Development of a new covering strategy in Indonesian coal mines to control acid mine drainage generation: a laboratory-scale result Hideki Shimada"*, Ginting Jalu Kusuma™®, Koh Hiroto", Takashi Sasaoka", Kikuo Matsui, Rudy Sayoga Gautama” and Budi Sulistianto” “Department of Earth Resources Engineering, Kyushu University, Fukuoka, Japan; "Department of Mining Engineering, Institut Teknologi Bandung, Bandung, Indonesia (Received 30 May 2011; final version received 22 July 2011) The waste dump of sulphide-containing rocks is one of the potential acid mine drainage sources, since it contains a huge amount of readily oxidised sulphide mineral, due to its exposure to air and water. The application of the dry cover system is regarded as one of the best practices since it prevents acid mine drainage of the waste rock dump at the surface coal mine. However, the implementation of the dry cover system in field practice has faced several obstacles due to the limited number of cover materials, The nature of geological condition is considered to be a controlled issue, whilst the problem is the mining method and equipment size. This article describes the acid generation mechanism and its control, application of cover system and the problems that are faced in Indonesian coal mines, whilst discussing the preliminary laboratory results of multi-layer cover systems. It finally proposes a new covering strategy in an attempt to overcome the problem. Keywords: surface coal mine; multi-layer cover system; acid mine drainage; prevention 1. Introduction Acid mine drainage (AMD) occurs when sulphide minerals, such as pyrite and marcasite, are exposed to oxidising conditions and, in the absence of alkaline materials, these are oxidised in the presence of water and oxygen to form highly acidic and sulphate-rich drainage [I]. AMD may cause a number of potential problems for the environment and for human beings. For example, the impact on the aquatic life and communities in the downstream environment, which results in acidity and dissolved metals, and the effect on groundwater quality and difficulties in re-vegetation. The neutralisation of the excess of AMD may take a long time, ranging from weeks, years and even decades after the mining activity has ceased. The threat of AMD has been faced not only by the metal mines associated with sulphide deposit, but also by many of the coal mines, especially surface coal mines. Although coal-bearing rock typically has lower sulphide content compared to the metal ores, due to its low durability characteristics, AMD problems at the coal mines are as severe as in the metal mines. This is also exacerbated by the nature of coal “Corresponding author. Email: shimada@ -kyushu-u.ac.jpdeposits, a typical characteristic of sedimentary-type deposit, which requires a larger area to be exposed in order to mine the coal seam as well as the waste rock dumping area that is usually larger than the mines’ mining base metal deposits. ‘The AMD potential and generation from a mine site are unique due to the site- specific conditions, geochemical composition of rocks and the climatic condition of the site. Thus, the natural size of the associated risk and feasibility of mitigation will vary from site to site. In general, there are two approaches in the AMD management: prevention and mitigation/treatment. Prevention method is an effort that could be conducted in an attempt to reduce the oxidation of sulphide mineral, while mitigation is the effort intended to treat the generated AMD into an environmentally acceptable level, by direct or indirect addition of alkaline material into the AMD stream. At the mine site, considering the fact that waste rock dump is the place in which the readily oxidised sulphide-mineral containing materials were stored for a long time, careful attention should be made for this area in an attempt to mitigate the AMD generation. The chemistry and volume of AMD seepage waters emanating from sulphide waste rock dump are largely influenced by the properties of the waste materials. The AMD development in the waste rock dump occurs via complex weathering reactions. The different rates of the various weathering reactions within the waste may cause temporal changes to the drainage chemistry. In this study, the AMD generation mechanism and its control strategy, the application of cover system and problems being faced in the Indonesian coal mines are described, and the results of the proposed solution technique to overcome the problem is discussed. 2. AMD generation and its control Among the coal-bearing rocks, pyrite is regarded as the most common reactive sulphide mineral. Pyrite oxidation and the factors affecting the kinetics of oxidation have been the focus of many researchers (2]-[4], and genera i of pyrite is in accordance to the following series of rea‘ FeS) + 7/20) + H20 — Fe”+ + 280}° + 2H* (Ll) Fe** + 1/402 + H+ — Fe*+ + 1/2H: (2) Fe** + 3H20+ — Fe(OH),(s) + 3H* QB) FeS; + 14Fe** + 8H20 — 15Fe** + 28077 + 16H* (4) Iron sulphide is oxidised and releases ferrous iron (Fe**), sulphate and acid (H*), thus lowering the pH of the solution. In a sufficiently oxidising environment, ferrous iron will be oxidised by molecular oxygen to form ferric iron (Fe**) (Equation (2)).. Ferric iron can then be hydrolysed to form ferric hydroxide (Fe(OH);) in a solution with pH above 3.5 [6] and acidity (Equation (3)), or it can directly attack pyrite when the pH of the solution is less than 3.5 and acts as a catalyst in generating much16 greater amounts of ferrous iron, sulphate and acidity (Equation (4)). This ferrous iron can then be oxidised to form ferric iron (Equation (2)) which in turn may oxidise more pyrites. Thus, a continuous self-propagating pyrite oxidation cycle is formed and may continue until the pyrite or oxidant supply is exhausted Under many conditions, due to low rate conversion of ferrous iron to ferric iron at pH values below 5 and under abiotic conditions, Equation (2) may act as the rate- limiting step in pyrite oxidation. However, the existence of Fe-oxidising bacteria, especially Thiobacillus, in pH below 3.5 greatly accelerates this reaction [4]. The secondary ferric hydroxide which is precipitated may form coating [7,8] and cemented layers [9,10], which can decrease the pyrite oxidation rate. Carbonate minerals, such as calcite (CaCO) and dolomite (CaMg(COs)s) which may co-exist in the sulphide mineral containing rocks, can neutralise the acidity generated by the oxidation of pyrite (Equation (5)), as well as the silicate minerals (Equation (6)), although these minerals have a slower reaction rate than carbonates (1). CaCO; + H* — Ca?* + H20 + CO; (5) MeAISiO,(s) + H* (aq) + 3H20 — Me** (aq) + AP (aq) + H4SiOgjaq) + 3H (aq) (6) (Me = Ca, Na, K, Mg, Mn or Fe) ‘The presence or absence of calcareous carbonate and silicate minerals in the rock sample/materials is extremely important in the AMD generation [12,13]. If the amount of these minerals in rocks is sufficient to offset the acid producing potential of the material, the acid drainage will not eventuate due to the neutralisation process, as long as the reaction rates of the respective materials are similar. Furthermore, these minerals would also inhibit pyrite oxidation by buffering the pH at a level where iron, released by pyrite oxidation, precipitates as ferric hydroxide rather than oxidising additional pyrite (Equation (3)). Based on the above reaction series, the characteristics of drainage as the result of sulphide mineral oxidation is dependent on many factors, both of inherent stoichiometry of the acid generating and neutralising mineral, as well as the external factors such as oxygen and water supply, temperature and presence of bacteria, which make the AMD generation characteristic site specific. However, it can be noted that the inhibition of the AMD generation in waste rock dumps depends on the control of oxygen and water supply as well as on maintaining the pH of the leachate over the compatible acid environment of metallogenium bacteria. Wet and dry covers are the two commonly suggested methods to achieve those required conditions of prevention/reduction oxidation of sulphide minerals. Wet cover works by sinking the sulphide-containing rocks at a certain minimum depth underneath water surface and it is the most effective method to prevent oxidation [14], since the oxygen diffusion rate in water is very low. However, from the economical and flexibility point of view, dry cover method is more promising and can be applied at various places at the mine sites [15], especially at the waste rock dump sites.11 At the mining area, in addition to the primary purposes, such as oxygen and water supply minimisation, dry cover system is also expected to be resistant to erosion and provide a suitable rooting zone for the vegetation to grow up as part of mine closure requirement [16]. The types of dry cover system vary from a single layer of earthen material to several layers of different material types. However, in general it can be classified into three categories: (1) resistive cover, (2) store and release cover and (3) mixed/novel type cover. Resistive cover relies on a very low permeability layer, which is commonly made of clay or compacted clay/natural soil, to minimise the percolation of water and the oxygen ingress. Store and release covers rely on the storage capacity of the certain layer of cover in which the precipitation in the wet season was stored and will be released as evaporated moisture in the dry season and therefore prevent further water percolation into the waste. The last type of cover is a combination of resistive and store and release cover, and the application of oxygen consuming layer (such as the organic waste or low reactive material); thus, it prevents the oxygen ingress into the protected waste underneath, in addition to the limited water deep infiltration. Local environmental characteristics, such as the climatic conditions and availability of the materials greatly affect the determination of the applied dry cover, in addition to the specific construction objectives which should be achieved. Therefore, in the tropical area where precipitation is far greater than the evaporation, the resistive cover is recommended over the store and release cover type, which is more suitable for the arid/semi-arid areas where the precipitation is less than the potential evaporation and where there are usually distinct wet and dry seasons. ‘The more complex cover design is usually expected to yield better performance over the simple cover. However, the more complex designs can increase the susceptibility of the performance to failure and such complex designs usually mean higher construction and maintenance costs. 3. Application of dry cover system and problems being faced in the Indonesian coal mines In general, coal-bearing rocks consist of sedimentary rocks such as clays/mudstones, siltstones and sandstones. Lithostratigraphie sequence of coal-bearing rocks which also reflect the geochemical change sequence, vary from order less than 1 m up to more than 10 m, while the dip rock layers vary from 5 m to 90 m. Due to the low uniaxial compressive strength and its slaking-swelling behaviour when contacted with water, many of the rocks associated with the coal deposits in Indonesia are classified as soft rocks, based on the classification by Bieniawski [17,18]. The nature of geological conditions coupled with the mechanical behaviour of rocks will affect the whole aspect of mining operations such as the mining method selection, unit operation and mine closure strategies. Currently, more than 90% of coal production in Indonesia was produced from surface coal mines. The combination of truck and shovel method is mostly used due to its compatibility with the complex geological conditions in the Indonesian coal fields. The coal mining production ranges from less than 1 million to more than 1 billion tons of material removed per year. Most of the Indonesian coal mines are located on the borders of forest areas which need to be reforested, as required by the mine closure.18 In terms of acid management by preventive method, the application of cover system is limited to several large-scale mines, due to some obstacles such as the budget constraints as well as to a lesser extent, the lack of knowledge and awareness, in addition to the limited capability of the equipment. Up to now, the most comprehensive and documented effort on the implementa- tion of dry cover system in Indonesian coal mine is the one conducted by PT Kaltim Prima Coal (KPC). For example, in the exploration stage, the geochemical characterisation (by Net Acid Generation (NAG) test) of each rock to be mined was conducted to construct the spatial geochemical distribution of the potentially acid-forming (PAF) rocks as well as non-acid-forming (NAF) rocks. The constructed block model is then used as guidance for mine planning to develop day-to-day overburden excavation and dumping strategy to achieve selective dumping that has been planned to prevent AMD generation by encapsulating the PAF with the NAF rock [19]. Moreover, an effort with attempt to verify/confirm the exact type of overburden (PAF or NAF) before being dumped of blasted materials in the mining block was also conducted by sampling the blast hole cutting [20]. The encapsulation scenario depends on the available composition of PAF and NAF material in the mining sequence. When the NAF material is available, as much as 10 m to 20 m of cover layer thickness was placed due to less cost and better geotechnical stability, instead of compaction of 2 m of the NAF layer or 1 m clay layer when the NAF material was insufficient (see Figure 1a and b), because of its potential to failure due to the desiccation of the material. The application of this cover system, coupled with the liming during the construction stage, has been proven that it could give good results of water effluent [21] ‘The PT KPC is such a large-scale mining operation that the construction of the cover layer up to 10 m thickness is usually not a big problem if the NAF material is available. However, for small mines which have limited equipment capability, the task of constructing the 10 m cover would be very difficult, Even at the PT KPC mine, it has been reported that there were potential shortages of the NAF material at the end of the mine operation. To deal with these problems, the research on mixing the NAF and PAF material to produce relatively inert cover material has been conducted. Preliminary results by using a column leach test under field conditions revealed that a site-specific mixture up to 30% PAF material was still able to provide a neutral leaching environment [22] for 1 year after the start of the monitoring programme. However, this method is not easy to be implemented on the field scale due to the lack of adequate blending techniques available, which can provide homogenous blending performance by using the truck and shovel method. The mixing and spreading of the material using a belt conveyor yields better results [23] ‘Therefore, research on the placement alternative of covering methods to produce an inert cover layer material that is more reliable to be constructed economically by smaller equipment is needed. 4. Laboratory study A laboratory study using the column test which involved the layering of NAF and PAF rock in order to evaluate the effectiveness of the cover system was conducted. The results of the study were expected to be considered one of the dump cover system alternatives.19 © =Soil:1m % =NAFLayer;10-20m- =PAFwasterock (a) li = Compacted Clay {1:m)/NAF(2m}Layer AN = PAF wasterock (b) Figure 1. Encapsulation scenario being implemented at KPC coal mine. 4.1. Materials and methods 4.1.1. Sample and geochemical characterisation In this study, two mudstone samples (T3 and Q10) were taken from a coal mine, PT Berau Coal, which was located in the Berau Regency, East Kalimantan Province, Indonesia. ‘A powder pattern X-ray diffraction (XRD) analysis using a Rigaku RINT 2000 X-ray diffractometer indicated that both T3 and Q10 rock samples consisted of mineral quartz (SiO), illite (Ko Al3 (Si,AI) 40 o(OH)s » nH30), kaolinite (AlpSi0 (OH),), albite Pyrite (FeS;) and dolomite (CaMg(CO3)>. The siderite (FeCO,) and calcite (CaSO,) were detected only on T3 sample. According to the XRD result, pyrite, dolomite, siderite and calcite are the minerals that significantly took part in the fast dissolution process since quartz, illite, kaolinite and albite have low reactivity and cation exchange capacity.80 The elemental composition obtained from X-ray fluorescence (XRF) analysis by Rigaku RIX 3100 XRF spectrometer for major element is presented in Table 1. It shows that the samples are enriched predominantly with silica (SiO,), alumina (Al:03), iron oxide (Fe,03), MgO, KO and sulphur, in addition to small amounts of CaO, NaxO, and MnO. Both samples have lower neutralising capacity of carbonate mineral than the silicate mineral, which is indicated by the lower content of CaO than MgO and Al;O, contents. By comparing the total alkali content in both the samples, it was revealed that the QO sample had quite a higher amount of alkali clement. Furthermore, the total sulphur content of T3 was higher than that of the Q10 sample which was 1.92% and 0.36%, respectively Static geochemical test analysis was conducted in order to determine the base composition of samples related to the acid and neutralising production potential. The analysis consists of acid-neutralising capacity (ANC) test [24] and also net acid generating (NAG) test [25]. ANC reflects the inherent acid buffering of sample with regard to acid neutralising minerals contained within the sample. The NAG test reflects the acid forming potential of sulphidic samples with contents of readily available neutralising minerals indicated by the NAG pH value. The results of geochemical analysis of samples are shown in Table 2. ‘The ANC of T3 and QIO samples were —9.6 kg H2SO,/ton and 46.06 H2SOx/ ton, respectively. The result of negative value meant that the T3 sample does not have neutralising capacity while the Q10 sample has quite high neutralising capacity. As the net acidity potential production (NAPP) is the theoretical calculation which represents the balance between the capacity of sample to generate acid (based on maximum total sulphur content subtracts by ANC), it can be said that the T3 sample Table 1. Major bulk elemental composition of sample. Mineral 13 SiO 57.66 TiO 0.87 ALOs 16.65 FeO; 5.67 MnO 0.06 MgO 215 CaO 0.42 Na,O 0.35 K,0 2.13 P20s 0.09 H,0+Lol 1191 Ss 192 Table 2. Geochemical characteristics of the samples. Sample MPA* ANC NAPP* NAG-pH NPR ‘Type Quo 10.99 46.06 —35.07 7.42 4.19 NAF T3 58.82 —9.604 68.43 261 0.16 PAF Note: Calculation: Maximum Potential Acid (MPA) = 30.6 x %Total Sulfur (S); Net Acid Producing Potential (NAP) = MPA-ANC; Neutralization Potential Ratio (NPR) = ANC/MPA. Classification: NAF = NAGpu > 4; NAPP <0; NPR > 4; PAF = NAG u < 4; NAPP > 0; NPR < 2. *KgH,SO,/ton81 has the capacity to produce as much as 68.43 kg H>SO,/ton while the Q10 sample has the capacity to neutralise acid up to 35.07 kg HaSOq/ton. Considering this result, it was confirmed that the Q10 sample was classified as an NAF material while the T3 sample as a PAF material. It was strengthened by the classification based on NAG pH [25] and the neutralising potential ratio (NPR) [26]. The NAG pH and NPR for T3 samples were 7.42 and 4.19, while for QUO samples are 2.61 and —0.16, respectively. 4.1.2. Column test The column system was conducted by using a plastic tube which was 45 mm in diameter and 140 mm in height (Figure 2). A total amount of 100 g of sample with a grain size diameter range of 1-2 mm was used on each column. Considering the static test result, several layering scenarios of PAF-NAF material were performed on eight columns (see Figure 3). In addition, compaction up to 20% volume spot lamp |, Rock sample (100 g) Filter (501m) Glass bed Filter (30m) Figure 2. Column apparatus. wa) wi) wid tH) ) Figure 3. Layering configuration of each column,82 decreasing was conducted in the lower NAF layer in some configuration in order to accomodate the circumtances in the field. The rock sample was placed on the column and 100 mL of distilled water was poured and drained fully until the next day for leachate sampling. Afier drying naturally for two days and a further 6 hours of drying by light in the fourth and fifth day, the water pouring process for the next eycle is repeated. The detailed procedure of column testing on each cycle is shown in Table 3. ‘The effluents from the columns following flushing were collected and sampled, which were then measured for pH as well as major anion (SO, ) and cation (Ca?* and Mg**) after filtering through 0.45 mm filters. pH and anion-cation were measured by using the TOA-DK pH meter HM-2IP series and the Dionex ICS-90 Jon Chromatography, respectively. 4 As the leachate quality is influenced by the total elemental content in the column, it is important to consider the static geochemical balance due to the layering configuration scenario of column test. It is conducted by calculating the acid-base composition, based on static acid base accounting, proportional to the amount and ratio of the sample (Table 4). As a control, two approaches are used in the calculation of the static acid-base equilibrium, i.e. based on NAPP and NPR. However, based on NPR criteria, all multi-layer columns revealed to be acid producing since the values are lower than 2. Results and discussion Table 3. Column test scenario. Daily activity of each n cycle Day 1 Water pouring a Day 2 Leachate sampling 2 Day 3 Natural drying o Day 4 6 hours drying by light Day 5 6 hours drying by light Table 4. Static geochemical MPA ANC NAPP Column No. T3 QIO (KgHsSO,/ton) (KgH;SO,/ton) (KgH,SO,/ton) NPR i I 10.99) 46.06 ~35.07 4.19 1 58.83 9.60 68.43 0.16 02 08 20.56 34.93 ~1437 1.70 02 08 20.56 34.93 = 14.37 170 05 05 34.91 18.23 16.68, 0.52 05 05 34.91 18.23 16.68 0.52 02 08 20.56 34.93 ~1437 170 0.2 08 20.56 34.93 -1437 1.70 i 04 06 30.13 23.79 6.33 0.79 x 04 06 30.13 23.79 6.33 0.7983 The results of control column test were shown in Figure 4. The leachate of single NAF layer (column i), like the static test result, has neutral pH with consistently low content of sulphate as well as calcium and magnesium. On the other hand, the pH of the single PAF layer leachate (column ii) is low (around 3) with high content of sulphate (> 30 mmol/L) and magnesium (> 10 mmol/L), while the calcium content is low along the experiment. It is revealed that oxidation has occurred upon the sulphide mineral which was exposed due to the degradation of rock sample as the result of drying and wetting cycles. On the double layer simulation (see Figure 5), the effect of PAF placement was evaluated. When a 20% PAF layer is placed over the NAF and directly made contact to the surface environment (column iii), PAF material will easily oxidise, Hence the leachate was continuously infiltrated along the NAF layer underneath with limited capacity to neutralise due to a short contact time; this resulted in a low pH of around 4 up to the sixth cycle. This is also the reason for why the concentration of Dw. w Doo 10 8 2 § g) som -e-anne-me en] 9 E 8 Ew : £2 soe 6 ~e 0% x 6 S80 sz 8B wo. SB gre : gre : a x» ; g » a i i 10 i = 2 2 8 ° Cycle Cycle Figure 4. Column test of control column; pH and anion-cation of the leachate. i 4 10 oo 10 Bol a Sta-encae-e-Mazewe- ph [os § ag] ASK = 8 Cate o—maze emp Lo Eo 1 Eo. hee a se 2 33 wo1— ——- a sz S80 sk gE ol = = © BE» 4 a 19° 3 Boe i ° i 0. o 2 4 o oye Yoo o a ~~ S042: 0 - Cade — 6 Mga pH ‘Anion/Cation Concentration mmol/L eS88sesss - pH Figure 5. Column test result of double layer; pH and anion-cation of the leachate,84 sulphate and magnesium was relatively erratic, A similar condition was revealed on the higher PAF composition (50% PAF) in the column v. However, when compaction was performed to the NAF layer underneath, pH was increased while the sulphate content was decreased after cycle 3. Although the mechanism that influenced the sulphate decrease is not yet evaluated, it is supposed to be controlled by a secondary mineral precipitation, since the low permeability of the NAF layer will give more contact time and a more uniform flow to react with the alkaline minerals contained in the NAF layer. The pH of the column of NAF over PAF layer (column iv), was increased gradually from pH 3.05 at the first cycle to pH 6.5 at the seventh cycle. It was in line with the significant decrease of sulphate content in the leachate, from more than 40 mmol/L in the first cycle to less than 10 mmol/L in the second cycle, meaning that the oxidation of sulphide was significantly inhibited. Triple layer system was performed in such a way where the PAF material was sandwiched between two NAF layers; the column result was shown in Figure 6. The sulphate content of columns vii and viii, which contained 20% of PAF sample, was decreased from more than 26.7 mmol/L to a value less than 3 mmol/L at the sixth cycle. Similarly, the sulphate content of the leachate from column ix and x, which contained 40% PAF material, was decreased from more than 15 mmol/L to around 6 mmol/L. However, due to the alkaline content in the material, the pH value of the column vii and viii were gradually increased from 4.5 to almost 6.1 on the third cycle, consequently. The pH value continued to increase up to 7 in the last cycle for column viii which experienced compaction at the bottom NAF layer, while the pH value remained steady at value 6 for column vii. Significant decrease of sulphate content after the fifth cycle is supposed to be responsible for the continued increase of pH in leachate, in addition to the compaction effect. Similar trends were observed in vii) viii) re 10 ” a i Eo» 2 § ep wwoece ag oman] ia ihe £ ow. s #gs ‘ 8300. sz S80 sé BF» 4 gk» ‘ 2. x 2 A re 14% ‘ in » 0. 20 £ on. 2 § z ; §e ‘0. = <0 ay: a ee tt, bee 5 a BE ay 8 2 8 x i i gw = o =o. cycle Figure 6. Column test result of triple layer; pH and anion-cation of the leachate.85 columns ix and column x which contain 40% of the PAF sample. The pH values of those last two columns were slightly lower than in the range of 5-6 as the consequences of higher PAF content. Compared to the uncovered scenario of PAF material in the column i, the sulphate concentration on these scenarios was lower, meaning that the existence of the upper layer of NAF had succeeded to inhibit the sulphide oxidation. In general, the presence of NAF rocks on the column has an effect on the increase of pH in the leachate. This was indicated by a higher pH value in the multi-layer column compared with the pH values in the single column. Sulphate concentration in the leachate is also comparable with the composition of PAF in each column, especially at the initial cycle. However, calcium and magnesium concentrations were higher in the column with higher composition of PAF. This is most likely caused by the dissolution process of calcite and dolomite as well as silicate which are more intense at a lower pH 27]. In the further cycle, the interaction of physical and geochemical properties was dominantly controlled the leachate quality rather than the geochemical composition itself. Physical weathering, initially induced by wet and dry cycle conditions, may occur in the upper layer, resulting in the decrease of the particle size of the rock matrix [28]. Consequently, the total reactive surface area will increase, which may accelerate the chemical process — both the oxidation rate of sulphide and the neutralising reaction [29]. On the other hand, a fine grain material resulting from weathering will fill the pore space and subsequently decrease the permeability. In addition, the existence of fine grain in the pore space potentially armours the previously available reactive surface area, which may reduce the weathering rate in advance [30]. Meanwhile, the decreased weathering rate means decreasing oxidation, which was reflected by the low content of sulphate as revealed in the column test. 4.3. The proposed new alternative covering system In the management of AMD, the lack of NAF-materials becomes an important concern. NAF material, in spite of being used to construct cover material over the waste dump area, is also used as mine infrastructure construction material such as the roads and drainage system. In other words, the NAF-material is expected to be laid over the overall surface area, which is going to be rehabilitated with an attempt to minimise the AMD load. Therefore, a proper management planning of the used NAF-materials is required in order to make sure that all areas which are in need of rehabilitation can get sufficient NAF-material for cover purpose. Based on the experimental results of this study, the concept of the new covering technique with a multi-layer system (Figure 7a) which takes the rock geochemical composition into consideration as an alternative to the current common practice of cover layer (Figure 7b) in coal mines, is developed. This developed method is proposed as a form of compromise in dealing with problems of limited NAF- material availability and the achieved cover performance. The proposed triple layer cover, without losing its function, has the following several advantages compared the conventional single layer NAF cover: (1) reduces the need of NAF layer, and (2) can be done with smaller capacities of mining equipment. In the scenario of triple layer cover, the upper NAF layer serves as a protective layer for the underlying layers from the threat of erosion and as an oxygen and water86 @) Figure 7. (a) Proposed scenario of new cover system. (b) Conventional cover system, infiltration limiter, especially for the PAF layer underneath, in addition to its function to support above topsoil for vegetation purpose. Thus itis expected that the oxidation of sulphide minerals and leachate product from the PAF layer would be limited. Furthermore, the PAF layer, in addition to the main function as a replacement material to reduce the requirement of NAF material, also functions as an oxygen consuming layer pursuant to the presence of oxidation reaction in this layer. Indirectly, this reaction will restrict further diffusion of oxygen into the lower layer, especially for the PAF waste rock at the bottom, and hence prevents its further oxidation. The second NAF layer, with its inherent ANC, serves as a source of neutralising alkali to buffer the leachate pH greater than 4.5 before further deeper percolation to the PAF waste rock. The pH value in the leachate as low as 4.5 is suggested to be used as an acceptable cut off since this value is considered to be a pH level that can distinguish among acid and near neutral drainage as a consequence of AMD. Moreover, this pH is also the level where the metal content such as dissolved Fe begins to precipitate to form ferric hydroxide [6] that may affect the decreasing oxidation rate reactions due to coating effect in the reactive surface. In addition, it is also the pH region in which the buffering capacity is dominated by the abundant silicate mineral, as shown by [31] who evaluate the non-carbonate neutralisation potential of the waste rock sample. ‘The similar result was found by [32] who evaluate the long-term kinetic test data of hard rock and coal mine. The evaluation shows that silicate mineral took an important role in the neutralisation after the carbonate mineral exhausted and maintained the pH over 4. In addition, pH 4.5 is a common cut off on determining the classified PAF waste rock by NAG test. Considering those suggested cut offs, the treatment of any leachate in the waste rock dump may still be needed in the initial construction period to some extent to meet the mandatory effluent threshold at near-neutral pH. However, the neutralization needed is expected to be a minor effort since the high potential of in-situ neutralisation as shown in the result of this study.87 5. Conclusion In terms of preventive management of AMD, dry cover system has been implemented on waste rock dump areas in several Indonesian large-scale coal mines with satisfactory results. However, there are some problems being faced with regard to the availability of NAF waste rock as cover material caused by nature of geological conditions and mining sequences. In this research, simulation of the multi-layer covering system by using the PAF and NAF waste rock was performed as another alternative method to substitute the single NAF waste rock layer cover that had been applied, in addition to the alternative of PAF and NAF waste rock blending. ‘The results of this study show that the construction cover system of triple layer cover, which consists of a certain amount of PAF layer and considering the geochemical equilibrium condition at each layer, has been resulting in similar performance with the single layer of material NAF, However, a scale-up simulation to field condition is necessary to investigate the influence of the external conditions, such as local climate, to the performance of cover scenario. Acknowledgements ‘The authors wish to extend their gratitude to PT Berau Coal and PT Kaltim Prima Coal in Indonesia for the acceptance of visiting mine sites and their cooperation in providing the samples and materials used in this study. Any opinions and comments expressed in the study are those of the authors’ themselves and not necessary those of the institutions and the mine. References [I] J. Skousen, A. Rose, G. Geidel, J. Foreman, R. Evans, W. Hellier, and W. Virginia, Handbook of Technologies for Avoidance and Remediation of Acid Mine Drainage, The National Mine Land Reclamation Center, West Virginia University, Morgantown, West Virginia, 1998, p. 131 [2] P.A. Weber, W.A. Stewart, W.M. Skinner, C.G. Weisener, J.E. Thomas, and R. St. C. Smart, Geochemical effects of oxidation products and framboidal pyrite oxidation in acid mine drainage prediction techniques, Appl. Geochem. 19 (2004), pp. 1953-1974 [3] J.E. Thomas, R. St. C. Smart, and W.M. Skinner, Kinetic factors for oxidative and non- oxidative dissolution of iron sulfides, Miner. Eng. 13 (2000), pp. 1149-1159, [4 K. Sasaki, M. Tsunekawa, T. Ohtsuka, and H. Konno, The role of sulfur-oxidizing bacteria Thiobacillus thiooxidans in pyrite weathering, Colloids Surf. A Physicochem. Eng. Asp. 133 (1989), pp. 269-278. [5] W. Stumm and J.J. Morgan, Aquatic Chemistry: An Introduction Emphasizing Chemical Equilibria in Natural Waters, 3rd ed., John Wiley and Sons, New York, 1996. [6] R-V. Nicholson, R.W. Gillham, and E.J. Reardon, Pyrite oxidation in carbonate-buffered solution: 2. Rate control by oxide coatings, Geochim. Cosmochim. Acta. 54 (1990), pp. 395-402, [7] ES. Ribeiro, J.W.V. Mello, W.A.P. Abrahao, L.E. Dias, and W.L. Daniels, Oxidation rate of pyrite from Brazilian coal as affected by CaCOs, oxygenation, and erystal size. Int. J. Min, Reclamat. Environ. 18 (2004), pp. 289-298. [8] D.M.C. Huminicki and J.D. Rimstidt, fron oxyhydroxide coating of pyrite for acid mine drainage control, Appl. Geochem. 24 (2009), pp. 1626-1634 [9] D.W. Blowes, C.J. Ptacek, J.L. Jambor, and C.G. Weisen Mine Drainage, Heinrich D. Holland and Karl K. Turekia on Geochemistry, Pergamon, Oxford, 2003, pp. 149-204. [10] B. Dold, C. Wade, and L. Fontbote, Water management for acid mine drainage control at the polymetallic Zn-Pb-( Ag-Bi-Cu) deposit of Cerro de Pasco, Peru. J. Geochem. Explor 100 (2009), pp. 133-141. The Geochemistry of Acid Editor(s)-in-Chief, Treatise88 [11] A.D. Paktune, Mineralogical constraints on the determination of neutralization potential and prediction of acid mine drainage, Environ. Geol. 39 (1999), pp. 103-112. [12] K.B.C. Brady, E.F. Perry, RL. Beam, D.C. Bisko, M.D. Gardner, and J.M. Tarantino, Evaluation of acid-base accounting to predict the quality of drainage at surface coal mines in Pennsylvania, USA, International Land Reclamation and Mine Drainage Conf., Vol 1 USBM SP-06A-94. Pittsburgh, PA, 1994, pp. 138-147. [13] RS. Gautama and GJ. Kusuma, Evaluation of Geochemical Tests in Predicting Acid Mine Drainage Potential in Coal Surface Mine, N. Rapantova, and Z. Hrkal, eds., Mine Water and the Environment — Paper #48; Ostrava (VSB — Technical University of Ostrava). Available at http://imwa.info [14] A. Mac, G. Robertson, and J.P. Barton-Briges, Management option for the short and long- term control of acid mine drainage Vol. Il. Proceedings from AMIC Environmental workshop, Darwin, Australia, 1988. [15] G.W. Wilson, Why are we still struggling with ARD?, 6th Australian Workshop on Acid and Metalliferous Drainage, Burnie, Tasmania, 2008a. [16] International Network for Acid Prevention (INAP), Evaluation of the long-term performance of dry cover systems-final report No, 684-02 (2003). Prepared by O'Kane ‘Consultant Inc [17] S. Kramadibrata, K. Matsui, M.A. Rai, and H. Shimada, Properties of soft rock with particular reference to Indonesian mining conditions. Proceedings of the ISRM Symposium — 2nd Asian Rock Mechanics Symposium, Beijing, 2001 [18] K. Matsui, H. Shimada, L.A. Sadisun, S. Kramadibrata, and I. Arif, Development and rehabilitation of open cut coal mines in Indonesia. Proceedings of Symposium on Environmental Issues and Waste Management in Energy and Mineral Production, Cagliari, aly, 2002, pp. 581-586, [19] C. Nugraha and H. Agoessusilo, Management of acid mine drainage with cover systems method at PT. Kaltim Prima Coal. 2nd National Workshop of Acid Mine Drainage, Institut Teknologi Bandung, Bandung, Indonesia (in Bahasa Indonesia), 2004. [20] RS. Gautama and S. Hartaji, Improving the accuracy of geochemical rock modelling for acid rock drainage prevention in coal mine, Mine Water Environ, 23 (2004), pp. 100-104. 21) C. Nugraha, I. Manege, T. Sasaoka, H. Shimada, M. Ichinose, and I. Matsui, Acid rock drainage management at PT. Kaltim Prima Coal and its relationship with water quality regulations in Indonesia, Proceedings of International Symposium on 16th Mine Planning, and Equipment Selection and Environment Issues and Waste Management in Energy and Mineral Production, CD-ROM, Bangkok, Thailand, 2007, [22] C. Nugraha, H. Shimada, T. Sasaoka, M. Ichinose, K. Matsui, E.V. Tular, and 1 Manege, Mixed acid and non-acid waste rocks associated with acid mine drainage generation: A preliminary result of colunm test experiment. Proceedings of the 2nd International Symposium on Novel Carbon Resource Science, Bandung, Indonesia, 2009, pp. IV-12-1V-18, [23] S. Miller, Y. Rusdinar, R. Smart, J. Adrina, and D. Richards, Design and construction of limestone blended waste rock dumps — Lessons learned from a 10-year study at Grasherg, Proceedings of the 7th International Conference on Acid Rock Drainage, St Louis, Missouri, 2006, pp. 1287-1301 [24] A.A. Sobek, W.A. Schuller, J.R. Freeman, and RM. Smith, Field and Laboratory ‘Methods Applicable to Overburdens and Mine soits, US Environmental Protection Agency, Cincinnati, Ohio, 1978, pp. 47-50, EPA-600/2-78-054, [25] S. Miller, A. Robertson, and T. Donahue, Advances in acid drainage prediction using the net acid generation (NAG) test. Proceedings of the 4th International Conference on Acid Rock Drainage, Vancouver, BC, 1997, pp. 533-549. [26] British Columbia AMD Task Force, Draft acid rock drainage technical guide, British Columbia Acid Mine Drainage Task Force Report, by Steffen Robertson and Kirsten (B.C.), Norecol Environmental Consultants, and Gormely Process Engineer- ing, British Columbia Ministry of Energy, Mines and Petroleum Resources, Canada, 1989. [27] EJ. Sherlock, R.W. Lawrence, and R. Poulin, On the neutralization of acid rock drainage ‘by carbonate and silicate minerals, Environ. Geol. 25 (1995), pp. 43-54.89 [28] P.W. Birkeland, Soils and Geomorphology, Oxford Universit 1999, p. 430. [29] G.B. Davis and A.LM. Ritchie, A model of oxidation in pyrite mine waste: part 3: importance of particle size distribution, App! Math Model. 11 (1987), pp. 417-422. (30) C. Nugraha, H. Shimada, T. Sasaoka, M. Ichinose, K. Matsui, and I. Manege, Geochemistry of waste rock at dumping area, Int. J. Min. Reclamat. Environ. 23 (2009), pp. 132-143. [31] B. Mattson, Assessing the availability and source of non-carbonate neutralization potential by pre-treatment of kinetic test samples. Proceedings of the 8th International Conference ‘on Acid Rock Drainage (8 ICARD), Skellefted, Sweden, 2009. Available at http://www. proceedings-stfandicard-2009.com) [32] S.D. Miller, W.S. Stewart, Y. Rusdinar, R.E. Schumann, J.M. Ciccarelli, J. Li, and RC. Smart, Methods for estimation of long-term non-carbonate neutralization of acid rock drainage, Sci. Environ, 408 (2010), pp. 2129-2135. Press, Inc., New York,