Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation USE OF LITHOLOGIC DESCRIPTIONS FOR WASTE ROCK CHARACTERIZATION CASE STUDIES F-ROIVI THE ESKAY CREEK PROJECT C. J. Stewart, Ministry of Environment, Lands and Parks, Smithers T. W. Higgs, T.W. Higgs Associates Ltd., Vancouver W. A. Napier, Homestake Canada Inc., Vancouver Abstract This paper outlines the program conducted at Eskay Creek to assess acid generation potential with a focus on the use of lithology and geochemical tests for classification and project planning. The paper discusses variation within rock types and in particular differences in behaviour of massive and brecciated forms of Rhyolite. A primary recommendation is that geological site-specific information be merged with the geochemical sampling, analyses and interpretation work. Other recommendations for ARD programs for new projects are also included. Introduction This paper describes the process used to develop an Acid Rock Drainage (ARD) program for the evaluation and characterization of waste rock and ore for the ESKAY Creek Project, which is operated by Prime Resources Group Inc. The project involves mining of a small complex gold/silver ore body and shipment of raw ore to smelters located in Canada and abroad. The mine is located approximately 83 air km north of Stewart, B.C. and 465 km north of Smithers, B.C. as shown on Figure 1. Figure 1 Location Map - Eskay Creek Proj ect 62 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation The known probable mining reserves are estimated to include 1.08 million tonnes at 65.5 grams per tonne gold and 2,930 grams per tonne silver. The Eskay Creek deposit, as currently known, contains 2.2 million ounces of gold and 102 million ounces of silver. Underground mining of the ore body will be completed by a drift and fill method. During the life of the mine approximately 380,000 tonnes of waste rock will be generated. In comparison to most mining operations with a 10-year life this is a relatively small amount of waste rock. Program Development At the early stages of the project, it was recognized that waste rock characterization would be an important component of the Mine Development Assessment Process. As a consequence, after submission of the initial prospectus in April 1990, the Mine Development Assessment Committee suggested that the Company consider forming a committee to oversee the ARD aspects of the project. In 1990, Prime Resources Group, its consultants, and members of the Mine Development Review Committee formed an ARD sub-committee to characterize the potential waste rock and ore to be mined at Eskay Creek in terms of acid generation characteristics. A key objective of the sub-committee was the development of appropriate abatement plans that would prevent and control acid generation. The program design was developed by the consultant and distributed to the committee. A meeting, either face to face, or via teleconference call, was then held to fine-tune the draft and develop the scope of work. As data became available, the sub-committee reviewed it and made comments on further investigations. Communal review and interpretation, combined with the shared experience of the committee, was particularly useful in the design of each segment of the test program. The ARD program evolved into five interrelated phases as summarized in Table 1. Table 1. ARD Characterization Program: Summary Phase 1 Phase 3 - Static testing - Ore characterization - Kinetic testing - Waste Pad testing Phase 4 - Neutralization capacity - Waste Rock pile leachate characterization Phase 2 - Drift static testing Phase 5 - Rhyolite characterization - In-situ waste rock disposal monitoring - Leachate mobility 63 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation Geological Overview The Eskay Creek mineral deposit is classified as a volcanogenic massive sulphide deposit which has been enriched with silver and gold. The ore body is hosted within a calc-alkaline island arc volcanic complex of early to middle Jurassic Hazelton Group, which is approximately 180 million years old. The Eskay Creek stratiform mineralization is hosted in the Salmon River Formation within a mudstone and breccia unit separating basal Rhyolites and overlying Andesitic volcanics. For the purposes of this discussion, the deposit geology can be simplified into three broad units: ƒ Hangingwall (H/W) Andesite ƒ Mudstone (Hosts ore body) ƒ Footwall (F/W) Rhyolite A drawing illustrating stratigraphy and the lithologie units of the 21 B Zone is provided in Figure 2. Figure 2 Stratigraphie Column - Eskay Creek Deposit 64 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation The andesitic volcanic unit varies from massive pillowed flows to hyaloclastites and sills and also contains argillaceous sedimentary units. The Andésite Unit lies stratigraphically above the ore zone. The stratiform mineralization of the Eskay Creek ore deposit occurs in the contact mudstone and breccia between the basal Rhyolite and the overlying Andesitic volcanics. The breccia contains mudstone, altered felsic volcanic fragments and less common sulphide fragments hosted in an often chaotic muddy matrix. F/W Rhyolite is variably massive, flow banded and brecciated. The "rhyolite breccia" or "brecciated rhyolite" is a laterally extensive, chloritized and serialized fragmentai unit underlying much of the ore zone. This altered fragmentai zone may represent a fluid recharge or mineralizing discharge vent zone for the ore deposit. The more massive felsic volcanic below and lateral to the rhyolite breccia are generally pervasively silicified or altered to potassic feldspar. Dominant alteration minerals in the deposit area include sericite, chlorite, quartz, calcite and lesser dolomite, potassic feldspar, graphite and barite. The Hangingwall Lens is a relatively thin zone hosted in a mudstone layer lying stratigraphically 5 to 40m above the ore zone which is a massive sulphide zone with significantly higher levels of pyrite and copper as chalcopyrite. The 109 zone lies in the F/W Rhyolite beneath the contact mudstone. Mineralization is commonly coarse-grained sphalerite, pyrite, galena and visible gold in milky quartz veins or stockwork, or disseminated in silicified and pyrite altered felsic volcanics. The vast majority of the waste rock will be from the F/W Rhyolite unit. Based on the results of the ARD assessment for this property, waste rock storage will be subaqueous. To properly assess the acid generating potential of a given ore deposit, a complete geological characterization is required. Mineralogy, alteration assemblages, depositional history, structure, weathering characteristics and component spatial relationships etc. can contribute significant information to the assessment of ARD potential. Combining this information with geochemical data from static and kinetic tests, and mine plans can provide a useful tool for project development. Eskay Creek ARD Test Program As described under Program Development, an assessment of ARD potential at Eskay Creek began with the creation of the ARD sub-committee. An Interim Report (1) summarized the comprehensive static and kinetic ARD test results for the main geological units of the ore zone and formed the framework to further define the potential for ARD at the Eskay Creek minesite. A lithologie description defines the physical characteristics of a rock, particularly in hand specimens and outcrops. Features such as colour, mineralogy, alteration, weatherability, grain size, structure and other 1 Contained in Final Report. "Eskay Creek Project, Acid Generation Characteristic of Waste Rock and Ore" June, 1993, by T.W. Higgs Associates Ltd., Vancouver, B.C. 65 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation physical features would be included. The geologist is extremely well-equipped to provide this information from the early exploration phase of a program. The information, routinely collected by geologists during exploration, should be incorporated into the ARD prediction process. The Committee used a holistic approach which incorporated these aspects of geology and mining: ƒ Geological identification and description of sample. ƒ Mineralogy. ƒ Alteration assemblages. ƒ Weathering potential. ƒ Sample location and relationship to the ore body. ƒ Static and kinetic test results. ƒ NP to MPA Ratios ƒ Ore geometry and depositional history ƒ Mine development plan. By considering this information, conclusions were arrived at which influenced the final mine plan and focused resources on specific key areas of concern. In this instance, the focus became F/W Rhyolite. To illustrate the importance of evaluating a wide range of factors in ARD prediction, the first case study presented is H/W Andesite. Case Study 1 - H/W Andesite The Interim Report concluded that the H/W Andesite had an overall positive net neutralizing potential and that it could ultimately be considered for infrastructure construction purposes and/or as an ARD mitigation medium. As discussed earlier, the Andésite Unit lies stratigraphically above the ore zone, however various forms of mineralization such as massive sulphide horizons, disseminations and stockwork zones are encountered within it. During Phase 1, ABA analyses were conducted on 60 samples consisting of one metre sections of drill core from a variety of drill intersections in the Andésite Unit. In addition, four kinetic tests were conducted using H/W Andesite - two humidity cells and two column tests using composite samples. A frequency distribution plot of Net NP for the Andesite ABA samples is provided in Figure 3. This figure illustrates the central tendency of this rock type in the positive Net NP range but also demonstrates that a significant percentage of the samples (42%) were theoretically acid generating. 66 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation The results of the static and kinetic tests on Andésite can be summarized as follows. ƒ Vertical/spatial distribution "indicated a strong trend between depth and Net NP. Net NP's for Andesite were highest at surface decreasing to negative values in samples collected directly above the Contact Zone". ƒ Kinetic tests confirmed both the acid generating (worst case condition) and acid neutralizing (mean condition) aspects of the Andesite unit. ƒ The results for Andesite "Worst Case" Composite #1 confirmed the ABA theoretical prediction in that sulfide present was reactive and acid generation rate exceeded neutralization capacity. Leachate pH was low, reaching pH 4.0 at Week 17. Composite #1 had a Net NP of -38.4 kg CaCO3/tonne and an NP/MPA of 0.38. ƒ The results for Andesite "Mean Case" Composite #2 confirmed the ABA prediction in that its capacity to neutralize acid exceeded its acid generation rate. The humidity cell for Composite #2 did not generate any appreciable acid. pH remained neutral at pH 7.4 to 7.8. Composite #2 had a Net NP of +91.9 kg CaCO3/tonne and an NP/MPA ratio of 8.76 Although Andésite was identified as an overall acid consumer, with 72% exceeding an NP to MPA ratio of 3:1, portions of this rock type, particularly near the ore zone, were theoretically acid generating (NP/MPA <1:1). The next step was to plot the ABA results on sections through the property and relate the information to mine development scenarios. 67 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation Drill holes were chosen from 4 different sections along the south/north trend of the ore body and profiles were roughly drawn illustrating the sample depth, rock type, NP:MPA ratio, %S and whether or not it was potentially acid generating. This data was then interpreted in conjunction with deposit geometry, mode of deposition and structure. The interpretation indicated a general trend of increasing potential acid generation in the Andésite unit from south to north and confirmed that the Andésite proximal to the ore body did not contain the expected buffering capacity but, rather, would potentially contribute to ARD development. The final step was to relate the proposed mine plan to the ARD potential of the material which would actually be encountered during development. It was apparent that the amount of potentially acid generating or consuming Andésite which would be handled during the mine life would be directly related to the mining plan. By this process it was determined that: ƒ Underground development would primarily encounter and expose, altered, low NP:MPA Andesite that could contribute to ARD, and not unaltered, potentially neutralizing Andesite. ƒ Based on the current mine plan, acid consuming Andesite would therefore not be available for pad/infrastructure construction and mitigation of ARD unless it was specifically quarried. The results of this assessment had implications in terms of mine development costs. As this issue was addressed early, the Company was able to react in a cost effective manner. Case Study 2 - Massive and Brecciated Rhyolite The Rhyolite Unit comprises the dominant footwall material to the ore deposit. Samples of rhyolite were collected during three different periods in the program - drill core (Phase 1), exploration adit (Phase 2) and underground (Phase 2). The ABA results for all rhyolite samples are summarized in Table 2. 68 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation The conclusions from the static tests on drill core were that Rhyolite was a theoretical acid producer. Static tests indicated that the majority of the drill core Rhyolite samples would potentially generate acid with 88% less than 1:1 NP to MPA ratio. Similarly the Adit samples were classified as theoretical acid producers with 64% less than 1:1 NP to MPA ratio. However the humidity cell tests on drill core concluded that the "worst case" and "mean case" composites "did not contain an appreciable amount of reactive sulphur and therefore did not generate significant amounts of sulphate." These results conflicted with observations from the static tests. Therefore, additional evaluation work was required since the mine plan indicated that Rhyolite would comprise the vast majority of waste rock materials during mining. In addition, it was observed during development of the adit that much of the waste rock was brecciated and highly sericitised. The angular rhyolitic fragments were supported in a very fine grained dark gray matrix with pyrite throughout. Brecciated rhyolite was highly altered, fractured and deemed susceptible to accelerated weathering. This fracturing would increase exposure of the sulphides to water and oxygen. By comparison, the massive silicified Rhyolite would not be expected to weather as quickly. This was confirmed in the field within one year of exposure when it was observed that brecciated rhyolite was extensively weathered (crumbling) while massive rhyolite was significantly more competent. As with the Andesite, it was important to consider a variety of parameters in assessing the conclusions derived from the geochemical work. In this instance, the most important were the physical and chemical alteration of the Rhyolite and its location relative to the ore body. Composition of the Rhyolite unit varied from being silicified and very competent to very weak where intense sericitic alteration was prevalent. It would be expected that highly silicified Rhyolite would oxidize slowly whereas the weak, brecciated sericitic Rhyolite would be conducive to rapid physical breakdown of the matrix and subsequent exposure/oxidation of the sulphides. Phase 2 of the program provided information on underground sampling which included sample locations, Net NP and rock type. When compared with the underground geology it was noted that the: ƒ Outline of the ore body appears to correspond with the brecciated Rhyolite in the footwall. ƒ Massive Rhyolite dominates away from the ore zone whereas the brecciated Rhyolite would be associated with the ore zone. ƒ Net NP is shown to shift dramatically from positive to negative when the massive Rhyolite is compared to the brecciated Rhyolite. Based on the above interpretation of the geochemical results, combined with geological interpretation, Phase 2 focused on brecciated Rhyolite. Thirty brecciated samples were collected from underground and a representative split from each crushed sample was analyzed for ABA, as well as sulphate and sulphide, by both Leco fusion and wet digestion and carbonate whole rock and 32 element ICP analysis. This 69 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation sample set is listed as "Underground" in Table 2. The mean net NP for these Rhyolite breccia samples was -56.8 kg CaCO3 /tonne with a mean of 2.26% sulphur. Based on the information generated from the ABA tests, two composite samples were generated; "mean case" (-56.8 Net NP ) and "worst case" (-166 Net NP). The sulphate generation rates from the drill core rhyolite and brecciated rhyolite humidity cell tests are plotted in Figure 4 while comparable leachate pH data is provided in Figure 5. For the worst case breccia, an accelerated oxidation rate occurred at Week 9, with sulphate generation peaking on Week 10 at 0.57g SO4/kg/day before plateauing at 0.lg SO4/kg/day by week 17. The "worst case" brecciated Rhyolite was considerably more reactive than the "worst case" massive Rhyolite composite from drill core which had a maximum sulphate generation rate of only 0.09 gSO4/kg/day. Approximately 16.8% of the sulphide content of brecciated rhyolite "worst case" had oxidized by the time the test was terminated at Week 29. By comparison with "worst case" massive rhyolite, only 0.7% of the sulphide had oxidized before termination of the humidity cell test. The "mean case" brecciated rhyolite did not exhibit the same accelerated oxidation rate as "worst case", with the sulphate generation rate stabilizing at 0.02gSO4/kg/day. However, the rates were still significantly higher than the drill core composites. Approximately 8.8% of the sulphide content of "mean case" breccia had oxidized when the test was terminated at Week 29. By comparison with "mean case" drill core rhyolite, only 1.0% of the sulphide had oxidized before termination of the test. 70 Proceedings of the 18th Annual British Columbia Mine Reclamation Symposium in Vernon, BC, 1994. The Technical and Research Committee on Reclamation Recommendations The following recommendations have been derived from the experience at Eskay Creek. ƒ Combine a number of interpretive and predictive tools to assess potential for ARD at each site. Emphasis should be placed on compiling geological and geochemical information to support program design and interpretation. ƒ Utilize plan views and section plots to illustrate drill holes, ore zone, geology, alteration halos, sample location ABA data, location of existing and proposed development and other pertinent information. These plots will assist in establishing an accurate spatial relationship between mine development and waste generation. ƒ Conduct petrographical, chemical and physical tests on the target waste and ore zones using the section plots and correlate this information with deposit geology and proposed mine plan(s). ƒ Collect and archive samples along with detailed physical and chemical alteration information. This will allow the test results to be confirmed and re-visited should the mine plan change while the ARD program is in progress. ƒ Evaluate disposal options for all waste materials early in the project, with emphasis on sub- aqueous disposal of waste rock and tailings in natural or man-made impoundments if these materials are potentially acid generating. Program Post Mortem The Committee approach worked well. By using the iterative approach to problem design and data interpretation, concerns were addressed early, key issues were identified and the work was priorized. The committee members were intimately involved in program development, therefore the result was a stream- lined ARD assessment program. The process required that the parties be open and willing to discuss all relevant issues. The geological approach is especially beneficial in estimating waste quantities, general characteristics and in predicting ARD potential. The use of geological information is cost effective and this information remains relevant throughout the life of the project. The co-operative approach is an excellent example of Government agencies and mining companies utilizing resources in determining solutions to common problems. 71