provide expert knowledge to system users. (Zadeh, 1987) By simulating the thought process of an expert, intelligent decisions can be made by all users. In much the same way as humans use the concepts of hot, warm and cold, or may presume a probable decision, using inexplicit information, heuristics, and incomplete information within an expert system can help to decide on goals despite the presence of uncertainty (Fisher, 1986). A system is developed using the same terminology as the expert, thereby incorporating the jargon and vocabulary understood by those working in the field. Why is an Expert System useful for A R D Remediation? ARD remediation involves a multi-disciplinary field of study that includes: biology (Thiobaccilus ferrooxidans bacteria facilitates sulphide oxidation leading to ARD.); chemistry (involving oxidizing ARD reactions and water-solid reactions between the physical environment, minerals, organisms and weathering elements); mining (knowledge of mining practices and engineering); ecology (understanding interactions within the site ecosystem); and hydrology (knowledge of surface and groundwater flow), to name but a few. Expert systems can hold the knowledge and decision-making capabilities of multiple experts. New technologies are being developed almost daily by many people within the ARD research community. Those involved include government agencies, mining companies, private consulting firms and academic institutions. Much of this new information is eventually published 15 A Fuzzy Expert System on Acid Rock Drainage Site Remediation in the literature. An expert system can assist in transferring such new knowledge and information by making it available through continued addition to the program. Although there is no universally agreed upon method to deal with mitigation of ARD, an expert system can assist by developing a logical decision-making framework. How Are Expert Systems Built? Expert systems are built by a knowledge engineer working with an expert (or many experts) within a field of study. The project begins with a choice of domain for the system. This is important as a small domain may not be adequate or useful enough to be beneficial whereas one that is too large can may be too broad leading to unfinished areas of the system. Interviews with the expert and information gathered from other sources follows on from this beginning. Rules are developed through the interviews to capture the expert's knowledge into the program. The rules can link together facts and data required to decide on a goal. The system is then tested, re-evaluated by the expert, and revised as necessary in a continuing, on-going fashion. In ARDx, the decision was made to focus specifically on remediation leaving the field of prediction for future development. Components of the ARDx Expert System Keyword Triplets (k.w.t.) were used to represent facts and variables within the system. This was achieved by assigning an attribute and value to each object (e.g, "surface.runoff.high"). 16 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Each triplet has an associated Degree of Belief (DoB) - the value of which is determined by the input data which can be either quantitative or qualitative information. Data in k.w.t. variables can be stored as strings, floating point or integer numbers, dates, time or as a logical variable (Meech, 1992). For example: "site.topography. steep" has a degree of certainty in the concept of a steep slope. Inputs of site details and characteristics were placed into Fuzzy Sets. The concept of fuzzy sets was first introduced in 1965 by Lofti A. Zadeh (Zadeh, 1987). Through these fuzzy sets, quantitative inputs are assigned a membership value in a set or a DoB in a concept such as "low", "high", and/or "medium" (Meech, 1992, 1995)(Zadeh, 1987). Rules (LF-AND-OR-THEN-ELSE) are constructed from these keyword triplets to allow inferencing about sub-goals and goals in the system as it drives from input data through to final conclusions. These rules can be derived from interviews with experts using 2- or 3-dimensional F A M maps. Figure 6 shows an example of a rule within the knowledge base. 17 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Rule Editor rngiixi Add Edit Del Copy OK Cancel Exit Help NAME I mobility_h3 Priority T | H DWhy •interference 1 Life 1 Time Comment RULE IF contaminent.level.high AND species.form.neutral THEN mobility.factor.high is T CF=70.00 THEN mobility.factor.med is T CF=30.00 endRule • < I • Figure 6. Example of a rule within the knowledge base. A F A M map is a convenient way to depict associated rules that are used to determine the degree of belief in a concept from a number of variables (Meech, 1995). They are used to assess the input information and decide upon an appropriate cover choice. FAM maps are used to acquire a degree of belief in an output from two variables that are themselves determined through other FAM maps or from user input to the system. The DoBs of relative concepts such as "sensitive", "slightly sensitive" and "resistant" need not add to 100 as there may be an overlap in the belief of each concept and so rules can be assigned Certainty Factors (CF) as indicated within the FAM map. As the number of variables necessary to decide on a concept increases, the size and complexity of a FAM map also increases resulting in large 3-dimensional maps of a decision making process. Each position within a FAM map becomes a separate rule within the system. 18 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Procedures allow a series of events to be played out in a pre-defined sequence. The system accesses a particular procedure when called upon by a rule that has fired successfully. Procedures are used to perform calculations, assign variables or other specified tasks. Figure 7 depicts a procedure within the cost module of the system written in Comdale/X. Procedure Editor rrTBim Add Edit Del Copy OK Cancel Exit Help NAME | NPV DO NPV.top.@float = -1 + POW ((1 + real.rateofreturn.©float ), operating.years.@fl sat ) NPV.bottom.©float = real.rateofretum.©float * POW ((1 + real.rateofreturn.@flo at ), operating.years.@fl oat ) NPV.factor.@float = NPV.top.@float / NPV.bottom.@float totalNPV.operatingcosts.@float = total.operatingcosts.©float * NPV.factor.@float i | COMMENT Figure 7. Example of a Procedure within the cost module. The User Interface consists of pop-up forms, text boxes and hypertext documents. Through "forms", data consisting of drainage characteristics and site specific facts are input by the user and stored as k.w.t. An example of a FORM is shown in figure 8 below. The system communicates with the user in the event of an inconsistent inputs using text boxes. Once a conclusion is reached, all output is displayed within hypertext documents. These later items are also used to provide help files. 19 A Fuzzy Expert System on Acid Rock Drainage Site Remediation CLIMATIC IMPACT ST Done Undo Help A number of Impacts can result due to the climate at your site Please rate the following impacts from 1-10 according to their likelynood of occuring. 1 = never 5 = rarely 10 = often Drought | Flooding Avalanches Earthquakes Steady hydrologies] conditions Freezing (ice in winter months) Thermal Overturn (in area lakes) Wave action (in area lakes) Figure 8. Example of a F O R M used as a user interface. 20 A Fuzzy Expert System on Acid Rock Drainage Site Remediation SYSTEM CONFIGURATION Overall Approach ARDx is a fuzzy expert system developed in this thesis with the intent to assist the mining industry by producing a standardized adaptable approach to the problem, providing quick advice to the user and equipped for training and teaching. Likely users of the system include engineers responsible for environmental problems, consultants, professors and students. In its entirety the ARDx system is designed to handle ARD problems ranging from prediction through prevention and monitoring to treatment. The scope of the project to date deals with decision-making tactics concerning prevention and treatment at a site during one of three mining stages: planning, operating or closure. The components of the system come together in the knowledge base and inference engine through the use of rules and fuzzy concepts; and with the explainer engine providing reasoning, explanations and answers to user questions via the user interface. Figure 9 shows the components of ARDx. (Meech, 1998) r Outside models - data base - trend analysis - monitoring 1 inference Engine - fuzzy sets Explainer Engine - advice, answers, and explanations LU L Knowledge Base - rules Hypertext documents - electronic, interactive textbook - to be developed in the future Figure 9. Basic A R D E X configuration flow-chart. 21 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Outside modules can be added and used to provide inputs to the system. For example, currently the input for ARD potential considers only the neutralization ratio NP/AP. (NP is the neutralization potential or the capacity to consume acid, while AP is the acid potential or the potential to generate acid from a fixed sulfur content.) Outputs from a module that predicts ARD and considers not only the NP/AP ratio but also trend analysis, kinetic tests, and other static tests can conclude better on ARD potential and be used as an input to the system in place of the current NP/AP ratio. The knowledge base consists of a main system module (ARDx main) and numerous sub-modules that include: 1. Water module subaqueous deposition includes disposal in man-made impoundments, disposal in lakes and oceans, and relocating waste to a flooded open pit; 2. Covers module dry cover capping including simple soil cover, multiple cover, clay and synthetic covers for waste rock and tailings; 3. Diversion module control of water flow; 4. Passive module passive treatment methods including anoxic limestone drains, aerobic wetlands and anaerobic wetlands; 5. Treatment module collection and treatment of contaminated water including various chemical options. ARDx operates within the COMDALE/X environment, a Canadian developed software tool for expert system building (Comdale Technologies Canada Inc., 1997). The system contains a number of sub-fields, giving the user the ability to control the system. Applications for ARD 22 A Fuzzy Expert System on Acid Rock Drainage Site Remediation prediction, design for prevention, closure and site remediation have been considered. The system is able to consider design requirements, quality assurance standards, the extent of the potential problem, available resources and site-specific heuristics. Output provides quick advice to probable treatment options and an initial cost analysis for implementation. Development of an expert system requires: 1. a clear definition of the problem and domain of the system 2. knowledge acquisition 3. system development (programming steps) 4. testing and verification of the system. (Meech and Kumar, 1992) The first step often poses the most obstacles. Initially the system domain was designed to include all treatment possibilities. During knowledge acquisition however, it became clear that the chosen domain was extremely large. Rather then change domains, it was decided to focus on two aspects, an overall structure to the decision-making process (developed through ARDx main) and a more detailed evaluation of separate treatment options (sub-modules). In this way a working system can be setup while separate modules containing new and additional treatment options can be added, revised or discarded as required in the future. ARDx main is the seed to the development of a larger and more complete system. Figure 10 shows the basic ARDx flowchart that has been instrumental in the organization and development of the system. 23 A Fuzzy Expert System on Acid Rock Drainage Site Remediation U S E R I N T E R F A C E Information Planning Operating Closure Waste Rock Tai l ings I Mine Workings Open Pit W A T E R C O V E R S l A C T I V E M E T H O D S • Covers • Sulphide Reduct ion* • Col lect & Treat • Bacter ic ides* F L O O D I N G * • Seal ing* M I G R A T I O N C O N T R O L I A C T I V E M E T H O D S • Col lect & Treat N O P A S S I V E M E T H O D S I Have methods been looked at together? YES N O C O M P A R E • cost • probabil ity o f success Is a suitable method available? * yet to be completed N O YES D E S I G N Select Col lect & Treat Figure 10. Bas ic A R D x Flowchart Outputs for each method (possibility of success) and cost (calculated separately), are compared through FAM maps to give a degree of belief (DoB) in the application of each particular treatment option. Initially the knowledge acquisition phase included choosing and interviewing the expert as well as conducting a literature search on the topic of ARD treatment. Through the interviews, the framework of the system was established. Mory Ghomshei served as the expert in the system development. Mory Ghomshei has worked on ARD mitigation for the Samatosum mine site in B.C., Canada (Ghomshei, 1997). Expertise was also taken from numerous and sometimes contradictory publications. The challenge of building a system on ARD has been to acquire rules in a field where decisions are presently being made by trial and error. Case studies were used to 24 A Fuzzy Expert System on Acid Rock Drainage Site Remediation attempt to mimic the actual decision-making process. Acquiring expertise is an ongoing process as the system develops and expands into new areas. System development and programming began by describing all concepts acquired through the knowledge acquisition phase according to 2-4 states, with most concepts having 3 states (eg: high, medium, low). Rules were developed using Fuzzy Associative Memory maps. Three dimensional FAM maps are generated in cases where two inputs combine to generate a single output. Two inputs each having 3 states generate a maximum of 9 rules each with a possibility of up to 3 conclusions based on the output variable having 3 states. Figure 11 shows two variables, socio-environmental impact and mobility of the pollutant, coming together to generate one output for environmental sensitivity. Up to three different conclusions are available for each circumstance. ENVIRONMENTAL SENSITIVITY Socio-environmental Impact L M H H s = 30 s = 70 s= 100 ss = 70 ss = 40 ss = 0 Mobility r= 10 r= 0 r = 0 of the M s = 0 s = 20 s = 60 Pollutant ss = 50 ss=90 ss = 50 r = 60 r= 10 r=0 L s = 0 s = 0 s = 30 ss = 0 ss = 40 ss = 70 r= 100 r = 70 r= 10 H = high, M = medium, L = low; s = sensitive, ss = slightly sensitive, r = resistant Figure 11. Example o f a F A M map. If a third variable is necessary to generate a single output then FAM maps were created for the three states of the first two variables for each state of the third variable. As more variables are introduced, the FAM maps become increasingly more complex to work with. With four or more variables needed to decide on a conclusion, a single fuzzy-neural A Fuzzy Expert System on Acid Rock Drainage Site Remediation rule using an inference equation was adopted as shown in equation 7 (Veiga, 1994). The weights for each variable state were defined by the expert as were the certainty factors for the rule conclusions within the FAM maps. DoB of output variable = £ [(DoB of each input variable state) * weight] 7. The DoB of the output variable was then mapped into fuzzy sets for each state of the variable. Figure 12 shows the fuzzy set which maps the DoB acquired for the output variable "dob.envirofactors.@f into the concept of high. @source = dob.envirofactors.@float Value Rank 5 0 7 50 10 100 Figure 12. Fuzzy set for "enviro.factors.high" 26 A Fuzzy Expert System on Acid Rock Drainage Site Remediation KNOWLEDGE BASE MODULES The flow charts in this section detail the FAM map inputs and outputs for each of the sub-modules. Two, three or four states for each input were used to create rules using 2-dimensional or 3-dimensional maps. Each rule in a map contains at least one state of one output variable but in some cases, multiple states are concluded with less than full certainty. Appendix A contains each FAM map showing the expertise regarding the DoB in each output state. Also indicated are those relationships which have been characterized using a fuzzy-neural approach. Appendix B details the specific fuzzy set definitions for each variable state. A R D x Main Module The ARDx main module connects with the sub-modules and drives the system. All interaction with the user is also done through this module by requesting the user for site specific data input and eventually presenting system advice and conclusions. The user begins the dialog by choosing the purpose of the current consultation. The options include: - examination of a specific method - detailed cost analysis - complete evaluation and cost analysis of the site (default) - examination of hypertext documentation on ARD mitigation This gives the system flexibility in tailoring the session to specific needs. According to the users' choice for the session, appropriate forms are called upon for data input. Inputs can be 27 A Fuzzy Expert System on Acid Rock Drainage Site Remediation quantitative where the user provides a measured value of a particular variable, or qualitative means of providing information allow an evaluation of mitigation based on heuristics or missing data. ARDx moves through the sub-modules to determine if an appropriate treatment option can be found. When a complete evaluation and cost analysis is selected, ARDx "main" decides on the final recommended treatment options for the site by rating the technical success (degree of belief), capital costs required and the net present value costs to implement each option. The overall rating for each option is determined based on the weighted equation below. Overall Rating = (T W T + Cc W C C + NPV WNP V ) / ( W T + W C C + WNP V ) 7b. Where: T = technical rank W T = technical weighting Cc = capital costs rank Wc c = capital costs weighting NPV = net present value rank WNPV = net present value rating The user is asked to provide the weightings used in calculating the overall rating for each option with the default ratings set at 75, 50 and 50 for technical, capital costs and NPV respectively. So essentially, the system provides a combined ranking that accounts for the technical ranking as being 50% more important than either capital costs and NPV rankings, but if the users wishes, these weight can be adjusted to suit unique situations. The output hypertext display lists the recommended treatment options (of which there may be more then one), the probability of success, ratings and the costs of each option. The user can "click" through the document to obtain justification for each recommendation and for 28 A Fuzzy Expert System on Acid Rock Drainage Site Remediation information on how the decision was made. A break down of costs for each option can be seen by "clicking" on "COSTS". As well, a detailed explanation / justification module of the reasoning behind the technical rating is available for examination. This module provides a window into the rules and fuzzy inference structure of the entire system. The output shows the linkage between each FAM map and by clicking on each FAM map block, the fuzzy inputs and outputs can be seen. Waste Rock and Tailings Water Module The water module deals with subaqueous deposition. Both rules and procedures were used in the decision-making process. The FAM maps were created using a weighted average method within a fuzzy-neural rule. The options evaluated by the module include submarine disposal within a lake or ocean, flooding within a man-made enclosure (tailings dam), and the removal of waste to an open pit for flooding. Decisions are made by examining site characteristics, environmental concerns, the quality of waste and failure risks. Most inputs provide the user with a choice of answering either quantitatively (eg. an annual precipitation) or qualitatively (eg. moderate precipitation). A number of strictly qualitative inputs to the module are provided by the user through a rating system that places each factor somewhere on a scale from high to low. Inputs that determine mobility and toxicity are provided in this manner. The presence of a contaminant in the waste and its form whether reactive or non-reactive are rated by the user. The module takes a conservative approach in using this information and rates the overall contaminant level or species form as high when any of the listed contaminants are input 29 A Fuzzy Expert System on Acid Rock Drainage Site Remediation as high. Other inputs evaluated in this manner include events with climatic impact that are used to decide on the risk of failure to an enclosure. The module outputs a DoB in the possibility of using sub-aqueous deposition for a given scenario. Data exported to ARDx main also includes DoBs in the specific type of recommended option. Figure 13 shows the water module flowchart. 30 A Fuzzy Expert System on Acid Rock Drainage Site Remediation SETTLING PROPERTIES TURBIDITY - day distribution - settling properties w - size distribution - wave action SITE CHARACTERISTICS - water bodies available - lake depth ENVIRONMENTAL QUALITY PHYSICAL /CHEMICAL - increased turbidity - excessive nutrient additions - physical impact from placement on habitat - toxicity of reagents and heavy metals MOBILITY - contaminant level - species form SPECIES FORM - aluminum -arsenic - cadmium - chromium - copper - cyanide - lead - magnesium - manganese - mercury - uranium - Plutonium -other CLIMATIC IMPACT -wave action -ice -drought - flooding - avalanches - earthquakes - thermal overturn TOTAL PRECIPITATION - precipitation snow - precipitation rain TOXICITY - mobility - enhancement ENHANCEMENT - organics level - organics form CONTAMINANT LEVEL - aluminum - arsenic - cadmium - chromium - copper - cyanide - lead - magnesium - manganese -mercury - uranium - plutonlum -other TAILINGS QUALITY - present oxidation - add potential ENVIRONMENTAL SUSTAINABILITY - tailings quality - waste quality - environmental quality social - environmental quality physical / chemical SUB - MARINE DISPOSAL ENVIRONMENTAL QUALITY SOCIAL - social acceptance - political acceptance FLOOD TAILINGS WATER AVAILABLE -water table -total precipitation - catchment area - surrounding permeability WATER COVER OPTION FAILURE RISK - dam stability - climatic impact - maintain water cover WASTE QUALITY - present oxidation - add potential WASTE ROCK PILE SITE CHARACTERISTICS - pit / waste volume ratio - water available -pit stability WASTE TO PIT AND FLOOD PIT QUALITY - present oxidation - add potential Figure 13. Water module flowchart. 31 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Cover Module The covers module decides on dry cover capping. It consists of two separate decision-making processes: one for waste rock and one for tailings. The module exports a DoB in using dry covers as a control method as well as the DoB in the appropriate type of dry cover to use. Cove rs flowchart for waste rock The flowchart for waste rock evaluates for a simple soil cover, multiple layer cover, or a complex cover (with a synthetic or clay layer). A cover is decided upon based on the volume of effluent exiting the waste rock, the quality of cover needed and the environmental sensitivity of the area. Most inputs to the module allow the user to answer either quantitatively or qualitatively, while a few allow for strictly a qualitative answer. The inputs come together through FAM maps (set up as rules within the module) to decide upon sub-goals. The flowchart is shown in Figure 14 below. 32 A Fuzzy Expert System on Acid Rock Drainage Site Remediation POTENTIAL UNDERFLUX - underlying peuneablty - waste dump peuneablty -waste dump tner PRECIPITATION AVAILABLE FOR N FILTRATION -snowfal -rainfall TOTAL PRECIPITATION -snowtal SURFACE INFLUX - catchment ansa - permeabnty - total preulpKsUon -topography SUBSURFACE INFLUX -water table -topographical re lef -potential underflux EFFLUENT VOLUME -surface kiflux -subsurface kiflux -water liilttrauon WATER AVAILABLE FOR INFILTRATION - predplatlon avaDablB for InlBratlon -pondhg WATER INFILTRATION - waste dump permeabUy - waste surface area - water ava labia for bifftraHon PONDING - ratio: pondkig / surface area - local dknate COVER QUALITY - avaJaUe caplal -ARD potential - treatment plant capacity EFFLUENT MOBILITY -water table - water gradient direction -water bodies In area -scavenger material ENVIRONMENTAL SENSITIVITY - effluent mobHy - scdo-envlranmental Impact SOCOENVIRONMENTAL IMPACT - distance to populated area - protected regions COVER CHOICE - effluent volume - cover quaBy - envlonmental aensMvfty Figure 14. Covers module flowchart for waste rock. Covers flow sheet for tailings The tailings flow chart for dry covers evaluates for two options: vegetation cover and a multiple layer cover with a drainage layer. The rules dealing with tailings were created through weighted equations within procedures while some were created directly from FAM maps. The flowchart detailing the inputs and outputs is shown in Figure 15 below. 33 A Fuzzy Expert System on Acid Rock Drainage Site Remediation SULPHIDE REACTIVITY - pyrrotite -pyrite - chalcopyrite TOTAL PRECIPITATION - precipitation snow - precipitation rain MINEROLOGY - sulphur content - sulphide reactivity - buffering capacity HYDROLOGY - catchment area - topography - water table - underlying permeability - surrounding permeability - total precipitation TAILINGS CHARACTERISTICS - size of tailings pond - grain size SOURCE FACTORS - mineralogy - tailings characteristics DRY COVER - hydrology - source factors - environmental quality TAILINGS QUALITY - interstitial waters -toxins - acid potential ENVIRONMENTAL QUALITY - environmental sensitivity - tailings quality TOXICITY - mobility - enhancement ENHANCEMENT - organics level - organics form ENVIRONMENTAL SENSITIVITY - effluent mobility - socio-environmental impact MOBILITY - contaminant level - species form CONTAMINANT LEVEL - aluminum - arsenic - cadmium - chromium -copper - cyanide -lead - magnesium - manganese -mercury - uranium - plutonium -other SPECIES FORM - aluminum - arsenic - cadmium - chromium -copper - cyanide -lead - magnesium - manganese -mercury - uranium - plutonium -other EFFLUENT MOBILITY - water table - water gradient - water bodies in area - scavenger material SOCIO-ENVIRONMENTAL IMPACT - distance to populated area - population size - protected regions Figure 15. Covers module flowchart for tailings. The inputs and decision making process used to evaluate for toxicity are the same as those discussed previously in the water module. Like the covers flow-chart for waste rock, the tailings flow-chart allows the user a choice in answering qualitatively or quantitatively for most inputs. 34 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Treatment Module The treatment module investigates the need to collect and treat contaminated water and suggests the appropriate chemical to use including estimated quantity. The treatment module flowchart is displayed in Figure 16 below. DRY COVER - cover module output WATER TREATMENT - water module output EFFLUENT ARD -ARD potential - aluminum - ferric iron - ferrous iron - manganese PASSIVE TREATMENT - passive module output TREATMENT OPTION -effluent ARD - dry cover - water treatment - passive treatment REQUIRED pH - aluminum - manganese - ferric iron - ferrous iron NOT AN OPTION TREATMENT CHEMICAL - required pH - effluent flow - effluent addtty - manganese Figure 16. Treatment module flow sheet. The first step in developing the treatment module was to write a set of preliminary rules for the required pH modification based on user input of ARD characteristics. Inputs of metal concentrations and required discharge limits that must be met for the site are grouped together and placed into fuzzy sets. Through these fuzzy sets, inputs are assigned a membership value in various sets to calculate DoBs in the concept of "regulation", "low", "medium", and "high" 35 A Fuzzy Expert System on Acid Rock Drainage Site Remediation concentrations. A concentration described as "regulation" is at or below the discharge limit set by the user while a concentration described as "low" is just above the discharge limit set by the user. "Medium" and "high" concentrations follow. A graph of the fuzzy sets for aluminum concentrations is shown in Figure 17. The input to the sets is an equation that scales the concentration such that the set changes depending on the specific regulations as seen below. (Al conc.-regulation Al conc.)/regulation Al cone. = fractional difference from regulation value 8. This allows the system to respond and account for different regulations at different sites. FUZZY SETS FOR ALUMINUM CONCENTRATIONS Fract iona l di f ference from user input regulation value —«— regulation — a — low —A— medium o high Figure 17. Fuzzy Sets for aluminum concentrations based on the fractional difference from the user input regulation value. Defuzzification is done using a Weighted-Average Method. The certainty of the required pH change of low, medium and high is used to determine the required pH for treatment. Equation 9 shows the calculation done in ARDx using the weighted average method. 36 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Required pH change = Z [DoB(effluent.pH.*) ] / Z[DoB(effluent.pH.*)*(*.supremum.@f)]] 9. where * = Neutral, Slightly-Basic, and Basic for each k.w.t. Sulphide Reduction Sub-Module Sulphide reduction investigates options to control ARD through reducing the amount of sulphide exposed to oxidation. This primary type of control can be achieved through: 1. Removal of sulphides from the waste by flotation - This process is under development and may not be economically feasible as selective disposal of the concentrate is still required. 2. Encapsulating or coating of sulphides - Still in experimental stages 3. Segregation and/or blending of sulphide wastes - Success will depend on • water flows • the degree of mixing of different rock types • quantity of each rock type • reactivity of different materials (both acid generating and acid consuming) 37 A Fuzzy Expert System on Acid Rock Drainage Site Remediation The methods can be particularly useful for dealing with closure issues in an underground mine in which only partial flooding is likely to be practical (Lawrence, 1997). This module will be completed in the future. Bactericides Sub-Module These methods control ARD production through the application of bactericides. Bactericides work to inhibit biological oxidation of sulphide minerals. Although not a likely control method for long term it may be a possible short term control used in combination with other methods (Delaney T., D. Hochley, D. Sollner, 1997). Success of the method depends on: 1. type of sulphide mineral in the waste 2. level of biological influence 3. weather conditions (chemicals are removed by water requiring reapplication) Other considerations include: 1. bactericides do not inhibit chemical oxidation 2. degradation occurs over time requiring reapplication 3. method works best when applied to a fresh unoxidized surface 4. method can provide temporary relief until a permanent dry barrier is put in place The bactericides module will be completed in the future. 38 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Mine workings and open pit Flooding sub-module The flooding module investigates flooding an open pit or mine workings to control ARD, generally a closure option for a mine. Flooding of mine workings is often a good long-term closure option (Stephen et al, 1991). Success depends on: 1. strength and stability of plugs and sealants (risk of failure) 2. site hydrology (seasonal fluxes and discharge control) 3. extent of drilling in area (mountainous terrain may preclude sealing of drill holes) 4. wall rock quality, presence of paste backfill and/or other backfill material 5. completeness of the flooded workings Flooding of an open pit will depend on: 1. water balance and site hydrology 2. rock quality within the pit 3. presence of pit wall oxidation 4. seasonal fluctuations in water level 39 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 5. short term release of metals The flooding module will be completed in the future. Sealing sub-module The sealing module investigates the use of plugs and sealants to enclose mine workings. This module will be completed in the future. Treatment sub-module The treatment sub-module is used to deal with both waste and tailings treatment options. The only difference is that success depends on identifying and collecting all effluent flows. Diversion Sub-Module The diversion module investigates options to control water migration. It takes into consideration treatment and control options already recommended to decide on whether diversion of surface water or interception of ground water both upstream and downstream of the waste dump would be beneficial. In addition to considering recommended mitigation options, other inputs to the module include precipitation, topography of the area, the size of the catchment area and permeability. Where the user has chosen to look only at the water diversion option, the module adjusts to decide on water diversion using only the user inputs provided without looking at other options (dry cover, collect and treat, and passive treatment). A flowchart of the diversion module is shown in Figure 18 below. 40 A Fuzzy Expert System on Acid Rock Drainage Site Remediation TOTAL PRECIPITATION - precipitation snow - precipitation rain SURFACE FLOW - topography - catchment area - permeability - preclpitational total DRY COVER - dry cover module output SUB-SURFACE FLOW - topography - catchment area - permeability COLLECT AND TREAT - treatment module output UPSTREAM OF WASTE INTERCEPTION - dry cover - surface flow - subsurface flow UPSTREAM OF WASTE DIVERSION DITCH / BERMS - dry cover - surface flow - subsurface flow i t * ± PASSIVE TREATMENT - passive module output DOWNSTREAM OF WASTE INTERCEPTION - surface flow - subsurface flow - collect and treat - passive treatment DOWNSTREAM OF WASTE DIVERSION DITCH / BERMS - surface flow - subsurface flow - collect and treat - passive treatment Figure 18. Divers ion module flowchart. Passive Sub-Module The passive module considers a number of passive treatment methods including anoxic limestone drains, aerobic wetlands and anaerobic wetlands. It works in a 2-step method: first deciding on whether passive treatment is a likely option for the site, and secondly, moving through a decision-making process on the type of treatment necessary if the first decision is positive. The rules for the later part of the module were developed according to a flow-chart from Skousen et al. and can be viewed within ARDx by choosing the information for passive treatments option. The module can recommend one method or a series of techniques that may work best together. Suggestions on aerating the water or providing a settling pond are also made. The inputs to the module include hydrology information, the amount of land area available, 41 A Fuzzy Expert System on Acid Rock Drainage Site Remediation contaminant concentrations, seasonal freezing, and the amount of dissolved oxygen within the drainage. A flow sheet of the passive treatment module is shown in Figure 19. TOTAL PRECIPITATION - precipitation snow - precipitation rain HYDROLOGY - catchment area - topography - water table - underlying permeability - surrounding permeability - total precipitation PASSIVE TREATMENT - land area - hydrology - contaminated concentration - water flow - seasonal ice - manganese present NOT AN OPTION SETTLING POND -ARD potential ANOXIC LIMESTONE DRAIN -ARD potential - dissolved oxygen - ferric iron - aluminum AEROBIC / ANAEROBIC WETLAND -ARD potential - dissolved oxygen - ferric Iron AERATE - water pH -ARD potential - dissolved oxygen - ferric iron AEROBIC - water pH -ARD potential - dissolved oxygen -ferric iron Figure 19. Passive treatment module flowchart. 42 A Fuzzy Expert System on Acid Rock Drainage Site Remediation COST MODULE Introduction To Cost Module A separate cost module is accessible by all sub-modules as required. Calculated costs for an option can be input for each module. The cost for each remediation option is calculated using unit prices (MEND, 1995) and site-specific information. Detailed information on cost calculations for each sub-module is provided within the following tables. The itemized costs are described individually following their corresponding table. Water Cover Costs Water cover costs are calculated for four different scenarios: flooding of taillllings within a man-made enclosure, relocating waste rock to an open pit and flooding, and sub-marine disposal within an ocean or lake for both tailings and waste rock. Tables 1 through 4 below provide detailed information for each item considered. WATER COVER - FLOOD TAILINGS CAPITAL COSTS Item KeyWord Triplet Value Equation 1 unitprice.construct_dams.@f ($ / m3) 12 unitprice.construct_dams.@f * construct dam.volume.(S!f 2 unitprice.constructdamsother.@f (%) 16 construct.dams_fills.@f * unitprice.constructdamsother.(S3f /100 3 unitprice.regrade_surface.@f ($/m 3) 2 unitprice.regrade_surface.@f * tailings.surfacearea.@f * regrade.area.@f /100 * regrade. depth. @f /100 4 unitprice.coverw.@f ($/m 2) 6 unitprice.cover.@f * tailings.surfacearea.@f 5 unitprice.spillways.(2)f ($) 100.000 unitprice. spill ways. (SJf OPERATING COSTS 6 unitprice.totalmaintenanceinspection.@f (%) 15 unitprice.totalmaintenanceinspection.@f / 100 * total. capital costs. @i Table 1. Cost details for flooding o f tail ings. 43 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Costs for flooding of tailings are compiled from MEND Report 5.8.1, Table la. Itemized costs are as follows: Item 1 - Construction of perimeter dams. Item 2 - Other costs incurred for construction of perimeter dams including site & foundation preperation, bedrock grouting and foundation treatment. These additional costs are estimated as a percent of the cost for actual dam construction (Item 1). Item 3 - Regrading of the tailings surface for the preperation of a water cover. Item 4 - Water cover costs including a sand layer over the tailings. Item 5 - Provision of spillways. Unlike the other costs, this cost is assumed at a fixed number ($100,000). The number is not scaled to be sitespecific as it accounts for no more then 0.7% of the capital costs (most conservative estimate as seen in MEND Report 5.8.1) Item 6 - An operating cost for maintenance and inspection is calculated as a percent of the capital costs. The item is scaled to each specific site according to the value of the calculated capital costs. WATER COVER - IN PIT FLOODING OF WASTE ROCK CAPITAL COSTS Item KeyWord Triplet Value Equation 9 unitprice.haulage.@f ($ / tonne) 1.5 unitprice.haulage.@f * waste.tonnes.@f 8 nmtprice.other.@f (%) 30 unitprice.other.f^ f * tailiog.haulage.(S}f /100 5 unitprice.spillways.(S)f ($) 100,000 unitprice.spillways.@f OPERATING COSTS 6 unitpn^ e.totalmaintenanceinspection.@f (%) 15 unitprice.totalmaintenanceuispection.@f / 100 * total, capital costs.(2d' Table 2. Cost details for in-pit f looding o f waste rock. 44 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Costs for in-pit flooding of waste rock are compiled from MEND Report 5.8.1 Table B.5. Itemized costs are as follows: Item 5 - as above itemized listing Item 8 - Other costs incurred including restoration of disturbed areas, preparation works, and water treatment (lime addition). Calculated as a percent of the total costs for haulage. Item 9 - Haulage costs are given based on the waste rock being in close proximity to the disposal site. WATER COVER - SUB-MARINE DISPOSAL OF TAILINGS CAPITAL COSTS item KeyWord Triplet Value water table water table , , , , „ , water table deep shallow moderate Equation 7 unitpri ce.haulage.@f ($ / m3) 5 3 1.5 unitprice.haulage.@f * tailings.depth.@f * tailings, surfacearea. @f 8 unitprice.other.(3}f (%) 13 unitprice.other.@f * tailing.haulage.@f / 100 10 uni tprice.haul distance. @f ($/m3km) -1 unitprice.hauldistance.@f * waterbody.distance.@f Table 3. Cost details for sub-marine disposal o f tailings. Costs for the sub-marine disposal of tailings are calculated based on costs given for a typical tailings deposit depth of 10m. Itemized costs are as follows: Item 7 - Cost for haulage of tailings scaled to be site specific according to the depth of the water table. The costs are based on the tailings being a distance of one kilometer from the disposal site. Item 8 - as above itemized listing 45 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Item 10 - An additional cost adjustment to the haulage costs according to the distance (further then one kilometer) of the tailings from the disposal site. This cost is based on a 20m3 haul load at $200/100km. WATER COVER - SUB-MARINE DISPOSAL OF WASTE ROCK CAPITAL COSTS Item KeyWord Triplet Value Equarion 10 uni tprice.haul distance. @f ($ / tonnes km) 0.08 unitprice.hauldistance.@f * waterbody.distance.@f 8 unitprice.other.@f (%) 30 unitprice.other.@f * tailing.haulage.@f / 100 9 unitprice.haulage.@f ($ / tonne) 1.5 unitprice.haulage.@f * waste.tonnes.@f Table 4. Cost details for sub-marine disposal o f waste rock. Costs for sub-marine disposal of waste rock are similar to those for tailings. Itemized costs are as follows: Item 8 - as for tailings above Item 9 - Haulage costs are given based on the waste rock being in close proximity to the disposal site. Item 10 - As in tailings above, a cost adjustment (to item 9) according to the distance from the waste rock pile to the disposal site. Based on a 25 tonne load at $200/km. Dry Cover Costs Dry cover costs are calculated for six different scenarios: vegetative cover, simple soil cover and a multiple cover with a drainage layer for tailings; simple soil cover, multiple layer cover and complex cover for waste rock. Tables 5 & 6 below provide detailed information for each item considered. 46 A Fuzzy Expert System on Acid Rock Drainage Site Remediation TAILINGS DRY COVER CAPITAL COSTS Item KeyWord Triplet Value multiple soil complex cover simple soil Equation 5 unitprice. spill ways. (S?f ($) 70,000 unitprice.spillways.^ f 11 unitprice.regradeslopes.@f ($ / m3) 4 4 4 unitprice.regradeslopes.@f * regradepercent.slope.@f / 100 * tailings.surfacearea.@f * regrade.area.@f /100 * regrade.depth.(ffif / 100 12 unitprice.regrade_surface.@f ($ / m3) 2 2 2 umtprice.regrade_surface.@f* tailings.surfacearea.@f'1 regrade.area.fSJff 100 *regrade.depth.(SJ7100 13 unitprice.cover.(S)f ($ / m") 26 29 12 unitprice.cover.@f * tailings.surfacearea.@f 14 unitpri ce.coverslopes.@f (%) 20 20 25 unitprice.coverslopes.^ f / 100 * cover.surface.(3>f 15 unitprice.runoffchannel.^ f ($ / m) 40 40 40 unitprice.nmoffchaimel.f^ Prunorrchannel.length.^ f OPERATING COSTS 6 unitpri ce.totalmaintenanceinspection.@f (%) 15 unitprice.totalmaintenanceinspection.@f /100 * total, capital costs.(S}f Table 5. Cost details for tailings dry covers. Costs for tailings dry covers were compiled from MEND Report 5.8.1 Appendix A. These are based on Scenario 1; a base metal sulphide mine and Scenario 2; an active gold mine. Itemized costs are as follows: Item 5 - as for water cover above. Item 11 - regrading of sloped areas. Scaled to the site according to the volume of sloped material to be regraded. Item 12 - regrading of tailings surface. Scaled to the sited according to the volume of sloped material to be regraded. Item 13 - provision of a dry cover. Cost varies with different types of cover (i.e. simple, multiple, or complex). Item 14 - provision of a dry cover for sloped areas. Calculated as an additional cost based on a percentage of the cost to provide a cover for flat areas (item 13). 47 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Item 15 - provision of a runoff channel based on the length required. Item 6 - Described in section for water cover costs. WASTE ROCK DRY COVER CAPITAL COST Item KeyWord Triplet Value multiple soil complex cover simple soil Equation 11 unitprice.regradeslopes.(2>f ($ / m3) 3 3 3 unitprice.regradeslopes.@f * regradeslope.volume.@f 12. unitprice.regrade_surface.@f ($ /m3) 2 2 2 unitprice.regrade_surface.@f* waste.surfacearea.@f* regrade.area.(5>i/ 100 *regrade.depth.(3)i7100 13 unitprice.cover.(Slf ($ / m2) 25 30 7 iinitprice.cover.@f * waste.surfacearea.@f 14 unitprice.coverslopes.(S)f ($ / m2) 25 35 8 unitprice.coverslopes.@f / 100 * cover.surface.@f OPERATING COSTS 6 uiu^ rice.totalmaintenancemspection.@f (%) 15 unitpn .^totalniaintenanceinspection.@fv / 100 * total, capital costs. @f Table 6. Cost details for waste rock dry covers. Costs for waste rock dry covers were compiled from MEND Report 5.8.1. Based on Scenario 5 - a waste rock dump. Itemized costs are similar to those for tailings dry covers and have been described above. Treatment Costs Water treatment costs are calculated for convention lime treatment. Table 7 below provides detailed information for each item considered. WATER TREATMENT CAPITAL COST Item Key Word Triplet Value high acidity medium acidity low acidity Equation 17 unitprice.wwtpm.@f ($h / m3) unitprice.wwtpb.(2!f ($) m=0.00264 m=0.00169 m=0.00164 b=1.1448 b=0.9571 b=0.7158 unirprice.wwtpm.@f*efiluent.flow.@f+ unitprice.wwtpb.@f* 1.000,000 OPERATING COSTS 18 unitprice.sludge.(^ f (%) 15 unitprice.sludge.@f*treat.volume.{^ f 19 treatw. volume.@f user input value effluent.flow.@f * 2002.5 20 unitprice.treatw.f^ f 0.6 | 0.5 | 0.4 unitprice.treatw.@f*treatw. volume. @f 6 umtprice.totalmaintenanceinspection.@f (%) 15 unitprice.totalniaintenanceinspection.@f / 100 * total, capital costs.(ff)f Table 7. Cost details for water treatment. 48 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Costs for treatment were compiled from MEND Report 5.8.1 and Lawrence 1997. Itemized costs are as follows: Item 6 - Described in section for water cover costs. Item 17 - Capital costs including building & foundation, neutralization circuit, solid/liquid separation, polishing pond costs, instrumentation, engineering and overheads, spare parts, and contingency of 25%. (Lawrence, 1997). Costs are scaled to be site specific according to the acidity and flow rate of the effluent to be treated. The variables given in the table represent equations compiled from a graph of capital costs versus flow rate at various acidity concentrations. Item 18 - Sludge disposal costs calculated as a percentage of the water treatment costs. Item 19 - Calculation for treatment volume based on the effluent flow (gpm) and adjusted to m /year. Item 20 - Cost of water treatment based on the volume to be treated and acidity. Diversion Costs Water diversion costs are compiled form MEND Report 5.8.1. Details are presented in table 8 below. 49 A Fuzzy Expert System on Acid Rock Drainage Site Remediation WATER DIVERSION CAPITAL COST item KeyWord Triplet Value Equation 15 unitprice.seep_collectionditches.@f ($ / m) 1000 (unitprice.seep_coUectionditches.@f*collectionditch.lengt h.@f)+ (unitprice.seep_collectionditches.@f collectionditch.len)jth.@f * 0.050000 ) 16 unitprice.diversiondyke.@f ($ / m) 2000 unitpri ce.diversiondyke.@fltdiversiondyke.length.@f OPERATING COSTS 6 unitpri ce.totalmaintenanceirispection.@f (%) 15 unitprice.totalmaintenanceinspection.@f / 100 * total, capital costs.(5}f Table 8. Cost details for water diversion. Itemized costs are as follows: Item 6 - Described in section for water cover costs. Item 15 - Cost for providing seepage collection ditches including a grout curtain along the ditches and based on the length required. Item 16 - Cost for providing a diversion dyke based on the length required. Passive Treatment Costs Passive treatment costs are calculated for three different treatment methods: anoxic limestone drain, aerobic wetland and anaerobic wetland. Tables 9 & 10 below provide detailed information for each item considered. PASSIVE TREATMENT - ANOXIC LIMESTONE DRAIN CAPITAL COST Item KeyWord Triplet Value Equation 21 unitprice.ald.@f ($ / flow*acidity) 1.6*.0022 unitpri ce.ald.@f*1.6*0.0022*1000*effluent.flow.@f*efB uent.toxicity.fSJf OPERATING COSTS 6 unitprice. totabnaintenanceinspection. @f (%) 15 unitprice.totalmaintenanceinspection.@f /100 * total, capital costs.@f Table 9. Cost details for anoxic limestone drain. 50 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Costs for anoxic limestone drain are compiled from data in Skousen 1991. The drain is assumed to have a longevity of 20 years with a 90% CaCC>3 content and 75% limestone dissolution. Itemized costs are as follows: Item 6 - Described in section for water cover costs. Item 21 - Scaled according to the flow and acidity of the effluent by a graph of capital costs versus tons/year of acid (developed form data in Skousen 1991). Adjustments are made for unit conversions from input data. P A S S I V E T R E A T M E N T - A E R O B I C O R A N A E R O B I C W E T L A N D C A P I T A L C O S T Item Key Word Triplet Vaiue Equation 22 unitprice.aerobicsize.@f ($ / ac) 480,000 unitprice.aerobicsize.@f*effluent.flow.@f*(conc.fe.@f*0. 012/180+conc.nm.@f*0.012/9+efflumt.addity.@f*0.012/ 60) 23 unitprice.anaerobicsize.@f ($ / m 2) 120 unitprice.anaerobicsize.@f*(e£Quent.flow.@f*e£Quent.aci dity.fijf* 7.8377) O P E R A T I N G C O S T S 6 unitphce. totalmaintenanceinspection. @ f (%) 15 unitprice.totalmauitenanceinspection.@f / 100 * total, capital costs. (ai,i Table 10. Cost details for aerobic and anaerobic wetland. Item 6 - Described in section for water cover costs. Item 22 - Providing an aerobic wetland. The size of the wetland required is based on the effluent flow, acidity, iron concentration and manganese concentration (an equation according to the U.S. Bureau of Mines criteria for AML sites). The cost of $480,000/acre is calculated based on the capital cost of constructing the Z&F wetland. (Faulkner, B.B., and J.G. Skousen, 1995) Item 23 - Providing an anaerobic wetland. The size of the wetland is based on the effluent flow and acidity (equation according to the U.S. Bureau of Mines criteria for AML sites). The cost of 51 A Fuzzy Expert System on Acid Rock Drainage Site Remediation $120/m2 is calculated based on the capital cost of constructing the Z&F wetland. (Faulkner, B.B., and J.G. Skousen, 1995) Costs Calculations and Adjustments In order to account for future variations in these costs; the module updates all information based on Marshall & Swift (M&S) index values which are published monthly in Chemical Engineering journal. (Marshall, 1999) The module contains past M&S values for future reference and calculations. A form is available to input future values. Equation 10 below shows the method used for updating cost values. Cost1994 x = Cost2000 1 0. A 994 The economic evaluation of a treatment or control option is broken down into capital and operating costs. Total annual operating costs are calculated as a Net Present Value (NPV) using a rate of return value provided by the user (defaulted to 3.5 % if one is unavailable). Equations 11 and 12 show the NPV calculations. NPV = CF[F(i)]-Cc 11. Cc = Capital Costs Cash Flow (CF) = -Cop (1 + 0 - 1 F(i) = - 12. K J i(l + i)" 52 A Fuzzy Expert System on Acid Rock Drainage Site Remediation n = operating years. Where existing revenues are available to write-off costs for tax purposes, the module applies a straight-line depreciation to the capital costs according to equation 13 below. Cc D = — 13. n An annual cash flow is then calculated according to equation 14 below. AnnualCashFlow = (1 - t)(-Cop) + tD 14. t = tax rate. A FORM is available for user input of a tax rate with the defaulted tax rate being 50% if one is unavailable. 53 A Fuzzy Expert System on Acid Rock Drainage Site Remediation SYSTEM VALIDATION AND VERIFICATION Validation of the system is the confirmation that ARDx and the experts will arrive at similar conclusions when provided the same information. A preliminary validation was done on the covers sub-module of ARDx. Validation for ARDx involves gathering inputs from mine sites with an expert, running ARDx using those inputs and reviewing the outputs with the expert. The system design can then be refined as advised by the expert. Verification is the confirmation that ARDx is a working expert system constructed to the design parameters, or flow charts, set by the experts. Verification was achieved by running simulations with test data. The inputs for the simulation runs were the same, apart from changes in one aspect of the data, allowing for observations on the systems outputs. System Validation The covers sub-module has undergone a preliminary validation using Samatosum mine data. Samatosum is a silver mine located in British Columbia that ceased operations in 1992 and began reclamation. The ARDx system was used to assess mitigation options for the waste-rock dump and its drainage. The dump consists of 4 million tonnes of pyrite-rich rock with pH drainage as low as 2.7. Data for the test run was input with the expert. User inputs and system outputs are listed below. (Ghomshei et al., 1997) INPUTS 1. capital ($1,000,000.00) 2. flow rate (80L/s peak) 3. NP/AP (.45) 4. topography (low gradient 70%, med 30%) 5. dump liner (none) 54 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 6. waste (permeable) 7. underlying material (slightly permeable) 8. surrounding material (somewhat permeable) 9. surface area (20 hectares) 10. population (distant town) 11. precipitation rain (100mm) 12. precipitation snow (100mm) 13. water bodies in the area (many) 14. ponding on dump (none) 15. water table depth (7m) 16. parks in the area (some) 17. scavenger material (moderate) 18. catchment area (moderately large) OUTPUTS 1. water cover (DoB = 6) 2. compound clay cover (DoB = 76) 3. geo-membrane cover (DoB = 43) 4. multiple layer cover (DoB = 24) 5. simple soil cover (DoB = 0) 6. walk away (DoB = 0) The outputs of the module were reviewed with the expert. It was noted that indeed one of the options Samatosum is considering is a compound cover for the dump. System Verification Four test scenarios were run to assess the ARDx system. The tests were designed to study if all system modules work together properly to assess the inputs and produce reasonable outputs to assist the user in choosing the most suitable option for further investigation. The first two scenarios deal with a waste rock pile. The sites are similar except for water conditions at the site. The last two scenarios examine tailings dams. They are similar except for effluent run-off characteristics. The scenarios were designed in this manner to better assess how a change in one aspect of the data inputs changes the output or system conclusions. 55 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Waste Rock Pile: Run #1 & Run #2 The first two runs both deal with a waste rock pile of the same size and type. The effluent and site conditions are also the same apart from the following changes to the water conditions at the site: Run #1 (typifies a Southern Ontario climate) • winter snow • fall/spring rain • hydraulic gradient moving towards a populated area • low topography • low land availability Run #2 (typifies a dry or desert environment) • little or no snow • very little rain • hydraulic gradient moving away from a populated area • more permeable surrounding area • moderate topography • high land availability The scenarios were run for all remediation options and their associated costs were determined. A detailed list of all inputs and outputs for Run #1 and Run #2 can be found in Appendix D. A summary of the output is shown in Table 11 below. 56 A Fuzzy Expert System on Acid Rock Drainage Site Remediation RUN# 1 RUN #2 Technical DoB Capital Costs NPV Costs Technical DoB Capital Costs NPV Costs Dry Cover 35 $8,791,000 $10,110,000 30 $8,791,000 $10,110,000 Water Diversion 75 $1,916,000 $2,203,000 53 $1,916,000 $2,203,000 Passive Treatment* 39 $1,701,000 $1,882,000 $1,956,000 $2,164,000 44 $213,000 $236,000 $245,000 $271,000 Water Treatment 89 $798,000 $4,869,000 89 $770,000 $3,608,000 Water Cover 45 $8,583,000 $9,870,000 44 $8,583,000 $9,870,000 *Costs are shown for aerobic and anaerobic wetlands respectively. Table 11. Summary of Results; Run #1 vs Run #2 Water Cover: The degree of belief in using water cover decreases slightly in Run #2 in comparison with Run #1. The higher precipitation and increased water conditions used as inputs in Run #1 should have made water cover a more viable option. The availability of land for wetlands in Run #2 however is much greater and this factor would have contributed to the degree of belief for water cover. The cost of subaqueous disposal increased dramatically from $25 million for Run #1 to $259 million for Run #2. This large increase is due to the large distance to the closest water source in Run #2 as compared to Run #1. Dry Cover: The degree of belief in covering the waste rock dump with a multiple soil cover increased in Run #2. Likewise the degree of belief in using a compound soil cover also increased though very marginally. The dryer conditions would make the dry cover a more viable option, as less water would be available to move through the cover. The effect of the dry cover would be increased. The cost of covering the waste rock dump remains the same however, as the volume of waste is unchanged. A Fuzzy Expert System on Acid Rock Drainage Site Remediation Diversion: Less water at the site would result in a reduced need to divert the water away from the site or intercept the effluent run off. The degree of belief in diversion decreased in Run #2. The cost of diversion remained the same as the inputs gave the same length of diversion ditch required for the site. Water Treatment: The degree of belief of treating the contaminated effluent with chemicals to eliminate metals and increase the pH remains the same in both runs at 89%. In both cases the effluent and metal contaminants were set at the same levels. In Run #2 however, due to lower water availability, the flowrate to be treated is decreased. The operating costs decrease in Run #2, thereby decreasing the net present value of treating the effluent. Passive Treatment: The degree of belief in using passive treatment is lower in Run #1 as compared to Run #2. This run has less land available that can be used for passive treatment. The costs for passive treatment are lower for Run #2. Reduced availability of water means a lower effluent flow, so the requirement is for a smaller treatment pond area. Tailings: Run #3 & Run #4 The last two scenarios are similar in the site conditions, climate and water availability. Differences in the tailings are as follows: Run #3 (less contaminated tailings) • higher pH • low acidity • higher buffering capacity of tailings 58 A Fuzzy Expert System on Acid Rock Drainage Site Remediation • lower metal and contaminant levels Run #4 (highly contaminated tailings) • low pH • high acidity • low buffering capacity of tailings • excessive nutrient additions • higher metal and contaminant levels The above scenarios were run to investigate all options and their associated costs. Detailed inputs and outputs for Run #3 and Run #4 can be found in Appendix D. A summary of the outputs is shown in Table 12. R U N #3 R U N #4 Technical D o B Capital Costs N P V Costs Technical D o B Capi ta l Costs N P V Costs Dry Cover 99 $4,184,000 $4,812,000 99 $4,184,000 $4,812,000 Water Diversion 95 $846,000 $973,000 95 $846,000 $973,000 Passive Treatment* 38 $14,506,000 $141,079,008 $16,682,000 $162,240,000 39 $58,128,000 $564,316,000 $66,848,000 $648,963,000 Water Treatment 54 $1,148,000 $9,175,000 84 $1,418,000 $10,576,000 Water Cover 45 $956,000 $1,099,000 45 $956,000 $1,099,000 *Costs are for aerobic and anaerobic wetlands respectively. Table 12. Summary o f Results; R u n #3 vs R u n #4 Water Cover: The degree of belief for water cover stayed the same in both Run #3 and Run #4 at 45%. A change in the level of contaminants from one run to the other has not made much of an effect on the outcome in this case. A further look at the effect of these contaminants and how the system deals with them seems necessary to refine or tune the system output. The 59 A Fuzzy Expert System on Acid Rock Drainage Site Remediation cost of implementing a water cover for the tailings is also equal in both cases which is expected, since the amount of tailings to be flooded or placed under water is the same. Dry Cover: The degree of belief for using a dry cover has remained the same in both runs at 99%. ARDx has indicated that a vegetative cover would be acceptable in both cases with a high degree of belief. This is expected since the site conditions, water availability and the tailings dam size remained constant. The cost of implementing the vegetative cover was also the same. Diversion: The degree of belief to divert the effluent water downstream remains the same for Run #4 and Run #3. This is due to the constant water conditions for the Runs. The cost of diverting the water around the waste dump also remains the same in both runs since the length of diversion ditches or berms are unchanged. Water Treatment: The degree of belief in requiring water treatment with chemicals has increased in run #4 to 84% compared to 54% in. The higher contaminant levels, lower pH and higher acidity all indicate a greater need for water treatment in Run #4. The cost of treatment is also increased due to the higher amount of chemicals required to bring the effluent in Run #4 to regulation standards. This increase in NPV cost is reflected in the operating costs and not in the up-front capital costs since the plant treats the same amount of effluent in both cases. Passive Treatment: The degree of belief in passive treatment has remained the same in both Run #3 and Run #4 at about 39%. Although the system accounts for the level of contaminants in the effluent to assess the need for passive treatment, this has not caused a significant difference in the final outcome of the two runs. The cost of passive treatment in the 60 A Fuzzy Expert System on Acid Rock Drainage Site Remediation second of the two runs has increased due to the increase in contaminants and acidity that would necessitate a larger wetland area for treatment. Summary of Runs The overall rating of one option compared to another has been calculated with user-input weights of 75 for technical (Degree of Belief), 50 for capital costs and 50 for NPV costs. The combined ranking that the system has given in each of the runs is shown in Table 13 following: T e c h n i c a l D o B C a p i t a l Costs N P V Costs O v e r a l l R a n k i n g R u n # 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 D r y C o v e r 5 5 1 1 5 5 4 4 5 5 3 3 5 5 2.4 2.4 W a t e r D ive rs ion 2 2 2 2 2 3 1 1 1 2 1 1 1.7 2.3 1.4 1.4 Passive T rea tment 4 3 5 5 3 1 5 5 2 1 5 5 3.1 1.9 5 5 W a t e r T rea tment 1 1 3 3 1 2 3 3 3 3 4 4 1.6 1.9 3.3 3.3 W a t e r C o v e r 3 3 4 4 4 4 2 2 4 4 2 2 3.6 3.6 2.9 2.9 Table 13. Ranking o f The Reclamation Options For Each Test Run . Based on the system outputs for waste rock Run #1, water treatment is the best option to look at with water diversion a very close second. Although water treatment is ranked first for technical issues and capital costs, water diversion is better in terms of NPV costs and ranks second in the other areas. For waste rock pile Run #2, water treatment and passive treatment are tied in the overall ranking. In this case, water treatment ranks first from a technical viewpoint but has higher capital and NPV costs then passive treatment which ranked third technically. Dry cover ranks last in all categories and is not a viable option for the waste rock pile runs. Based on the recommendations for the tailing runs (Run # 3 & Run #4), water diversion is the best option to look at for the site based on the overall ranking. Implementing a dry cover is 61 A Fuzzy Expert System on Acid Rock Drainage Site Remediation the next best option overall yet is considered the best one to look at in terms of technical issues. Water cover, water treatment and passive treatment are options that should be discarded for this case. The selected options to be considered together with the calculated combined capital costs and NPV costs for all runs are shown in Table 14 below. It is quite likely that when two options are implemented together that an overall reduction in costs for one or the other can be achieved. For example if water diversion is implemented ahead of water treatment, the total effluent flow can be significantly reduced leading to a reduction in the treatment plant requirements. Similarly, water diversion can lead to significant reduction in the cost of the passive system since the area of the wetlands can be reduced as the flowrate is decreased. The impact of water diversion on the cost of dry cover is less clear. Diversion can reduce ground and surface water flows from the waste or tailings thus reducing the need for expensive dry cover options. Furthermore, the reverse may be the case, in which a dry cover may lead to a reduction in the cost of diversion. Material Waste Rock Tai ing Run 1 Run 2 Run 3 Run 4 Processes Water Diversion Water Diversion Water Diversion Water Diversion Water Treatment Passive Treatment Dry Cover Dry Cover Cap. costs $2,714,000 $ 2,366,000 5,030,000 5,030,000 NPV costs $ 7,072,000 $ 2,720,000 5,785,000 5,785,000 Table 14. Recommended Reclamation Options 62 A Fuzzy Expert System on Acid Rock Drainage Site Remediation CONCLUSIONS 1. A prototype expert system has been developed using Fuzzy Logic and Fuzzy-Neural equations to provide recommendations on ARD mitigation for waste rock dumps and tailing deposits. The system provides a logical, structural approach to decision-making which should contribute to assist in the development of a strategy to deal with ARD problems at an existing mine site or in consideration of a new project. 2. A method to combine the rating on process selection based on technical and economic factors has been developed. The method uses a weighted inference equation with the user able to adjust the weights to suit any site-specific factors. 3. In addition to the decision-making modules, the system also provides an on-line documentation to give new people to the field insight into the complexities of ARD and its remediation. Furthermore, the system provides the user with access to justification for outputs in which the user can click through a flow sheet of each option investigated to see the degree of belief for all input and internally-inferred variables. 4. The system has been designed to allow both new modules to be added in without the need for extensive reprogramming and for knowledge to be added into existing modules with realistic ease. Adaptation of the fuzzy sets in response to new knowledge can also be done relatively easily. 63 A Fuzzy Expert System on Acid Rock Drainage Site Remediation RECOMMENDATIONS Limitations and recommendations to the system include: 1. The system is not yet able to cycle back to redo a module after implementing another. (Example: After implementing the diversion module, changes in the water availability may alter the inputs and decisions made within the dry covers or treatment modules.) Adding a loop to cycle back through the system would allow proper assessment of using more then one option in parallel or series. 2. The system is not currently able to recalculate costs for combined options. Again, the cost of implementing an option may change when using two or more options together. (Example: Treatment costs would likely go down if a cover is added to the waste and/or the effluent flow is reduced.) Once the use of two or more options is investigated, cost recalculations can be done to get a more realistic picture of viable options for the site. 3. ARD is a field that is evolving and changing, wherein new laboratory and field testing provide on-going monitoring information. As technology advances and new options are developed, these can be added to the modules to update the system. 4. Currently ARDx investigates one area at a time for remediation. Future work should include a routine to deal with multiple-problems at a site. (Example: Waste rock pile effluent and tailings can be looked at together to find an appropriate solution to both rather than dealing with them separately.) 64 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 5. The first modules to be added to the system are those dealing with in-pit flooding, the use of bactericides, and the use of sulfate-reducing bacteria. 6. The water treatment module and the passive treatment module each require expansion to include a variety of optional processes and chemicals. 65 A Fuzzy Expert System on Acid Rock Drainage Site Remediation REFERENCES Bowen, A., 1995. Expert systems: Truth and Rumors. Canadian Mining J., Mining Sourcebook, pp. 8-12. Broughton, L.M., and P.M. Healey, 1992. Mine site decommissioning: Technology for assessment and control of acidic rock drainage. In: Environmental Issues and Mangement of Waste in Energy and Mineral Production, Volume 2, Balkema Publishers, Rotterdam, Netherlands, pp.789-797. Faulkner, B.B., and J.G. Skousen, 1995. Treatment of Acid Mine Drainage by Passive Treatment Systems. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia. 1995. Filipek, L., A. Kirk, W. Schafer, 1996. Control Technologies for ARD. Mining Environment Management, pp. 4-8, Dec. Delaney, T., D. Hockley and D, Sollner, 1997. Applications of Method for Delaying the Onset of Acidic Drainage. Proceedings of the 4th International Conference on Acid Rock Drainage. Vol. 2, pp. 798-801. Ghomshei, M., A, Holmes, E. Denholm, R. Lawrence, T. Carriou, 1997. Acid rock drainage from the Samatosum waste dump, British Columbia, Canada. Proceedings of the 4th International Conference on Acid Rock Drainage. Vol. 1, pp. 351-366. Gordon, Robb, and J. Robinson, 1995. Acid Mine Drainage Prediction and Remediation. In: Mining Environmental Management, pp 19-21, September 1995. 66 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Kalin, M., and M.P. Smith, 1997. Microbial Acid Reduction in Sediments - Concepts and Application. In: Fourth International Conference on Acid Rock Drainage Proceedings, Vancouver, B.C., Vol. 3, pp.1321-1333. Lawrence, R. W., 1997. A Course on Fundamental and Applied Aspects of Acid Rock Drainage. Course notes, Department of Mining and Mineral Process Engineering, UBC, Vancouver, Canada, p. 75, 1997. Marshall, Swift, 1999. Marshall & Swift Equipment cost index. Chemical Engineering, Vol 106,No.3, p. 170. Meech, J.A., 1995. AI Applications in the mining industry into the 21st Century. AP COM XXV Conference, pp. 93-101. Meech, J A., 1998. Intelligent Supervisory Control of Mineral Processing Plants. Course notes, Department of Mining and Mineral Process Engineering, UBC, Vancouver, Canada, p. 8, 41. Meech, J.A., C.A. Harris, 1992. Expert systems for gold processing plants. Randol Gold Forum. Vancouver, Canada, p. 15, March 25-27, 1992. Meech, J A. and S Kumar, 1992. Hypermanual on Expert Systems. Ed. CANMET, Ottawa, Ont, Canada. This is an electronic book on expert systems, 3200e.p. (electronic pages), topic 26. MEND, 1995. Economic evaluation of acid mine drainage technologies. MEND Report 5.8.1 Placer Dome Inc., 1997. Water covers and Underwater disposal to prevent ARD. In: Fourth International conference on Acid Rock Drainage, Short Course #1., p. 37, June 1997. A Fuzzy Expert System on Acid Rock Drainage Site Remediation Robb, G., and J. Robinson, 1995. Acid mine drainage prediction and remediation. In: Mining Environmental Management, pp. 19-21, Sep. 1995. Skousen, J., 1991. Anoxic Limestone Drains for Acid Mine Drainage Treatment. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia, pp. 211-216, 1995. Skousen, J., A. Sextone, K. Garbutt, and J. Sencindiver, 1992. Wetlands for Treating Acid Mine Drainage. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia, pp. 199-210, 1995. Skousen, J., B. Faulkner, and P. Sterner, 1995. Passive Treatment Systems and Improvements of Water Quality. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia, pp. 243-254, 1995. Stephen, Robertson, and Kirsten (B.C.) Inc. November 1991. Evaluation of ARD From Britannia Mine And The Options For Long Term Remediation of The Impact On Howe Sound. In: British Columbia Acid Mine Drainage Task Force, British Columbia, pp. 6-2 to 6-10, November 1991. Veiga, M.M. and J.A. Meech, 1994. Heuristic Approach to Hg Pollution in the Amazon. In: Proceedings International Symposium on Treatment of Wastes, 123rd Congress, San Francisco, California, pp 23-38,1994. 68 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Zadeh, L. A., 1987. The Role of Fuzzy Logic in the Management of Uncertainty in Expert Systems. In: Fuzzy Sets and Applications: Selected Papers by L. A. Zadeh, edited by R.R. Yager..[et al.], Wiley, New York. p. 414, 1987. Zadeh, L. A., 1987. Fuzzy Sets. In: Fuzzy Sets and Applications: Selected Papers by L. A. Zadeh, edited by R.R. Yager..[et al.], Wiley, New York. p. 30, 1987. 69 A Fuzzy Expert System on Acid Rock Drainage Site Remediation BIBLIOGRAPHY Bechard, G., and R.G.L. McCready, 1991. Microbiological Process for the Treatment of Acidic Drainage at the Halifax International Airport. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 3, pp. 171-183, Sep. 17-18, 1991. Bender, J. Dr., and Dr. P. Phillips, 1995. Biotreatment of mine drainage. In: Mining Environmental Management, pp. 25-27, Sep. 1995. Bennett, P.G., CR. Ferguson, and T.H. Jeffers, 1991. Biological Treatment of Acid Mine Waters - Case Studies. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Broughton, L.M., R.W. Chambers, and A.MacG. Robertson, 1992. Mine Rock Guidelines Design Control of Acid Drainage Water Quality. Steffen, Robertson, and Kristen (B.C.) Inc., Vancouver, B.C. Butcher, S.S., et al., Eds, 1992. Global Biogeochemical Cycles. Academic Press, Toronto, Ontario. Cambridge, M., 1997. Wheal Jane - The Long Term Treatment of Acid Mine Drainage. In: Fourth International Conference on Acid Rock Drainage Proceedings, Vancouver, B.C., Vol. 3. Dodds-Smith, M.E., C.A. Payne, and J.J. Gusek, 1995. Reedbeds at Wheal Jane. In: Mining Environmental Management, pp. 22-24, Sep. 1995. 70 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Dvorak, D.H., R.S. Hedin, S.P. Mclntire, and H.M. Edenborn, 1995. Treatment of Metal-Contaminated Water Using Bacterial Sulphate Reduction-Results from Pilot-Scale Reactors. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Eger, P., J.R. Wagner, and G. Melchert, 1997. The Use of Peat/Limestone System to Treat Acid Rock Drainage. In: Fourth International Conference on Acid Rock Drainage Proceedings, Vancouver, B.C., Vol. 3. Everell, M.D., and A J . Oliver, 1992. Environmental Technologies for the Minerals and Metals Sector - A Canadian Perspective. In: Minerals, Metals and the Environment. Elsevier Science Publishers Ltd., New York. Gould, W.D., G. Bechard, and L. Lortie, 1994. The Nature and Role of Microorganisms in the Tailings Environment. In: Short Course Handbook on Environmental Geochemistry of Sulfide Mine-Wastes, Mineralogical Association of Canada, Waterloo, Ontario, Chap. 7, May 1994. Gusek J.J., 1995. Passive-Treatment of Acid Rock Drainage: What is the Bottom Line? In: Mining Engineering, pp. 250-253, March 1995. Hendricks, A.C., 1991. The Use of Artificial Wetland to Treat Acid Mine Drainage. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 3, Sep. 17-18, pp.549-558, 1991. A Fuzzy Expert System on Acid Rock Drainage Site Remediation Herlihy, A.T., A.L. Mills, G.M. Hornberger, and A.E. Bruckner, 1987. The Importance of Sediment Sulfate Reduction to the Sulfate Budget of an Impoundment Receiving Acid Mine Drainage. In: Water Resources Research, Vol. 23, No. 2, pp. 287-292, Feb. 1987. Kalin, M., 1995. Biological Amelioration of Acidic Seepage Streams. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Kalin, M., and R.O. Van Everdinger, 1987. Ecological Engineering: Biological and Geochemical Aspects, Phase I Experiment. In: Acid Mine Drainage Seminar / Workshop Proceedings, Halifax, Nova Scotia. Kalin, M., M.P. Smith, and G. Landry, 1994. The Application of Ecological Engineering to Acid Coal Seepages in Eastern Canada. In: International Mine Reclamation and Mine Drainage Conference and Third International Conference on the Abatement of Acidic Drainage Proceedings, Vol. 2, April 24-29, 1994. Kalin M., et al. Passive Treatment Process for the Mineral Sector. In: Waste processing and recycling in mineral and metallurgical industries. Canada Institute of Mining Metallurgy and Petroleum, Montreal, Canada, pp. 363-375, 1995. Karathanasis, A.D., and Y.L. Thompson, 1995. Metal Speciation and Retention Patterns in a High Metal Load Acid Mine Constructed Wetland of Southeastern Kentucky. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 2, Sep. 17-18, pp. 485-498, 1991. A Fuzzy Expert System on Acid Rock Drainage Site Remediation Kozinski, J. A., and I. Makarow, 1995. Micro-encapsulation of toxic metals during thermal remediation of solid wastes. In: Waste processing and recycling in mineral and metallurgical industries. Canada Institute of Mining Metallurgy and Petroleum, Montreal, Canada, pp. 309-375, 1995. Kuyucak, N., P. St-Germain, and K.G. Wheeland, 1995. In Situ Bacterial Treatment of AMD in Open Pits. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Kwong Y.T.J. Generation, attenuation and abatement of acidic drainage in an abandoned minesite on Vancouver Island, Canada. In: Environmental Issues and Mangement of Waste in Energy and Mineral Production, Volume 2, Balkema Publishers, Rotterdam, Netherlands, pp.757-762. Lawrence, R.W. Dr., and Dr. A.MacG. Robertson, 1994. A Short Course on Acid Rock Drainage Understanding the Problems - Finding Solutions. The CLM District 6 Annual Meeting, Vancouver, B.C., Oct. 11, 1994. Lawrence, R.W., A.MacG. Robertson, 1994. Acid rock drainage Understanding the problems -finding solutions. CIM District 6 Annual General Meeting, Workshop No. 1. Moskalyk R. R., et al. Protective coating for retarding the natural leaching of mine waste rock. In: Waste processing and recycling in mineral and metallurgical industries. Canada Institute of Mining Metallurgy and Petroleum, Montreal, Canada, pp. 439-453, 1995. Paulik H. F., et al. Simulation of pit closure alternatives for an open pit mine. In: Tailings and Mine Waste 1995. Balkema, Brookfield, USA, pp. 67-77. 73 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Pesavento, B.G., 1987. Factors to be Considered in the Use of Wetlands to Treat Acid Mine Drainage. In: Acid Mine Drainage Seminar / Workshop Proceedings, Halifax, Nova Scotia. Plane and Associates Ltd., 1990. Assessment of Existing Natural Wetlands Affected by low pH, Metal Contaminated Seepages. Final Report, MEND project 3.12.1. Richardson, J.S., and C.J. Perrin, 1990. The Effect of Acid Mine Drainage on Stream Macroinvertebrates and Periphytic Algae. BCAMD project 4.30.c.4. Ritchie, A.I.M., 1994. The Waste Rock Environment. In: Short Course Handbook on Environmental Geochemistry of Sulfide Mine-Wastes, Mineralogical Association of Canada, Waterloo, Ontario, Chap. 5, May 1994. SENES Consultants Ltd., 1994. Acid Mine Drainage - Status of Chemical Treatment and Sludge Management Practices. MEND project 3.32.1. Sengupta, M., 1993. Environmental Impacts of Mining Monitoring, Restoration, and Control. Lewis Publishers, USA. Sextone, A.J., J.C. Sencindiver, J.P. Calabrese, D.K. Bhumbia, and G.K. Bissonnette, 1991. Application of Constructed Wetlands for the Removal of Iron from Acid Mine Drainage. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 3, Sep. 17-18, pp. 559-575, 1991. 74 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Shelp, G., G. Southam, G. Spiers, L. Liu, and W.Chesworth, 1994. An Evaluation of a Peat-Wood Chip-Microflora Admixture to act as an Ameliorant for Acid Mine Drainage. In: International Mine Reclamation and Mine Drainage Conference and Third International Conference on the Abatement of Acidic Drainage Proceedings, Vol. 2, April 24-29, 1994. Skermer N A, Rohrs RG 1992. Natural & Artificial Barriers To Seepage. In: Environmental Issues & Management of waste in energy and mineral production. A.A. Balkena Publishers, Brookfield, USA. Vol 1 1992. Skousen, J.G., 1995. Douglas Abandoned Mine Land Project: Description of an Innovative Acid Mine Drainage Treatment System. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia, pp. 231-239, 1995. Stark, L.R., W.R. Wenerick, P.J. Wuest, and F.M. Williams, 1995. Adding a Carbon Supplement to Simulated Treatment Wetlands Improves Mine Water Quality. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 2, Sep. 17-18, pp.465-483, 1991. Taddeo, F.J., and R.K. Weider, 1991. The Accumulation of Iron Sulphides in Wetlands Constructed for Acid Coal Mine Drainage (AMD) Treatment. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 3, Sep. 17-18, pp. 529-547, 1991. 75 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Wheeler, W.N., and M. Kalin, 1995. The Ecological Response of a Bog to Coal Acid Mine Drainage. Deterioration and Subsequent Initiation of Recovery. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 2, Sep. 17-18, pp. 449-464, 1991. Wieder, R.K., M.N. Linton, and S.T. Starr, 1991. Spatial and Temporal Patterns in Surface and Subsurface Water Chemistry in Wetlands Constructed for Acid Coal Mine Drainage. In: Second International Conference on the Abatement of Acidic Drainage Proceedings, Tome 3, Sep. 17-18, pp.507-514, 1991. Wildeman, T., G. Brodie, and J. Gusek, 1993. Wetland Design for Mining Operations. BiTech Publishing Co., Vancouver, B.C. Ziernkiewicz, P., J. Skousen, and R. Lovett, 1994. Open Limestone Channels for Treating Acid Mine Drainage: A New Look at an Old Idea. In: Acid Mine Drainage Control and Treatment, West Virginia University and the National Mine Reclamation Center, Morgantown, West Virginia, pp. 225-230, 1995. 76 A Fuzzy Expert System on Acid Rock Drainage Site Remediation APPENDIX A - FAM MAPS 77 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Cover Quality (control method) ard.potent ia l .h igh L ava i l ab le cost M MH H treatplant H G G H H capac i ty M G H H VH L G H VH VH ard.potent ia l .h igh L ava i l ab le cost M MH H treatplant H O K G G H capac i ty M O K G H VH L G H H VH ard.potent ia l .h igh L ava i l ab le cost M MH H treatplant H O K O K G H capac i t y M O K G G H L O K G H H G = g o o d , H=high, O K = o k , VH=ve ry high W a t e r , P a s s i v e , C o v e r M o d u l e Contaminant Level s e e water m o d u l e 78 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Environmental Sensitivity mobil i ty L Med H ss=50 r=60 s=30 soc io - '-env i ro r=70 ss=70 r=10 s=20 ss=100 imapct M ss=40 r=70 s=70 ss=40 r=10 H s=30 ss=70 s=60 ss=50 s=100 r=10 s=sens i t ive , ss=sl ight ly sens i t i ve , r=resistant C o v e r M o d u l e Effluent Vo lume S u r f a c e Influx h igh water infi l tration L M MH H subsurface H L=80 M=20 M=90 L=10 M=50 H=60 H=80 M=20 influx M o d L=20 M=80 M=80 H=20 M=50 H=60 H=90 M=10 i M=80 M=50 H=80 H=100 L H=20 H=60 M=20 S u r f a c e Influx M o d water infi l tration L M MH H subsurface H L=90 M=10 L=60 M=50 M=80 H=20 H=60 M=50 inf lux M o d L=80 M=20 M=80 L=20 M=60 H=50 H=80 M=20 i L=20 M=80 H=70 H=90 L M=80 L=20 M=30 M=10 S u r f a c e Influx low water infi l tration L M MH H subsurface H L=100 L=80 M=20 M=60 L=50 M=80 H=20 inf lux M o d L=90 L=60 M=90 H=60 M=10 M=50 H=10 M=50 1 L=80 M=90 M=50 H=80 l_ M=20 L=10 H=60 M=20 79 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Water Infiltration d u m p permeabi l i ty high ava i l ab le water for infiltration H igh MH M L o w was te La rge H H H MH sur face Mod- la rge H H MH M a r e a M H H MH M Sma l l MH MH M L d u m p permeabi l i ty m e d ava i l ab le water for infiltration H igh Mod-high Moderate L o w was te La rge H H MH M sur face Mod- la rge H H M M a r e a Modera te MH MH M L Sma l l M M L L d u m p permeabi l i ty low ava i l ab le water for infiltration H igh Mod-high Moderate L o w was te La rge MH MH M L sur face Mod- la rge MH M L L a r e a Modera te M L L L S m a l l L L L L H=high, MH=moderate high, ML=moderate-large, M=moerdate, L=low C o v e r M o d u l e Ponding Effect pond ing /su r face a r e a ratio H M L loca l hot sun H H M c l imate hot c loud H M L co ld sun H M L co ld c l oud M L L C o v e r M o d u l e Potential Underflux P under ly ing mater ia l M 1 d u m p N H M L l iner Y L L L 80 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Precipitation Available for Infiltration precip i tat ion s n o w H M L precipitation H M H M=80 L=30 rain M H H=80 MH=30 MH=80 M=30 M=50 L=60 M H=60 MH=50 M L=80 M=30 L M H M=80 L=30 L C o v e r , W a t e r , D i ve r s i on M o d u l e Total Precipitation precip i tat ion s n o w H M L precipitation H H H M rain M H H M M M M M L L M L L C o v e r M o d u l e Socio-environmental Impact d is tance far popu la t ion s i z e L M S l_l protected M L=40 L=60 M=70 M=40 reg ions ^ L=60 1 1 M=40 L L L L L L d is tance m e d popu la t ion s i z e L M s l_l protected M M L=40 M=70 reg ions M M L=40 L=60 M=70 M=40 i L=40 L=60 • L. M=70 M=40 d is tance c l o s e popu la t ion s i z e L M s I_I protected H H H=80 M=30 reg ions M H H=80 H=50 M=30 M=60 i H=70 H=40 M L. M=40 M=60 81 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Sub-Sur face Influx water tab le sha l l ow P topograph ica l relief LR MR H R potent ia l H M M H H under f lux M M M H H L M M M M water tab le sha l l ow P topograph ica l relief LR MR H R potential underflux L M=80 L=30 M=60 H=50 H M L L=50 M=60 M H=50 M=60 L L L M M water tab le sha l l ow P topograph ica l relief LR MR H R potent ia l H L L M M under f lux M L L L M L L L L L M = m o d e r a t e W a t e r , P a s s i v e , C o v e r M o d u l e S p e c i e s Form s e e water m o d u l e C o v e r M o d u l e Mobility (tails cover) contaminant level L Mod H species reactivity M=30 L=70 H H N L M H=20 M=30 N R L L H=70 M=30 82 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Cover Module Mobility water bodies high water bodies low hydro.gradient.away scavenger_material hydro.gradient.away scavenger_mater ial Plenty Mod L P M L Water S M M M Water S L L=80 M=30 L=80 M=30 Table M M=80 L=30 M=80 L=30 M Table M L L L Deep M=60 L=40 M=80 L=30 M Deep L L L water bodies high water bodies low hydro.gradient.level scavenger_material hydro.gradient.level scavenger_mater ial P M L P M L Water S H H H Water S L=80 M=30 L=60 M=50 L=60 M=50 Table M H=80 M=30 H=80 M=30 H Table M L L L=80 M=30 Deep M=60 H=40 H=60 M=40 H=80 M=30 Deep L L L water bodies high water bodies low hydro.gradient.toward scavenger_material hydro.gradient.toward scavenger_mater ial P M L P M L Water S H H H Water S M M=80 H=30 M=60 H=40 Table M H H H Table M M M=80 H=30 M=80 H=30 Deep H=80 M=30 H=80 M=30 H Deep M M M water bodies medium water bodies medium hydro.gradient.away scavenger_material hydro.gradient.toward scavenger_mater ial P M L P M L Water S L=80 M=30 L=60 M=40 L=60 M=40 Water S H H H Table M L=80 L=80 Table M H=80 H=80 H L M=30 M=30 M=30 M=30 Deep L L L Deep M=60 H=40 H=60 M=40 H=80 M=30 water bodies medium hydro.gradient.level scavenger_material P M L Water s M M M=80 H=30 Table M M M M Deep M=80 L=30 M M 83 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C o v e r M o d u l e Toxicity mobi l i ty factor H 0.5 M 0.25 L 0 e n h a n c e toxici ty H 0.5 M 0.25 L 0 Cover Module Surface Influx catchment area large precipitation high surrounding permeability P Med 1 catchment area large precipitation medium sunounding permeability P Med 1 catchment area large precipitation low surrounding permeability P Med 1 topographs H H H H topographic H M H H topographic H L M H relief M H H H relief M M H H relief M L M M L H H H L M M H L L L M catchment area moderate precipitation high surrounding permeability P Med 1 catchment area large precipitation medium surrounding permeability P Med 1 catchment area large precipitation tow surrounding permeability P Med 1 topographic H H H H topographic H M M M topographic H L L M relief M H H H relief M M M M relief M L L L L M H H L L M M L L L L catchment area small precipitation high surrounding permeability P Med 1 catchment area large precipitation medium surrounding permeability P Med 1 catchment area large precipitation low surrounding permeability P Med 1 topographic H H H H topographs H M M M topographs H L L L relief M M H H relief M L M M relief M L L L L L M M L L L M L L L L L = plateau, M=moderate Dive rs i on M o d u l e Downstream Diversion, Ditch/Berms sur face f low H 0.8 M 0.4 L 0 subsu r face f low H 0.2 M 0.1 L 0 D i v e r s i o n M o d u l e Downstream Diversion, Ditch/Berms (passive treatment) su r face f low H 0.6 M 0 .3 L 0 s u b s u r f a c e f low H 0.1 M 0 .05 L 0 p a s s i v e opt ion H 0.3 M 0 .15 L 0 84 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Divers ion M o d u l e Downstream Diversion, Di tch/Berms (active treatment) sur face f low H 0.6 M 0.3 L 0 subsu r face f low H 0.1 M 0 .05 L 0 treat opt ion H 0.3 M 0 .15 L 0 D ive rs ion M o d u l e Downstream Diversion, Interception sur face f low H 0.2 M 0.1 L 0 subsu r face f low H 0.8 M 0.4 L 0 D ive rs ion Modu le Downstream Diversion, Interception (passive treatment) su r face f low H 0.1 M 0 .05 L 0 subsu r face f low H 0.6 M 0 .3 L 0 p a s s i v e opt ion H 0.3 M 0 .15 L 0 D ive rs ion M o d u l e Downstream Diversion, Ditch/Berms (passive & active water treatment) su r face f low H 0.5 M 0 .25 L 0 subsu r face f low H 0.1 M 0.05 L 0 treat opt ion H 0.2 M 0.1 L 0 p a s s i v e opt ion H 0.2 M 0.1 L 0 D ive rs ion M o d u l e Downstream Diversion, Interception (active water treatment) sur face f low H 0.1 M 0 .05 L 0 s u b s u r f a c e f low H 0.6 M 0.3 L 0 treat opt ion H 0.3 M 0.15 L 0 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Dive rs ion M o d u l e Downstream Diversion, Interception (passive & active water treatment) su r face f low H 0.1 M 0.05 L 0 s u b s u r f a c e f low H 0.5 M 0 .25 L 0 treat opt ion H 0.2 M 0.1 L 0 p a s s i v e opt ion H 0.2 M 0.1 L 0 D ive rs ion M o d u l e Upstream Interception (dry cover) su r face f low H 0.1 M 0 .05 L 0 s u b s u r f a c e f low H 0.6 M 0.3 L 0 c o v e r opt ion H 0.3 M 0.15 L 0 D i ve rs i on M o d u l e Upstream Diversion, Ditch/Berms sur face f low H 0.8 M 0.4 L 0 subsu r face f low H 0.2 M 0.1 L 0 D i ve rs i on M o d u l e Upstream Interception sur face f low H 0.2 M 0.1 L 0 subsu r face f low H 0.8 M 0.4 L 0 86 A Fuzzy Expert System on Acid Rock Drainage Site Remediation D i v e r s i o n M o d u l e D ive rs ion M o d u l e Subsur face Flow sur round ing mater ia l P 3 M 1.5 1 0 topography relief H 4 M 2 L 1 P 0 water tab le D 0 M 1.5 S 3 Surface Flow ca tchment a r e a L 2 M 1 S 0 total precipi tat ion H 4 M 2 L 0 sur rounding mater ia l P 0 M 1 I 2 topography relief H 2 M 1 L 0.5 P 0 D i v e r s i o n M o d u l e Divert Opt ion downs t ream d ivers ion H M L ups t ream H H H H d i ve rs ion M H M M L H M L Dive rs i on M o d u l e Downstream Diversion downs t ream intercept ion H M L d o w n s t r e a m H H H H d i t ch /be rms M H M M L H M L D i v e r s i o n M o d u l e Upstream Diversion ups t ream intercept ion H M L ups t ream H H H H d i t ch /be rms M H M M L H M L C o v e r , W a t e r , D i ve r s i on M o d u l e Total Precipitation s e e c o v e r m o d u l e 87 A Fuzzy Expert System on Acid Rock Drainage Site Remediation W a t e r Modu le Water , P a s s i v e M o d u l e Dummy waste pit f lood H 0.05 M 0.1 L 0.6 W a t e r Modu le Environmental Factors, Tails i nc reased turbidity H 0 M 0 .125 L 0.25 nutrient addi t ions H 0 M 0.125 L 0.25 tox ins present H 0 M 0.125 L 0.25 habitat impact H 0 M 0.125 L 0.25 W a t e r Modu le Environmental Phys/Chem i nc reased turbidity H 0 M 0.125 L 0.25 nutrient addi t ions H 0 M 0.125 L 0.25 toxins present H 0 M 0.125 L 0.25 habitat impact H 0 M 0.125 L 0.25 Enhancement organ ics leve l H 5 M 2.5 L 0 o rgan ics form Me th 5 N o n - M e t h 0 W a t e r Modu le Environmental Factors, Waste nutrient add i t ions H 0.05 M 0.1 L 0.2 tox ins present H 0.05 M 0.1 L 0.2 habitat impac t H 0.05 M 0.1 L 0.2 soc ia l a c c e p t a n c e G 0.2 F 0.1 P 0.05 W a t e r Modu le Environmental Sustainability env i ronmenta l soc ia l H 0.3 M 0.15 L 0 tai ls2 qual i ty H 0.3 M 0.15 L 0 env i ronmenta l H 0.3 p h y s / c h e m M 0.15 L 0 88 A Fuzzy Expert System on Acid Rock Drainage Site Remediation W a t e r Modu le W a t e r Modu le Environmental SustainabilityW envi ronmenta l soc ia l H 0.3 M 0.15 L 0 waste quali ty H 0.3 M 0.15 L 0 env i ronmenta l H 0.3 phys /chem M 0.15 L 0 W a t e r Modu le Leaching magnes ium present H 0.125 M 0.25 L 0.5 c a d m i u m present H 0.125 M 0.25 L 0.5 Failure Risk d a m stabi l i ty H 0 M 0.1 L 0.2 c l imat ic impac t H 0.2 M 0.1 L 0 main ta in water H 0.15 M 0.3 L 0.6 W a t e r Modu le Physical Characteristics, Tails fa i lure risk H 0.05 M 0.1 L 0.2 water ava i l ab le H 0.6 M 0.3 L 0.15 topograph ic relief H 0 .025 M 0.05 L 0.1 P 0.2 W a t e r M o d u l e Physical Characteristics, Waste pit / was te ratio H 0.3 M 0.15 L 0 .075 water ava i l ab le H 0.3 M 0.15 L 0 .075 pit stabil i ty G 0.25 M 0.1 P 0 89 A Fuzzy Expert System on Acid Rock Drainage Site Remediation W a t e r Modu le Rock Quality, Pit A R D potential H 0 .05125 M 0 .125 L 0 .25 envi ro factors R 0 .05125 A 0 .125 L 0 .25 A R D current H 0 .05125 M 0 .125 L 0.25 W a t e r Modu le Rock Quality, Waste A R D potential H 0 .05125 M 0 .125 L 0 .25 enviro factors R 0 .05125 A 0 .125 L 0 .25 A R D current H 0 .05125 M 0 .125 L 0 .25 W a t e r Modu le Rock Quality, Tails settl ing propert ies G 0.2 F 0.1 P 0.05 A R D potential H 0.05 M 0.1 L 0.2 leach ing H 0.05 character is t ics M 0.1 L 0.2 env i ro factors R 0.05 A 0.1 L 0.2 A R D current H 0.05 M 0.1 L 0.2 W a t e r Modu le Ta i ls l Quality envi ro soc ia l H 0.2 M 0.1 L 0 A R D potential H 0 M 0.1 L 0.2 env i ronementa l H 0.05 phys /chem M 0.1 L 0 A R D current H 0.05 M 0.1 L 0.2 90 A Fuzzy Expert System on Acid Rock Drainage Site Remediation W a t e r Modu le W a t e r Modu le T a i l s 2 Quality A R D potential H 0 M 0.25 L 0.5 A R D current H 0 M 0.25 L 0.5 W a t e r Modu le Toxicity mobil i ty factor H 0.5 M 0.25 L 0 e n h a n c e toxici ty H 0.5 M 0.25 L 0 Tails Quality A R D potential H 0 M 0.25 L 0.5 A R D current H 0 M 0.25 L 0.5 W a t e r Modu le Waste Pit Flood waste site G 0.6 character is t ics M 0.3 P 0 pit quali ty H 0.2 M 0.1 L 0 waste quali ty H 0.2 M 0.1 L 0 Water Module Water Available water table S 0.3 M 0.15 D 0.075 total precipitation H 0.2 M 0.1 L 0.05 catchment area L 0.1 M 0.05 S 0.025 surrounding material I 0.1 M 0.05 P 0.025 Water, Cover, Passive Module Totall Precipitation see cover module 91 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Water, Passive Module Water Module Mobility contaminant level H M L species R form H H L=70 M=30 N H=70 M L M=30 NR M=70 L=30 L L Water Module Sub-marine Disposal tails site characteristics G M P environmental H H M L sustainability M M M L L L L L Settling Properties size distribution C M F clays H present M L P P P F F P G F P Water Module Increased Turbidity settling properties G F P wave H action M L M H H L M H L L M Water Module Site Characteristics water bodies lake H water bodies M L D G M P lake depth M M=60 G=40 M P S M P=60 M=40 P water bodies ocean H water bodies M L D G M P lake depth M M P S M P Water Module Maintain Water water H H available M M L L Water Module Water Opt ion A R D site tailings H submarine disposal M L H H H H flood tails M H M M L H M L A R D site waste H submarine disposal M L H H H H waste pit flood M H M M L H M L 92 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Water, Passive, Cover Module Water, Passive, Cover Module Species Form NR N R aluminum reactivity arsenic reactivity i _ i cadmium reactivity UJ > < UJ z co UJ > chromium reactivity REACT! T < copper reactivity cyanide reactivity REACT! "RAL Wl IVE IS F REACTI lead reactivity z "RAL Wl IVE IS F REACTI magnesium reactivity O z T?, O >-manganese reactivity _ l 1 m ^ Z w >- cc < mercury reactivity < uranium reactivity z < other reactivity Contaminant Level aluminum arsenic cadmium chromium copper cyanide lead magnesium manganese mercury uranium other present present present present present present present present present present present P a s s i v e M o d u l e P a s s i v e Modu le Passive Option, Tai ls Hydrological Factor land a rea ava i l ab le L 0.25 ca tchment a r e a L 1 M 0.125 M 0.5 S 0 S 0.25 hydrological fac tors H 0.25 total precipi tat ion H 2.5 M 0.125 M 1.25 L 0 L 0 .625 contaminant H 0 topography relief H 1.5 concent ra t ion M 0.1 M 0.75 L 0.2 L 0 .375 P 0 .1875 water f low S 0.1 M V 0.05 water tab le S 3 V 0 M 1.5 D 0.75 seasona l i ce H 0 M 0.05 under ly ing mater ia l I 1 L 0.1 M 0.5 P 0.25 m a n g a n e s e present H 0.1 M 0.05 surrounding mater ia l I 1 L 0 .025 M 0.5 P 0.25 93 A Fuzzy Expert System on Acid Rock Drainage Site Remediation P a s s i v e M o d u l e Contaminant Concentration tox ins H H present M M L L P a s s i v e , W a t e r Modu le Mobility Factor s e e water modu le W a t e r , P a s s i v e , C o v e r M o d u l e Spec ies Form s e e water modu le P a s s i v e , W a t e r , C o v e r M o d u l e Toatal Precipitation s e e c o v e r m o d u l e W a t e r , P a s s i v e , C o v e r M o d u l e Contaminant Level s e e water m o d u l e P a s s i v e M o d u l e Enhancement s e e water m o d u l e 94 A Fuzzy Expert System on Acid Rock Drainage Site Remediation APPENDIX B - FUZZY SETS 95 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 96 A Fuzzy Expert System on Acid Rock Drainage Site Remediation CADMUIMREACTIVITY 120 0 5 10 CA DM U IM REACTIVITY (rating fromO to 10) C H R O M U I M P R E S E N T 120 , 0 5 10 CHROMUIM P R E S E N T (rating from 0 to 1 0) C L A Y S 120 0 5 10 C L A Y S P R E S E N T (rating from 0 to 10) C O P P E R R E A C T I V ITY 120 , 5 10 C O P P E R REACTIVITY (rating from 0 to 1 0) CHARACTERISTICS. 120 , DoB CHARACTERISTICS CHROMUIM RE ACTIVITY 1 20 CHROMUIM REACTIVITY (rating from 0 to 10) C O P P E R P R E S E N T . 120 , 0 5 10 C O P P E R P R E S E N T (rating from 0 to 1 0) CYANIDE 120 , 0 5 10 CYANIDE P R E S E N T (rating fromO to 10) 97 A Fuzzy Expert System on Acid Rock Drainage Site Remediation CYANIDEREACTIVITY 120 0 5 10 CYANIDE REACTIVITY (rating from 0 to 1 0) DROUGHT. 120 -, 0 5 10 DROUGHT RISK (rating from 0 to 1 0) E N H A N C E 120 ENHANCE TOXICITY (rating from 0 to 10) E N V I R O P H Y S C H E M 120 ) 5 10 DoB ENVIRO-PHYS/CHEM (rating from 0 to 10) DAM 120 5 10 E A R T H Q U A K E RISK (rating fromO to 10) E N V I R O F A C T O R S 120 , 0 5 10 ENVIRO-SOCIAL (rating from 0 to 10) 98 A Fuzzy Expert System on Acid Rock Drainage Site Remediation ENVIROSUST_ 0 5 E N VIRO -S USTA IN AB L E FLOODING RISK (rating fromO to 10) 0 5 10 S E A S O N A L ICE (rating fromO to 10) L A K E D E P T H _ 120 , M : 80 60 40 20 1 ;\ : \ 0 1 * \ 0 50 100 L A K E DEPTH (rating from 0 to 1 00) 0 5 DoB FAILURE RISK (rating from 0 to 10) HABITAT IMPACT (rating from 0 to 1 0) INTERSTITIAL W A T E R S . 0 5 10 INTERSTITIAL W A T E R S (rating fromO to 10) DoB LEACHING (rating from 0 to 1 0) . M 99 A Fuzzy Expert System on Acid Rock Drainage Site Remediation L E A D P R E S E N T . 120 5 10 LEAD P R E S E N T (rating fromO to 100) MAGNESIUM. 120 5 10 MAGNESIUM P R E S E N T (rating fromO to 10) MOBILITY 120 0 5 10 MOBILITY FACTOR (rating from 0 to 1 00) O R G L E V E L . 120 0 5 10 ORGANICS P R E S E N T (rating fromO to 100) L E A D R E ACTIVITY 120 0 5 10 L E A D REACTIVITY (rating from 0 to 1 0) MAGNESIUMREACTIVITY 120 5 10 MAGNESIUM REACTIVITY (rating from 0 to 1 0) NUTRIENT. 120 , 0 5 10 ORGANICS METHYLATING (rating from0 to 1 0) 100 A Fuzzy Expert System on Acid Rock Drainage Site Remediation O T H E R S P R E S E N T _ 0 5 10 ORGANICS P R E S E N T (rating fromOto 100) PITQUALITY. 0 5 10 ROCK QUALITY (rating fromO to 100) P O L I T I C A L A C C E P T A N C E _ POLITICAL A C C E P T A N C E (rating fromO to 10) SIZEDISTRIBUTION_ SIZE DISTRIBUTION (rating from 0 to 1 0) O T H E R S R E A C T I V ITY_ N NR O T H E R S REACTIVITY (rating fromO to 10) P I T W A S T E . DO O D 120 1 00 40 20 -10 10 30 PIT TO W A S T E RATIO ROCKQUALITY_ 0 5 10 ROCK QUALITY (rating fromO to 1 0) S O C I A L A C C E P T A N C E . 0 5 10 SOCIAL A C C E P T A N C E (rating fromO to 10) 101 A Fuzzy Expert System on Acid Rock Drainage Site Remediation DoB TAILS QUALITY (rating fromO to 10) TAILSSITE_ TAILS SIT CHARACTERISTICS (rating from 0 to 1 0) 0 5 DoB TOXICITY (rating fromO to 10) URANIUM REACTIVITY _ N NR URANIUM REACTIVITY (rating from 0 to 10) 0 5 10 DoB TAILS QUALITY (rating fromO to 10) T H E R M A L O V E R T U R N _ 0 5 10 THERMAL OVERTURN (rating from 0 to 1 0) URANIUMPRESENT_ ) 5 10 URANIUM P R E S E N T (rating from 0 to 1 0) W A S T E P I T F L O O D _ W A S T E - PIT FLOOD (rating fromO to 10) 102 A Fuzzy Expert System on Acid Rock Drainage Site Remediation WASTEQUALITY_ ROCK QUALITY f a1hB I r omOto 10) W A T E R A V A I L A B L E _ 0 5 10 W A T E R A V A I L A B I L I T Y (rating from 0 to 10) EN VIROQUALITY_ ill • 0 5 10 ENVIROQUALITY (rating from 0to 10) H Y D R O F A C T O R S _ 0 5 10 H Y D R O F A C T O R S ( r a t i n g f r o m 0 t o 1 0 ) W A S T E S I T E . W A S T E SITE C H A R A C T E R I S T I C S (rating from 0 to 1 0) W A V E A C T I O N _ 0 5 W A V E A C T I O N (rating from 0 to 1 0) 150 100 s K / - H O T SUN . COLD SUN . HOT C L O U D . C O L D C L O U D 0 500 1000 1500 EVAP M I N E R A L O G Y . P 0 50 100 M INEROLOGY( ra t ing f rom 0to10) 103 A Fuzzy Expert System on Acid Rock Drainage Site Remediation C H A R A C T E R I S T I C S . TAILS C H A R A C T E R I S T I C S (rating f rom Oto 10) tt • 0 500 1000 C E I L I N G C O S T ( -1000) UNDERMAT_ - G O O D . FAIR • P O O R - H I G H . MOD. -HIGH . M O D . " L O W UNDERLYING MATERIAL PERMEABILITY W A S T E SUR F A C E A R E A _ W A S T E S U R F A C E AREA (rating Irom 0 to 1 00) C O V E R C H O I C E _ 0 5 10 COVER C H O I C E (rat ing f rom Oto 10) E N V I R O _ D I S T A N C E _ 150 100 50 0 ENVIRO D I S T A N C E (rat ing f rom Oto 1000) W A S T E MAT_ W A S T E MATERIAL PERMEABILITY W A T E R T A B L E _ 120 , 100 -I •.-80/. h \ / , 40 / / 20 . 1 0-W A T E R T A B L E 104 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 150 100 so -j ;i li \ PRECIPITATION RAIN (rating from 0 to3000) - H I G H „ M O D - H . M O D . M E D - L O W TAILS QUALITY (rating! romOto 1 0) SULPHIDEREACTIVITY_ M SULPHIDEREACTIV !TY( ra t ing from 0 to 10) T A I L S _ S U R F A C E A R E A _ 0 50 100 TAILINGS S U R F A C E A R E A (*1000) PRECIPITATION S N O W ( r a t i n g from 0 to 3000) S O U R C E F A C T O R S _ 0 5 10 S O U R C E F A C T O R S (rating from 0 to 1 0) S U R R O U N D _ M A T _ 150 \ 50 S U R R O U N D I N G MATERIAL PERMEABIL ITY TR E A T P L A NT_ 0 500 1000 1500 2000 TREATING P L A N T C A P ACITY - H I G H S . M O D E R A T E S . L O W S 105 A Fuzzy Expert System on Acid Rock Drainage Site Remediation DIVERT D O W N S T R E A M D/B (rating from 0 to 1 00) PRECIPRAIN PRECIPITATION RAIN (rating from 0 to 3000) . M INTERCEPT D O W N S T R E A M (rating from 0 to 1 00) 0 50 100 S U B S U R F A C E FLOW (rating fromO to 100) H Y D R O L O G I C A L F A C T O R (rating from 1 to 10) 50 100 PASSIVE OPTION (rating from 0 to 100) L A N D A R E A _ LAND A R E A AVAILABLE (rating fromO to 100) W A T E R F L O W _ - S . M V V WATER FLOW (rating from 0 to 100) 106 A Fuzzy Expert System on Acid Rock Drainage Site Remediation DIVERT U P S T R E A M D/B (rating fromO to 1 00) D I S S O L V E D _ O X Y _ 120 100 0 50 100 D ISSOLVED O X Y G E N (rating fromOto 100) ARDPOTENTIAL OQ o O 80 H 60 a - ! M 40 - / ' L 20 - 11 • \ , 0— > V 0 5 ARD POTENTIAL (rating from -3 to 1 0) C O N C _ F E R R I C 90 190 290 FERRIC C O N C E N T R A T I O N (mg/L) INTERCEPT U P S T R E A M (rating from 0 to 1 00) FERRIC CONCENTRATION (rating fromO to 1 00) 0 90 190 290 ALUMINUM CONCENTRATION (mg/L) C O N C _ F E R R O U S 90 190 290 F E R R O U S CONCENTRATION (mg/L) 107 A Fuzzy Expert System on Acid Rock Drainage Site Remediation S U L P H A T E 120 1 00 - l _ . til o O 20 0 0 5000 10000 S U L P H A T E CONCENTRATION (mg/L) 120 100 80 S U R F A C E FLOW (rating Irom 0 to 1 00) EFFLUENT_ACIDITY_ 120 , 100 ""'. r 80 60 1 f 40 )'. 20 r. 0 - » • 500 1000 1500 2000 E F F L U E N T ACIDITY (mg/L) E F F L U E N T _ F L O W _ 190 390 E F F L U E N T FLOW (gpm) 40 90 T R E A T . OPTION 120 , 100 80 1 1 60 N 40 i A 20 1 n n TAILSSURF_ L A -1 0 10 30 50 70 TAILINGS S U R F A C E A R E A (M000 m3) 120 1 00 80 60 40 1 < t i 20 4 | oji. -1 0 90 190 M A N G A N E S E CONCENTRATION (mg/L) ENVIRO_DISTANCE_ •10 40 90 140 ENVIRO DISTANCE (km) 108 A Fuzzy Expert System on Acid Rock Drainage Site Remediation COVER_OPTION 109 A Fuzzy Expert System on Acid Rock Drainage Site Remediation APPENDIX C - HYPERTEXT DOCUMENTS n o A Fuzzy Expert System on Acid Rock Drainage Site Remediation Bactericides (Bacterhy.doc) \bt\ \nt[coverpage]\ \h2.Bacterial Treatment of A R D \ \t This module has been created to work wi th A R D X and facilitate the user in determining the most appropriate treatment option o f which one may be bacterial treatment. A t any point within the expert system the user has access to explanations and justif ications. Fo r the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definit ion o f a topic or word Ye l l ow text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Exit to ARDX\\c2\ \ j t [contents] .Module Contents\ \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard] .Introduction to A c i d R o c k Drainage ( A R D ) \ \c2\\ j t [ARDchemistry].Chemistry o f A R D \ \c2\\jt[bacteria].Bacterial Treatment \c2\\jt[types].A L is t o f Bactericides\ \c2\\jt[references] .References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. It is of significant concern and an increasing environmental challenge for mining industry today. The exposure o f mining wastes to weathering conditions increase the solubil i ty o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D w i l l impact watershed characteristics and create adverse effects in the ecosystem. \t M in ing Waste Dumps are often a location o f A R D generation. Covers can impede A R D generation by l imiting the amount o f oxygen and/or water entering the dump. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A 111 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt [ARDchemistry] \ \h2.Chemistry o f A c i d Rock Drainage\ Pyrite is very abundant in virtually al l sulphide ore bodies and is often the main culprit contributing to the creation o f A R D . The reactions taking place to cause the highly acidic metal contaminated drainage can be looked at through the oxidat ion o f pyrite. A s exposed pyrite comes into contact with oxygen and water it oxidizes to produce ferrous iron, sulphate and hydrogen ions. \ j t [ c h e m l ] . R E A C T I O N \ The reaction produces ferrous iron and sulphate increasing the total dissolved solids and H+ ions thereby resulting hi a decrease in p H . Trace metals such as copper, lead, and zinc that were bound up within the pyrite are released as the pyrite is oxid ized. The ferrous ions can further oxidize to the ferric state. \jt[chem2] . R E A C T I O N \ This normally slower reaction is facilitated by the presence of certain bacteria the most commonly known as Thiobaci l lus Ferrooxidans. Other bacteria include Thiobaci l lus Thiooxidans and Sulfolobus. A l l of these bacteria use energy from the above reaction for their l ife processes. The presence o f the ferric iron can now oxidize pyrite further in the fo l lowing fast reaction. \ j t [ chem3] .REACTION\ A vic ious cycle o f reactions is created constantly generating more acidity (for every mole of pyrite, 4 moles o f acidity are produced), decreasing p H , and creating optimal conditions for Thiobaci l lus bacteria to thrive and drive the reactions further. A s contaminated water continues down the f low path, further reactions complicate the issue leaching metals into solution, neutralizing acidity by calcium carbonate and hydrolyzing iron. \jt[contents].Back to Index.. A\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt [cheml]\ M i l . Pyri te in the presence o f oxygen and water is oxidizedA \ i2 \ \np.cheml.bmp\ Mi l .The reaction produces ferrous iron, sulphate, and hydrogen ionsA Mi l .Total dissolved solids increase while p H decreasesA \i4\ \ j t [ARDchemistry] .Back to T E X T . . A \ i0\ \bt\ \nt[chem2]\ Mil .Ferrous iron oxidizes further assisted by Thiobaci l lus Ferrooxidans\ \ i2 \ \np.chem2.bmp\ M4\ \ j t [ARDchemistry] .Back to T E X T . . A M0\ \bt\ \nt[chem3]\ Ferr ic i ron further oxidizes pyriteA \ i2 \ \np.chem3.bmp\ M4\ \ j t [ARDchemistry] .Back to T E X T . . A 112 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \ i0\ \bt\ \nt[bacteria]\ \h2.Bacterial Treatment o f A R D A \t Bactericides work by creating an unsuitable environment for the growth of Thiobaci l lus ferrooxidans bacteria thereby preventing the increased rate of oxidation of sulphides due to bacterial activity. Bactericides do not stop the inorganic oxidation o f sulphides. Al though not a l ikely control method for long term it may be a possible short term control used in combinati ion with other methods. \t Success o f the method depends on: \c2\-The type o f sulphide in the waste \c2\-The level o f biological activity \c2\-Weather conditions (chemicals are removed by water requireing reapplication) \t Other considerations include: \c2\-Degradation occurs over time requiring reapplication \c2\-The method works best when applied to a fresh unoxidized surface \c2\-The method can provide temporary rel ief until a perminant barrier is put in place \c3\-Appl icat ion may contravene environmental regulations in the area \c3\-Appl icat ion to surfaces is easy while effectiveness decreases with depth \c3\-Fined grained material w i l l exhibit more uniform penetration o f the bactericide with depth \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[types]\ \h2.A L is t o f Bactericides\ \t This is a short list of some types o f Bactericides. \c3\Benzonate Compounds \c3\Sorbate Compounds \c3\Anion ic Surfactants (Sodium Laury l Sulphate (SLS) - \dt[equity].Equity Si lver M i n e \ in B .C . ) \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[equity]\ \t Equi ty Si lver mine condusted lab tests and a mall f ield test using S L S . There was little effect on the water quality, however further lab tests showed that this may have been due to a concentration that was too l o w (16 times higher concentration in lab tests inhibited bacterial action). \bt\ \nt[application]\ \h2.Appl icat ion o f Bactericides\ \t Bactericides normally come in the form o f a powder (requiring reapplication every 3-6 months) although they are also available as pellets (time released for approximately 7 years). 113 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \t Appl icat ion can be achieved by: \c2\spraying - The powder is first mixed with water to form a slurry \c2\mixing - The bactericide is mixed into the waste during construction \c3\Benzonate Compounds \c3\Sorbate Compounds \c3\Anion ic Surfactants (Sodium Lauryl Sulphate) \c3\Phosphate Compounds \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[references]\ \h2.References\ Broughton, L . M . , R .W. Chambers, and A . M a c G . Robertson, 1992. M i n e Rock Guidel ines Design Control of A c i d Drainage Water Quality. Steffen, Robertson, and Kr isten (B .C. ) Inc., Vancouver, B . C . Delaney, T., D. Hock ley, D. Sollner, 1997. Appl icat ions o f Method for Delay ing the Onset o f A c i d i c Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , pp.797-810, V o l . 2. Z iemkiewicz, Pau l , 1995. A c i d M ine Drainage Control Technologies. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia University and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia. , p. 40. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T Dry Covers (Covershy.doc) \bt\ \nt[coverpage]\ \ h 2 . C O V E R S \ \t This module has been created to facilitate the user in determming the most appropriate cover option through the attached expert system. A t any point within the expert system the user has access to explanations and justifications. For the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definition o f a topic or word Y e l l o w text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Covers Expert System\\c2\\jt[contents].Module Contents\ 114 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard].Introduction to A c i d Rock Drainage\ \c2\\jt[introduction].Introduction to Dry Covers\ \c2\\jt[vegitation].Vegitative Cover\ \c2\\jt[simple].Simple Soi l Cover\ \c2\\jt[multiple].Multiple Layer Cover\ \c2\\jt[Complex_clay].Complex Clay Cover\ \c2\\jt[Complex_synthetic].Complex Synthetic Cover\ \c2\\jt[materials].Construction Materials\ \c2\\jt[hydrology].Site Hydrology Characteristics\ \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. It is o f significant concern and an increasing environmental challenge for mining industry today. The exposure o f mining wastes to weathering conditions increase the solubil i ty o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D wi l l impact watershed characteristics and create adverse effects in the ecosystem. \t M in ing Waste Dumps are often a location o f A R D generation. Covers can impede A R D generation by l imit ing the amount o f oxygen and/or water entering the dump. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[introduction]\ \h2.Introduction to Covers\ \t There are various covers now employed by the mining industry as control techniques impeding A R D generation. A large portion o f the seepage out o f a waste dump is due to infiltration o f precipitation through the dump. D ry covers are used to restrict the entry of water into the waste thereby decreasing the volume out o f the dump. They also contribute by decreasing the oxygen entering the waste dump. Wet covers or saturated covers inhibit oxygen entry to a far greater extent then dry covers. Often oxygen diffusion through a wet cover versus a dry cover can be a magnitude lower or more. 115 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Types o f Dry covers include: \jt[vegitation].Vegitative Covers\ \jt[simple].Simple So i l Covers\ \jt[multiple].Multiple So i l Covers\ \ jt[complex_clay].Complex C lay Covers\ \jt[complex_synthetic].Complex Synthetic Covers\ \jt[contents].Back to Contents...\\t\t\et[exit].Back to Kj iowledge Base..A \bt\ \nt[simple]\ \h2. Simple Soi l Covers\ \t Simple soil covers consist o f a single layer o f a lower permeability soil . This may be a compacted t i l l , clay, silt, or another material that may be local ly available on site. The cover is effective in decreasing the amount of infiltration of precipitation into the waste dump. \t Soi ls with fine grain sizes are more susceptible to frost damage, erosion, settlement, burrowing o f animals and root action. The efficiency o f the cover is increased through the addition o f a vegetative layer that w i l l help protect the cover from wind and rain erosion. \t A simple cover w i l l not be as effective in areas with longer snow thawing season. In such areas infiltration can be reduced by 5% to 10% o f the precipitation. \t Lower permeability material used in a simple cover may also be subject to cracking due to seasonal changes in precipitation and moisture content. This w i l l increase the permeability of the material and thereby decrease the efficiency o f the cover. \t The cost implementation o f a simple soi l cover can be anywhere from $2.00 to $6.00 per square meter. \jt[introduction].Back to Introduction to Covers\ \bt\ \nt[multiple]\ \h2.Mult iple Soi l Covers\ \t Mul t ip le soi l covers consist o f several layers with specif ic purposes. These covers are often used in higher rainfall areas and are more effective in restricting oxygen from entering the waste dump. \t A n upper most layer consists of revegetated so i l ; it protects from wind and rain erosion, l imits desiccation of the underlying compacted layer, and retains moisture. \t A coarse break layer that consists o f a coarser material, provides lateral drainage o f the infiltrating water, and prevents the loss of water f rom the underlying compacted layer. \t The compacted layer may consist o f a clay or other l ow permeability soi l . It serves to prevent the infiltration of water and the diffusion o f oxygen into the waste rock below. \jt[introduction].Back to Introduction to Covers\ \bt\ \nt[complex_clay]\ \h2.Complex C lay Covers\ \t Complex clay covers; very similar to the multiple soi l cover and also consist o f several layers with specific purposes including a clay layer. These covers are often used in higher rainfall areas and are more effective in restricting oxygen from entering the waste dump. \t A n upper most layer consists o f re-vegetated so i l ; it protects from wind and rain erosion, l imits desiccation of the underlying compacted layer, and retains moisture. 116 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \t A coarse break layer that consists of a coarser material, provides lateral drainage of the infiltrating water, and prevents the loss o f water from the underlying compacted layer. \t The compacted layer consists o f a low permeability clay. It serves to prevent the infiltration o f water and the diffusion o f oxygen into the waste rock below. \t A coarse lower break layer is made of coarser material and may include added neutralizing material. It prevents desiccation o f the overlying layer by preventing contaminated water from moving up from the waste rock into the compacted layer and assists in neutralizing contaminated waste water entering the break layer. $20.00 to $30.00 per square meter \jt[introduction].Back to Introduction to Covers\ \bt\ \nt[complex_synthetic]\ \h2.Complex Synthetic Covers\ \t The complex synthetic layer can provide a more permanent and effective solution (has a life span of approximately 100 years). \t The cover is constructed the same as the multiple soil cover or the complex clay cover with synthetic material (liner) added below the compacted layer. \jt[introduction].Back to Introduction to Covers\ \bt\ \nt[hydrology]\ \h2.Site Hydro logy Characteristics\ \t O n choosing appropriate control methods for the site, considerations to the hydrology o f the site are always necessary. A high volume o f water moving through a waste dump wi l l increase the amount o f effluent that can be generated. O n the other hand; a site with minimal water f low through the dump can lead to a decrease in effluent volume and an increase in acid and metal concentrations within the effluent. \t Site hydrology involves an understanding o f the water balance o f the waste dump or other area of concern (tailings pond). A waste rock dump water balance generally consists of: \c2\ - Water infi ltration through the surface o f the dump due to precipitation including rain and snowfall. \c2\ - Water runoff f rom surrounding higher topographical regions into the waste dump. \c2\ - Ground water movement into the dump due to a high water table. \cOWt Other factors inf luencing the site hydrology include evaporation of water from the surface of the dump leaving less water available for infi ltration, a long seasonal freshet (spring snow melt) period increasing infiltration, runoff and water table height due to. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[vegitation]\ \h2.Vegitative Cover\ \t A vegitative cover consists of a moisture retention layer to support growth and a vegitation top layer. The cover does not l imit the water infi ltration into the tailings but enhances evapotranspiration. Its function is to prevent erosion due to wind and water, to facilitate regeneration o f the area and to prepare it for future land use. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ 117 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \nt[materials]\ \h2.Construction Materials\ \t D r y covers can be constructed using many different materials. Materials found on or near the site can often be used and are generally a less expensive option. \t Cover materials include: \c3\\dt[soil].soil\ \c3\\dt[clay].clay\ \c3\\dt[till].compacted t i l l \ \c3\\dt[concrete].concrete and asphaltA \c3\\dt [synthetic]. synthetic materials\ (Lawrence, 1997) \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[soil]\ \h2.Soi l \ \t So i l is used as a humic top layer. This layer acts as a moisture retention zone to support vegitative growth. \bt\ \nt[clay]\ Vh2.Clay\ \t Compacted Clay, although low in permeability ( 10EXP-9 to 10EXP-11 m/s)is subject to erosion, cracking and root penetration; and may not be readily available in al l areas. C lay caqn be a good sealing material i f it is we l l maintained. \bt\ \nt[till]\ \h2.Compacted T i l l \ \t Compacted t i l l has similar characteristics as compacted clay although it is less permeable at 1 0 E X P - 7 to 1 0 E X P - 9 m/s. \bt\ \nt[concrete]\ \h2.Concrete and AsphaltA \t Concrete and asphalt are both subject to cracking, frost and mechanical damage. \bt\ \nt[synthetic]\ \h2.Synthetic Materials\ \t Synthetic materials are highly impermeable albeit they require the necessary bedding and cover to protect f rom mechanical and root penetration. \bt\ 118 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \nt[references]\ \h2.References\ Delaney, T., D . Hockley, D. Sollner, 1997. Applicat ions of Method for Delay ing the Onset o f A c i d i c Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , pp.797-810, V o l . 2. Lawrence, R . W . , 1997. A Course on Fundamental and App l ied Aspects o f A c i d Rock Drainage. Course notes, Department o f M in ing and Minera l Process Engineering, U B C , Vancouver, Canada. Z iemkiewicz, Pau l , 1995. A c i d M ine Drainge Control Technologies. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia Universi ty and the National M ine Reclamation Center, Morgantown, West V i rg in ia . , p39. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T Water Diversion (Divershy.doc) \bt\ \nt[coverpage]\ \h2.Diversion o f Water to prevent A R D \ \t This module has been created to work with A R D X and facilitate the user in detennining the most appropriate treatment option. A t any point within the expert system the user has access to explanations and justif ications. For the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definition of a topic or word Y e l l o w text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Exit to ARDX\\c2\\ j t [contents].Module Contents\ \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard].Introduction to A c i d Rock Drainage\ \c2\\jt[diversion] .Water Diversion\ \c2\\jt[surface].Diversion o f Surface Water\ \c2\\jt[interception].Interception of Ground Water\ 119 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \c2\\jt[references] .References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. It is of significant concern and an increasing environmental challenge for mining industry today. The exposure of mining wastes to weathering conditions increase the solubi l i ty o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D wi l l impact watershed characteristics and create adverse effects in the ecosystem. \t M i n i n g Waste Dumps are often a location o f A R D generation. Covers can impede A R D generation by l imit ing the amount of oxygen and/or water entering the dump. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[diversion]\ Vh2.Water Divers ion\ \t Control o f water migration is achieved through: \c3\\jt[surface].diversion of surface water f lows\ \c3\ and \jt[interception].interception of groundwater f lows\ \t The goal is to: \c2\ - decrease the amount of water entering an acid producing site such as a waste rock pi le thereby decreasing the amount of contaminated effluent exiting the site \c2\ - prevent the release o f contaminated effluent into the surounding environment \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[surface]\ \h2.Diversion of surface Water\ \t Divers ion o f surface water is achieved through: \c4\ditches \c4\berms \c4\surface contouring \t Divers ion can be implemented both upstream o f waste to prevent water from entering the site and downstream to collect contaminated effluent. \t Maintainance and inspection o f diversion structures would be required for long term and can become a significant cost factor. Proper mine planning and site selection for waste rock piles and other acid generating areas can avoid natural drainage channels and runoff areas. The need for man-made diversion structures can then be eliminated and costs decreased. (Broughton, et. A l . , 1992) 120 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[interception]\ \h2.Interceptiion of Groundwater \t Interception of Groundwater is achieved through: \c3\ proper site selection - to avoid natural groundwater discharge areas \t and interception through: \c3\ slurry walls \c3\ grout curtains \c3\ deep drainage ditches \c3\ boreholes or wells \c3\ impermeable cut-off wal ls \t Short term interception that requires ongoing mantainnance may include: \c3\ gravel f i l led ditches \c3\ wells and pumps (Broughton, et. A l . , 1992) \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[references]\ \h2.References\ Broughton, L . M . , R .W. Chambers, and A . M a c G . Robertson, 1992. M i n e Rock Guidel ines Design Control of A c i d Drainage Water Quality. Steffen, Robertson, and Kr is ten (B .C. ) Inc., Vancouver, B . C . Broughton, L . M . , P . M . Healey, 1992. M i n e Site Decommissioning: Technology for Assessment and Control of A c i d Rock DrainageQuality. In: Environmental Issues and Waste Management in Energy and Minerals Production. Singhalet al.(eds) Balkema Publishers, Ba lkema, Rotterdam, Netherlands. Pp789-797. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T Passive Treatment (Passivhy.doc) \bt\ \nt[contents]\ \h2.Document Contents\ B y cl icking on a topic you can have direct access to further information in that area. The Forward-Browse button w i l l a l low access to the topics page by page. \c2\\j t[overview]. Overv iew o f Documen t 121 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \c2\\jt[introduction].Introduction to Passive Treatments\ \c2\ \ j t [ARDchemistry] .Acid R o c k Drainage Chemistry\ \c2\\jt[passive].Passive Treatment Systems\ \c2\\jt[aLkalimty].Alkalinity Generation\ \c2\\jt[bacteria].The Ro le o f Bacter ia in A lka l in i ty Generation\ \c2\\jt[microbes].The Ro le o f Microorganisms in Meta l Remova l \ \c2\\jt[algae].The Ro le o f A lgae in Meta l Removal \ \c2\\jt[plants].Reactions W i t h Plants\ \c2\\jt[organic].Adsorption to Organic Substrate\ \c2\ \ j t [ALD] .Anox ic Limestone Drains\ \c2\\jt[aerobic].Aerobic Ce l ls \ \c2\\jt[anaerobic] .Anaerobic Ce l ls \ \c2\\jt[ald_con].Construction o f an Anox i c Limestone Dra in\ \c2\\jt[aerobic_con].Construction o f an Aerob ic C e l l \ \c2\\jt[anaerobic_con].Construction o f an anaerobic Ce l l \ \c2\\jt[param].Design Parameters\ \c2\\jt[site_con].Site Considerations\ \c2\\jt[cell_design].Cell Des ign\ \c2\\jt[f low].Flow Pattern\ \c2\\jt[costs].Costs\ \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[overview]\ \h2.Manual on Passive Treatments for A c i d Rock Drainage\ \ i l \ Th i s manual has been prepared to assist the user in understanding Passive Treatment processes for the remediation o f A c i d R o c k Drainage sites. \ i l \ Some useful features include: White Text - provides a short definit ion o f a topic or word Y e l l o w text - jumps you to a different topic or displays equations for further clarif ication \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[ard]\ A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering of pyrite and other sulfide minerals. The weathering conditions increase the solubil i ty of heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D w i l l impact watershed characteristics and create adverse effects in the ecosystem. \bt\ \nt[weather]\ Weathering is the breakdown o f rocks and minerals at the Earth's surface by the action o f physical and chemical processes. The process can oxid ize pyrite, breaking it down and releasing iron, sulfide, and other minerals that may be locked into the matrix o f the rock. \bt\ \nt[introduction]\ 122 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \h2.Min ing A n d A c i d Rock Drainage\ \dt[ard].Acid Rock Drainage\ ( A R D ) occurs when pyrite and other sulphide minerals are exposed to \dt[weather].weathering\ f rom the mining o f mineral resources. The problem is ongoing in coal as we l l as metal mines. Once exposed, waste rock generating acidic and metal contaminated drainage may continue to do so for centuries. Furthermore, A R D may take time to generate and the problem may not be recognized for may years. Chemical treatment wi th l ime is the accepted form of dealing with A R D . Treated effluent can be released safely into the environment or recycled to the process. A s A R D continues to be generated in an active or abandoned mine, chemical treatment w i l l often need to continue for hundreds o f years. This may become a demanding and costly long term commitment. M i n e sites may be located in remote locations, fluctuations in f low rates, large volumes o f l ow intensity sludge, and the unknown stability o f the chemical sludge are al l unwanted factors affecting the long term treatment o f A R D using chemical methods. Thus the mining industry has been faced with the ongoing challenge of f inding an economic and sustainable solution to the A R D problem. \jt[passive].Passive Treatment systems may be a step toward f inding this solution. For further information on the chemistry o f A R D cl ick \ j t [ARDchemistry] .HERE\. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt [ARDchemistry] \ \h2.Chemistry o f A c i d Rock Drainage\ Pyrite is very abundant in virtually al l sulphide ore bodies and is often the main culprit contributing to the creation o f A R D . The reactions taking place to cause the highly acidic metal contaminated drainage can be looked at through the oxidat ion o f pyrite. A s exposed pyrite comes into contact with oxygen and water it oxidizes to produce ferrous iron, sulphate and hydrogen ions. \ j t [ chem l ] .REACTION\ The reaction produces ferrous iron and sulphate increasing the total dissolved solids and H+ ions thereby resulting in a decrease in p H . Trace metals such as copper, lead, and zinc that were bound up within the pyrite are released as the pyrite is oxidized. The ferrous ions can further oxidize to the ferric state. \ j t [ chem2] .REACTION\ This normal ly slower reaction is facilitated by the presence o f certain bacteria the most commonly known as Thiobaci l lus Ferrooxidans. Other bacteria include Thiobaci l lus Thiooxidans and Sulfolobus. A l l o f these bacteria use energy f rom the above reaction for their l ife processes. The presence of the ferric i ron can now oxidize pyrite further in the fo l lowing fast reaction. \ j t [ chem3] .REACTION\ A v ic ious cycle of reactions is created constantly generating more acidity (for every mole o f pyrite, 4 moles o f acidity are produced), decreasing p H , and creating optimal conditions for Thiobaci l lus bacteria to thrive and drive the reactions further. A s contaminated water continues down the f low path, further reactions complicate the issue leaching metals into solution, neutralizing acidity by calcium carbonate and hydrolyzing iron. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt [cheml]\ \h l .Pyr i te in the presence o f oxygen and water is oxidizedA \ i2\ \np.cheml.bmp\ Mi l .The reaction produces ferrous i ron, sulphate, and hydrogen ionsA \h l .To ta l dissolved solids increase whi le p H decreasesA \i4\ \ j t [ARDchemistry] .Back to T E X T . . A \i0\ 123 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[chem2]\ \h l .Ferrous iron oxidizes further assisted by Thiobacil lus Ferrooxidans\ \ i2\ \np.chem2.bmp\ \ i4\ \ j t [ARDchemistry].Back to T E X T . . A \ i0\ \bt\ \nt[chem3]\ Ferr ic i ron further oxidizes pyriteA \i2\ \np.chem3.bmp\ \ i4\ \ j t [ARDchemistry].Back to T E X T . . A \ i0\ \bt\ \nt[passive]\ \h2.Passive Treatment Systems\ Passive treatment systems are self sustaining systems that can be attractive alternatives to conventional l ime treatment. Treatment of A c i d Rock Drainage includes the generation o f \jt[alkalimty].alkalmity\ and elimination of contaminants. M a n y processes such as \jt[microbes].bacteria\, \jt[algae].algae\, \jt[plants].plants\, and \jt[organic].organic substrate\ are at work to facilitate the removal o f contaminants. Passive systems include: \ i3 \ \ j t [ALD] . Anox i c Limstone Drains\ (ALD 's ) \jt[aerobic].Aerobic\ cells \jt[anaerobic].Anaerobic\ cells \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[alkalinity]\ \h2.Alkal in i ty Generation\ Passive treatment w i l l generate alkalinity through a number o f processes o f wh ich two o f the most dominant are limestone and \jt[bacteria].sulfate reductionV The limestone layer in an anoxic limestone drain or as a bed in an anaerobic cel l can react with hydrogen ions to decrease the acidity o f the water. \ j t [ a l k l ] .REACT ION\ This process occurs only in anaerobic conditions. Once exposed to an ox id iz ing environment the limestone effectiveness is inhibited and the effective buffering capacity decreases dramatically do to the formation o f an iron hydroxide coating on the limestone surface. The dissolution o f the limestone cannot occur, thus hindering the alkalinity generation capability of the cel l . \ j t [ALD] .ALD 's \ are one treatment system that uses this process. \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ 124 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \nt[alkl]\ \ i2\ \np.aLkl.bmp\ \ i4\ \jt[alkalinity].Back to T E X T . . A \i0\ \bt\ \nt[bacteria]\ \h2.The Role o f bacteria in passive treatment systems\ Bacteria play a role in treating acid rock drainage by assisting in both alkalinity generation and the removal o f \jt[microbes].metals\. Sulphate reduction by bacteria contributes to the generation o f alkalinity. The process occurs due to Sulfate Reducing Bacteria (SRB) that use simple organic compounds as an energy source. These compounds are usually products of fermentation as the SRB 's must rely on other microbes for the bacterial degradation of proteins, carbonates, and other dead biomass. \ j t [ bac t l ] .REACTION\ The system wi l l generate permanent alkalinity as H 2 S gas is released into the atmosphere through the generation o f H S - ions. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[bactl]\ \h2.Reduction of sulphate with lactate is shown hereA \ i2\ \np.bactl.bmp\ \i4\ \jt[bacteria].Back to T E X T . . A \i0\ \bt\ \nt[microbes]\ \h2.The Role of Microorganisms in Meta l Removal \ Numerous different bacteria assist in metal removal. Sulfate reducing bacteria are instrumental in removing metals by producing hydrogen sulfide causing metal precipitation in an anaerobic environment. These bateria are also instrumental in \jt[bacteria].alkalinity\ generation. Other bacteria also play a role in metal removal. Iron can be ox id ized with the help o f a family of bacteria known as the Ferrooxidans. These bacteria can survive an environment wi th a p H o f less than three whi le feeding on the inorganic nutrients o f Fe(JT), C 0 2 , NH4+. It has been found through experimentation that Ferrooxidans have the highest populations in the aerobic layer o f substrate in a wetland. \ t \ t \ j t [micr l_2] .REACTIONS\ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[micrl_2]\ 125 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Bacteria derive energy by converting ferrous iron to ferric through the fo l lowing reaction... \i2\ \np.micr l .bmp\ \ i0\ The ferric iron now undergoes a hydrolysis reaction... \i2\ \np.micr2.bmp\ \ i0\ The reaction produces ferric hydroxide and the iron precipitates out of the system. \ i2\ \jt[microbes].Back to T E X T . . A \ i0\ \bt\ \nt[algae]\ \h2.The Ro le of Algae in Meta l Remova l \ Few studies have been done on the role o f algae in an A R D environment. Their effects usually go unnoticed, introduced into passive treatment systems unintentionally (with the introduction o f other plants) through secondary sources. Algae however, can accumulate metals within their body structure. They can act as biological filters trapping sediments and nutrients high in metals. Intracellular crystals have been observed in suspended algal cells under a microscope. Iron and manganese are used by algae as essential micronutrients, and algae have been found to accumulate manganese up to 56 g/kg of plant tissue. Algae are capable o f reproducing rapidly and may also benefit to increase organic matter in a cel l as dead biomass. Mic rob ia l mats made up o f blue-green algae and bacteria have been developed and used in treatment of acid coal and metal mine drainage. B i o m a t T M s have been shown to remove M n at a rate o f 6.5g/m2/d wi th inf low levels o f 7.5mg/L M n and a f low rate o f 16L/min. B i o m a t T M promotes high p H levels and can incorporate metals into organic matter through cation exchange, adsorption, precipitation & co-precipitation, and complexation or chelation. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[plants]\ \h2.Reactions Wi th Plants\ Passive treatments rely on plants to perform a number o f functions to ameliorate A R D . Dur ing past years it was thought that that metal uptake by plants was a significant and important process. Recently, however, studies have shown that plants actually accumulate very little metals accounting for less than 1% annual i ron removal by Typha in one case. Al though plants may not be directly responsible for significant metal uptake, they aid in the process of metal removal by assisting in other ways. The stems and leaves o f plants increase the surface area available for the attachment of microbes. In this way the plants presence impacts the population of microbes that are necessary within the system. On an annual basis new exchange sites are made available through the death of o ld biomass as new biomass is produced. The stem and root systems o f the plant play an important role in the transportation o f atmospheric gasses. Oxygen is introduced into the system through the roots and creates an aerobic region. Oxygen surrounds each root hair supporting further microbial populations. Plant growth itself becomes instrumental in many passive systems. Photosynthesis is responsible for carbon f ixing and the accumulation o f new biomass. O l d biomass is al lowed to die and decompose providing a constant source of carbon. Mic rob ia l processes are sustained as the organic carbon is ut i l ized as an energy source o f electrons. \jt[contents].Back to Index...\\t\t\et[exifj.Back to Knowledge Base..A 126 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[organic]\ \h2.Adsorpt ion to Organic Substrate\ Metals can be accumulated in a wetland through an exchange process with humic and f luvic acids in soil . Organic soils (peat) have a cation exchange capacity ( C E C ) that increases at lower p H values. The C E C allows peat to absorb posit ively charged ions. \ j t [ o r g l ] . R E A C T I O N \ The carboxyl ic group ( - C O O H ) attached to the inert organic portion o f the substrate ( R ) wi l l dissociate al lowing the carboxyl ion ( C O O - ) to react with the metal ion (M) in the solution. The C E C o f the substrate w i l l go down as metal ions react wi th the exchange sites and the organic substrate becomes saturated with metals. N e w exchange sites w i l l only become available on an annual basis through the introduction o f new organic matter as the wetland plants grow and die. This complexation of metal ions can work against the intended removal o f metals from A R D waters. A soi l wi th a high concentration of metals that comes into contact with acidic water with lower concentrations can cause the reaction to reverse, releasing the metals back into solution. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[orgl]\ The metal ions are exchanged with other ions according to the fo l lowing reaction. A \ i2\ \np.orgl .bmp\ \i4\ \jt[organic].Back to T E X T . . A \ i0\ \bt\ \n t [ALD] \ \h2.Anoxic Limestone Drains\ Anox i c Limestone Drains use limestone to raise p H and add alkalinity to A R D waters. A L D ' s have been shown to generate up to 300mg/L o f alkalinity when installed correctly and water quality parameters comply (Skousen, 1991). The Drains must be constructed such that anoxic conditions prevai l and therefore are buried in underground channels or trenches such that oxygen is not available. \dp[ald].ald.bmp\ \ i4 \ \ j t [a ld_con] .DESIGN & C O N S T R U C T I O N o f an Anox i c L imeston Drain.. . \jt[contents].Back to Index.. A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[aerobic]\ \h2.Aerobic C e l l \ Aerob ic cells are built to encourage the oxidation processes. They are used to collect water and remove metals as oxides by providing residence time within the cel l . In A R D waters iron is primari ly removed within the aerobic cel l . Other metals removed here are A l , M g , Arsenic, cyanide and mercury. The shal low construction o f the cel l increases the surface area exposed to oxygen. Water treatment occurs on the surface sometimes requiring large areas for treatment depending on the input water quality and the demands o f the output quality. The shal low nature of the aerobic cells can result in unwanted freezing during the winter months. \dp[aerobic].aerobic.bmp\ 127 A Fuzzy Expert System on Acid Rock Drainage Site Remediation T o further increase the aerobic conditions, vegetation (reed beds) or green algae (release oxygen during photosynthesis) are added. Oxygen is introduced through the root structure of the vegetation and facilitates metal precipitation. The vegetation also serves as a filter for suspended materials. The roots can hold together the substrate. This prevents the formation of channels and serves to increase the residence time. Furthermore, vegetation has the benefit o f added aesthetic value for the treatment area. \jt[aerobic_con] . D E S I G N & C O N S T R U C T I O N of an aerobic cell...\t\jt[contents].Back to Index.. A\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[anaerobic]\ \h2.Anaerobic Ce l l \ Anaerobic cells work pr imari ly to generate alkalinity by promoting reducing conditions. Treatment and thus water f low occurs in the subsurface o f the cel l . Since the anaerobic cells are deeper then the aerobic, unlike the aerobic they are capable o f functioning in sub-freezing climates. Sulfate reducing bacteria (SRB's ) such as desulfouibrio and desulfotomaculum play a large rol l by ut i l iz ing organic matter as a carbon source. S R B ' s can generate up to 200 mg/L o f sulfides. B y using sulfate as an electron acceptor for l i fe processes, SRB 's convert sulfate into hydrogen sulfide. \dp [anaerobic] .anaero.bmp\ \ j t [anaerobic_con].DESIGN & C O N S T R U C T I O N ; of an anaerobic cel l . \ i0 \ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[anaerobic_con]\ \h2.Design and Construction of an Anaerobic Ce l l \ Anaerobic cells are designed for sub-surface flows. These are cells that encourage the growth o f S R B ' s and thus need the addit ion o f organic matter. Organic matter may include potato waste or spent mushroom compost and should be about 12 to 18 inches in depth, with most cells being closer to the later end (Skousen, et a l . , 1992). Biocompatabi l i ty is important. Presence o f S R B ' s and other microbes within the chosen organic substrate w i l l be a good indication that populations w i l l grow there. T o ensure optimal conditions for the growth o f microbes it is ideal to maintain an influent p H o f about 5 or above (Dodds-Smith, et a l . , 1995). A Vegetation cover can assist in creating good conditions for microbes and can be added as a floating mat to reduce surface mix ing, provide a continuous input of new organic matter and to sustain a larger microbial population. A n example o f the construction o f a vegetation mat uses mats made up o f 4 " diameter P V C tubing with plastic fencing as support. The rafts are filled with a 10cm layer o f peat and planted with cattail seedlings. (Ka l in , 1997) Other covers such as synthetic ones can also be used to cover the cel l . L imestone is sometimes added to the cel l . Th is would represent a layer at the bottom o f the cel l for the purpose o f adding further alkalinity to the system. A s in the A L D , the particle size w i l l be important in deterrnining good hydraulic condit ion and unobstructed f low through the system. A range of sizes can give the benefits o f a larger surface area for reaction as wel l as larger particles for better f low conditions. Hydraul ic conductivity is important since the cel l is usually set up for up or down f low through the organics (and limestone i f present). Experience has shown that down f lowing cells are less prone to problems and require less head in the system to push the water through. Figure 5.3a and Figure 5.3b show two different cells for anaerobic treatment. Volumetr ic and area-loading factors w i l l determine the size o f the sell design. Typical ly cells are designed for a volumetric factor o f 0.3 moles o f metal loading per m3 o f cel l volume and an area-loading factor o f 20m2 o f surface area per L/sec o f f low (Gusek, 1995). A loading factor based on 5g/m2/d of iron removal may also be used (Skousen, et a l . , 1992). \jt[anaerobic].To Anaerobic Cel l . .A \t\jt[contents].Back to Index..A\t\et[exit].Back to Knowledge Base..A 128 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[ald_con]\ \h2.Design and Construction of an Anox ic Limestone Drain\ The size of an anoxic limestone drain is dependent on the f low rate, the drainage water quality and the amount of treatment needed. Higher flow rates would necessitate a larger system in order to increase retention time o f the contaminated water. A n average retention times vary and can be anywhere between 15-20 hours (Faulkner, and Skousen, 1995) to 48 hours (Gusek, 1995). Max imum flow rates are usually about 100 gpm due to size and area limitations unless there is a low mineral acidity. The drain should be designed to be completely f i l led wi th water year round and al low for overflows. There must be enough alkalinity added to counter the acidity of the inlet water. M o r e acidic water w i l l require a larger A L D to attain equivalent results. A L D ' s generate approximately 300 mg/L o f alkal inity, and therefore w i l l not treat drainage of more than 300 mg/L of acidity (Skousen, 1991). A low dissolved oxygen content is required to prevent armouring o f the limestone. The iron should be mainly in a ferrous state such that it w i l l not precipitate at the p H level o f the drain. A s wel l aluminum concentrations should be less than 25 mg/L otherwise the aluminum can precipitate in the drain and can cause armouring and clogging (Skousen, e ta l , 1992) \ i4 \ \ j t [a ld_ l imes tone ] .L IMESTONEADDITIONS. . . \ \jt[ald].To Anox ic Limestone Drain..A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[ald_limestone]\ \h2.Limestone Addi t ions to an Anox ic Limestone Dra in\ The limestone quality w i l l affect the level o f alkalinity generated by the drain, therefore, a min imum grade o f 9 0 % calcium carbonate is recommended. The size o f the limestone particles used within the drain should be large enough to al low for uninterrupted flow yet small enough to provide a large enough surface area for dissolution and alkalinity generation. A range o f particle sizes would provide the best results. Recommendations are between 1.5 to 4 inches in diameter (Skousen, 1991). In addition, the drain should be able to hold limestone to generate 20 years o f alkalinity. The amount o f limestone necessary for a particular l i fe o f the drain and other important calculations are shown below: \t 1. A c i d load o f the drainage (tons/year), A tons/year o f acid = f low (gpm) * acidity (mg/L) * 0.0022 2. Estimated life o f the drain (years), B tons o f limestone over years of l ife = A (tons/year) * years of l i fe 3. Calc ium carbonate content o f limestone (%), C tons = B (tons) / calcium carbonate content (%/100) 4. Amount of dissolution o f limestone (%), D total tons = C (tons) / dissolution (%/100) \ i0\ A plastic layer placed on the top of the limestone serves to keep oxygen out o f the ce l l . Other covers include hay bales separated from the limestone by a filter fabric and enclosed with plastic (Skousen,1991). This type o f cover also serves to impede the movement o f oxygen into the cel l . The plastic in both cases is overlain wi th a compacted fine soil matrix such as clay to further hinder the diffusion of oxygen. \jt[ald_con].Back to A L D Construction..A \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[aerobic_con]\ \h2.Design and Construction o f an Aerobic Ce l l \ A n aerobic cell is built to promote surface flow. It is a shallow cel l o f approximately 0.3 meters in depth wi th a length to width ratio of approximately 10:1 (Sengupta, 1993). The size must accommodate the volume o f the precipitate that w i l l come out o f solution. Approximately 2 to 11 g/day/m3 o f i ron can be removed from the surface area o f the cell (Gusek, 1995). Siz ing is also often based on 20g/m2/d o f i ron removal (Skousen, et a l . , 1992). The 129 A Fuzzy Expert System on Acid Rock Drainage Site Remediation cell should be able to retain 20 years of precipitate and w i l l then need to be dredged. Precipitate volume can be calculated for iron in the fol lowing example (Dodds-Smith, et a l . , 1995): A (mg/L) o f iron * B (L/sec) of f low = C (m3) o f hydroxide precipitate per year Aerobic cells usually require a large surface area. The cel l must therefore be built to accommodate what may result in significant rainfall in order to prevent overf low o f untreated water. A rea required for treatment in an aerobic cel l is dependent on the p H o f the drainage water entering. Pretreatment to give a p H value of 5 resulted in a decrease in the required area o f the anaerobic cel l by 7 5 % at the Wheal Jane site (Dodds-Smith, et al . , 1995). Residence time can be increased through the use of cells in series. Other features such as cascades between cells can increase aerobic mix ing and introduce more oxygen into the cell thereby increasing the productivity. Plants must be introduced. When choosing the right species its tolerance to local watershed conditions and the often harsh acid drainage must be considered. A diversity o f species may be used as poly-cultures offer the benefits of a better chance o f survival and usually better treatment. Another consideration should be bioaccumulation. If the plant has a tendency to accumulate certain toxins these may be introduced through the plant into the food chain. The plants should be introduced and allowed to establish themselves before drainage water is introduced to al low for a better chance o f survival. Common plants used in an aerobic cel l include Sphagnum sp., Typha sp., and Phragmites sp. \jt[aerobic].To Aerobic Cell..AM: \jt[contents].Back to Index..A\t\et[exit].Back to Knowledge Base..A \bt\ \nt[design_over]\ \h2.Overview of Design Criteria\ Design begins with choosing the appropriate passive treatment system for the specif ic contaminated drainage site. Design parameters can only be set once the proposed function o f the system is identified. A n analysis of the drainage water to determine the necessary processes needed to reach effluent standards w i l l determine the function o f the system. Among these may be included metal removal, alkalinity generation and p H increase. Passive treatments are not an exact science and involve inter-disciplinary work. Imprecise decisions based on an incomplete understanding of the many processes in passive systems as we l l as the chemistry o f the drainage itself are involved. It is not surprising then that some discrepancies arise between various sources. When choosing an appropriate passive system it is necessary to look at many parameters and characteristics o f the drainage. \ i3 \ \jt[param].Design Parameters\ \jt[site_con].Site Considerations; \jt[cell_design].Cell Design\ \ jt[f low].Flow Pattern\ \jt[costs].Costs\ \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[param]\ \h2.Design Parameters; The fo l lowing is a list o f some o f parameters that should be measured in order to understand the drainage characteristics and be able to start choosing an appropriate system for treatment. \ i l \ 1) F lows with account for f low variations 2) p H 3) acidity and alkalinity 130 A Fuzzy Expert System on Acid Rock Drainage Site Remediation 4) iron and manganese concentrations \ i0\ Measurements should be taken ass close to the A R D outlet as possible. If results show a net acidity in the water an A L D may be of consideration and further parameters must be measured. \ i l \ 1) Dissolved oxygen content (this may be measured according to the E h with Eh=0 or less showing low oxygen content ( < 2 mg/L )) (Skousen, 1991) 2) ferrous iron [Fe+2] and aluminum concentrations \ i0\ \i2\\jt[design_over].Back to Overv iew of Design Criteria.. A \jt[site_con].Site Considerations..A \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[site_con]\ \h2.Site Considerations\ A n appropriate system to treat acid rock drainage must be site specific. Topography, climate and local weather variabil i ty, and watershed characteristics al l play a role in a passive treatment system. (Wi ldeman, et al . , 1993) Site considerations include: \ i3 \ \dt [source]. Sources\ \dt[flow_rate].Flow Rate Variabi l i ty\ \dt[fluid].Fluid Co l lec t ion \dt[contain].Containment Structures\ \ i0\ \i2\\jt[design_over].Back to Overv iew o f Des ign Criteria.. A \jt[cell_design].Cell Design..A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[source]\ Sources - The source o f A R D may be diverse and must be identif ied. Some typical sources include adit/tunnel portals, waste rock piles, tailings ponds, open pits, shafts and seeps (either naturally occurring or due to mining activity). \bt\ \nt[flow_rate]\ F l o w Rate Variabi l i ty - Determining the f low rate and its variabi l i ty is as important task to designing. Histor ical data (at least one year previous) may be extremely beneficial to determining seasonal highs and lows. F l o w variabil i ty of streams within a watershed can provide further information. Surge flows can dramatically effect the passive treatment system. Risks of a surge event may dictate the construction of features such as water holding cells or impoundments. \bt\ \nt[fluid]\ F lu id Col lect ion - In order to prevent metal hydroxide formation in conveyances f lu id interception should be done as close to the source o f drainage as possible. I f impoundments are necessary they may include underground impoundments or open ponds. \bt\ 131 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \nt[contain]\ Containment Structures - The slope of the site, area available, watershed characteristics and groundwater f low are all useful in the design o f the structures and may cause problems i f forgotten or not considered. \bt\ \nt[cell_design]\ \h2.Cel l Design\ A modular design for the cells within a passive system wi l l provide the most f lexibi l i ty for the system. A large single cell has numerous drawbacks including diff icult control, slow response to changing conditions, and difficulty in adapting and redesign i f necessary. A system made up o f many small modules instead o f one large one w i l l generate problems when balancing water levels, f lows and other conditions. The optimal system wi l l consist a number of modules that are easy enough to handle yet not so numerous as to pose regulatory problems. The \jt[size].size o f the cel ls\ w i l l be governed by the f lux to set the lower \ minimum size l imit, while overall treatment requirements and site characteristics w i l l set the upper \ maximum size limit. Modules o f cells can be set up in various configurations based on requirements of the system and site limitations. Three categories of cel l arrangements al l provide different benefits to the system. \ i3 \ \dt[parallel] .Parallel Modu le Arrangement \dt[series].Series C e l l Ar rangement \dt[combo].Combination Arrangement \ i0\ \i2\\jt[design_over].Back to Overv iew o f Design Criteria..A \jt[flow].Flow...\ \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[size]\ \h2.Cel l Number and SizeA \t\Further criteria that may be used to determine the size and number of cells include: \ i l \ 1) Maintenance - addit ion o f organics and long term reliabil ity 2) Hydro logy and F l o w - control o f large f lows may be distributed amongst a number of cells 3) Site - topography and hydrology o f a site can l imit size or serve to assist with f low 4) Need for Process Cel ls - separation o f different functions to separate cells for better performance (separate cell for aerobic and anaerobic functions) 5) Mater ials Handl ing A r e a - for storage or handling of new or old substrate \ i0\ \jt[cell_design].Back to C e l l Design..A \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[parallel]\ Paral lel Modu le Arrangement - This type o f arrangement contains a number o f cells making up equal and parallel streams. Usua l ly f ive to ten cells are used wi th more cells becoming harder to control. The system is more reliable then a single cel l system. A s one cel l may be unavailable due to maintenance the others can serve as temporary backup cells. \bt\ . \nt[series]\ 132 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Series Ce l l Arrangement - Cel ls in series are made up of two or more cells connected by f low that moves through one and into the next. Drainage water is treated by all the cells within the series. This allows for special ized cells that can treat one aspect o f the drainage whi le the next cel l may treat another. It offers better results then a single large cel l system. \bt\ \nt[combo]\ Combinat ion Arrangement - The combination arrangement offers the benefits of both the parallel modules and the series cells such that f low can be routed and divided up into a number of streams and also have the benefit o f treatment through a number o f specific cells. This type o f system usually requires a fairly substantial area o f land thus site l imits might restrict the use of such a system. A combination arrangement also has the benefit o f being able to evolve f rom a test scale system into a larger system to treat a larger amount of drainage through the addition o f more cel ls. \bt\ \nt[flow]\ \h2.Flow\ F l o w through the cells o f a passive treatment system can be designed using three different types o f f low patterns: plug f low, step feed, and recirculation. P lug f low is used most often and requires minimum piping, energy, operation and maintenance (Sengupta, 1993). Water f lows through the system from the inlet to the outlet. Water in a step feed system can move into treatment cells through a number o f inlets along its path. Recirculat ion al lows water to be re-routed back through the system for further treatment. \i2\\jt[design_over].Back to Overv iew of Design Criteria.. A \jt[costs].Costs...\ \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[costs]\ \h2.Costs\ Passive treatments have a wide variety o f construction conditions and are built on a number o f different site locations. It is diff icult then to give an estimate of costs for the construction of these systems as these would vary greatly f rom one treatment system to another. A n overview of cost components and a comparison o f several implemented passive treatments to conventional treatment costs w i l l be reviewed. Cost components are usually typical from site to site and may be divided into Capital & Construction costs, and Operating Costs. \ i3 \ Capi tal & Construct ion Costs \dt[eng_test] .Engineering and Testing\ \dt[enviro].Environmental Basel ine Studies\ \dt[land].Land Equis i t ion\ \dt[righ_of].Right o f Way Access\ \dt[f inal].Final Engineering Design & Construction; \dt [construction]. Construction; Operating Costs ;dt[nonunal].Nominal Maintenance Tasks; ;dt[major].Major Overhauls; 133 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \jt[compare].Passive Treatment versus Conventional \i2\\jt[design_over].Back to Overview of Design Criteria.. A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[eng_test]\ Engineering and Testing - This includes all preliminary work for feasibility studies and treatment processes. Activities such as the testing of source water; substrate materials; field testing in "mini-cells" for substrate permeability, loading rates and flux; preliminary design; pilot scale testwork; and hydrologic investigations. \bt\ \nt[enviro]\ Environmental Baseline Studies & Permits - These are studies required to be done for siting and permitting. They include air & water quality, climate, geology, cultural, vegetation & wildlife, land use as well as socio-economic impacts. \bt\ \nt[land]\ Land Acquisition - Private land holdings and large area requirements can create significant costs. \bt\ \nt[right_of]\ Right of Way Access - This pertains to systems that may not be close to the source and water must be routed through pipes or open channels over land that may not be owned. \bt\ \nt[final]\ Final Engineering Design and Construction Specifications - Final details and construction specifications including text and drawing and contractor pay for building the project. \bt\ \nt[construction]\ Construction - Some costs may include mobilization of personnel, equipment and supplies, site preparation, earthwork, source control, linings, vegetation and etc. \bt\ \nt[nominal]\ Nominal Maintenance Tasks - Periodic inspections, sampling, cleaning, flow adjustments, and balancing will be required on a timely basis. \bt\ \nt[major]\ Major Overhauls - Removal, treatment, transportation and disposal of old substrate, preparation and installation of new substrate, and temporary rehandling of untreated effluents. \bt\ \nt[compare]\ \h2.A Look at Passive Treatment Costs Versus Conventional Chemical Treatment Costs\ 134 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Passive Treatment costs stand up wel l to costs o f chemical treatment. Even in cases where post conventional treatment was necessary passive treatment has lowered costs versus conventional treatment alone. A study performed by the Pennsylvania Department o f Environmental Resources on 73 constructed wetlands treating coal mining sites concluded that passive treatment is the Best Ava i lab le Technology ( B A T ) economically achievable for seeps with acidity up to 300 mg/L as C a C 0 3 (Hell ier, et a l . , 1994). Seep chemistry, f low volume and site characteristics can all adversely affect the cost o f a passive treatment system from one site to another. Cost benefit analysis of passive treatment at the Wheal Jane site is reported as marginal due to large volumes of water that needs treatment. (Cambridge, 1997) The Keister Wetland incurred capital costs o f $225,000 whi le chemical treatment would have cost an estimated $8,500 to $9,700 for caustic soda and ammonia respectively. Al though the chemical treatment may have worked out to be cheaper, the remote location o f the site, poor access roads, and intermitted drainage flows rendered chemical treatment inefficient (Faulkner and Skousen, 1994). Another site found the $110,000 construction costs for the passive treatment system was worth the investment versus the $22,000 per annum chemical treatment with caustic soda that could have been easily doubled once attendant labour and sludge handling costs are included (Faulkner and Skousen, 1994). Gusek compared passive treatment systems with l ime treatment and found that although capital costs for passive treatment may be similar to l ime treatment the operating costs are 7 1 % to 3 1 1 % higher (Gusek, 1995). \jt[costs].Back to Costs\\t\t\et[exit].Back to Knowledge Base\ \bt\ \nt[references]\ \ h 2 . R E F E R E N C E S \ Bechard, G . , and R . G . L . McCready , 1991. Microb io log ica l Process for the Treatment o f Ac i d i c Drainage at the Hal i fax International Airport. In: Second International Conference on the Abatement o f Ac i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Bender, J . Dr., and Dr. P. Phi l l ips, 1995. Biotreatment o f mine drainage. In: M i n i n g Environmental Management, pp. 25-27, Sep. 1995. Bennett, P . G . , C R . Ferguson, and T . H . Jeffers, 1991. B io log ica l Treatment o f A c i d M i n e Waters - Case Studies. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Broughton, L . M . , R .W. Chambers, and A . M a c G . Robertson, 1992. M i n e Rock Guidel ines Design Control o f A c i d Drainage Water Quality. Steffen, Robertson, and Kr is ten (B .C . ) Inc., Vancouver, B . C . Butcher, S.S. , et al . , Eds, 1992. Global Biogeochemical Cyc les. Academic Press, Toronto, Ontario. Cambridge, M . , 1997. Wheal Jane - The Long Term Treatment o f A c i d M i n e Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , V o l . 3. Dodds-Smith, M . E . , C . A . Payne, and J .J . Gusek, 1995. Reedbeds at Wheal Jane. In: M i n i n g Environmental Management, pp. 22-24, Sep. 1995. Dvorak, D . H . , R.S. Hedin, S.P. Mclnt i re , and H . M . Edenborn, 1995. Treatment o f Metal-Contaminated Water Us ing Bacterial Sulphate Reduction-Results from Pi lot-Scale Reactors. In: Second International Conference on the Abatement o f Ac id i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Eger, P., J .R. Wagner, and G . Melchert, 1997. The Use o f Peat/Limestone System to Treat A c i d Rock Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , V o l . 3. Evere l l , M . D . , and A J . Ol iver, 1992. Environmental Technologies for the Minera ls and Metals Sector - A Canadian Perspective. In: Minerals, Metals and the Environment. E lsev ier Science Publishers L td . , N e w York . 135 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Faulkner, B . B . , and J .G . Skousen, 1994. Treatment o f A c i d M i n e Drainage by Passive Treatment Systems. In: International M ine Reclamation and M i n e Drainage Conference and Thi rd International Conference on the Abatement o f Ac id i c Drainage Proceedings, V o l . 2, A p r i l 24-29, 1994. Faulkner, B . B . , and J .G . Skousen, 1995. Treatment o f A c i d M i n e Drainage by Passive Treatment Systems. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Universi ty and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia . 1995. Gould , W . D . , G . Bechard, and L. Lort ie, 1994. The Nature and Ro le o f Microorganisms in the Tai l ings Environment. In: Short Course Handbook on Environmental Geochemistry of Sulf ide Mine-Wastes, Mineralogical Associat ion of Canada, Waterloo, Ontario, Chap. 7, M a y 1994. Gusek J .J . , 1995. Passive-Treatment o f A c i d Rock Drainage: What is the Bottom L ine? In: M in ing Engineering, pp. 250-253, March 1995. Hendricks, A . C . , 1991. The Use o f Ar t i f i c ia l Wet land to Treat A c i d M ine Drainage. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Herl ihy, A . T . , A . L . M i l l s , G . M . Hornberger, and A . E . Bruckner, 1987. The Importance o f Sediment Sulfate Reduction to the Sulfate Budget o f an Impoundment Receiv ing A c i d M ine Drainage. In: Water Resources Research, V o l . 23, N o . 2, pp. 287-292, Feb. 1987. Ka l i n , M . , 1995. B io log ica l Amel iorat ion o f A c i d i c Seepage Streams. In: Second International Conference on the Abatement of Ac i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Ka l i n , M . , and M . P . Smith, 1997. M ic rob ia l A c i d Reduct ion in Sediments - Concepts and Appl icat ion. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , V o l . 3. Ka l i n , M . , and R.O. V a n Everdinger, 1987. Eco log ica l Engineering: B io log ica l and Geochemical Aspects, Phase I Experiment. In: A c i d M i n e Drainage Seminar / Workshop Proceedings, Hal i fax, N o v a Scotia. Ka l i n , M . , M . P . Smith, and G . Landry, 1994. The Appl icat ion o f Eco log ica l Engineering to A c i d Coa l Seepages in Eastern Canada. In: International M i n e Reclamat ion and M i n e Drainage Conference and Thi rd International Conference on the Abatement o f A c i d i c Drainage Proceedings, V o l . 2, Ap r i l 24-29, 1994. Karathanasis, A . D . , and Y . L . Thompson, 1995. Me ta l Speciation and Retention Patterns in a H igh Meta l Load A c i d M ine Constructed Wetland o f Southeastern Kentucky. In: Second International Conference on the Abatement of Ac id i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Kuyucak, N . , P. St-Germain, and K . G . Wheeland, 1995. In Situ Bacterial Treatment o f A M D in Open Pits. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Lawrence, R .W. Dr. , and Dr . A . M a c G . Robertson, 1994. A Short Course on A c i d Rock Drainage Understanding the Problems - Finding Solutions. The C I M Distr ict 6 Annua l Meet ing, Vancouver, B . C . , Oct. 11, 1994. Lawrence, R .W. , 1997. A Course on Fundamental and App l i ed Aspects o f A c i d Rock Drainage. Course notes, Department o f M in ing and Minera l Process Engineering, U B C , Vancouver, Canada. Pesavento, B . G . , 1987. Factors to be Considered in the Use o f Wetlands to Treat A c i d M i n e Drainage. In: A c i d M ine Drainage Seminar / Workshop Proceedings, Hal i fax, N o v a Scotia. 136 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Plane and Associates L td . , 1990. Assessment o f Exist ing Natural Wetlands Affected by low p H , Metal Contaminated Seepages. F ina l Report, M E N D project 3.12.1. Richardson, J .S . , and C . J . Perr in, 1990. The Effect o f A c i d M i n e Drainage on Stream Macroinvertebrates and Periphytic A lgae. B C A M D project 4.30 .C .4 . Ri tchie, A . I . M . , 1994. The Waste Rock Environment. In: Short Course Handbook on Environmental Geochemistry of Sulf ide Mine-Wastes, Minera logica l Associat ion o f Canada, Waterloo, Ontario, Chap. 5, M a y 1994. Robb, G . , and J . Robinson, 1995. A c i d mine drainage prediction and remediation. In: M in ing Environmental Management, pp. 19-21, Sep. 1995. S E N E S Consultants L td . , 1994. A c i d M i n e Drainage - Status of Chemical Treatment and Sludge Management Practices. M E N D project 3.32.1. Sengupta, M . , 1993. Environmental Impacts of M in ing Moni tor ing, Restoration, and Control. Lewis Publishers, U S A . Sextone, A . J . , J .C . Sencindiver, J .P. Calabrese, D .K . Bhumbia , and G . K . Bissonnette, 1991. Appl icat ion of Constructed Wetlands for the Removal o f Iron from A c i d M i n e Drainage. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Shelp, G . , G . Southam, G . Spiers, L . L i u , and W.Chesworth, 1994. A n Evaluation of a Peat-Wood Chip-Microf lora Admixture to act as an Amel iorant for A c i d M ine Drainage. In: International M ine Reclamation and M ine Drainage Conference and Th i rd International Conference on the Abatement o f A c i d i c Drainage Proceedings, V o l . 2, Ap r i l 24-29, 1994. Skousen, J . , 1991. A n o x i c Limestone Drains for A c i d M i n e Drainage Treatment. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Universi ty and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia . 1995. Skousen, J . , A . Sextone, K . Garbutt, and J . Sencindiver, 1992. Wetlands for Treating A c i d M i n e Drainage. In: A c i d M ine Drainage Contro l and Treatment, West V i rg in ia Universi ty and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Skousen, J . , B . Faulkner, and P. Sterner, 1995. Passive Treatment Systems and Improvements o f Water Quality. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia Universi ty and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Skousen, J . G . , 1995. Douglas Abandoned M i n e Land Project: Descr ipt ion o f an Innovative A c i d M ine Drainage Treatment System. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia University and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Stark, L .R . , W . R . Wener ick, P .J . Wuest, and F . M . Wi l l iams, 1995. Add ing a Carbon Supplement to Simulated Treatment Wetlands Improves M i n e Water Quality. In: Second International Conference on the Abatement of A c i d i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Taddeo, F .J . , and R . K . Weider, 1991. The Accumulat ion o f Iron Sulphides in Wetlands Constructed for A c i d Coa l M i n e Drainage ( A M D ) Treatment. In: Second International Conference on the Abatement of Ac id i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. 137 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Wheeler, W . N . , and M . Ka l i n , 1995. The Eco log ica l Response o f a B o g to Coa l A c i d M ine Drainage. Deterioration and Subsequent Initiation of Recovery. In: Second International Conference on the Abatement o f Ac i d i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Wieder, R .K . , M . N . L inton, and S.T. Starr, 1991. Spatial and Temporal Patterns in Surface and Subsurface Water Chemistry in Wetlands Constructed for A c i d Coa l M i n e Drainage. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Wi ldeman, T., G . Brod ie , and J . Gusek, 1993. Wetland Design for M in ing Operations. B iTech Publ ishing C o . , Vancouver, B . C . Z iemkiewicz , P., J . Skousen, and R. Lovett, 1994. Open Limestone Channels for Treating A c i d M i n e Drainage: A N e w Look at an O l d Idea. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Universi ty and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia . 1995. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T Sulphide Reduction (Sulphhy.doc) \bt\ \nt[coverpage]\ \h2.Removal & Encapsulation of Sulphides\ \t This module has been created to facilitate the user in deterrnining the most appropriate option for removal and encapsulation o f sulphides. A R D X considers both these options in finding a possible solution to your A R D challenges. A t any point within the expert system the user has access to explanations and justifications. For the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definit ion o f a topic or word Y e l l o w text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Exit to ARDX\\c2\\ j t [contents] .Module Contents\ \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard].Introduction to A c i d Rock Drainage\ \c2\\jt[removal].Removal o f Sulphides\ \c2\\jt[blending].Segregation and/or Blending o f Sulf ide Wastes\ 138 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. It is o f significant concern and an increasing environmental challenge for mining industry today. The exposure of mining wastes to weathering conditions increase the solubil ity o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D wi l l impact watershed characteristics and create adverse effects in the ecosystem. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[removal]\ \h2.Removal o f Sulphides\ \t This method is usually not very effective or cost efficient. It may be a consideration for a new mine during the design stage. Removal of sulphides f rom waste is achieved through floatation. The process involves crushing the waste to liberate the sulphides prior to floatation. \c2\ Tai l ings - sulphide content is much less then the original waste \c2\ Sulphide concentrate - needs special disposal for A R D control \cO\(Broughton, 1992) \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[blending]\ \h2.Segregation and/or Blending of Sulphide Wastes\ \t Segregation, blending or selective handling o f waste works by balancing the production and consumption o f acidity. The goal is to achieve: \c2\ - predominantly alkaline leachate throughout the dump \c2\ - internal consumption o f acidity Ac2\ - internal precipitation o f dissolved metals \t Success o f this method depends on: \c2\-water f lows \c2\-degree of mix ing o f different rock types \c2\-quantity o f each rock type \c2\-reactivity o f different materials (both acid consuming and acid generating) \t Layer ing o f wastesis an incomplete form of blending. It is based on acid base accounting and can be achieved through: 139 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \ c l \ an acid consuming cap layer \c5\ an acid consuming base layer \ c l \ an oxygen consuming cap layer \c5\ multiple layers \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[references]\ \h2.References\ Broughton, L . M . , R .W. Chambers, and A . M a c G . Robertson, 1992. M i n e Rock Guidel ines Design Control o f A c i d Drainage Water Quality. Steffen, Robertson, and Kristen (B .C . ) Inc., Vancouver, B . C . , pp.6-4, 6-11. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[exit]\ E X I T Water Treatment (Treathy.doc) \bt\ \nt[coverpage]\ \ h 2 . W A T E R T R E A T M E N T \ \t This module has been created to facilitate the user in detennining the most appropriate water cover or f looding option through the attached expert system. A t any point within the expert system the user has access to explanations and justif ications. For the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definition of a topic or word Y e l l o w text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Exit to ARDX\\c2\\ j t [contents].Module Contents\ \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard].Introduction to A c i d Rock Drainage\ \c2\\jt[treat].Introduction to Water Treatment \c2\\jt[l ime].Lime\ \c2\\jt[limestone].Limestone\ 140 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \c2\\jt[causticsoda].Caustic Soda\ \c2\\jt[sodaash].Soda Ash \ \c2\\jt[ammonia].Ammonia\ \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. It is o f significant concern and an increasing environmental challenge for ni ining industry today. The exposure of mining wastes to weathering conditions increase the solubil i ty o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D w i l l impact watershed characteristics and create adverse effects in the ecosystem. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[treat]\ \h2.Introduction to Water Treatment \t Col lect ion and treatment of effluent is required in al l cases where water discharge l imits are not met by other means. Water treatment is a continuous process o f adding energy, manpower and alkaline chemicals to decrease acidty, precipitate out metals and increase the p H of the effluent. The objective for treatment is "to produce an effluent that can be recycled to the process or released into the environment". (Lawrence, 1997) \t The selection of a treatment system and chemical necessary for treatment is based on the flowrate, sulphate concentration and metal concentrations (in particular that o f ferrous iron). \t Some of the more commonly used chemicals are: \c2\\jt[lime].Lime\ \c2\\jt[limestone].Limestone\ \c2\\jt[causticsoda].Caustic Soda\ \t Other chemicals that may be used are: \c2\\jt[sodaash].Soda A s h \ \c2\\jt[ammonia].Ammonia\ \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A 141 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[lime]\ \h2.Lime\ \t Quick l ime (calcium oxide) and hydrated l ime (calcium hydroxide)are commonly used for acid neutralization in conventional treatment. Quick l ime is converted to hydrated lime by a slaking process prior to use. (Broughton, 1992) \t Benefits and conditions o f its use are as fol lows: \c3\- less expensive then sodium \c3\- good for large treatment systems \c3\- aeration and mix ing are required (for oxidatioon o f ferrous to ferric iron) \c3\- can increase the p H to 10 and above \c3\- can treat effluent containing M n and ferrous iron \c3\- a high sulphate concentrat ion above 2000mg/l) can cause reactions forming anhydrite and insoluble gypsum. \c3\- requires careful handling \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[limestone]\ \h2.Limestone\ «• \t Limestone (calcium carbonate) is used in conventional treatment of A R D . It comes in lump, crushed or ground form. \t Benefits and conditions o f its use are as fol lows: \c3\- costs the least \c3\- good for large treatment systems \c3\- aeration and mix ing are usually required \c3\- higher C a limestone is more reactive then Dolomi t ic limestone \c3\- most useful for p H ranges 4 to 6 \c3\- the sludge produced is more granular and dense \c3\- a high sulphate concentrat ion above 2000mg/l) can cause reactions forming anhydrite and insoluble gypsum. \c3\- good for effluent where ferric iron is main problem \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[causticsoda]\ \h2. Caustic Soda\ \t Caustic Soda (sodium hydroxide) is usually uneconomical for use in large treatment systems and is considered only in smaller systems. \t Benefits and conditions o f its use are as fol lows: \c3\- can be used in smaller treatment systems \c3\- reagent costs are higher then for l ime \c3\- needs careful handling \c3\- aeration and mix ing are usually not required \c3\- good for high sulphate (greater then 2,500mg/l) f lows \c3\- can treat effluent containing M n and ferrous iron \c3\- good for increasing p H to above 10 142 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[sodaash]\ \h2.Soda A s h \ \t Soda ash (sodium carbonate) is not widely used but has been used in some coal mines in the U.S. \t Benefits l imitations and conditions o f its use are as fol lows: \c3 \ - can be used in smaller treatment systems \c3\ - reagent costs are higher then for l ime \c3\ - o.k. for f lows with a low ferrous iron concentration \c3\ - can raise the p H up to 8.5 \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[ammonia]\ \h2.Ammonia\ \t A m m o n i a is not often used due to the environmental concerns that accompany its use. \t Benefits and limitations o f its use are as fol lows: \c2\- good for increasing the p H to 9 or 10 \c2\- high reactivity \c2\- questionable due to environmental concerns: \c3\ - un- ionized ammonia can be present in neutral p H water \c3\ - effluent can be toxic to downstream fish and aquatic l i fe \c3\ - \dt[nitrification].nitrification\ can increase nitrate concentrations and acidity in the downstream environment \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[nitrification]\ \t The b io logical conversion of ammonia to nitrate by bacteria. \bt\ \nt[references]\ \h2.References\ Broughton, L . M . , R . W . Chambers, and A . M a c G . Robertson, 1992. M ine Rock Guidelines Design Contro l o f A c i d Drainage Water Quali ty. Steffen, Robertson, and Kr isten (B.C. ) Inc., Vancouver, B . C . Lawrence, R . W . , 1997. A Course on Fundamental and App l ied Aspects o f A c i d Rock Drainage. Course notes, Department o f M i n i n g and Minera l Process Engineering, U B C , Vancouver, Canada. Skousen, J . , 1988. Treatment of A c i d M i n e Drainage. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Univers i ty and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia . 1995. 143 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Skousen, J., R. L i l l y , and T. Hi l ton, 1993. Special Chemicals for Treating A c i d M i n e Drainage. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia University and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia . 1995. Skousen, J . , et. A L , 1990. A c i d M ine Drainage Treatment Systems: Chemicals and costs. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia University and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[exit]\ E X I T Water Covers (Waterhy.doc) \bt\ \nt[coverpage]\ \ h 2 . S U B - A Q U E O U S D I S P O S A L \ \t This module has been created to facilitate the user in determining the most appropriate water cover or f looding option through the attached expert system. A t any point within the expert system the user has access to explanations and justifications. For the student, the module has access to hypertext information that can be used as a study or reference guide. Some useful features include: White Text - provides a short definition o f a topic or word Y e l l o w text - jumps you to a different topic or displays equations providing direct access to further information. \c6\\et[exit].Exit to ARDX\\c2\\ j t [contents].Module Contents\ \bt\ \nt[contents]\ \ h 2 . M O D U L E C O N T E N T S \ \c2\\jt[ard].Introduction to A c i d Rock Drainage\ \c2\\jt[water].Introduction to Sub-Aqueous D i s p o s a l \c2\\jt[floodt] .Flooding o f Tai l ings\ \c2\\jt[wpflood].Flooding o f Waste in an Open PitA \c2\\jt[mine].Flooding o f M i n e Workings\ \c2\\jt[submarine]. Sub-Marine D i s p o s a l 144 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ \bt\ \nt[ard]\ \h2.Introduction to A c i d Rock Drainage\ \t A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering of pyrite and other sulfide minerals. It is of significant concern and an increasing environmental challenge for mining industry today. The exposure o f mining wastes to weathering conditions increase the solubil i ty o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D wi l l impact watershed characteristics and create adverse effects in the ecosystem. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[water]\ \h2.Introduction to Sub-Aqueous D i s p o s a l \t Sub-aqueous disposal (or water coverage) is one of the most promissing methods for controling A R D . A water cover l imits the amount of oxygen that comes in contact with the waste due to the low solubil i ty o f oxygen in water and its low aqueous diffusivity. The solubility of oxygen in water is normally below 1 l m g / L , decreasing with increased temperature. A t IOC, the level is 5.8 mg/L. This low solubi l i ty leads to a low transport o f oxygen to the waste. The majority of this dissolution takes place in the upper meter o f the water cover (epi-l imnion) and so without significant mixing, high 0 2 levels at depth are not possible. Di f fus ion o f oxygen is four times lower in water then in air (Broughton, 1992). Wi th the absence o f oxygen, the rate o f oxidation of waste material is reduced to a level such that A R D essentilly does not occur. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[wpflood]\ \h2.Flooding o f Waste in an Open PitA \t F looding o f an open pit containing waste rock requires the re-handling and relocating o f waste from its established location to an open pit that can be flooded to prevent exposure to oxygen. The nature and quantity o f the awste and the quality o f the pit wal l rock must be considered. The pit must be able to adaquately hold the volume of waste rock to be disposed of. The swell factor during mining may result in a volume o f rock that cannot be retained by the pit. Waste rock or pit wal l rock that is already oxidized and currently producing A R D may not be appropriate for f looding. The cost of transporting the waste material to the pit can be high making the distance from the waste rock pile to the pit an important consideration. In most cases however, the waste rock is usually located in good proximity to the pit. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[submarine]\ \h2.Sub-Marine D i s p o s a l \t Disposal in an existing lake or ocean. 145 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \t Effective control of lake disposal requires an understanding Of: \c3\seasonal fluctuations in water levels - Water level fluctuations must insure a continuous min imum water cover of one meter in order to l imit oxygen diffusion. \c3\wave action - Wave action can introduce oxygen into the water facilitating oxidation of the underlying wastes. \c3\convection/seasonal lake overturn - Overturn can introduce oxygen to the wastes through a shift in the oxygenated top water layer to the bottom o f the lake. \t Disposing o f tailings or waste rock in an existing lake or ocean is also associated with many environmental issues including: \c2\- the toxicity of reagents and heavy metals f rom the m i l l process \c2\- eccessive nutrient additions from the use o f explosives \c2\- increased turbidity due to suspended solids \c2\- direct impact of placing wastes on the present habitat within the lake or ocean \c2\- public and government acceptance o f sub-marine disposal \t The proximity of the lake or ocean is another important consideration in sub-marine disposal. Costs may increase greatly due to the haulage distance to the disposal site. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[floodt]\ \h2.Flooding o f Tail ings\ \t F looding o f tailings can provide the el imination or near el imination of acid generation. It also eliminates the release o f air born particles and dusting that would be present with exposed tailings. \t F looding within a man made enclosure requires consideration of: \c2\stability o f the tailings enclosure - R i sk o f structural failure must be very remote to ensure long term durability and prevention of damage due to release o f tail ings into the environment. \c2\rock r immed basins - Rock r immed tail ings basins provide for low seepage rates o f contaminated effluent into the ground water. \c2\contaminant release levels - The release o f contaminants in case o f a f looding event should be within accepted loading. \c2\climate and water balance - Site hydrology and loca l cl imatic conditions need to support the maintainance of a one meter water cover. \c2\present oxidation - Present oxidation w i l l introduce contaminants and acidity into the water making f looding unsuitable. L im ing before f looding should be a consideration. \jt[contents].Back to Contents..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[mine]\ \h2.Flooding o f M i n e Workings\ \t F looding o f mine workings involves the sealing o f openings to the mine workings fol lowed by f looding to prevent the further introduction of oxygen to exposed rock. F lood ing is often a good long-term closure option. \t Success o f this method w i l l depend on: \c3\- The strength and stability o f plugs and sealants (risk o f failure). \c3\- Site hydrology including seasonal f luxes and discharge control. \c3\ - Extent o f dri l l ing in the area. \c3\ - wal l rock quality and the presence o f paste or other backf i l l material. \c3\- The completeness of the f looded workings. 146 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[references]\ \h2.References\ Lawrence, R .W. , 1997. A Course on Fundamental and App l ied Aspects o f A c i d Rock Drainage. Course notes, Department o f M i n i n g and Minera l Process Engineering, U B C , Vancouver, Canada. Ludgate, L R . Et .a l . , 1997. Decommisioning o f Denison Mines Tai l ings Management Areas. In: Fourth International Conference on A c i d M i n e Drainage Proceedings, Vancouver, B . C . , pp.917-933, V o l . 2. \jt[contents].Back to Contents...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T Help Document (Hyperard.doc) \bt\ \nt[contents]\ \h2.Document Contents\ B y cl icking on a topic you can have direct access to further information in that area. The Forward-Browse button w i l l a l low access to the topics page by page. \c2\\jt[overview].Overview o f Documen t \c2\\jt[introduction].Introduction to Passive Treatments\ \c2\\ j t [ARDchetnistry].Acid Rock Drainage Chemistry\ \c2\\jt[passive].Passive Treatment Systems\ \c2\\jt[alkalinity] .Alkal in i ty Generation\ \c2\\jt[bacteria].The Ro le o f Bacter ia in A lka l in i ty Generation\ \c2\\jt[microbes].The Ro le o f Microorganisms in Meta l Removal \ \c2\\jt[algae].The Ro le o f A lgae in Meta l R e m o v a l \c2\\jt[plants].Reactions W i t h Plants\ \c2\\jt[organic].Adsorption to Organic Substrate\ \c2\\ j t [ALD] .Anox ic Limestone Drains\ \c2\\jt[aerobic].Aerobic Cel ls \ \c2\\jt[anaerobic] .Anaerobic Cel ls \ \c2\\jt[ald_con].Construction o f an A n o x i c Limestone Drain\ \c2\\jt[aerobic_con].Construction o f an Aerob ic C e l l \ \c2\\jt[anaerobic_con].Construction o f an anaerobic Ce l l \ \c2\\jt[param].Design Parameters\ \c2\\jt[site_con]. Site Considerations\ \c2\\ j t [cel l_design].Cel lDesign\ \c2\\jt[flow].Flow Pattern\ \c2\\jt[costs].Costs\ \c2\\jt[references].References\ \c6\\et[exit].Back to Knowledge Base\ 147 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[overview]\ \h2.Manual on Passive Treatments for A c i d Rock Drainage\ \ i l \ T h i s manual has been prepared to" assist the user in understanding Passive Treatment processes for the remediation o f A c i d Rock Drainage sites. \ i l \ Some useful features include: White Text - provides a short definit ion o f a topic or word Y e l l o w text - jumps you to a different topic or displays equations for further clarification \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[ard]\ A c i d Rock Drainage ( A R D ) is contaminated acidic drainage due to the spontaneous weathering o f pyrite and other sulfide minerals. The weathering conditions increase the solubil ity o f heavy metals, radionuclides, sulfate, and acidity, and reduce p H . A R D wi l l impact watershed characteristics and create adverse effects in the ecosystem. \bt\ \nt[weather]\ Weathering is the breakdown of rocks and minerals at the Earth's surface by the action of physical and chemical processes. The process can oxidize pyrite, breaking it down and releasing iron, sulfide, and other minerals that may be locked into the matrix of the rock. \bt\ \nt[introduction]\ \h2.Min ing A n d A c i d Rock Drainage\ \dt[ard].Acid Rock Drainage\ ( A R D ) occurs when pyrite and other sulphide minerals are exposed to \dt[weather].weathering\ from the mining o f mineral resources. The problem is ongoing in coal as wel l as metal mines. Once exposed, waste rock generating acidic and metal contaminated drainage may continue to do so for centuries. Furthermore, A R D may take time to generate and the problem may not be recognized for may years. Chemical treatment with l ime is the accepted form o f dealing with A R D . Treated effluent can be released safely into the environment or recycled to the process. A s A R D continues to be generated in an active or abandoned mine, chemical treatment w i l l often need to continue for hundreds o f years. This may become a demanding and costly long term commitment. M i n e sites may be located in remote locations, fluctuations in f low rates, large volumes o f l ow intensity sludge, and the unknown stability o f the chemical sludge are al l unwanted factors affecting the long term treatment o f A R D using chemical methods. Thus the mining industry has been faced with the ongoing challenge o f finding an economic and sustainable solution to the A R D problem. \jt[passive].Passive Treatment systems may be a step toward finding this solution. For further information on the chemistry o f A R D cl ick \ j t [ARDchemist ry ] .HERE\. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt [ARDchemistry] \ \h2.Chemistry o f A c i d Rock Drainage\ Pyrite is very abundant in virtually al l sulphide ore bodies and is often the main culprit contributing to the creation o f A R D . The reactions taking place to cause the highly acidic metal contaminated drainage can be looked at through 148 A Fuzzy Expert System on Acid Rock Drainage Site Remediation the oxidation o f pyrite. A s exposed pyrite comes into contact with oxygen and water it oxidizes to produce ferrous iron, sulphate and hydrogen ions. \ j t [ chem l ] .REACTION\ The reaction produces ferrous iron and sulphate increasing the total dissolved solids and H+ ions thereby resulting a decrease in p H . Trace metals such as copper, lead, and zinc that were bound up within the pyrite are released as the pyrite is oxidized. The ferrous ions can further oxidize to the ferric state. \ j t [ chem2] .REACTION\ This normally slower reaction is facilitated by the presence of certain bacteria the most commonly known as Thiobaci l lus Ferrooxidans. Other bacteria include Thiobacil lus Thiooxidans and Sulfolobus. A l l o f these bacteria use energy from the above reaction for their l ife processes. The presence o f the ferric iron can now oxidize pyrite further in the fol lowing fast reaction. \ j t [ chem3] .REACTION\ A vicious cycle of reactions is created constantly generating more acidity (for every mole o f pyrite, 4 moles of acidity are produced), decreasing p H , and creating optimal conditions for Thiobaci l lus bacteria to thrive and drive the reactions further. A s contaminated water continues down the f low path, further reactions complicate the issue leaching metals into solution, neutralizing acidity by calcium carbonate and hydrolyzing iron. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base...\ \bt\ \nt[cheml]\ M i l . Pyrite in the presence o f oxygen and water is oxidizedA \ i2 \ Mip.cheml.bmp\ Mi l .The reaction produces ferrous iron, sulphate, and hydrogen ionsA Mil .Total dissolved solids increase while p H decreasesA M4\ \ j t [ARDchemistry].Back to T E X T . . A M0\ \bt\ Mit[chem2]\ Mil.Ferrous iron oxidizes further assisted by Thiobaci l lus Ferrooxidans\ M2\ \np.chem2.bmp\ M4\ \ j t [ARDchemistry].Back to T E X T . . A M0\ \bt\ \nt[chem3]\ Ferr ic iron further oxidizes pyriteA M2\ Mip.chem3.bmp\ \ i4\ \ j t [ARDchemistry].Back to T E X T . . A \ i0\ \bt\ Mit[passive]\ Mi2.Passive Treatment Systems\ 149 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Passive treatment systems are self sustaining systems that can be attractive alternatives to conventional l ime treatment. Treatment of A c i d Rock Drainage includes the generation o f \jt[alkalinity].alkalinity\ and elimination o f contaminants. Many processes such as \jt[microbes].bacteria\, \jt[algae].algae\, \jt[plants].plants\, and \jt[organic].organic substrate\ are at work to facilitate the removal o f contaminants. Passive systems include: \ i3 \ \ j t [ALD] .Anox ic Limstone Drains\ (ALD 's ) \jt[aerobic].Aerobic\ cells \jt[anaerobic].Anaerobic\ cells \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[alkalinity]\ \h2.Alkal ini ty Generation\ Passive treatment w i l l generate alkalinity through a number o f processes o f wh ich two o f the most dominant are limestone and \jt[bacteria].sulfate reductionV The limestone layer in an anoxic limestone drain or as a bed in an anaerobic cel l can react wi th hydrogen ions to decrease the acidity of the water. \ j t [ a l k l ] .REACTION\ This process occurs only in anaerobic conditions. Once exposed to an oxid iz ing environment the limestone effectiveness is inhibited and the effective buffering capacity decreases dramatically do to the formation o f an iron hydroxide coating on the limestone surface. The dissolution o f the limestone cannot occur, thus hindering the alkalinity generation capability o f the cel l . \ j t [ALD] .ALD 's \ are one treatment system that uses this process. \jt[contents].Back to Index...\\t\t\et[exif].Back to Knowledge Base..A \bt\ \nt[alkl ] \ \i2\ \np.alk l .bmp\ \i4\ \jt[alkalinity].Back to T E X T . . A \ i0\ \bt\ \nt[bacteria]\ \h2.The Ro le o f bacteria in passive treatment systems\ Bacteria play a role in treating acid rock drainage by assisting in both alkalinity generation and the removal o f \jt[microbes].metals\. Sulphate reduction by bacteria contributes to the generation o f alkalinity. The process occurs due to Sulfate Reducing Bacteria (SRB) that use simple organic compounds as an energy source. These compounds are usually products o f fermentation as the SRB 's must rely on other microbes for the bacterial degradation o f proteins, carbonates, and other dead biomass. \ j t [ bac t l ] .REACTION\ The system w i l l generate permanent alkalinity as H 2 S gas is released into the atmosphere through the generation o f H S - ions. 150 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[bactl]\ \h2.Reduction o f sulphate with lactate is shown hereA \ i2\ \np.bactl .bmp\ \ i4\ \jt[bacteria].Back to T E X T . . A \ i0\ \bt\ \nt[microbes]\ \h2.The Ro le o f Microorganisms in Meta l Removal \ Numerous different bacteria assist in metal removal. Sulfate reducing bacteria are instrumental in removing metals by producing hydrogen sulfide causing metal precipitation in an anaerobic environment. These bateria are also instrumental in \jt[bacteria].alkalinity\ generation. Other bacteria also play a role in metal removal. Iron can be oxid ized wi th the help o f a family o f bacteria known as the Ferrooxidans. These bacteria can survive an environment wi th a p H o f less than three whi le feeding on the inorganic nutrients of Fe(U), C 0 2 , NH4+ . It has been found through experimentation that Ferrooxidans have the highest populations in the aerobic layer o f substrate in a wetland. \ t \ t \ j t [micr l_2] .REACTIONS\ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[micrl_2]\ Bacteria derive energy by converting ferrous iron to ferric through the fo l lowing reaction... \ i2\ \np.micr l .bmp\ \ i0\ The ferric iron now undergoes a hydrolysis reaction... \ i2\ \np.micr2.bmp\ \ i0\ The reaction produces ferric hydroxide and the iron precipitates out o f the system. \ i2\ \jt[microbes].Back to T E X T . . A \ i0\ \bt\ \nt[algae]\ \h2.The Ro le o f Algae in Meta l Removal \ Few studies have been done on the role o f algae in an A R D environment. Their effects usually go unnoticed, introduced into passive treatment systems unintentionally (with the introduction o f other plants) through secondary sources. A lgae however, can accumulate metals wi th in their body structure. They can act as b io logical filters trapping sediments and nutrients high in metals. Intracellular crystals have been observed in suspended algal cells under a microscope. Iron and manganese are used by algae as essential micronutrients, and algae have been found to accumulate manganese up to 56 g/kg of plant tissue. 151 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Algae are capable o f reproducing rapidly and may also benefit to increase organic matter in a cel l as dead biomass. Microb ia l mats made up o f blue-green algae and bacteria have been developed and used in treatment o f acid coal and metal mine drainage. B ioma tTM 's have been shown to remove M n at a rate o f 6.5g/m2/d with inf low levels of 7.5mg/L M n and a f low rate o f 16L/min. B i o m a t T M promotes high p H levels and can incorporate metals into organic matter through cation exchange, adsorption, precipitation & co-precipitation, and complexation or chelation. \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[plants]\ \h2.Reactions W i t h Plants\ Passive treatments rely on plants to perform a number o f functions to ameliorate A R D . During past years it was thought that that metal uptake by plants was a significant and important process. Recently, however, studies have shown that plants actually accumulate very little metals accounting for less than 1% annual iron removal by Typha in one case. Although plants may not be directly responsible for significant metal uptake, they aid in the process of metal removal by assisting in other ways. The stems and leaves o f plants increase the surface area available for the attachment of microbes. In this way the plants presence impacts the population o f microbes that are necessary within the system. On an annual basis new exchange sites are made available through the death o f o ld biomass as new biomass is produced. The stem and root systems o f the plant play an important role in the transportation of atmospheric gasses. Oxygen is introduced into the system through the roots and creates an aerobic region. Oxygen surrounds each root hair supporting further microbial populations. Plant growth itself becomes instrumental in many passive systems. Photosynthesis is responsible for carbon f ixing and the accumulation o f new biomass. O l d biomass is al lowed to die and decompose providing a constant source o f carbon. M ic rob ia l processes are sustained as the organic carbon is ut i l ized as an energy source of electrons. \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[organic]\ \h2.Adsorption to Organic Substrate\ Metals can be accumulated in a wetland through an exchange process with humic and f luvic acids in soi l . Organic soils (peat) have a cation exchange capacity ( C E C ) that increases at lower p H values. The C E C allows peat to absorb posit ively charged ions. \ j t [ o r g l ] . R E A C T I O N The carboxylic group ( - C O O H ) attached to the inert organic portion o f the substrate ( R ) w i l l dissociate allowing the carboxyl ion ( C O O - ) to react wi th the metal ion ( M ) in the solution. The C E C o f the substrate w i l l go down as metal ions react with the exchange sites and the organic substrate becomes saturated with metals. N e w exchange sites w i l l only become available on an annual basis through the introduction o f new organic matter as the wetland plants grow and die. This complexation o f metal ions can work against the intended removal o f metals from A R D waters. A soi l wi th a high concentration o f metals that comes into contact with acidic water with lower concentrations can cause the reaction to reverse, releasing the metals back into solution. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base... \ \bt\ \nt[orgl]\ 152 A Fuzzy Expert System on Acid Rock Drainage Site Remediation The metal ions are exchanged with other ions according to the fol lowing reaction.A \ i2\ \np.orgl .bmp\ \ i4\ \jt[organic].Back to T E X T . . A \ i0\ \bt\ \ n t [ALD] \ \h2.Anoxic Limestone Drains\ A n o x i c Limestone Drains use limestone to raise p H and add alkalinity to A R D waters. A L D ' s have been shown to generate up to 300mg/L o f alkalinity when installed correctly and water quality parameters comply (Skousen, 1991). The Drains must be constructed such that anoxic conditions prevail and therefore are buried in underground channels or trenches such that oxygen is not available. \dp[ald].ald.bmp\ \ i4 \ \ j t [a ld_con] .DESIGN & C O N S T R U C T I O N o f an Anox i c Limeston Drain... \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[aerobic]\ \h2.Aerobic Ce l l \ Aerob ic cells are built to encourage the oxidation processes. They are used to collect water and remove metals as oxides by providing residence time within the cel l . In A R D waters iron is primari ly removed within the aerobic cel l . Other metals removed here are A l , M g , Arsenic , cyanide and mercury. The shal low construction of the cel l increases the surface area exposed to oxygen. Water treatment occurs on the surface sometimes requiring large areas for treatment depending on the input water quality and the demands o f the output quality. The shal low nature o f the aerobic cells can result in unwanted freezing during the winter months. \dp[aerobic].aerobic.bmp\ T o further increase the aerobic conditions, vegetation (reed beds) or green algae (release oxygen during photosynthesis) are added. Oxygen is introduced through the root structure o f the vegetation and facilitates metal precipitation. The vegetation also serves as a filter for suspended materials. The roots can hold together the substrate. This prevents the formation o f channels and serves to increase the residence time. Furthermore, vegetation has the benefit o f added aesthetic value for the treatment area. \ j t [aerobic_con].DESIGN & C O N S T R U C T I O N o f an aerobic cell...\t\jt[contents].Back to Index.. A\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[anaerobic]\ \h2.Anaerobic CeliA Anaerobic cells work pr imari ly to generate alkalinity by promoting reducing conditions. Treatment and thus water f low occurs in the subsurface o f the cel l . Since the anaerobic cells are deeper then the aerobic, unlike the aerobic they are capable o f functioning in sub-freezing climates. Sulfate reducing bacteria (SRB's ) such as desulfouibrio and desulfotomaculum play a large rol l by uti l iz ing organic matter as a carbon source. S R B ' s can generate up to 200 mg/L of sulfides. B y using sulfate as an electron acceptor for l i fe processes, S R B ' s convert sulfate into hydrogen sulfide. \dp[anaerobic].anaero.bmp\ \ j t [anaerobic_con].DESIGN & C O N S T R U C T I O N o f an anaerobic cel l . \ i0\ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A 153 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[anaerobic_con]\ \h2.Design and Construction o f an Anaerobic Ce l l \ Anaerobic cells are designed for sub-surface flows. These are cells that encourage the growth of S R B ' s and thus need the addition o f organic matter. Organic matter may include potato waste or spent mushroom compost and should be about 12 to 18 inches in depth, with most cells being closer to the later end (Skousen, et a l . , 1992). Biocompatabi l i ty is important. Presence o f SRB 's and other microbes within the chosen organic substrate w i l l be a good indication that populations w i l l grow there. T o ensure optimal conditions for the growth of microbes it is ideal to maintain an influent p H o f about 5 or above (Dodds-Smith, et al . , 1995). A Vegetation cover can assist in creating good conditions for microbes and can be added as a floating mat to reduce surface mixing, provide a continuous input of new organic matter and to sustain a larger microbial population. A n example o f the construction of a vegetation mat uses mats made up o f 4" diameter P V C tubing with plastic fencing as support. The rafts are f i l led with a 10cm layer o f peat and planted with cattail seedlings. (Ka l in , 1997) Other covers such as synthetic ones can also be used to cover the cel l . Limestone is sometimes added to the cel l . This would represent a layer at the bottom o f the cel l for the purpose o f adding further alkalinity to the system. A s in the A L D , the particle size w i l l be important in determining good hydraulic condit ion and unobstructed f low through the system. A range of sizes can give the benefits o f a larger surface area for reaction as wel l as larger particles for better f low conditions. Hydraul ic conductivity is important since the cell is usually set up for up or down f low through the organics (and limestone i f present). Experience has shown that down flowing cells are less prone to problems and require less head in the system to push the water through. Figure 5.3a and Figure 5.3b show two different cells for anaerobic treatment. Volumetr ic and area-loading factors w i l l determine the size of the sell design. Typ ica l ly cells are designed for a volumetric factor o f 0.3 moles o f metal loading per m3 of cel l volume and an area-loading factor o f 20m2 o f surface area per L/sec o f f low (Gusek, 1995). A loading factor based on 5g/m2/d of iron removal may also be used (Skousen, et a l . , 1992). \jt[anaerobic].To Anaerobic Cel l . .A \t\jt[contents].Back to Index..A\t\et[exit].Back to Knowledge Base..A \bt\ \nt[ald_con]\ \h2.Design and Construction o f an Anox ic Limestone Drain\ The size of an anoxic limestone drain is dependent on the f low rate, the drainage water quality and the amount o f treatment needed. Higher f low rates would necessitate a larger system in order to increase retention time o f the contaminated water. A n average retention times vary and can be anywhere between 15-20 hours (Faulkner, and Skousen, 1995) to 48 hours (Gusek, 1995). Max imum f low rates are usually about 100 gpm due to size and area limitations unless there is a low mineral acidity. The drain should be designed to be completely f i l led with water year round and a l low for overflows. There must be enough alkalinity added to counter the acidity of the inlet water. M o r e acidic water w i l l require a larger A L D to attain equivalent results. A L D ' s generate approximately 300 mg/L o f alkal inity, and therefore w i l l not treat drainage o f more than 300 mg/L o f acidity (Skousen, 1991). A l ow dissolved oxygen content is required to prevent armouring of the limestone. The iron should be mainly in a ferrous state such that it w i l l not precipitate at the p H level of the drain. A s wel l aluminum concentrations should be less than 25 mg/L otherwise the aluminum can precipitate in the drain and can cause armouring and clogging (Skousen, et a l , 1992) \ i4 \ \ j t [a ldJ imes tone ] .LEVIESTONEADDITIONS. .A \jt[ald].To Anox i c Limestone Drain..A \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A 154 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[ald_limestone]\ \h2.Limestone Addit ions to an Anox ic Limestone Drain\ The limestone quality w i l l affect the level of alkalinity generated by the drain, therefore, a min imum grade of 9 0 % calcium carbonate is recommended. The size of the limestone particles used within the drain should be large enough to al low for uninterrupted f low yet small enough to provide a large enough surface area for dissolution and alkalinity generation. A range of particle sizes would provide the best results. Recommendations are between 1.5 to 4 inches in diameter (Skousen, 1991). In addition, the drain should be able to hold limestone to generate 20 years o f alkalinity. The amount o f limestone necessary for a particular l ife of the drain and other important calculations are shown below: \t 1. A c i d load o f the drainage (tons/year), A tons/year o f acid = f low (gpm) * acidity (mg/L) * 0.0022 2. Estimated life of the drain (years), B tons of limestone over years o f l i fe = A (tons/year) * years o f l ife 3. Calc ium carbonate content of limestone (%), C tons = B (tons) / calc ium carbonate content (%/100) 4. Amount o f dissolution of limestone (%), D total tons = C (tons) / dissolution (%/100) \ i0\ A plastic layer placed on the top of the limestone serves to keep oxygen out o f the cel l . Other covers include hay bales separated from the limestone by a filter fabric and enclosed with plastic (Skousen,1991). This type o f cover also serves to impede the movement of oxygen into the cel l . The plastic in both cases is overlain with a compacted fine soi l matrix such as clay to further hinder the diffusion o f oxygen. \jt[ald_con].Back to A L D Construction...\ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[aerobic_con]\ \h2.Design and Construction of an Aerobic Ce l l \ A n aerobic cel l is built to promote surface flow. It is a shallow cel l o f approximately 0.3 meters in depth with a length to width ratio of approximately 10:1 (Sengupta, 1993). The size must accommodate the volume o f the precipitate that w i l l come out o f solution. Approximately 2 to 11 g/day/m3 o f i ron can be removed from the surface area o f the cel l (Gusek, 1995). Siz ing is also often based on 20g/m2/d o f i ron removal (Skousen, et al . , 1992). The cell should be able to retain 20 years of precipitate and w i l l then need to be dredged. Precipitate volume can be calculated for iron in the fol lowing example (Dodds-Smith, et a l . , 1995): A (mg/L) o f iron * B (L/sec) of f low = C (m3) o f hydroxide precipitate per year Aerobic cells usually require a large surface area. The cell must therefore be built to accommodate what may result in significant rainfall in order to prevent overflow of untreated water. A rea required for treatment in an aerobic cell is dependent on the p H o f the drainage water entering. Pretreatment to give a p H value of 5 resulted in a decrease in the required area o f the anaerobic ce l l by 7 5 % at the Wheal Jane site (Dodds-Smith, et al . , 1995). Residence time can be increased through the use o f cells in series. Other features such as cascades between cells can increase aerobic mixing and introduce more oxygen into the cel l thereby increasing the productivity. Plants must be introduced. When choosing the right species its tolerance to local watershed conditions and the often harsh acid drainage must be considered. A diversity o f species may be used as poly-cultures offer the benefits o f a better chance o f survival and usually better treatment. Another consideration should be bioaccumulation. If the plant has a tendency to accumulate certain toxins these may be introduced through the plant into the food chain. The plants should be introduced and allowed to establish themselves before drainage water is introduced to al low for a better chance o f survival. Common plants used in an aerobic cel l include Sphagnum sp., Typha sp., and Phragmites sp. \jt[aerobic].To Aerobic CelL.Wt \jt[contents].Back to Index.. A\t\et[exit].Back to Knowledge Base..A 155 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[design_over]\ \h2.Overview o f Design Criteria\ Design begins with choosing the appropriate passive treatment system for the specif ic contaminated drainage site. Design parameters can only be set once the proposed function o f the system is identified. A n analysis of the drainage water to determine the necessary processes needed to reach effluent standards w i l l determine the function o f the system. Among these may be included metal removal, alkalinity generation and p H increase. Passive treatments are not an exact science and involve inter-disciplinary work. Imprecise decisions based on an incomplete understanding of the many processes in passive systems as we l l as the chemistry o f the drainage itself are involved. It is not surprising then that some discrepancies arise between various sources. When choosing an appropriate passive system it is necessary to look at many parameters and characteristics o f the drainage. \ i3\ \jt[param].Design Parameters\ \jt [site_con]. Site Considerations\ \jt[cell_design].Cell Design\ \jt[f low].Flow Pattern\ \jt[costs].Costs\ \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[param]\ \h2.Design Parameters\ The fol lowing is a list o f some o f parameters that should be measured in order to understand the drainage characteristics and be able to start choosing an appropriate system for treatment. \ i l \ 1) F lows with account for f low variations 2) p H 3) acidity and alkalinity 4) iron and manganese concentrations \ i0\ Measurements should be taken ass close to the A R D outlet as possible. If results show a net acidity in the water an A L D may be of consideration and further parameters must be measured. \ i l \ 1) Dissolved oxygen content (this may be measured according to the E h with Eh=0 or less showing low oxygen content ( < 2 mg/L )) (Skousen, 1991) 2) ferrous iron [Fe+2] and aluminum concentrations \ i0\ \i2\\jt[design_over].Back to Overv iew o f Design Criteria.. A \jt [s i tecon] . Site Considerations.. A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[site_con]\ \h2.Site Considerations\ A n appropriate system to treat acid rock drainage must be site specific. Topography, climate and local weather variability, and watershed characteristics al l play a role in a passive treatment system. (Wi ldeman, et a l . , 1993) Site considerations include: \ i3\ 156 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \dt[source]. Sources\ \dt[fiow_rate].Flow Rate VariabiliryA \dt[fiuid].Fluid C o l l e c t i o n \dt[contain].Containment Structures\ \ i0\ \i2\\jt[design_over].Back to Overv iew of Design Criteria.. A \jt[cell_design].Cell Design..A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[source]\ Sources - The source o f A R D may be diverse and must be identified. Some typical sources include adit/tunnel portals, waste rock piles, tail ings ponds, open pits, shafts and seeps (either naturally occurring or due to mining activity). \bt\ \nt[flow_rate]\ F l ow Rate Var iabi l i ty - Determining the f low rate and its variabil ity is as important task to designing. Historical data (at least one year previous) may be extremely beneficial to determining seasonal highs and lows. F l ow variabil ity of streams within a watershed can provide further information. Surge f lows can dramatically effect the passive treatment system. R isks o f a surge event may dictate the construction o f features such as water holding cells or impoundments. \bt\ \nt[fluid]\ F lu id Col lect ion - In order to prevent metal hydroxide formation in conveyances fluid interception should be done as close to the source o f drainage as possible. If impoundments are necessary they may include underground impoundments or open ponds. \bt\ \nt[contain]\ Containment Structures - The slope o f the site, area available, watershed characteristics and groundwater f low are al l useful in the design o f the structures and may cause problems i f forgotten or not considered. \bt\ \nt[cell_design]\ \h2.Cel l Design\ A modular design for the cells wi th in a passive system w i l l provide the most flexibility for the system. A large single cell has numerous drawbacks including diff icult control, s low response to changing conditions, and diff iculty in adapting and redesign i f necessary. A system made up o f many small modules instead of one large one wi l l generate problems when balancing water levels, flows and other conditions. The optimal system wi l l consist a number of modules that are easy enough to handle yet not so numerous as to pose regulatory problems. The \jt[size].size o f the cells\ w i l l be governed by the f lux to set the lower \ minimum size l imit, while overall treatment requirements and site characteristics w i l l set the upper \ maximum size limit. Modules of cells can be set up in various configurations based on requirements of the system and site limitations. Three categories o f cel l arrangements al l provide different benefits to the system. \ i3\ \dt[paraIlel].ParaIlel Modu le Arrangement \dt[series].Series C e l l Ar rangement \dt[combo].Combination Arrangement 157 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \i0\ \i2\\jt[design_over].Back to Overv iew of Design Criteria..A \jt[flow].Flow..A \jt[contents].Back to Index..A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[size]\ \h2.Cel l Number and SizeA \t\Further criteria that may be used to determine the size and number of cells include: \ i l \ 1) Maintenance - addition o f organics and long term reliabil ity 2) Hydro logy and F l o w - control o f large f lows may be distributed amongst a number of cells 3) Site - topography and hydrology o f a site can l imit size or serve to assist with f low 4) Need for Process Cel ls - separation of different functions to separate cells for better performance (separate cel l for aerobic and anaerobic functions) 5) Mater ials Handl ing Area - for storage or handling o f new or old substrate \ i0\ \jt[cell_design].Back to C e l l Design..A \jt[contents].Back to Index.. A\t\t\et[exit].Back to Knowledge Base.. A \bt\ \nt[parallel]\ Paral lel Modu le Arrangement - This type o f arrangement contains a number of cells making up equal and parallel streams. Usua l l y five to ten cells are used with more cells becoming harder to control. The system is more reliable then a single cel l system. A s one cell may be unavailable due to maintenance the others can serve as temporary backup cel ls. \bt\ \nt[series]\ Series Ce l l Arrangement - Ce l ls in series are made up o f two or more cells connected by f low that moves through one and into the next. Drainage water is treated by al l the cells within the series. This allows for specialized cells that can treat one aspect o f the drainage while the next cel l may treat another. It offers better results then a single large cell system. \bt\ \nt[combo]\ Combinat ion Arrangement - The combination arrangement offers the benefits o f both the parallel modules and the series cells such that f low can be routed and divided up into a number o f streams and also have the benefit of treatment through a number o f specific cells. This type of system usually requires a fairly substantial area o f land thus site l imits might restrict the use of such a system. A combination arrangement also has the benefit of being able to evolve f rom a test scale system into a larger system to treat a larger amount o f drainage through the addition o f more cells. \bt\ \nt[flow]\ \h2.Flow\ F l ow through the cells o f a passive treatment system can be designed using three different types of f low patterns: plug f low, step feed, and recirculation. P lug f low is used most often and requires minimum piping, energy, operation and maintenance (Sengupta, 1993). Water f lows through the system from the inlet to the outlet. 158 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Water in a step feed system can move into treatment cells through a number o f inlets along its path. Recirculat ion allows water to be re-routed back through the system for further treatment. \i2\\jt[design_over].Back to Overv iew of Design Criteria.. A \jt[costs].Costs...\ \jt[contents].Back to Index.. A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[costs]\ \h2.Costs\ Passive treatments have a wide variety of construction conditions and are built on a number o f different site locations. It is diff icult then to give an estimate o f costs for the construction o f these systems as these would vary greatly from one treatment system to another. A n overview of cost components and a comparison o f several implemented passive treatments to conventional treatment costs w i l l be reviewed. Cost components are usually typical from site to site and may be divided into Capital & Construction costs, and Operating Costs. \ i3 \ Capital & Construction Costs \dt[eng_test].Engineering and Testing\ \dt[enviro].Environmental Basel ine Studies\ \dt[land].Land Equis i t ion\ \dt[righ_of].Right o f W a y Access\ \dt[final].Final Engineering Design & Construct ion \dt [construction] .Construct ion Operating Costs \dt[nominal] .Nominal Maintenance Tasks\ \dt[major].Major Overhauls\ \jt[compare].Passive Treatment versus Convent iona l \i2\\jt[design_over].Back to Overv iew of Design Criteria.. A \jt[contents].Back to Index.. A\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[eng_test]\ Engineering and Testing - This includes al l preliminary work for feasibility studies and treatment processes. Act iv i t ies such as the testing o f source water; substrate materials; field testing in ' 'mini-cel ls" for substrate permeability, loading rates and flux; preliminary design; pilot scale testwork; and hydrologic investigations. \bt\ \nt[enviro]\ Environmental Basel ine Studies & Permits - These are studies required to be done for siting and permitting. They include air & water quality, climate, geology, cultural, vegetation & wi ldl i fe, land use as we l l as socio-economic impacts. \bt\ \nt[land]\ L a n d Acquis i t ion - Private land holdings and large area requirements can create significant costs. 159 A Fuzzy Expert System on Acid Rock Drainage Site Remediation \bt\ \nt[right_of]\ Right o f W a y Access - This pertains to systems that may not be close to the source and water must be routed through pipes or open channels over land that may not be owned. \bt\ \nt[final]\ F inal Engineering Design and Construction Specifications - F inal details and construction specifications including text and drawing and contractor pay for building the project. \bt\ \nt[construction]\ Construction - Some costs may include mobil ization o f personnel, equipment and supplies, site preparation, earthwork, source control, l inings, vegetation and etc. \bt\ \nt[nominal]\ Nomina l Maintenance Tasks - Periodic inspections, sampling, cleaning, f low adjustments, and balancing w i l l be required on a timely basis. \bt\ \nt[major]\ Major Overhauls - Removal , treatment, transportation and disposal o f o ld substrate, preparation and installation of new substrate, and temporary rehandling o f untreated effluents. \bt\ \nt[compare]\ \h2.A L o o k at Passive Treatment Costs Versus Conventional Chemical Treatment Costs\ Passive Treatment costs stand up wel l to costs of chemical treatment. Even Ln cases where post conventional treatment was necessary passive treatment has lowered costs versus conventional treatment alone. A study performed by the Pennsylvania Department o f Environmental Resources on 73 constructed wetlands treating coal mining sites concluded that passive treatment is the Best Avai lable Technology ( B A T ) economical ly achievable for seeps wi th acidity up to 300 mg/L as C a C 0 3 (Hell ier, et al . , 1994). Seep chemistry, f low volume and site characteristics can al l adversely affect the cost of a passive treatment system from one site to another. Cost benefit analysis o f passive treatment at the Wheal Jane site is reported as marginal due to large volumes o f water that needs treatment. (Cambridge, 1997) The Keister Wetland incurred capital costs o f $225,000 whi le chemical treatment would have cost an estimated $8,500 to $9,700 for caustic soda and ammonia respectively. A l though the chemical treatment may have worked out to be cheaper, the remote location o f the site, poor access roads, and intermitted drainage f lows rendered chemical treatment inefficient (Faulkner and Skousen, 1994). Another site found the $110,000 construction costs for the passive treatment system was worth the investment versus the $22,000 per annum chemical treatment with caustic soda that could have been easily doubled once attendant labour and sludge handling costs are included (Faulkner and Skousen, 1994). Gusek compared passive treatment systems with l ime treatment and found that although capital costs for passive treatment may be similar to l ime treatment the operating costs are 7 1 % to 3 1 1 % higher (Gusek, 1995). \jt[costs].Back to Costs\\t\t\et[exit].Back to Knowledge Base\ \bt\ \nt[references]\ \ h 2 . R E F E R E N C E S \ 160 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Bechard, G . , and R . G . L . McCready , 1991. Mic rob io log ica l Process for the Treatment of Ac i d i c Drainage at the Hal i fax International Airport. In: Second International Conference on the Abatement of Ac id i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Bender, J . Dr. , and Dr. P. Phi l l ips, 1995. Biotreatment o f mine drainage. In: M in ing Environmental Management, pp. 25-27, Sep. 1995. Bennett, P . G . , C R . Ferguson, and T . H . Jeffers, 1991. B io log ica l Treatment of A c i d M i n e Waters - Case Studies. In: Second International Conference on the Abatement o f Ac i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Broughton, L . M . , R .W. Chambers, and A . M a c G . Robertson, 1992. M i n e Rock Guidelines Design Control o f A c i d Drainage Water Quality. Steffen, Robertson, and Kr is ten (B .C. ) Inc., Vancouver, B . C . Butcher, S.S. , et a l . , Eds, 1992. Globa l Biogeochemical Cycles. Academic Press, Toronto, Ontario. Cambridge, M . , 1997. Wheal Jane - The Long Term Treatment o f A c i d M i n e Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , V o l . 3. Dodds-Smith, M . E . , C . A . Payne, and J.J . Gusek, 1995. Reedbeds at Whea l Jane. In: M in ing Environmental Management, pp. 22-24, Sep. 1995. Dvorak, D . H . , R .S . Hedin, S.P. Mcln t i re , and H . M . Edenborn, 1995. Treatment o f Metal-Contaminated Water Us ing Bacterial Sulphate Reduction-Results from Pi lot -Scale Reactors. In: Second International Conference on the Abatement o f Ac i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. Eger, P., J.R. Wagner, and G . Melchert, 1997. The Use o f Peat/Limestone System to Treat A c i d Rock Drainage. In: Fourth International Conference on A c i d Rock Drainage Proceedings, Vancouver, B . C . , V o l . 3. Evere l l , M . D . , and A . J . Ol iver, 1992. Environmental Technologies for the Minerals and Metals Sector - A Canadian Perspective. In: Minerals, Metals and the Environment. Elsevier Science Publishers L td . , N e w York . Faulkner, B . B . , and J .G . Skousen, 1994. Treatment o f A c i d M i n e Drainage by Passive Treatment Systems. In: International M ine Reclamation and M i n e Drainage Conference and Thi rd International Conference on the Abatement o f Ac id i c Drainage Proceedings, V o l . 2, A p r i l 24-29, 1994. Faulkner, B . B . , and J .G . Skousen, 1995. Treatment o f A c i d M i n e Drainage by Passive Treatment Systems. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Universi ty and the National M i n e Reclamation Center, Morgantown, West Vi rg in ia. 1995. Gou ld , W . D . , G . Bechard, and L. Lort ie, 1994. The Nature and Ro le o f Microorganisms in the Tai l ings Environment. In: Short Course Handbook on Environmental Geochemistry of Sulf ide Mine-Wastes, Mineralogical Associat ion o f Canada, Waterloo, Ontario, Chap. 7, M a y 1994. Gusek J .J . , 1995. Passive-Treatment o f A c i d Rock Drainage: What is the Bot tom L ine? In: M in ing Engineering, pp. 250-253, M a r c h 1995. Hendricks, A . C . , 1991. The Use o f Ar t i f ic ia l Wetland to Treat A c i d M i n e Drainage. In: Second International Conference on the Abatement o f Ac id i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Her l ihy, A . T . , A . L . M i l l s , G . M . Hornberger, and A . E . Bruckner, 1987. The Importance of Sediment Sulfate Reduction to the Sulfate Budget of an Impoundment Receiv ing A c i d M i n e Drainage. In: Water Resources Research, V o l . 23, N o . 2, pp. 287-292, Feb. 1987. 161 A Fuzzy Expert System on Acid Rock Drainage Site Remediation K a l i n , M . , 1995. B io log ica l Amel iorat ion o f A c i d i c Seepage Streams. In: Second International Conference on the Abatement of Ac i d i c Drainage Proceedings, Tome 1, Sep. 17-18, 1991. K a l i n , M . , and M . P . Smith, 1997. M ic rob ia l A c i d Reduct ion in Sediments - Concepts and Appl icat ion. In: Fourth International Conference on A c i d R o c k Drainage Proceedings, Vancouver, B . C . , V o l . 3. K a l i n , M . , and R .O . V a n Everdinger, 1987. Eco log ica l Engineering: B io log ica l and Geochemical Aspects, Phase I Experiment. In: A c i d M i n e Drainage Seminar / Workshop Proceedings, Hal i fax, N o v a Scotia. Ka l i n , M . , M . P . Smith, and G . Landry, 1994. The Appl icat ion o f Eco log ica l Engineering to A c i d Coa l Seepages in Eastern Canada. In: International M i n e Reclamat ion and M i n e Drainage Conference and Third International Conference on the Abatement of A c i d i c Drainage Proceedings, V o l . 2, A p r i l 24-29, 1994. Karathanasis, A . D . , and Y . L . Thompson, 1995. Me ta l Speciation and Retention Patterns in a H igh Meta l Load A c i d M ine Constructed Wet land o f Southeastern Kentucky. In: Second International Conference on the Abatement o f Ac id i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Kuyucak, N . , P. St-Germain, and K . G . Wheeland, 1995. In Situ Bacterial Treatment of A M D in Open Pits. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 1, Sep. 17-18,1991. Lawrence, R .W. Dr . , and Dr . A . M a c G . Robertson, 1994. A Short Course on A c i d Rock Drainage Understanding the Problems - Finding Solutions. The C L M Distr ict 6 Annua l Meet ing, Vancouver, B . C . , Oct. 11, 1994. Lawrence, R .W. , 1997. A Course on Fundamental and App l ied Aspects o f A c i d Rock Drainage. Course notes, Department of M in ing and Minera l Process Engineer ing, U B C , Vancouver, Canada. Pesavento, B . G . , 1987. Factors to be Considered in the Use o f Wetlands to Treat A c i d M i n e Drainage. In: A c i d M i n e Drainage Seminar / Workshop Proceedings, Ha l i fax , N o v a Scotia. Plane and Associates L td . , 1990. Assessment o f Exis t ing Natural Wetlands Affected by low p H , Meta l Contaminated Seepages. F ina l Report, M E N D project 3.12.1. Richardson, J .S. , and C . J . Perr in, 1990. The Effect o f A c i d M i n e Drainage on Stream Macroinvertebrates and Periphytic Algae. B C A M D project 4.30.C.4. Ri tchie, A . I . M . , 1994. The Waste Rock Environment. In: Short Course Handbook on Environmental Geochemistry of Sulfide Mine-Wastes, Minera log ica l Associat ion o f Canada, Waterloo, Ontario, Chap. 5, M a y 1994. Robb, G . , and J . Robinson, 1995. A c i d mine drainage prediction and remediation. In: M in ing Environmental Management, pp. 19-21, Sep. 1995. S E N E S Consultants L td . , 1994. A c i d M i n e Drainage - Status o f Chemical Treatment and Sludge Management Practices. M E N D project 3.32.1. Sengupta, M . , 1993. Environmental Impacts o f M i n i n g Moni tor ing, Restoration, and Control. Lewis Publishers, U S A . Sextone, A . J . , J .C . Sencindiver, J .P. Calabrese, D . K . Bhumbia , and G . K . Bissonnette, 1991. Appl icat ion of Constructed Wetlands for the Remova l o f Iron f rom A c i d M i n e Drainage. In: Second International Conference on the Abatement o f Ac i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Shelp, G . , G . Southam, G . Spiers, L . L i u , and W.Chesworth, 1994. A n Evaluat ion o f a Peat-Wood Chip-Micro f lora Admixture to act as an Amel iorant for A c i d M i n e Drainage. In: International M i n e Reclamation and M i n e Drainage 162 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Conference and Thi rd International Conference on the Abatement of Ac id ic Drainage Proceedings, V o l . 2, A p r i l 24-29, 1994. Skousen, J . , 1991. A n o x i c Limestone Drains for A c i d M i n e Drainage Treatment. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia University and the Nat ional M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Skousen, J . , A . Sextone, K . Garbutt, and J . Sencindiver, 1992. Wetlands for Treating A c i d M ine Drainage. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia Universi ty and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Skousen, J . , B . Faulkner, and P. Sterner, 1995. Passive Treatment Systems and Improvements o f Water Quali ty. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia Universi ty and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Skousen, J . G . , 1995. Douglas Abandoned M i n e Land Project: Description of an Innovative A c i d M i n e Drainage Treatment System. In: A c i d M i n e Drainage Control and Treatment, West V i rg in ia University and the National M ine Reclamation Center, Morgantown, West V i rg in ia . 1995. Stark, L .R . , W . R . Wener ick, P .J . Wuest, and F . M . Wi l l iams, 1995. Adding a Carbon Supplement to Simulated Treatment Wetlands Improves M ine Water Quality. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Taddeo, F . J . , and R . K . Weider , 1991. The Accumulat ion of Iron Sulphides in Wetlands Constructed for A c i d Coa l M i n e Drainage ( A M D ) Treatment. In: Second International Conference on the Abatement of Ac id i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Wheeler, W . N . , and M . K a l i n , 1995. The Ecologica l Response o f a Bog to Coal A c i d M ine Drainage. Deterioration and Subsequent Initiation o f Recovery. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 2, Sep. 17-18, 1991. Wieder, R . K . , M . N . L in ton, and S.T. Starr, 1991. Spatial and Temporal Patterns in Surface and Subsurface Water Chemistry in Wetlands Constructed for A c i d Coal M i n e Drainage. In: Second International Conference on the Abatement o f A c i d i c Drainage Proceedings, Tome 3, Sep. 17-18, 1991. Wi ldeman, T., G . Brod ie , and J . Gusek, 1993. Wetland Design for M in ing Operations. B iTech Publ ishing C o . , Vancouver, B . C . Z iemkiewicz , P. , J . Skousen, and R. Lovett, 1994. Open Limestone Channels for Treating A c i d M i n e Drainage: A N e w Look at an O l d Idea. In: A c i d M ine Drainage Control and Treatment, West V i rg in ia Universi ty and the Nat ional M i n e Reclamat ion Center, Morgantown, West Vi rg in ia . 1995. \jt[contents].Back to Index...\\t\t\et[exit].Back to Knowledge Base..A \bt\ \nt[exit]\ E X I T 163 A Fuzzy Expert System on Acid Rock Drainage Site Remediation APPENDIX D - INPUTS AND OUTPUTS 164 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Run #1 and Run #2 INPUTS Units Runl Run2 Waste / tails W W Capital available $ $1,000,000 $1,000,000 Treat plant capacity gpm 150 150 How many bodies of water very few very few Distance to populated area km 50 50 Annual rain mm 600 20 Annual snow mm 1000 none Parks & protected regions in area some some Size of closest population town town Hydraulic gradient relative to toward60 parallel40 parallel40 away80 Scavenger material available low amount low amount Hydraulic conductivity / permeability of surrounding area cm/s somewhat permeable 20 impermeable80 somewhat permeable 60 impermeable40 Size of catchment area moderate small A R D potential (NP/AP ratio) 0.45 0.45 Topography of area low gradient 70 mod.gradient 30 low gradient 30 mod.gradient 70 Hydraulic conductivity of waste m/s permeable permeable Hydraulic conductivity of underlying material cm/s slightly permeable impermeable Is dump liner present (Y/N) (Y/N) No No Surface area of waste dump (hectares) / tailings pond (m2) hectare / m2 20 20 Is there ponding on the waste dump (Y/N) No No Tonnes of waste rock Tonne 4,000,000 4,000,000 Water table m 7 deep level reactivity level reactivity Aluminum 1-10 1 1 1 1 Arsenic 1-10 1 1 1 1 Cadmium 1-10 1 1 1 1 Chromium 1-10 1 1 1 1 Copper 1-10 1 1 1 1 Cyanide 1-10 1 1 1 1 Lead 1-10 1 1 1 1 Magnesium 1-10 5 1 5 1 Manganese 1-10 6 6 Mercury 1-10 1 1 1 1 Uranium 1-10 1 1 1 1 Others 1-10 6 3 6 3 Is clay present in waste 1-10 4 4 Size distribution mesh Physical impact on habitat within lake/ocean 1-10 3 5 Social acceptance of underwater deposition 1-10 5 3 Political acceptance of underwater deposition 1-10 3 3 165 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Excessive nutrient additions 1-10 1 1 Level of organics present 1-10 3 3 Organic present methylating 1-10 1 1 Tail ings dam stability 1-10 1 1 Land availability (for wetland sludge disposal) 1-10 2 8 Water bodies closest to area Ocean / Lake lake lake Volume occupied by waste rock m3 7,700,000 7,700,000 Volume of the pit m3 3,600,000 3,600,000 Pit stability moderate moderate Drought 1-10 1 8 F lood 1-10 1 1 Avalanche 1-10 3 1 Earthquake 1-10 3 1 Steady hydro conditionss 1-10 2 8 Freezing (ice in winter) 1-10 10 1 Thermal overturn in area lakes 1-10 5 1 Wave action 1-10 2 1 Lake depth moderate shallow p H runoff 7.3 5 Dissolved oxygen mg/L 1000 1000 Ferrous iron mg/L 1.5 1.5 Ferr ic iron mg/L 1.5 1.5 Total iron mg/L 3.37 3.37 Aluminum mg/L 0.01 0.01 Manganese mg/L 29.3 29.3 Sulphate mg/L 5500 5500 Acid i ty mg/L 25 25 F l o w gpm 80 10 p H regulation 7 7 Ferrous iron mg/L 0.1 0.1 Ferr ic iron mg/L 0.1 0.1 Total iron mg/L 0.1 0.1 Manganese mg/L 0.01 0.01 A luminum mg/L 0.01 0.01 Rea l rate of return actual value 0.035 0.035 Operating years (treatment) yr 20 20 M S Index value / year 1062.4/1999 1062.4 / 1999 Cost revenues available to write off N o N o % o f surface area requiring regrading % 2 2 Average depth of earth requiring regrading m 0.5 0.5 Volume of slope material m3 40,000 40,000 Pump & pumphouse present ( Y / N ) N o N o Distance to nearest water body k m 50 500 Length of collection ditches m 600 600 Length of dyke required m 600 600 Length of runoff channel m 100 100 166 A Fuzzy Expert System on Acid Rock Drainage Site Remediation OUTPUTS Runl Run2 Waste to pit & flood $9,869,796 $9,869,796 cost sub-marine disposal $25,471,656 $259,395,056 simple soil 0% $2,007,208 0% $2,007,208 Multiple soil cover w/ drainage layer 0% $7,834,367 30% $7,834,367 Compound soil cover w/ plastic liner 35% $10,109,640 36% $10,109,640 Diversion $2,202,749 $2,202,749 Treatment atpH 10 3.52 tons caustic soda atpHIO 0.44 tons caustic soda DoB Rating 1 DoB Rating Dry cover 35 3 more capital required 30 4 more capital required Water Diversion 75 34 more capital required 53 24 more capital required Passive Treatment 39 20 more capital required 44 181 Excess capital available Water Treatment 89 18 more capital required 89 25 more capital required Water cover 45 5 more capital required 44 4 more capital required Cover Module Capital Operating Total Capital Operating Total complex soil DoB 35% $8,791,000 $93,000 $10,110,000 36% $8,791,000 $93,000 $10,110,000 Multiple soil cover w/ drainage layer DoB 0% $6,813,000 $72,000 $7,835,000 30% $6,813,000 $72,000 $7,835,000 simple soil DoB 0% $1,746,000 $19,000 $2,008,000 0% $1,746,000 $19,000 $2,008,000 Water Diversion $1,916,000 $21,000 $2,203,000 $1,916,000 $21,000 $2,203,000 Downstream Interception of Effluent 48 53 Downstream Ditch or Berm 75 50 Upstream Interception of Effluent 40 47 Upstream Ditch or Berm 74 43 Passive Treatment Aerate then aerobic/anaerobic wetland Anoxic limestone Drainage $7,000 $1,000 $8,000 $1,000 $1,000 $1,000 Aerobic Wetland $1,701,000 $18,000 $1,956,000 $213,000 $3,000. $245,000 Anaerobic Wetland $1,882,000 $20,000 $2,164,000 $236,000 $3,000 $271,000 Water Treatment $798,000 $279,000 $4,869,000 $770,000 $192,000 $3,608,000 Ammonia 0% 0% Caustic soda 100% 100% Hydrated Lime 0% 0% Limestone 0% 0% Quick Lime 0% 0% Water Cover waste to pit and flood 45 $8,583,000 $91,000 $9,870,000 44 $8,583,000 $91,000 $9,870,000 sub-marine disposal 40 $22,150,000 $234,000 $25,472,000 40 $225,560,992 $2,381,000 $259,396,000 167 A Fuzzy Expert System on Acid Rock Drainage Site Remediation R u n #3 a n d R u n #4 INPUTS Units Run3 Run4 Waste / tails T T Capital available $ $8,000,000 $8,000,000 Treat plant capacity gpm 0 0 How many bodies of water some some Distance to populated area km 10 10 Annual rain mm 200 200 Annual snow mm 300 300 Parks & protected regions in area yes yes Size of closest population city city Hydraulic gradient relative to population away away Scavenger material available moderate moderate Hydraulic conductivity / permeability of surrounding area cm/s somewhat permeable somewhat permeable Size of catchment area small small ARD potential (NP/AP ratio) 3 0.5 Topography of area low grad low grad Hydraulic conductivity of tailings m/s SP70. P30 SP70, P30 Buffering capacity of tailings high low Sulphur content low moderate pyrrhotite low (none) moderate pyrite low high Chalcopyrite moderate moderate Hydraulic conductivity of underlying material cm/s SP 80, P 20 SP80, P20 Is dump liner present (Y/N) (Y/N) yes yes Surface area of waste dump (hectares) / tailings pond (m2) hectare / m2 800,000 800,000 Average depth of tailings m 6 6 Interstitial waters description semi-hazardous hazardous Water table m 1 1 level reactivity level reactivity Aluminum 1-10 1 |_ 1 5 1 Arsenic 1-10 1 1 1 1 Cadmium 1-10 1 1 1 1 Chromium 1-10 1 1 1 1 Copper 1-10 1 4 1 Cyanide 1-10 1 1 1 1 Lead 1-10 1 1 1 1 Magnesium 1-10 1 1 6 Manganese 1-10 1 1 5 1 Mercury 1-10 1 1 1 1 Uranium 1-10 1 1 1 1 Others 1-10 2 3 8 6 Is clay present in waste 1-10 3 3 Size distribution mesh 80% 400 mesh (fine) 80% 400 mesh (fine) Physical impact on habitat within lake/ocean 1-10 6 9 Social acceptance of underwater deposition 1-10 6 6 Political acceptance of underwater deposition 1-10 6 6 168 A Fuzzy Expert System on Acid Rock Drainage Site Remediation Excessive nutrient additions 1-10 3 8 Level of organics present 1-10 2 5 Organic present methylating 1-10 1 1 Tai l ings dam stability 1-10 9 9 L a n d availabil i ty (for wetland 1-10 7 7 sludge disposal) Water bodies closest to area Ocean / Lake lake lake Drought 1-10 1 1 F lood 1-10 1 1 Avalanche 1-10 1 1 Earthquake 1-10 1 1 Steady hydro conditionss 1-10 3 3 Freezing (ice in winter) 1-10 10 10 Thermal overturn in area lakes 1-10 8 8 Wave action 1-10 8 8 Lake depth moderate moderate p H runoff 5.5 2.3 Dissolved oxygen mg /L 1000 80 Ferrous iron m g / L 1 6 Ferr ic i ron m g / L 1 9 Total iron m g / L 1 15 A l u m i n u m m g / L 0.1 2 Manganese m g / L 0.5 2 Sulphate m g / L 1500 1500 Ac id i t y mg /L 500 2000 F l o w gpm 300 300 p H regulation 7 7 Ferrous iron m g / L 0.1 0.1 Ferr ic i ron m g / L 0.1 0.1 Total i ron m g / L 0.1 0.1 Manganese m g / L 0.01 0.01 A l u m i n u m m g / L 0.01 0.01 Rea l rate of return actual value 0.035 0.035 Operat ing years (treatment) yr 20 20 M S Index value / year 1062.4 / 1999 1062.4 / 1999 Cost revenues available to write N N off taxes % of surface area requir ing % 1 1 regrading % o f regrading sloped area % 0 0 Average depth of earth requir ing m 0.5 0.5 regrading Ta i l ings completely enclosed ( Y / N ) n n I f N O Vo lume of dams/dykes m3 57000 57000 required Pump & pumphouse present ( Y / N ) n n Distance to nearest water body k m 5 5 Length o f collection ditches m 600 600 Length of dyke required m 100 100 Length o f runoff channel m 50 50 169 A Fuzzy Expert System on Acid Rock Drainage Site Remediation OUTPUTS Run3 Run4 cost flooding $1,098,920 $1,098,920 cost sub-marine $21,207,840 $21,207,840 disposal Veg. Cover $4,811,378 $4,811,378 Simple soil cover $14,847,176 $14,847,176 Multiple soil cover w/ $30,786,386 $30,786,386 drainage laver Treatment pH9.14 264 tons caustic soda 2 process 1056 tons caustic soda Diversion $972,872 $972,872 DoB Rating DoB Rating 1 Dry cover 99 164 Adequate capital available 99 164 Adequate capital available Water Diversion 95 778 Excess capital available 95 778 Excess capital available Passive Treatment 38 18 More capital required 39 5 More capital required Water Treatment 54 47 Adequate capital available 84 64 Adequate capital available Water cover 45 328 Excess capital available 45 328 Excess capital available Cover Module Capital Operating Total Capital Operating Total Cover w/ drainage 0% $26,771,000 $283,000 $30,787,000 0% $26,771,000 $283,000 $30,787,000 simple soil cover 0% $12,911,000 $137,000 $14,848,000 0% $12,911,000 $137,000 $14,848,000 Vegitation cover 99% $4,184,000 $45,000 $4,812,000 99% $4,184,000 $45,000 $4,812,000 Water Diversion $846,000 $9,000 $973,000 $846,000 $9,000 $973,000 Downstream 70 76 Interception of Effluent Downstream Ditch or 67 73 Berm Upstream 95 95 Interception of Effluent Upstream Ditch or 91 91 Berm Passive Treatment Aerate then aerobic/anaerobic wetland Anoxic limestone $480,000 $6,000 $552,000 $1,920,000 $21,000 $2,208,000 Drainage Aerobic Wetland $14,506,000 $154,000 $16,682,000 $58,128,000 $614,000 $66,848,000 Anaerobic Wetland $141,079,008 $1,489,000 $162,240,992 $564,316,000 $5,956,000 $648,963,000 Water Treatment $1,148,000 $553,000 $9,175,000 $1,418,000 $630,000 $10,576,000 Ammonia 0% 0% Caustic soda 70% 70% Hydrated Lime 0% 0% Limestone 0% 0% Quick Lime 0% 0% Water Cover Flood tailings 45 $956,000 $11,000 $1,099,000 45 $956,000 $11,000 $1,099,000 sub-marine disposal 20 $18,442,000 $195,000 $21,208,000 20 $18,442,000 $195,000 $21,208,000 170 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089908/manifest

Comment

Related Items