Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterization of arsenic in a tailings impoundment under post depositional conditions Meilleur, Desiree 2004

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

Item Metadata

Download

Media
831-ubc_2004-0559.pdf [ 32.92MB ]
Metadata
JSON: 831-1.0081126.json
JSON-LD: 831-1.0081126-ld.json
RDF/XML (Pretty): 831-1.0081126-rdf.xml
RDF/JSON: 831-1.0081126-rdf.json
Turtle: 831-1.0081126-turtle.txt
N-Triples: 831-1.0081126-rdf-ntriples.txt
Original Record: 831-1.0081126-source.json
Full Text
831-1.0081126-fulltext.txt
Citation
831-1.0081126.ris

Full Text

Characterization of Arsenic in a Tailings Impoundment Under Post Depositional Conditions by Desiree Meilleur B.A.Sc, University of Waterloo, 2001 A THESIS SUBMITTED IN PARTIAL FULFILMET OF THE REQUIREMENTS FOR THE DEGREE OF Master of Applied Science in THE FACULTY OF GRADUATE STUDIES (Mining and Mineral Process Engineering) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA JUNE 2004 ©Desiree Meilleur 2004 Library Authorization In presenting this thesis in partial fulfil lment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my writ ten permission. Name of Author (please print) Date (dd/mm/yyyy) Title of Thesis: C > > O J n x < ^ W i^zAarv cA A r ^ e f W c_ i r\ cx^  TovWc. Degree: f Y n ^ c >r o£ Apc^W^ ^ C A e ^ c t Y e a r : Z*3oM Department of fY\\r\^cx a ^ , _ _ rY\i*\* r<x.\ §>ce>C£^ The University of British Columbia ^ Vancouver, BC Canada Abstract Arsenic is found naturally in the gold ores extracted f rom the Red Lake Mine in Balmertown, Ontario. Milling of ores produced arsenic-rich tai l ings that have been d isposed of in different locations around the site through the over 50 years of mine operat ions. The composi t ion of the tail ings depended upon the mineral processing methods (cyanidation, sulphide f lotat ion, and roasting) used at the t ime. The primary object ive of this thesis was to character ize the forms, stability, and mobil i ty of arsenic in the different wastes located around the mine site. In addit ion, a potential method to reduce arsenic transport f rom tail ings pond sediments into the overlying pond waters was invest igated. Who le rock analysis, Scanning Electron Microscopy (SEM), Rietveld refined powder X-ray diffraction, sequential extractions, and X-ray Absorpt ion Near Edge Structure (XANES) using a Synchrotron light source were used to mineralogical ly character ize the var ious tail ings types. Exper iments were conducted to investigate the behaviour of arsenic in several tai l ings types under var ious condit ions with the objective of determining if arsenic could be stabil ized under reducing condit ions, ideally in the form of arsenopyri te or arsenian pyrite. High aqueous arsenic concentrat ions were found to be associated with oxidized arsenic bear ing mineral phases (i.e. arsenic bearing iron oxyhydroxides conta ined in historical roaster-derived tail ings and tail ings pond sediments). Mobil ization of a signif icant fraction of the arsenic contained within these wastes is likely to cont inue as a result of the reducing condit ions that these wastes are stored under. Arsenic is found at relatively low concentrat ions in the freshly produced tail ings, primarily in the form of arsenopyri te, and is likely to remain immobi le as long as saturated condit ions exist. In order to minimize arsenic mobil ization f rom solid wastes, oxidized arsenic bearing phases (e.g. arsenic bearing iron oxyhydroxides) should be kept dry and dissolved organic carbon influxes should be limited. Reduced arsenic bearing mineral phases (e.g. arsenopyri te, arsenian pyrite), however, should be maintained under saturated condit ions. During f ield and laboratory exper iments it appears as if arsenic was immobi l ized as a reduced sulphide phase under strongly reducing condit ions, however further work is required to determine the mechanism and stability of the produced phase. Table of Contents 1.0 Introduction 1 2.0 Background Information 3 2.1 Historic Operation and Tail ings Deposit ion 3 2.2 Historical Surface Water Chemistry 10 2.3 Historical Groundwater Information 13 3.0 Literature Review 17 4.0 Solids Identification 26 4.1 Sample Collection and Preparation .26 4.1.1 Sample Collection 26 4.1.2 Sample Preparation 31 4.2 Analytical Methods Performed on Samples 32 4.2.1 Whole Rock Analysis 32 4.2.2 Scanning Electron Microscopy 32 4.2.3 Rietveld Refined X-Ray Powder Diffraction 33 4.2.4 Sequential Extractions 34 4.2.5 X-ray Absorption Near Edge Structure 36 4.3 Results and Discussion 40 4.3.1 Whole Rock Analysis 40 4.3.2 Rietveld Refinement 41 4.3.3 Scanning Electron Microscopy 42 4.3.4 Sequential Extractions 61 4.3.5 Synchrotron X-Ray Absorpt ion Spectroscopy 70 4.4 Conclusions 78 5.0 In situ and Laboratory Experiments 82 5.1 In situ Column Experiment 83 5.1.1 Methods 85 5.1.2 Column Experiment Results 88 5.2 Designed Laboratory Experiments 96 5.2.1 Methods 99 5.2.2 Results 103 5.3 Conclusions 121 6.0 Conclusions 123 7.0 Recommendat ions 128 8.0 References 132 Appendix I: Historical Surface Water Data 137 Appendix II: Whole Rock Analysis Results 147 Appendix III: Rietveld Refinement Reports 149 Appendix IV: Sequential Extraction Data and Calculations 168 Appendix V: Column Experiment Results 172 Appendix VI : Designed Laboratory Experiment Results 188 Appendix V l l : Design Ease Results 193 iii List of Tables Tab le 1 : Summary of Whole Rock Analysis Results 41 Tab le 2: Rietveld Refinement Results (wt %) 42 Tab le 3 : Sequential Extraction Data 62 Tab le 4: Percent of Total Arsenic Removed in Each Extraction Step 63 Tab le 5: Amount of Arsenic Removed in Each Extraction Step 64 Tab le 6: Percent of Total Iron Removed in Each Extraction Step 68 Tab le 7: Amount of Iron Removed in Each Extraction Step 68 Tab le 8: Semi-quantitative Arsenic Distribution (+/- 10%). Calculated from linear least-squares fitting of the As K-edge Synchrotron XANES spectra. ..71 Tab le 9: Design Matrix for Laboratory Experiments 101 Tab le 10: Average Sulphate Concentration (mg/L) for Each Tailings Type 105 Tab le 1 1 : Average Dissolved Organic Carbon Concentrat ion (mg/L) for Each Tail ings Type 106 iv List of Figures Figure 1 : Tail ings Deposit ion History 6 F igure 2: Time Series Arsenic Data for Primary Pond Discharge 11 F igure 3: Time Series Arsenic Data for Secondary Pond Discharge 11 F igure 4: Groundwater Well and Tail ings Sampling Locations in Active TMA. .114 F igure 5: Groundwater Well and Tail ings Sampling Locations in Inactive TMA 115 F igure 6: Backscattered Electron Image for RLM-1 . Particle (a) is pyrrhotite, particles (b) and (c) are iron oxide, and particle (d) is arsenopyrite 44 F igure 7: X-Ray Spectrum for RLM-1 , showing pyrrhotite pattern 45 F igure 8: X-Ray Spectrum for RLM-1 , showing iron oxide pattern 45 F igure 9: X-Ray Spectrum for RLM-1 , showing arsenopyrite pattern 46 F igure 10: Backscattered electron image for RLM-2-1 47 F igure 1 1 : Backscattered Electron Image for RLM-2-2 48 F igure 12: Backscattered Electron Image for RLM-3-1 49 F igure 13: Backscattered Electron Image for RLM-3-2 50 F igure 14: X-Ray Spectrum for RLM-3-2 51 F igure 15: Backscattered Electron Image for RLM-4 53 F igure 16: Backscattered Electron Image for RLM-5 54 F igure 17: X-Ray Spectrum for RLM-5, showing arsenic precipitate 54 F igure 18: Backscattered Electron Image for RLM-6-1 55 F igure 19: Backscattered Electron Image for RLM-6-1 56 F igure 20: Backscattered Electron Image for RLM-6-2 57 Figure 2 1 : X-Ray Spectrum for RLM-6-2 57 F igure 22: Backscattered Electron Image for Primary Pond Sediment 59 F igure 23: Percent of Total Arsenic Removed in Each Extraction Step 67 F igure 24: mg As/kg Sediment Removed in Each Extraction Step 67 F igure 25: Percent of Total Iron Removed in Each Extraction Step 69 F igure 26: Arsenic K-edge XANES spectra of three different model compounds with different oxidation states 71 F igure 27: Arsenic K-edge XANES spectra of RLM-2 series samples 73 F igure 28: Fitted Arsenic K-edge spectrum of RLM 2-1 73 F igure 29: Fitted Arsenic K-edge spectrum of RLM 2-2 74 F igure 30: Fitted Arsenic K-edge spectrum of RLM 2-3 74 F igure 3 1 : Diagram of an installed column 86 F igure 32: Dissolved Arsenic Concentrat ion at the Bottom Sampl ing Port 89 F igure 33: Concentrat ion of Dissolved Iron at the C3 Bottom Sampling Port....91 F igure 34: Concentrat ion of Dissolved Arsenic at the Middle Sampl ing Port ....92 F igure 35: Concentrat ion of Dissolved Arsenic at the Surface of the Co lumns.94 F igure 36: Designed Experiment Laboratory Setup Photograph 103 F igure 37: Flask #6 after Day 21 104 F igure 38: Arsenic Results for Flasks Containing Secondary Pond Sediment 107 F igure 39: Arsenic Results for Flasks Containing RLM-2 Tailings 108 F igure 40: Arsenic Results for Flasks Containing RLM-5 Tailings 109 F igure 4 1 : Arsenic Results for Flasks Containing RLM-7 Tail ings 110 F igure 42: Set 1 - Day 42 Interaction Graph for AD 113 vi Figure 43: Set 1 - Day 42 Interaction Graph for A F 114 F igure 44: Set 1 - Day 21 Interaction Graph for CE 115 F igure 45: Set 2 - Day 42 Interaction Graph for BE 118 F igure 46: Set 2 - Day 42 Interaction Graph for CE 118 vii 1 1.0 Introduction The Red Lake Mine, owned and operated by Goldcorp Inc., is located in Northwestern Ontario, near the community of Balmertown. Gold milling operations have been carried out at the Red Lake Mine since the 1940's. Up until 1980 a roaster was used to make the refractory gold (mostly associated with arsenopyrite) amendable to cyanidation. Prior to the 1970's tailings f rom the process were discharged directly into Balmer Creek and Balmer Lake. After this t ime a tailings management area (TMA) was constructed for the disposal of the tailings. From 1996 to 2000 operations were suspended due to a labor dispute, during this t ime a new 600 tpd mill was commissioned, and modifications were made to the TMA. The current tailings management area at the Red Lake Mine consists of a primary pond, a secondary pond, and Balmer Lake that acts as a tertiary polishing pond. The discharge f rom Balmer Lake represents the point where compliance with the current certificate of approval, and the Ontario Municipal Industrial Strategy for Abatement (MISA) must be met. Goldcorp Inc. would like to remove Balmer Lake from the TMA by generating a secondary pond effluent that is consistently compliant with the effluent criteria in effect for the discharge of Balmer Lake. Arsenic and ammonia have been identified as the parameters of greatest concern in achieving this goal. Arsenic levels in the Secondary Pond are commonly greater than 0.5 mg/L (the compliance concentration) and have been found to exceed 3 mg/L on occasion. Several sources of arsenic to the TMA have been identified including underground waters, mill process waters, subaerial tailings, subaqueous tailings, groundwater, and mine site run-off. It has been estimated that contributions of arsenic from subaqueous tailings represent the largest arsenic source to the Secondary Pond. Tail ings derived from the roasting process, used in the past, 2 are believed to be located on the bottom of the Secondary Pond. Data collected previously suggest that arsenic is released from the tailings sediments via reductive dissolution. This has been hypothesized to occur during the winter when an ice cover limits atmospheric oxygen diffusion into the pond, and during the summer when an increase in sediment oxygen demand occurs as a result of plankton growth and microbial activity. Historical operations give rise to many issues at the Red Lake Mine that do not exist at newer mines. The long history of the mine and the many changes made to the milling process and tailings management methodologies make it difficult to determine exactly what types of tail ings are stored around the property. To effectively mange the various tailings deposits at the Red Lake Mine site, it is necessary to identify the forms of arsenic found in each of the tailings storage areas. Once the forms of arsenic have been identified and quantif ied the tailings can be managed appropriately to minimize arsenic dissolution into the surrounding pore waters and surface waters. In this thesis, the fate and transport of arsenic in the tailings impoundment at Goldcorp's Red Lake Mine are investigated. The primary focus of the study is to quantify the form and stability of arsenic in each of the tailings deposits at the Red Lake Mine. This will enable a better understanding of the sources and geochemical processes controll ing arsenic behaviour in the tailings and will allow for recommendat ions to be made on the most effective way to manage each of the tailings areas. The overall goal of the project is to provide Goldcorp Inc. with information that will enable the development of methods to reduce arsenic levels in the water leaving the tailings impoundments, and assist in planning for closure. 3 2.0 Background Information The Red Lake Mine has a long history of operations since the 1940's. As a result of the past milling and tailings management practices, elevated surface and groundwater arsenic levels have developed around different areas of the mine site. As an initial step to understanding the cause of the elevated arsenic concentrat ions it is necessary to compile available historical information on the milling process f low sheets that were used over the years. The different milling processes resulted in the production of several types of tailings solids, that were deposited in various areas around the mine site. Information was collected from various reports and from mine personnel in an attempt to gain an understanding of the tailings management practices that have taken place at the mine. In this section the historical milling and tailings management practices are discussed and historical surface and groundwater chemistry data are reviewed. It is shown that there are three primary types of tail ings that were produced throughout the operation of the mine: Low sulphide content cyanidation/flotation tailings, high sulphide content cyanidation tailings, and roaster-derived tailings. It is expected that the arsenic in each of the tailings types will be contained in different forms and associated with different minerals. The tailings solids were deposited in two main areas. Initially solids were deposited in the vicinity of Balmer Creek, and at a later date solids were deposited at different locations throughout the currently active tailings management area. Elevated groundwater arsenic concentrations exist in areas where roaster-derived tailings were reported to have been deposited. 2.1 H i s to r i c O p e r a t i o n a n d T a i l i n g s D e p o s i t i o n The Red Lake Mine property was first staked in 1926 as part of the Red Lake Gold Rush. Diamond drilling began in March 1945, and shaft sinking started in 1946. In December of 1948 milling began at Dickenson's Red Lake Mine. Initially 100 tons of ore per day passed through the processing plant, which consisted of grinding in cyanide solution followed by being treated in a cyanide 4 circuit. Late in 1950 a mineral jig was placed in the grinding circuit and the concentrate from the jig was amalgamated. The tailings produced during this t ime period are likely to be high sulphide content cyanidation tailings. The milling rate increased to 300 tons per day in early 1951 after another grinding circuit was installed. Flotation equipment and a roaster were also added. In 1951 the f lowsheet consisted of grinding the ore in water with free gold recovered with a mineral jig and amalgamated. Following grinding, the slurry passed through a flotation circuit. The flotation concentrate was roasted, and calcine (roaster product) and flotation tailings were subjected to cyanidation. Milling capacity was increased to 450 tons per day in 1954 by the addition of a standard cone crusher into the circuit. In addit ion, the grinding was conducted in cyanide, and was converted to two stage grinding. Free gold was recovered following secondary grinding via corduroy blankets and a mineral j ig. Concentrate from the blankets and jig was amalgamated, and blanket tail ings reported to cyanidation. Gold was recovered from the cyanide solution via precipitation with zinc dust. Cyanide tailings were sent to the flotation circuit, flotation concentrate was roasted and the calcine leached along with the blanket tailings. In 1956 the processing plant had a capacity of 450 tons per day and the process consisted of crushing, grinding in cyanide, amalgamation of free gold recovered with a mineral j ig, cyanidation of the grinding slurry along with calcine residue, and finally flotation of the leach product. Flotation concentrate was roasted. The roaster had a capacity of 16 tons per day. The above information was taken from a memo prepared by M.G. Sveinson, Mill Metallurgist, on September 17 t h , 1956. The roaster operated continuously from 1951 until it was shutdown in 1980 for environmental, health and safety reasons. It is speculated that the roaster was shutdown due to new regulations concerning air emissions from roasters that were scheduled to come into effect July 1 s t , 1980 (Barr, 1983). In 1982 the 5 milling capacity was increased to 700 tons per day. The flotation circuit continued to operate until November 1990, with concentrate being sold to custom smelters until 1989 and stockpiled after that. During the t ime period from 1980 -1990 it is likely that the tailings produced were low sulphide content cyanidation tailings. In 1994 the mill operated at approximately 1000 tons per day (rated capacity was 850 tons/day). In February of 1995 the milling rate was reduced to 680 tons/day in order to facilitate increased exploration costs. Milling ceased operation in June of 1996, the mill was demolished and construction of a new mill was completed in June 2000. High sulphide content cyanidation tailings were likely produced from the t ime flotation operation ceased (November 1990) until the mill shutdown in 1996. The new mill consists of a standard crushing/grinding circuit with gravity recovery using a Knelson Concentrator. Knelson concentrate is tabled and refined into bullion on site. Slurry f rom the grinding circuit passes through a leaching and carbon in pulp (CIP) circuit. The carbon is stripped and reactivated, and the gold is recovered from the pregnant solution by electrowining. CIP tailings pass through a cyanide detoxification circuit and ferric sulphate is added in an attempt to precipitate dissolved arsenic. The CIP tailings then pass through a flotation circuit where refractory sulphides are concentrated. The concentrate is currently stockpiled on site and shipped to a custom smelter for further refining. Flotation tailings are sent to the paste backfill plant. Approximately 5 0 % of the tailings are used underground as backfill and the remainder of the low sulphide content cyanidation/flotation tailings are piped to the tailings management area (TMA). Throughout the operation of the mine, mill tailings were disposed of in several different areas. Figure 1a shows a schematic of the area prior to operation of the mine. Prior to 1960 there were no control structures in place to attempt to retain the tailings in any way. Tail ings were spilled on the ground and f lowed into and along Balmer Creek (Figure 1 b). Tail ings located on both sides of Balmer Creek 8 consist of material subjected to cyanidation alone (1948 - 1951), and material subjected to cyanidation, flotation, and roasting (1951 - 1960). In 1960 a dyke was constructed and the tailings discharge point was moved in an attempt to divert tail ings from directly entering Balmer Creek. The tailings f lowed into a horseshoe - shaped depression, f lowing through a road culvert making their way through a small creek to Balmer Lake (Figure 1c). It was reported that retention t ime in the horseshoe - shaped area was minimal, allowing tailings solids to reach Balmer Lake. In 1962, classification of the tailings began and 6 5 % by weight of the tailings (coarse fraction) were used as fill underground, with the remaining material being pumped to the horseshoe - shaped area. In 1979 a certificate of approval was issued, giving permission to build two Dams, #1 and #2. Dam #1 was constructed to separate a primary pond from a secondary pond, and Dam #2 was constructed to separate the secondary pond from Balmer Lake. Both Dams were constructed to be permeable, allowing water to flow relatively unimpeded through them but retaining solids behind them. The bottoms of the newly created Primary and Secondary Ponds contained tailings subjected to cyanidation, flotation and roasting (Figure 1d). In 1984 the Primary Dam (Dam #1) was raised and clay was placed on the upstream side to make the dam relatively watertight. A spillway was constructed with a Gabion weir to discharge water from the Primary Pond to the Secondary Pond. In April of 1985 a 600-foot long section of the Primary Dam failed, resulting in the release of solids and solution into the Secondary Pond. This breach in turn caused a fifty-foot section of the Secondary Dam to fail resulting in the release of tailings slurry into Balmer Lake. Consequently, some non-roaster tailings were deposited in the Secondary Pond. In 1986 wick drains were installed in the Primary Dam in an attempt to stabilize the dam. In 1986 a flow control weir was completed at Balmer Creek (L2). In 1987 the tailings pipeline was extended to its now current discharge location, approximately 5000 ft upstream of the Primary Dam (Figure 1e). In addit ion, the crests of both the Primary and Secondary Dams were raised approximately 1.5 feet. In 1988, the crest of the Primary Dam 9 was raised by an additional 1-2 feet. The Secondary Dam failed again on April 2 3 r d , 1990. High water levels in the Secondary pond overtopped the dam, washing out a 150 foot section. An emergency spil lway was installed in the dam to prevent an additional washout from occurring in the future. Around 1994 two splitter dykes, constructed out of waste rock, were placed in the Primary Pond. Splitter Dyke #1 was constructed just downstream of the tailings discharge point, and Splitter Dyke #2 was built approximately 1000 ft upstream of the Primary Dam (Figure 1f). The dykes were built in an attempt to retain more solids further back in the tailings system and to increase the residence time in the Primary Pond. The dykes function reasonably well, with the majority of the solids being retained upstream of Splitter Dyke #2. Improvements were made to the permeable rock fill dams in the Secondary Pond during 1995 and 1996. Filter layers were added to the upstream side of the Dams to turn them into water retaining embankments. An engineered stop log weir was constructed to allow for controlled release of water from the pond. In May of 1997 a Diversion Channel was completed that carried water from the Beaver Pond (upstream of the tailings management area) around the tailings management area, discharging into the Secondary Pond downstream of the Primary Pond Spillway. This water acts as a significant source of dilution in the Secondary Pond. Figure 1 (a-f) depicts the historical evolution of the tailings management areas at the Red Lake Mine. 10 2.2 H is to r i ca l Sur face Water C h e m i s t r y Water chemistry data for the Primary Pond, and G1 (effluent of Secondary Pond) is available since 1991. Grab samples were collected from the Primary Pond and G1 and were analyzed at a number of laboratories throughout the years (since 2000 all samples for environmental compl iance have been analyzed at Envirotest Laboratories Inc. in Thunder Bay Ontario - a CAEAL certified lab). Complete tabulated data can be found in Appendix I. Figures 2 and 3 show the t ime series of arsenic data for the Primary Pond, and for G 1 , respectively. The t ime series can be broken down in to three segments: Before shutdown (Jan-91 -April-96), shutdown (May-96 - Oct-00), and after shutdown (Nov-00 - present). Before Shutdown Prior to the shutdown, Primary Pond waters were characterized by high concentrat ions of arsenic and cyanide. The average concentrat ions of total arsenic and total cyanide were 1.58 mg/L and 24 mg/L respectively. The arsenic concentration was consistently above 0.5 mg/L and a trend of increasing arsenic concentrat ions before the shutdown was observed. The total cyanide concentration never fell below 2 mg/L. In addit ion, the total suspended solids concentrat ion was 52 mg/L, and the Cu, Ni, Zn, and Fe concentrat ions were also elevated, averaging 2.92, 1.65, 1.35, and 6.1 mg/L, respectively. The average ammonia concentration was relatively low at 4.8 mg/L, and the average pH was 8.3. The concentration of arsenic at G1 was consistently above 0.5 mg/L before the shutdown with concentration peaks occurring predominately in late summer (Aug - Sep) and early winter (Dec). The average concentrat ions of total arsenic, total iron and total cyanide for this t ime period were 1.2 mg/L, 3.27 mg/L and 5.79 mg/L, respectively. The total iron and cyanide concentrat ions were consistently above 1.0 mg/L and 0.5 mg/L, respectively. The pH of G1 water averaged 7.95, the total suspended solids concentration was 25 mg/L, and the average ammonia 12 concentration was 6.86 mg/L. In addit ion, Cu, Ni, and Zn concentrations averaged 1.26, 0.87, and 0.4 mg/L, respectively. During Shutdown During the shutdown all contaminant concentrations dropped dramatically in the Primary Pond. The average arsenic concentration during the shutdown was calculated to be 0.54 mg/L. In general, peaks in the arsenic concentration were seen in August. Outside of the summer peaks, the arsenic concentration was for the most part below 0.5 mg/L, with a trend of increasing arsenic concentration observed as t ime progressed. The average Cu, Ni, and Zn were low at 0.06, 0.07, and 0.02 mg/L, respectively, and the iron concentration decreased significantly to 0.66 mg/L. The average total cyanide concentration during the shutdown was 0.36 mg/L. The average total suspended solids concentration was <15 mg/L, the pH averaged 7.8, and the ammonia concentration dropped to 1.53 mg/L. Significant peaks in the arsenic concentration at G1 during Aug/Sep and Jan/Feb occurred throughout the shutdown. For the first t ime, arsenic concentrations below 0.5 mg/L were seen. The diversion channel, completed in May 1997, added dilution to the Secondary Pond, which likely aided in reducing arsenic concentrations. The average concentrations of arsenic, iron and cyanide during the shutdown were 0.72 mg/L, 0.66 mg/L and 0.07 mg/L, respectively. The average pH was 7.63, while the average total suspended solids and ammonia concentrat ions decreased to <10 mg/L, and 0.43 mg/L, respectively. After Shutdown Under the new mill operation (after the shutdown) the average arsenic concentration in the Primary Pond was calculated to be 1.84 mg/L, with peaks observed in august of 2001 and 2002. The concentration of arsenic usually 13 exceeded 1 mg/L. The introduction of an I N C O - S 0 2 cyanide destruction process in the mill greatly reduced the cyanide concentration in the Primary Pond. The average cyanide concentration after the shutdown was 2.27 mg/L, with the lowest concentrat ions being observed in the summer months. The ammonia concentrat ion, however, increased dramatically, averaging 27.7 mg/L. Cu, Ni, Zn, and Fe concentrations were significantly lower than before the shutdown averaging 0.12, 0.019, 0.013, and 0.36 mg/L, respectively. The average total suspended solids concentration was 5 mg/L, and the pH averaged 7.9. Since July 2 0 0 1 , the arsenic concentration at G1 has not dropped below 0.5 mg/L, peaks in arsenic concentration occurred in August of 2001 and 2002, and the average concentration was determined to be 0.736 mg/L. The average concentrat ions of cyanide and ammonia at G1 after the shutdown were 0.09 mg/L, and 7.96 mg/L respectively. The cyanide concentration rose since the shutdown ended, however the concentration was almost always below 0.2 mg/L. Cu, Ni, Zn, and Fe concentrations remained much lower than before the shutdown, averaging 0.03, 0.02, 0 .01 , and 0.47 mg/L, respectively. The average total suspended solids and pH were 5.84 mg/L and 7.7. 2.3 H is to r i ca l G r o u n d w a t e r I n f o r m a t i o n Initial groundwater work was conducted at the site in 1993. Wells were installed in the active TMA, the inactive TMA, and the Mil l /Headframe area. As of 2002 only 5 of the eleven wells were still functional. Twelve additional groundwater monitoring wells were installed in March of 2002 and all operational wells were sampled in 2002. Figures 4 and 5 show the groundwater well locations in the active and inactive TMA's, respectively. Elevated arsenic concentrations are seen in all of the tailings deposits surrounding the Red Lake Mine site. The highest arsenic concentrations occur in mm -UJ4-U*_l*^ * ngJ • M rt_ilrt»«il mt U n a - U J * - U # _ U ™ * r*>li m r t _ U r t w i l mj 16 the historical tailings deposit south of the mine site, on the east side of Balmer Creek. In the active TMA it was found that groundwater f lows horizontally f rom southeast to northwest, and f lows vertically upward. Groundwater concentrat ions of dissolved arsenic in the historical tailings in the active TMA (in Primary Pond, and under the Primary and Secondary dams) are significantly elevated. These arsenic levels are significantly lower than what is observed in the tailings deposit located near Balmer Creek, however the concentrations are still significantly higher than background levels. Vertical movement of groundwater from the tailings layer upward to the overlying pond waters, and horizontal seepage through the dams provide the most significant groundwater contributions of arsenic to the surface water bodies. It was found that groundwater moves horizontally f rom northeast to southwest and vertically downward in the inactive TMA located southeast of the mine. Arsenic concentrations in the tailings layer are extremely elevated, and increase closer to the creek. These are the oldest tailings located on the site, and they were produced at a t ime when roasting was a part of the milling process. This area was revegetated in the late 1980's. Starting in 1997 nitrogen and iron concentrations in groundwater were seen to increase. At the same t ime, a more than 10 fold increase in arsenic concentrations was observed. It is speculated that the large increase in the concentration of ammonia nitrogen was from fertilizer used during the revegetation of the area. The applied nitrogen f rom the fertilizer appears to have migrated downward into the groundwater likely resulting in the increased arsenic concentrat ions that have been observed. 17 3.0 Literature Review Arsenic has a complex chemistry and is readily mobile under many condit ions. As a result of gold mining activities in Canada, thousands of tonnes of arsenic bearing rock are brought to the surface, crushed and ground to a small particle size, chemically and mechanically treated, and disposed of into tailings impoundments. The largest source of arsenic in gold mining ore is arsenopyrite (FeAsS). Arsenopyrite is stable deep beneath the earth surface, however as a result of milling processes high concentrat ions of arsenic may be released into solution, and less stable secondary minerals containing arsenic may be created. There are various methods available to remove arsenic from solution, however it is the stability of the solid phase by-product of the treatment process that is of the greatest concern. The preferred arsenic treatment method of both the US EPA and the Canadian metallurgical industry is co-precipitation with ferric iron (Riveros et al., 2001). High iron arsenical ferrihydrite has been shown to be stable for many years under the correct storage condit ions (Krause and Ettel, 1985, 1987, 1988, 1989). Unfortunately the correct storage condit ions involve maintaining an oxidizing environment, which can be difficult to do in natural tailing impoundment systems. McCreadie et al. (2000) report that oxidized arsenic phases occurring in tailings deposits, as a result of oxidation of arsenopyrite during processing, are potentially susceptible to dissolution under saturated conditions. The combination of organic matter and a water cover can lead to the onset of microbially mediated reducing condit ions in tailings impoundments. Water covers are used to limit oxygen flux to tailings surfaces as a method to prevent the onset df acid rock drainage. By limiting the flux of oxygen, reducing condit ions can be maintained. W h e n arsenic bearing oxyhydroxide phases are present in a tailings pond the onset of reducing condit ions can lead to the release of arsenic from the sediments. 18 Bacteria play a major role in the development of reducing conditions. Bacteria use dissolved organic carbon as an electron donor in order to reduce various chemical species for energy. There is a well known sequence of reactions that occur when water becomes anaerobic (Stumm and Morgan, 1995). The following equations describe the sequence of steps involved in the development of biologically mediated reducing conditions. C H 2 0 + 4 / 5 N 0 3 " - » HCO3" + 2/5N 2 (g) + 2/5 H 2 0 + 1/5H+ C H 2 0 + 1/2NCV - » HCO3" + 1 /2NH 4+ + 1/2 H 2 0 C H 2 0 + 2 M n 0 2 + 3 H+ 2 M n 2 + + HCO3" + 2 H 2 0 C H 2 0 + 4 F e O O H + 7 H+ - » 4 F e 2 + + H C 0 3 " + 6 H 2 0 C H 2 0 + I/2SO42" -> HCO3" + 1/2 HS" + 1/2H+ C H 2 0 + 1/2 H 2 0 - » 1 /2HC0 3_ + 1/2CH 4(aq) + 1/2H+ Initially all 0 2 is reduced ( C H 2 0 gives off electrons that oxygen takes, organic carbon is oxidized and oxygen is reduced). Once all of the oxygen is consumed, nitrate ( N 0 3_ ) is reduced to N 0 2_ and the gases N 2 0 and N 2 . Solid phase manganese and iron oxides are reduced to M n 2 + and F e 2 + . This is fol lowed by the reduction of sulphate ( S 0 22 " ) to sulphide (S 2 _ ) . Fermentat ion and methanogenesis occur next, resulting in the production of C H 4 . Finally, nitrogen gas (N 2 ) is reduced to N H 4+ . It is thought that the reduction of A s 5 + to A s 3 + occurs after the reduction of iron but before sulphate reduction (Smedley and Kinniburgh, 2002). 19 The exact mechanism of release of arsenic from iron oxide phases is not completely understood. A combination of reductive dissolution of the iron oxide ( F e 3 + being reduced to F e 2 + resulting in the solubilization of the iron oxide phase and the release of sorbed arsenic) and the direct reduction of arsenate to arsenite is thought to occur (Smedley and Kinniburgh, 2002). Arsenite has a much lower affinity to iron oxide phases at near neutral pH than arsenate (Jain et al., 1999). McCreadie et al. (2000) propose the following equation to describe the release of arsenic via reductive dissolution f rom roaster-derived ferric oxides. 2Fe 2 03»xH 3 As03 + C H 2 0 + 7 H+ -> 4 F e 2 + + HCOY + 4 H 2 0 + 2 x H 3 A s 0 3 Where C H 2 0 represents a model dissolved organic carbon molecule and x is the amount of non-structural arsenic associated with hematite grains in the roaster tailings. There are numerous studies that document the increase in aqueous arsenic concentrations following the development of anaerobic condit ions in sediments containing arsenic bearing iron oxyhydroxides. These studies include: Deuel and Swoboda (1972), McGeehan and Naylor (1994), Azcue and Nriagu (1995), McCreadie et al. (2000), and Martin and Pedersen (2002). Little evidence of arsenic removal as an arsenic sulphide species has been documented (Smedley and Kinniburgh, 2002), and it is unclear as to why high dissolved arsenic concentrations are observed under reducing condit ions in the presence of sulphide. McCreadie et al. (2000) saw increased arsenic concentrations in a sulphate reducing zone of the Campbel l Mine tailings impoundment. Meng et al. (2003) indicate that biotic reductions can convert arsenic and sulphide into arsenian pyrite, although there is limited evidence of this occurring in natural systems. Martin and Pedersen (2002) report that in the deeper sediments of Balmer Lake arsenic is consumed as an authigenic sulphide phase. In the shallow sediment, arsenic is released to the surface water due to 20 seasonal anoxia that develops in the near surface pond sediments (Martin and Pedersen, 2002). It is speculated that conditions may not be reducing enough in some situations to cause the formation of an arsenic sulphide phase, and or not enough sulphide is available (not enough sulphate available to be reduced to sulphide) to precipitate all of the arsenic (plus other dissolved metal species). Pyrite formation in low temperature sedimentary environments has been studied intensively, however there is only limited information available on arsenopyrite formation in low temperature sedimentary environments. Iron sulphide formation has been documented in oceans, lakes, moors, swamps and aquifers, and it has been suggested that pyrite may have played a crucial role in the origin of life on earth (Rickard et al., 1995). There are three major reactants involved in pyrite formation that can become limiting: metabolizable organic matter, dissolved sulphide, and reactive iron minerals (Morse, 1999). There are three essential processes in the formation of pyrite: production of hydrogen sulphide, formation of iron monosulphides, and formation of iron disulphides (i.e. pyrite). In the sequential order of redox processes, iron reduction has an overall higher energy yield than sulphate reduction, therefore, it is generally thought that significant sulphate reduction will not occur in the presence of ferric iron. Postma and Jakobsen (1996) suggest that the reduction sequence is better explained as a partial equil ibrium process rather than f rom the overall energy yield of the different reactions. From a thermodynamic point of view Fe(ll l) and sulphate reduction may proceed simultaneously over a wide range of environmental condit ions (Postma and Jakobsen, 1996). The stability of the iron oxyhydroxide species and the pH are the dominant factors in determining which reduction is energetically favoured (Postma and Jakobsen, 1996). Field data confirms that simultaneous reduction of Fe(ll l) and sulphate occurs and that the reduction order, as predicted by the overall energy yield, can even be reversed. The boundaries between Fe(ll l) and sulphate reduction are strongly 21 affected by the variability in the stability of iron oxyhydroxides present. The more variability that exists, the more blurred the boundaries will be (Postma and Jakobsen, 1996). Iron monosulphide formation is kinetically limited, and the amount of iron in sulphide bearing waters is higher than predicted by equil ibrium suggesting that significant amounts of iron may be maintained in solution as sulphide complexes (Rickard et al., 1995). Rickard et al. (1995) listed two compet ing pathways for iron (II) monosulf ide formation in solution: a) hydrogen sulphide pathway (direct precipitation of Fe(ll) monosulf ide) F e 2 + + H 2 S FeS + 2 H+ b) bisulphide pathway (involves the formation of complexes) F e 2 + + 2HS" Fe(SH) 2 (s) Fe(SH) 2 ( S ) FeS + H 2 S The bisulphide pathway is faster at neutral to alkaline pH and S 2 _ > 10" 3 M, while the hydrogen sulphide pathway dominates under acidic condit ions and S 2 ' < 1 0 * M, however the hydrogen sulphide pathway becomes more important at temperatures below 25°C (Rickard et al., 1995). In most sedimentary environments pyrite is the stable iron sulphide, and ultimately all iron and sulphide species will become pyrite, implying that there are multiple competing pathways for its formation (Rickard et al., 1995). Iron monosulphides are scarce in modern low temperature sedimentary systems. Pyrite is stable and forms rapidly under oxidized/reduced boundary condit ions, however the reaction mechanisms of pyrite formation at low temperatures is incomplete (Furukawa and Barnes, 1995). Rickard et al. (1995) describe three pathways for pyrite formation: 22 a) Polysulphide Pathway Involves the reaction of Fe(ll) monosulphide and polysulphide. Numerous experiments have shown that in order for iron monosulphides to be converted to pyrite, a sulphur source of an intermediate oxidation state (elemental sulphur, thiosulphate, or polysulphides) must be present. The early studies (such as Berner (1970)) suggest that pyrite forms due to the addition of a sulphur atom to the precursor monosulphide rather than through the loss of iron. Furukawa and Barnes (1995) show theoretically that in order for the change in volume of the solids to be negative (a requirement of thermodynamics), pyrite formation must proceed through the loss of iron and not the addition of sulphur. Furukawa and Barnes (1995) suggest that intermediate sulphur species, required in the experiments, act as oxidizing agents rather than as a source of sulphur. Intermediate sulphur species are efficient oxidizing agents, however other species may be able to act as oxidizing agents which may explain why many natural pyrite forming systems are not found to contain a high concentration of intermediate sulphur species (Furukawa and Barnes, 1995). Wilkin and Barnes (1996) used sulphur isotope ratios to show experimentally that pyrite formation proceeds via the loss of iron from the precursor monosulphide rather than via the addition of zero-valent sulphur. b) FeS Oxidation Pathway Involves a progressive oxidation mechanism starting with the aging of amorphous FeS to Mackinawite (FeSo.94). Mackinawite is t ransformed to greigite (Fe3S4) under slightly oxidizing condit ions and greigite is t ransformed to pyrite. The individual steps are not completely understood, but it is stated that the transformation to mackinawite is slow. This pathway is really the same as in a), as it has been shown that polysulphides act as oxidants. 23 c) H 2 S Pathway It has been shown that iron monosulphides could react with H 2 S to form pyrite at 100°C in a few days (FeS + H 2 S -> F e S 2 + H 2( 9)). The rate of the reaction is suggested to depend on pH, FeS, surface area, temperature, and hydrogen sulphide concentration. This reaction however, has not been observed in a number of carefully controlled experiments (Wilkins and Barnes (1996), Berner (1970), Schoonen and Barnes (1991), etc.). In these experiments it was shown that pyrite formation only proceeded at significant rates when an oxidant other than H 2 S was present. Benning et al. (1999) showed that if iron monosulphides are kept in a reducing environment with no other reactant than H 2 S, the formation of pyrite is inhibited (over a wide range of temperature and pH) and mackinawite is the stable phase. It was shown that only oxidation of the aqueous sulphur species or of the precursor iron monosulphide species resulted in pyrite formation (Benning et al., 1999). Crystal growth of sedimentary iron disulfides becomes important once F e S 2 nuclei are formed. The nuclei can grow from solution once they are formed. The F e S 2 nuclei are formed as a result of porewaters in reduced sediments being close to saturation with respect to iron monosulphide phases (iron monosulphides are about nine orders of magnitude more soluble than iron disulphides) (Rickard et al., 1995). The rate of dissolution of iron monosulphides may exceed the growth of pyrite, leading to increased aqueous iron concentrations. The low concentrations of polysulphides in natural waters suggest that the polysulphide pathway may not be a viable mechanism under all condit ions (Rickard et al., 1995). Benning et al. (1999) showed that pyrite will only form at low temperatures if some degree of oxidation is present (the importance of oxygen vs polysulphides 24 is unknown), and that pyrite forms at negligible rates in H 2 S solutions. Sulphur compounds with an oxidation state greater than S 2 _ have the potential to act as terminal electron acceptors in biologically mediated oxidation of organic material (Neal et al., 2001). Morse and W a n g (1997), and Morse (1999) suggest that at low pH's pyrite formation proceeds via the faster H 2 S pathway, however at near neutral (and higher) pH's the formation of pyrite proceeds via the much slower polysulphide pathway. Morse and W a n g (1997) showed that high DOC concentrations could significantly reduce the rate of pyrite formation possibly due to the complexation of dissolved iron with organic mater making it unavailable for iron sulphide mineral formation. At higher pH the Fe(ll) is not readily available due to the competing reaction of Fe(l l l) oxyhydroxide formation (Wei and Osseo-Asare, 1996). Al though pyrite is by far the most abundant sulphide mineral, other metals can react with sulphide to form sulphide minerals, and/or metals can coprecipitate and adsorb onto iron sulphides (Morse and Luther, 1999). There exists the potential for trace metals to be immobil ized by incorporation into sulphide minerals (Di Toro et al. (1992), and Morse (1994)). Huerta-Diaz and Morse (1992), and Huerta-Diaz et al. (1998) state that reactions occurring at the surfaces of iron sulphides play a major role in metal retention, mobility, and bioavailability. Arsenic species must first be reduced to arsenite before they can be incorporated into sulphide minerals. Next to mercury, arsenic is the most likely metal to be incorporated with pyrite (Morse and Luther, 1999). Moore et al. (1988) found that diagenetic sulphides were important sinks for metals and arsenic in the reduced sulphidic sediments of the Milltown Reservoir. It can be assumed that arsenic will be incorporated into pyrite or will f rom arsenopyrite under similar conditions that favour pyrite formation. Pyrite represents a relatively stable sink for toxic trace metals, however pyrite is susceptible to dissolution by iron oxidizers (such as Thiobacillus ferrooxidans) if condit ions are al lowed to be come oxidizing (Neal et al., 2001). If sulphides are 25 moved into oxidizing environments, trace elements may be released into solution (Moore et al., 1988). It is essential to know the forms of arsenic present in a tailings impoundment to apply appropriate management strategies. 26 4.0 Solids Identification To effectively manage the mine site, it is essential to understand and characterize the form of the arsenic in the Red Lake Mine tailings. The long history of the Red Lake Mine Site, and the multitude of milling methods used over the years have resulted in the production of various different types of tailings that have been deposited in several areas around the mine site. In addition, tailings have been subject to a number of different post depositional condit ions. Each tailings type is expected to contain different types of arsenic bearing solids, in varying quantit ies. Accordingly, a major effort was undertaken to characterize the tailings material in various deposit ional environments. The objective of this work was to identify the type of arsenic in each of the various tailings samples. Type of arsenic refers to the minerals that arsenic is associated with, whether it is part of the mineral structure or adsorbed to the surface of the mineral, and the speciation of the arsenic (XANES). It was also desired to determine the amount of each type of arsenic in the samples (via the Rietveld method, XANES, and sequential extractions). Based on this information, recommendat ions for optimal long term storage condit ions for each type of tailings are made. 4.1 S a m p l e Co l l ec t i on a n d Prepara t ion Tailings samples were collected in the fall of 2002 f rom 9 locations around the Red Lake Mine site. The samples included fresh tailings, historical tailings, historical tailing produced at a t ime when roasting was used in the milling process, and pond sediments. Figure 4 shows the sampling locations. 4.1.1 S a m p l e C o l l e c t i o n A soil corer was used to collect the samples. In some cases a sample was taken at the surface using the corer and a pit was dug so that deeper samples could be 27 obtained. Tail ings samples were labelled RLM-1 - RLM-7, with samples from different depths at the same location being identified by a second number (e.g. RLM-2-1 , RLM-2-2, and RLM-2-3). The samples and sample locations are described below. RLM - 1: Downstream of SD#1, north of the culvert The tailings at this location are either new tailings (after the shutdown), or historical non-roaster tailings produced before SD#1 was constructed. The tailings were soft (it would not have been possible to sample the tailings if the ground was not frozen), dark grey and fully saturated. Core from the surface down to approximately 1 ft was recovered. RLM - 2: Downstream of SD#2, on tailings beach Roaster tailings were co-disposed with cyanidation and flotation tailings in this area from 1960 until roaster operation ceased some time around late 1979 to early 1980. After roaster operation ceased, tailings that had been subjected to cyanidation and flotation were deposited in this area until 1987 when the tailings pipeline was extended to its current location. The exact f low path that the tailings took between 1960 and 1987 is unknown, therefore it is not clear whether the tailings sampled at this location were derived from the roaster. The tailings were solid and unsaturated for approximately 3 feet. The corer had to be pounded into the ground with a s ledgehammer in order to obtain samples. Samples were taken from three depths at this location. 28 RLM-2-1 Core was obtained from the ground surface down to approximately 6 inches. The tailings were visibly oxidized, with several different-coloured layers present. The top of the tailings was orange, fol lowed by a thin grey layer, fol lowed by a red layer. The tailings in this sample were unsaturated. RLM-2-2 Core was taken from approximately 1.5 feet to 2 feet below the ground surface. The tailings were grey and unsaturated. RLM-2-3 Core was obtained from approximately 2.5 feet to 3 feet below the ground surface. The tailings were grey and saturated. RLM - 3: Old tailings on south side of access road Roaster tailings were co-disposed with cyanidation and flotation tailings in this area from 1960 until roaster operation ceased some t ime around late 1979 to early 1980. After roaster operation ceased, tailings that had been subjected to cyanidation and flotation were deposited in this area until 1987 when the tailings pipeline was extended to its current location. It is likely that the cyanidation/flotation tailings overlie the roaster tailings in this area therefore the tailings sampled (at least in the shallow depth samples) are not likely to be roaster derived tailings. The tailings were solid and unsaturated for approximately 3 feet. The corer had to be pounded into the ground with a s ledgehammer in order to obtain samples. Samples were taken from three depths at this location. 29 RLM-3-1 Core was taken from the ground surface down to approximately one foot. The tailings were unsaturated and were noticeably oxidized (orange in colour). RLM-3-2 The sample was taken at the top of the grey layer (underlying a red layer), f rom approximately 1.5 feet to 2.5 feet below the ground surface. The tailings were grey but still unsaturated. RLM-3-3 The sample was taken at the top of the saturated zone, from approximately 3 feet to 3.5 feet below the ground surface. RLM - 4: Upstream of SD #1 in f low path of new tailings The tailings at this location are newly deposited, fresh tailings. The tailings were soft but were covered by a layer of ice making it possible to walk on them. A shovel was used to take the sample as the tailings were too wet to use the corer. RLM - 5: End of Pipe Several buckets of tail ings were collected from the tailings pipeline discharge while the Paste Backfill Plant was not operating. The tailings were al lowed to settle in the field and the water was decanted off. The tailings were then filtered in a pressurized filter apparatus. RLM - 6: Revegetated Tailings east of Balmer Creek, in between groundwater wells DK-93-3 and BH-9 Tailings were deposited in this area from 1948 - 1960. From 1951 - 1960 the tailings that were deposited were derived from the roasting process (i.e. a large portion of the sulphides likely unde rwen t oxidation during the milling process). Prior to sampling, thick vegetation and a thin sand/soil layer were removed to 30 expose the top of the tailings layer. The tailings were solid and unsaturated for approximately 4 feet. Dead wood was encountered throughout the tailings, making sampling quite difficult. Samples were taken f rom two depths. RLM-6-1 Core was taken from the surface down to approximately two feet. Tail ings were a brownish orange colour and were unsaturated. RLM-6-2 Core was taken at the top of the saturated zone, approximately 4 feet below the ground surface. The tailings were a brownish orange colour. RLM - 7: CIP tailings from carbon safety screen To obtain a sample of tailings that had not gone through the Detox and Ferric circuits, a bucket of CIP tailings was collected from the carbon safety screen. The tailings were filtered in a pressurized filter apparatus. Secondary Pond Sediment A hole was augured through the ice on the Secondary Pond (in the vicinity of the Limnocorrals) and a dredge sampler was used to collect a sample of the tai l ings/sediments on the bottom of the Secondary Pond. Primary Pond Sediment A hole was augured through the ice on the Primary Pond (in the vicinity of the columns) and a dredge sampler was used to collect a sample of the tai l ings/sediments on the bottom of the Primary Pond. Primary Pond Backhoe 31 During installation of a column experiment in the Primary Pond, a sample of the tailings in the pond was taken from the Backhoe bucket. The tailings were sticky and grey. 4.1.2 S a m p l e Prepara t ion Air temperatures were below zero at the t ime of sampling, ensuring that tailings samples would freeze upon exposure to the air. Samples were stored frozen, and were shipped frozen via refrigerated truck to Vancouver, where they were stored in a freezer. These precautions were taken in order to minimize potential oxidation of the solids. Several of the tests to be conducted required that the samples be dry. After some deliberation, it was decided that in order to minimize oxidation during drying the samples should be freeze dried (as opposed to being dried at room temperature). A portion of each sample was dried for approximately 8 days in a freeze dryer, the remaining portion of each sample was returned to the freezer. The dried portion of each sample was subsampled for the various analyses to be conducted on it, and was shipped to the appropriate facilities. 32 4.2 Ana ly t i ca l M e t h o d s P e r f o r m e d o n S a m p l e s A variety of analytical methods were used to characterize the nature of the tailings and in particular the form and nature of the arsenic in the tailings. The methods used in this study include whole rock analysis, scanning electron microscopy, Rietveld refined X-ray powder diffraction, sequential extractions, and X-ray Absorption Near Edge Structure using a Synchrotron light source. 4.2.1 Whole Rock Analysis Whole rock analysis is the simplest method of determining the total amount of each element in a sample. Total element concentrations, on their own, only provide general information on the types of minerals that may be present in a sample. A portion of each tailings type was dried for approximately 8 days in a freeze dryer and a representative subsample from each dried material was submitted to ALS C H E M E X in Vancouver for whole rock analysis using a four acid near total digestion. A 25-element scan was done on the samples that been HF- HNO3-HCIO4 acid digested and HCI leached. In addit ion, total organic carbon analysis was also conducted. 4.2.2 Scanning Electron Microscopy Using Scanning Electron Microscopy (SEM), mineral phases can be identified and the amount of each phase present can be determined qualitatively. The sample is bombarded by a beam of electrons, some of the electrons are adsorbed by the sample while some are scattered off the sample surface (backscattered electrons). The backscattered electrons create a greyscale image that can be used to identify individual mineral crystals. The heavier the unit weight of the mineral phase the brighter the mineral appears on the backscattered image. It is also possible to examine the texture of the individual grains. Most SEM's contain an energy dispersive spectrometer (EDS) that is used to collect the X-ray spectra emitted by the sample when it is hit by the beam 33 of electrons. The X-ray spectra can be analysed to determine the grain's elemental composit ion. The position of each peak in the spectrum identifies the elements that are present, while the relative height of each peak gives an indication of the concentration of each element in the grain. The detection limit for each element in the X-ray spectrum is about 1 wt%. Electron Microprobes have the ability to determine the quantitative chemical formula of minerals in a sample, however the amount of each mineral present can still only be determined qualitatively. A portion of each tailings type was dried for approximately 8 days in a freeze dryer and a representative subsample from each dried material was submitted to Vancouver Petrographies Ltd. for polished thin section preparation. Vancouver Petrographies Ltd. was instructed to prepare 26 x 46 mm sections using the submitted material without screening or pulverization. They were also instructed not to use water during the preparation of the sections in order to minimize potential oxidation and the dissolution of water-soluble minerals that may have been present in the samples. The polished thin sections were examined by a Phillips XL-30 scanning electron microscope (SEM) located in the Department of Earth and Ocean Sciences at the University of British Columbia. For most of the analysis the beam was set at a current of 20 kV in order to distinguish arsenic peaks, and a count of 60 seconds was used. 4.2.3 Rietveld Refined X-Ray Powder Diffraction X-Ray powder diffraction is also used to determine which mineral phases are present in a sample, however used alone it is not possible to quantitatively determine the amount of each phase present. Quantitative phase analysis using Rietveld refined X-Ray powder diffraction data is the most versatile method of quantitative phase analysis. Crystalline matter is composed of periodic arrays of 34 atoms in three dimensions. The crystal structure is determined using X-ray powder diffraction by passing a beam of monochromatographic X-rays through a crystal and recording the intensities and angles of the diffracted beams (Raudsepp and Pani, 2003). An X-ray diffraction pattern is produced with peaks that are a function of the size and symmetry of the crystalline unit cell of the substance, and with peak intensities that are a function of the atomic arrangement within the unit cell (Raudsepp and Pani, 2003). By comparing the positions and intensities of the peaks to a reference database the identities of the minerals contributing to the powder-diffraction pattern can be determined. The Rietveld method fits a simulated model to the diffraction pattern and uses a least-squares refinement to minimize the error between the modelled pattern and the actual pattern. The model that is fitted to the diffraction pattern is the sum of three models: a model for the shapes and widths of the diffraction peaks, a model for any aberrations in the shapes and positions of the peaks and a model for the background (Raudsepp and Pani, 2003). The models are obtained from a database, mineral phases are added into the model and the relative weight fraction of each phase is adjusted during the least squares reduction until the best fit is obtained. One weakness of the Rietveld method (and x-ray diffraction methods in general) is that it is not possible to identify amorphous phases; it is however possible to determine the quantity of amorphous phases present. The Rietveld method is most accurate when mineral phases are present in high weight percentages. The error increases as the weight percentage decreases and is possibly as high as 100% for percentages less than 1 wt% (Raudsepp and Pani, 2003). Freeze dried representative subsamples were submitted to M. Raudsepp of the UBC Department of Earth and Ocean Sciences for Rietveld-refined X-ray diffractometry. 4.2.4 Sequential Extractions A five-step sequential extraction procedure, designed specifically for arsenic, was conducted on 4 samples in duplicate (RLM-2-1, RLM-5, RLM-6-1 , and Secondary 35 Pond Sediments). Several arsenic extraction procedures were reviewed including Keon et al. (2001), Wenzel et al. (2001), Lombi et al. (1999), Gleyzes et al. (2001), and Loeppert et al. (2003). The extraction procedure selected was based on the method of Keon et al. (2001), slightly modif ied after the fourth step to reflect the methods of Wenzel et al. (2001), Lombi et al. (1999), and Gleyzes et al. (2001). The fractions in order were; S tep 1 : lonically Bound + Pore Water 1 M MgCI 2 , pH 8, room temperature, 2 hours (2 repetitions, 1 water wash) S tep 2: Strongly Adsorbed 1 M N a H 2 P 0 4 , pH 5, room temperature, 16 and 24 hours (1 repetition at each t ime, 1 water wash) S tep 3: Coprecipitated with acid volatile sulphides (AVS), Manganese oxides, and very amorphous iron oxyhydroxides 1 N HCI, room temperature, 1 hour (1 repetition, 1 water wash) S tep 4: Coprecipitated with amorphous iron oxyhydroxides 0.2 M ammonium oxalate/ 0.2 M oxalic acid, pH 3, room temperature in the dark (1 repetition, 1 water wash) S tep 5: Coprecipitated with crystalline iron oxyhydroxides 0.2 M ammonium oxalate/ 0.2 M oxalic acid/ 0.1 M ascorbic acid, pH 3, 30 minutes in water bath at 96 °C (1 repetition, 1 water wash) Reagents were prepared using disti l led-deionised water, and were de-aired in a nitrogen filled glove bag by bubbling nitrogen into the reagents. The pH of the reagents was adjusted inside an anaerobic chamber using environmental grade hydrochloric acid and sodium hydroxide. The extractions were carried out in 36 disposable 50 ml centrifuge tubes. Approximately 0.4 g (dry equivalent) of frozen sample was placed inside each tube and the first reagent was added to it inside the anaerobic chamber. Wet sediment was used as Keon et al. (2001), Buykx et al. (2000) and Zhang et al. (2001) indicated that all means of drying sediment can potentially result in changes in arsenic speciation. The tubes were sealed and shaken by hand, and then transferred to a shaker table for the duration of the extraction step. Tubes were centrifuged for 10 minutes at 3000 rpm at the end of each extraction step and then transferred back to the anaerobic chamber. The reagent was decanted from each tube and fi ltered, using 0.45 pm filters, into a sample bottle. The next reagent/water was added to each tube, the tubes were sealed and shaken and then transferred to the shaker table for the duration of the extraction step. Forty ml of reagent were used in each step, and 10 ml of de-aired, distilled-deionised water were used for the water rinses. For the water washes, the tubes were shaken by hand for several minutes after the water was added to them, the tubes were then centri fuged, transferred back to the anaerobic chamber, the water decanted and filter into a sample container, and the next reagent added. In step four, the tubes were covered with tinfoil and placed into a sealed box to exclude all light. In step five, the tubes were set in a test tube rack and placed in a water bath that had been preheated to approximately 96 °C. 4.2.5 X-ray Absorption Near Edge Structure X-ray absorption fine structure (XAFS) spectra were collected for all of the samples. X-ray absorption spectra were collected on April 28-29, 2003 and June 13-16, 2003 at the National Synchrotron Light Source located at Brookhaven National Laboratories, Upton, New York. The bending magnet beam line X11A (Navel Research Laboratory-Synchrotron Radiation Consort ium) was used. The 37 study was conducted through the Canadian Light source (CLS), and spectra were analyzed by CLS. A fixed exit double crystal monochromator with Si (111) crystals, detuned by approximately 15% to eliminate higher energy harmonics, was used to scan the energy region around the arsenic K-edge absorption energy ( E 0 = 11867 eV). The inflection point of the Au Lm-edge of a thin gold foil at 11919.7 eV was used as an internal energy scale reference. Both f luorescence yield (samples oriented at 45° with respect to the incident beam) using a Lytle detector and transmission modes (samples oriented perpendicular to the incident beam) were used for the collection of X-ray absorption spectra. Previously freeze dried samples were packed into a slit in a manufactured sample holder and covered on both sides with Kapton tape. XAFS spectra were collected over the photon energies from 11667 eV - 12825 eV, using 10 eV steps from 11667 - 11817 eV, and 0.75 eV steps from 11817 - 11917 eV (XANES region). At least 4 scans were collected for each sample and were averaged for the analysis. The f luorescence yield mode was used for all the tailings samples. All spectra were collected at ambient temperature and pressure. XAFS spectroscopy can be used to investigate the local coordination environment around the arsenic atom in a mineral phase, including the oxidiation state (Moldovan et al., 2003). In addition to Moldovan et al. (2003), McGeehan (1996), Rochette et al. (1998) and Reynolds et al. (1999) also used XAFS spectroscopy to speciate arsenic in soil solids. The technique has been shown to be effective in determining the molecular level speciation of arsenic over the concentration range of 50 mg/kg to several weight percent in mine tailings solids (Jiang, 2002). Synchrotron light sources are electron accelerators that confine high energy charged electrons traveling in a circular orbit at a speed close to that of light (relativistic speed) (Sham, 2002). When an electron is accelerated it produces 38 electromagnetic radiation. In a synchrotron, the electrons are accelerated centrifugally into a circular orbit using bending magnets. Synchrotron radiation is emitted tangential to the orbit as a result of the bending. A linear accelerator and a booster synchrotron are usually used to first accelerate the electrons to the desired energy of the storage ring. The pre-accelerated electrons are then injected into the storage ring that consists of straight and bending sections of stainless steel tubes kept under ultra-high vacuum (Sham, 2002). An atomic absorption edge occurs when an X-ray photon is absorbed in a single scattering event, resulting in the transfer of the photons energy to the production of a photoelectron escaping the atomic potential well (Jiang, 2002). Each element has specific binding energies of the atomic core level electrons (absorption edges, i.e. K-edge absorption energy). The binding energy for each element shifts slightly due to different oxidation states, in general the higher the oxidation state the higher the absorption edge energy. The absorption edges of different elements are well separated allowing the X-ray absorption spectra of different elements to be analyzed separately (Jiang, 2002). XAFS refers to the entire spectrum of absorption coefficient vs. photon energy. The region within approximately 50 eV of the absorption edge is referred to as the X-ray near edge structure (XANES) while the region above the near edge region is referred to as the extended X-ray absorption fine structure (EXAFS). A bound core level photoelectron is excited from the absorbing atom into a free electron state when an X-ray photon is absorbed. The excitation of the 1s (K edge) core state is one of the most commonly used absorption edges XANES is an element specific, non-destructive method that is very sensitive to the oxidation state, electronic structure and local symmetry of a mineral phase (Bancroft and Hallin, 2002). Model compounds are required to characterize the unknown oxidation states of the element of interest in a sample. By doing this the valence speciation and an estimate of composit ion can be made (Jiang, 2002). Data from multiple scans are overlain and averaged. The first step in the 39 data analysis is to remove the pre-edge background. Usually the pre-edge background is fit with a linear function of energy that is extrapolated into the post-edge region (Jiang, 2002). The linear function is then subtracted from the data. A kinetic energy zero point (E 0 ) must then be determined. In most cases the first peak of the derivative (first inflection point) is used for E 0 (Jiang, 2002). The data is then normalized using the determined E 0 value. The post-edge data background must be fit and subtracted from the data. A cubic spline is commonly used to fit the post-edge background. Three model compounds were used in the study: arsenopyrite (FeAsS; As 1 - ) , arsenic trioxide ( A S 2 O 3 ; A s 3 + ) , and iron arsenate (scorodite - FeAs04«2(H 2 0) ; A s 5 + ) . Principal Component Analysis (PCA), a linear algebraic technique, was used to semi-quantitatively determine the composit ion of the tailings samples (i.e. the number of unique components present within the spectra). The model compounds listed above were used along with a deconvoluted XANES spectra to semi-quantitatively determine the amount of arsenic present in each oxidation state in the various samples. The XANES spectra were deconvoluted using a linear least-squares fitting procedure (Kotzer, 2003). According to Kotzer (2003), linear least-squares fitting of XANES spectra has been shown to be a good technique to composit ionally determine the relative amounts of various oxidation states within complex materials. 40 4.3 Resu l t s a n d D i s c u s s i o n 4.3.1 W h o l e R o c k A n a l y s i s Table 1 shows a summary of the results from the ICP scan, complete results can be found in Appendix II. Arsenic concentrations range from 1180 mg/kg to 5690 mg/kg (0.12% - 0.57%). Historical tailings samples (RLM-2, RLM-3, and RLM-6) exhibit significantly higher arsenic concentrations than the samples containing newer tailings (RLM-1 , RLM-4, and RLM-5). The results from RLM-7 are somewhat irrelevant as the sample was taken prior to the removal of high arsenic content sulphides via flotation. The concentration of arsenic in the Primary and Secondary Pond sediments is 3000 mg/kg and 2950 mg/kg respectively, higher than in the new tailings samples. The high sulphur content in the RLM-2 samples, ranging from 1.86 - 2.63 %, indicates that these tailings were produced during a period of t ime when roasting and concentration of the sulphide portion of the ore had ceased. The remaining samples have relatively low sulphur contents, none greater than 1 % . The newer tailings have low sulphur contents as the majority of the sulphides are concentrated and removed from the tailings during the milling process for further gold recovery. The older historical tail ings likely have low sulphur contents as a result of the roasting process (sulphides were burnt off during roasting). Several interesting results are present in the Secondary Pond sediment data. The concentrat ions of copper, nickel and zinc are much higher in the Secondary Pond sediments than in the tailings samples. The copper concentration is 2330 mg/kg (0.2%) in the Secondary Pond sediment, the next highest Cu concentration in a tailings sample is 182 mg/kg at end of pipe (RLM-5). The nickel and zinc concentrations are 826 mg/kg, and 1860 mg/kg, respectively (the next highest Ni and Zn concentrations in a tailings sample are 192 mg/kg and 395 mg/kg). These results indicate that metals are concentrating in the pond sediments. The organic carbon concentration in the Secondary Pond sediments 41 is also relatively high, at 1.66%. The high organic carbon concentration is due to the vast amount of biological activity in the pond over the last few years. Tab le 1 : Summary of Whole Rock Analysis Results Sample As Cu '•Fe Ni Pb Zn S OC mg/kg mg,kg % mg.-kg mg/kg mg/kg % % RLM-1-1 1990 98 7.22 139 13 127 0.68 0.26 RLM-2-1 3230 104 9.77 174 68 179 1.86 0.18 RLM-2-2 4130 128 10.35 192 82 225 2.63 0.07 RLM-2-3 4100 125 9.93 176 114 258 2.34 0.09 RLM-3-1 3300 111 9.14 152 263 294 0.80 0.18 RLM-3-2 3210 119 9.16 142 347 300 0.80 0.18 RLM-3-3 2660 112 8.70 144 351 202 1.00 0.30 RLM-4 1180 111 7.71 138 146 148 0.72 0.11 RLM-5 1995 182 8.00 134 232 183 0.66 0.22 RLM-6-1 2630 121 9.01 171 66 395 0.99 0.25 RLM-6-2 2050 130 9.12 154 174 345 0.57 0.30 RLM-7 5690 133 8.03 168 142 335 1.34 0.02 RLM-SP 2950 2330 7.35 826 322 1860 0.58 1.66 RLM-PP 3000 491 7.32 320 393 1325 0.54 0.64 RLM-PP-BH 2150 303 6.34 104 140 184 0.74 0.18 4.3.2 R ie tve ld Re f i nemen t The results from the Rietveld refinement of the X-Ray powder diffraction data taken for the majority of the samples are presented in Table 2. The complete reports from the analysis including the Rietveld refinement plots can be found in Appendix III. The whole rock analyses indicate that none of the samples contain even as much as 1 wt% As, and it is therefore unlikely that the samples would contain 5% or more of an arsenic containing mineral phase that is necessary for accurate determination by Rietveld analysis. Some general conclusions can, however, be drawn from the analysis. The major mineral in all of the samples is quartz. Other major constituents include plagioclase, biotite, chlorite, dolomite and amphibole, with the new tailings containing considerably more dolomite and amphibole than the old 42 tailings. The unsaturated old tailings (RLM-2-1, RLM-3-1 , and RLM-6-1) contain gypsum which is an indication that sulphide oxidation has occurred. When sulphides are oxidized to sulphate, the sulphate will often precipitate with calcium present in the tailings to form gypsum. Tab le 2: Rietveld Refinement Results (wt %) Mineral RLM 1 RLM 2-1 RLM 2-3 RLM 3-1 RLM 3-3 RLM Ijjllj RLM \*'-' RLM '• 6-1 RLM ;6-2 RLM mmm 2' Pond Sed. 1' : Pond Sed. Quartz 36.5 43.5 48.7 49.2 50.5 33.6 32.1 46.9 42.9 32.3 37.3 36.0 Plagioclase 15.9 14.1 15.6 12.5 12.7 18.3 16.4 9.7 8.3 15.3 16.6 15.0 Biotite 4.0 5.8 4.8 6.5 7.3 3.5 5.7 5.0 6.6 5.3 9.3 10.7 Muscovite 3.6 4.7 5.9 Chlorite 3.5 8.0 5.1 14.8 8.8 4.5 7.9 10.4 12.8 8.0 18.0 15.1 Talc Gypsum 3.9 0.8 3.8 Calcite 0.8 1.1 3.0 0.9 2.4 2.0 Dolomite 24.1 9.1 10.6 9.1 11.9 22.0 21.2 10.2 14.9 19.0 11.3 13.2 Siderite 0.7 1.1 3.2 2.9 Amphibole 12.9 4.7 4.7 5.2 4.0 13.7 12.5 1.2 1.2 13.2 6.6 6.8 Pyrite 1.5 Arsenopyrite 1.9 0.8 0.2 1.6 1.7 1.2 0.2 0.3 Pyrrhotite 1.5 0.6 3.2 0.4 2.3 2.1 0.7 0.6 2.2 Titanite Hematite Magnetite 0.8 2.9 2.5 1.6 1.4 0.9 0.8 1.9 1.3 0.7 0.7 0.5 Goethite 1.4 0.2 0.7 0.5 0.4 Rutile 2.4 0.2 4.3.3 S c a n n i n g E lec t ron M i c r o s c o p y A scanning electron microscope (SEM) was used to identify the arsenic containing minerals in each tailings sample. After spending some t ime becoming familiar with the samples it was determined that all grains containing arsenic appeared bright under the SEM. After this was determined little t ime was spent looking at the less bright, grey particles. In general, there are three types of arsenic bearing minerals that appear bright: arsenopyrite being the brightest fol lowed by pyrite/pyrrhotite, and iron oxides (some containing arsenic). 43 RLM-1 - Downstream of SD#1 The only arsenic containing species detected in this sample was f ine-grained Arsenopyrite, found both as liberated grains and included with other minerals. Figure 6 shows a typical spot of the thin section. Particle (a) is pyrrhotite, particles (b) and (c) are iron oxide, and particle (d) is encapsulated arsenopyrite. The grains are clean, with no visible signs of weather ing. The grains are sparsely distributed throughout the sample, however, since arsenopyrite is 4 6 % As by mass (in the ideal formula), the total amount of arsenic found in the sample could be explained by the small percentage of arsenopyrite grains found (Jambor, 2003). Some iron oxide grains were located, however no arsenic was found to be associated with the iron oxide grains. These results were expected as the tailings are believed to have been non-roaster tailings that have remained saturated. 44 Figure 6: Backscattered Electron Image for R L M - 1 . Particle (a) is pyrrhotite, particles (b) and (c) are iron oxide, and particle (d) is arsenopyrite. Figures 7, 8 and 9 show the X-Ray spectrum patterns from RLM-1 for pyrrhotite, iron oxide, and arsenopyrite, respectively. Although these patterns were obtained from RLM-1 they are indicative of what was seen for these minerals in all of the samples. 45 Figure 7: X-Ray Spectrum for RLM-1 , showing pyrrhotite pattern. I^ X-ray Display i -•'File <r) .Edit view r) Setup window: M * — HLM_1_1A 16246 FS > Fe \ i L A n.n • B.O I 1 J 10.0 1S.0 20.0 !!_ ™ BJ £j v?L> ^ £j Coarse . KLW tines):' Snergy Wins) ' lin/tug ) Print) Figure 8: X-Ray Spectrum for RLM-1 , showing iron oxide pattern i^X-ray Display 1 ; FileVj Edit *•)• Wiew . Setup - Processing r) — * flL11_l_lL 15294 FS Fe 0 .^JL — J L I0.0 B.O 10.0 keV 1B.0 20.0 fas !: /start) _ Stop).! KLM iines } ; Energy »i JS L-11/..-.m s.int i 46 Figure 9: X-Ray Spectrum for RLM-1, showing arsenopyrite pattern ><X-ray Ditpl.iy 1 l l l i t l l l ^ 9296 FS * RLM_l_ld m 0.0 B.O 10 0 keV 1!> II 20. (J ^ v U - ^ ' ^ v*J <!i A ^ U Coarse " , '' \ \ .Jteft) Stop] =-_ KiM Lines) Energy Wins] _ M n / l o g ) ' P r i n t ) flL/W 2-7 - Downstream of SD#1, from 0-0.5 ft Weathering was observed on some grains, namely pyrrhotite and pyrite. Arsenic was found both as arsenopyrite and on the edges of weathered pyrrhotite/pyrite grains. Arsenic was also found in encapsulated particles that appeared to have been altered by cyanidation. These tailings were believed to have been derived from the roasting process, however upon examination it is more likely that the iron oxides in the sample were a result of oxidation of sulphides. Indeed, the quantity of iron oxide observed could easily have been produced by the oxidation of sulphides in the unsaturated tailings. Figure 10 shows a bright pyrite grain (at the center of the image) that has undergone oxidation and is surrounded by a rim of iron oxide. 47 Figure 10: Backscattered electron image for RLM-2-1 . Shows pyrite grain with a weathered rim of iron oxide (center of image) RLM-2-2 - Downstream of SD#1, from 1.5 - 2.0 ft Many large grains of arsenopyrite and pyrrhotite/pyrite were seen in this sample. The arsenopyrite and pyrrhotite/pyrite were found both as liberated and encapsulated grains. Arsenic was found predominately in the form of arsenopyrite. Some weathering of pyrrhotite/pyrite grains was visible, and a small amount of arsenic was found on the weathered edge of one grain. Figure 11 shows the large size of the sulphide grains. Particle (a) is arsenopyri te and is about 50 urn wide. Particle b is iron oxide, and particle c is pyrrhotite. Weathered edges can be seen all around particle (c). 48 Figure 1 1 : Backscattered Electron Image for RLM-2-2. Particle (a) is arsenopyrite, particle (b) is iron oxide, and particle (c) is pyrrhotite. The tailings in the RLM-2 samples were likely not roaster-derived. These tailings contain a higher than average amount of sulphide grains suggest ing that they were produced at a t ime when roaster operation had ceased and sulphide concentration was not being carried out. The upper few feet of these tailings have remained unsaturated and have undergone visible oxidation. Arsenic is predominately present as arsenopyrite, however, due to oxidation of the sulphides some arsenic has been released from the arsenopyrite and has become associated with iron oxide grains (iron oxides are produced during the sulphide oxidation process). 49 RLM -3-1 - Old tailings on north side of access road from 0-1 ft Arsenic was found predominately in small grained arsenopyrite. Spongy textured iron oxide particles with distinct rings were found, some of which contained arsenic. In addit ion, the outer ring of weathered pyrrhotite/pyrite was also found to contain a small amount of arsenic. Figure 12 shows a large mass of spongy material made up of arsenopyrite, iron oxides (both arsenic and non-arsenic bearing), pyrite and pyrrhotite. The spongy nature of this mass suggested that it was produced during roasting. F igure 12: Backscattered Electron Image for RLM-3-1 . Large spongy mass containing arsenopyrite, iron oxides, pyrite and pyrrhotite. 50 RLM -3-2 - Old tailings on north side of access road from 1.5-2.5 ft Many spongy textured iron oxide grains containing arsenic were found. Several arsenopyrite grains were also located. Figure 13 (a, b, and c) shows examples of the arsenic containing iron oxide material. The lighter coloured material in each of the particles is the arsenic bearing iron oxide, while the bright spots are arsenopyrite and pyrite. Figure 14 shows a typical X-Ray spectrum pattern for the arsenic bearing iron oxide material in this sample. The height of the arsenic peaks varied from grain to grain. F igure 13: Backscattered Electron Image for RLM-3-2. Particles (a), (b), and (c) show iron oxide grains that contain arsenic. 51 Figure 14: X-Ray Spectrum for RLM-3-2, showing arsenic bearing iron oxide pattern. ><X-iay Displ.iv 1 nlxl • f-Hfc r j _Cd1t View r) Setup v) 'Processing r) Wimtovi: _ 1 „ — * RLM_3_2 _4b 8232 FS Fi i ' As Ca I \ As 0 1 5 .0 • 10 .0 k«V 1 15 .0 20.0 <*J J j -±) -ZJ & * C*y K ! ^ ± | coarse' ; _ ' " S t a r t ) S t o p } L i n e s ) Energy wins) . t i n / L o g ) . P r i n t ) • . ' , ' - , ' , i , • . . . - . -RLM -3-3 - Old tailings on north side of access road from 3.0-3.5 ft A significant amount of arsenopyrite was detected along with some iron oxide grains. None of the iron oxide grains were found to contain arsenic, however upon further examination it is possible that arsenic containing iron oxide grains would have been found. This sample was saturated, where as RLM-3-1 and RLM-3-2 where not. This could explain why particles appeared unweathered and no arsenic containing iron oxides were observed. It is also possible that iron oxide grains may have undergone reduction as a result of the oxygen depleted conditions that exist in the saturated zone. Reduced iron oxide grains are soluble and would have dissolved into the surrounding groundwater. It appears as if the tailings located near the RLM-3 sample site contain material that has been roasted. Arsenic is present both as arsenopyrite and in arsenic 52 bearing iron oxides, however under reducing condit ions that develop in the saturated subsurface it is the arsenic associated with the iron oxides that is likely to be the most mobile and will cause the greatest problem in the near future. RLM-4 - Upstream of SD #1 in flow path of new tailings This sample contained relatively large grains with few small grains present. Most of the particles ranged between 50 - 200 um. The large particle size can be explained by the proximity of the sample location to the end of pipe. Indeed, larger particles settle out closer to the tailings discharge location while smaller particles travel further before being deposited. Arsenopyrite was found as liberated and encapsulated grains and was the predominant form of arsenic. Some small grains of highly heterogeneous and amorphous material containing As, Fe, Ca, Cl, S, O and other components were found. The arsenic was just barely detectable in these grains. It is speculated that these "junky" grains are secondary minerals formed during the milling process, either on their own or with the help of the addition of ferric in an attempt to form ferric arsenate. Figure 15 shows a typical section of the RLM-4 sample, where the bright spots are arsenopyrite, pyrite and pyrrhotite. 53 Figure 15: Backscattered Electron Image for RLM-4 RLM-5 - End of Pipe Discharge Small arsenopyrite particles were numerous in this sample. Small grains of iron oxide were also observed, but did not contain detectable arsenic. Some small particles of very heterogeneous and amorphous material containing As, Fe, Ca, Cl, S, O and other components were found, similar to the material seen in RLM-4, but contained more arsenic. This material was also found as a coating around pyrrhotite/pyrite grains. No iron arsenate grains were detected. Figure 16 (a) shows a mass of the "junky" arsenic material, while Figure 16 (b) shows a rim of the "junky" arsenic material surrounding a pyrrhotite grain. A typical X-Ray spectrum for this arsenic material can be seen in Figure 17. The higher arsenic content of the "junky" arsenic material in the RLM-5 sample compared to the RLM-4 sample may indicate that arsenic has mobil ized from the material located near RLM-4. 54 Figure 16: Backscattered Electron Image for RLM-5, (a) showing arsenic precipitate, and (b) showing rim of arsenic containing substance on pyrrhotite. F igure 17: X-Ray Spectrum for RLM-5, showing arsenic precipitate. \ X-ray Display 1 File ri Edit r ) view «•) Sew jjxjl Witiduw: KLM 5-4a 5393 FS Si Ca As 0.0 5.0 , 1 1 H .«•) . M i 10.0 keV n/t.09 i Print i 15.0 20.oa 55 RLM-6-1 - Revegetated Tailings east of Balmer Creek from 0-2 ft This sample consisted of relatively large grained material, however most of the arsenopyrite was found as small grains. Many particles with neatly defined rims where found in this sample. The rims contained arsenic, at relatively high concentrations in some cases, along with numerous other components including Fe, Ca, S, and O. Figures 18 and 19 show two nice examples of benign material (quartz, chlorite, dolomite, etc.) surrounded by a rim of arsenic containing material. The particles in Figures 18 and 19 are approximately 100 urn wide. Moderate amounts of iron oxide material were found both within a loose "spongy" state and in a more solid state. Much more arsenic was contained in the rims and in the iron oxide material in this sample then in any of the previous samples. Figure 18: Backscattered Electron Image for RLM-6-1, arsenic bearing iron oxide coating around quartz particle. 56 Figure 19: Backscattered Electron Image for RLM-6-1 , showing arsenic bearing iron oxide coating around quartz particle. RLM-6-2 - Revegetated Tailings east of Balmer Creek from 4-5 ft This sample was similar to RLM-6-1 in that a moderate amount of iron oxide particles were found, however the concentration of arsenic in these particles was much lower than in RLM-6-1 . Arsenopyrite was found mostly as small grains. An iron oxide particle with a relatively high arsenic concentrat ion was found, the digital image and X-Ray pattern can be seen in Figures 20 and 2 1 , respectively. In Figure 20, the bright particle near the center of the image is pyrrhotite, the particle to the left of it is a typical arsenic bearing iron oxide particle, while the large particle towards the bottom of the image is the iron oxide particle containing a high concentration of arsenic. Some iron oxide was found included in quartz, biotite, and other minerals, indicating that these particles have been roasted. 57 Figure 20: Backscattered Electron Image for RLM-6-2, showing pyrrhotite, and two arsenic bearing iron oxide particles. Figure 21: X-Ray Spectrum for RLM-6-2, showing high arsenic content iron oxide pattern. BO Ell f i l e - E d i t =- View ~ ) Setup r ) P r o c e s s i n g r ] Window: ;'j I As — BLM6_2_3b 2434 £047 FS Fe 58 The material found in RLM-6 was produced from the roasting process. Substantial arsenic is associated with iron oxide material and will provide a large source of readily mobile arsenic in an oxygen depleted environment (i.e. under reducing condit ions). RLM-7-1 - CIP Tailings Arsenic was found only as arsenopyrite. Large and small particles of arsenopyrite were found as well as l iberated and encapsulated grains. The large amount of arsenopyrite found in this sample was expected as at this point in the milling process the sulphides had not been removed from the tailings stream via flotation. A moderate amount of pyrrhotite/pyrite was also found, and as expected only a small amount of iron oxide material was observed. Primary Pond Sediment This sample was made up of f ine grained material. A small amount of arsenopyrite was found, along with a greater amount of pyrrhotite/pyrite. Some iron oxide grains were seen, however few contained arsenic. A few grains of iron oxide had a brighter ring surrounding them that contained arsenic. Figure 22 (a) shows arsenic bearing iron oxide material (lighter area's within the bright material), and in (b) a rim of arsenic oxide material surrounds an iron oxide particle. 59 Figure 22: Backscattered Electron Image for Primary Pond Sediment, showing arsenic bearing iron oxide particles in (a) and arsenic bearing iron oxide coating in (b). Primary Pond Backhoe This sample was similar to the Primary Pond Sediment sample, however one noticeable difference was that the grains were a bit larger. No iron oxide grains containing arsenic were found, arsenic was found only as arsenopyrite. Secondary Pond Sediment As in the Primary Pond Sediment, the grains were small (most less than 10 urn). Most of the sample appeared as the same shade of grey and consisted predominately of quartz, and chlorite. Both pyrite and pyrrhotite were found. A small amount of iron oxide was seen, however no arsenic was detected in it. The only arsenic bearing mineral detected was arsenopyrite. Work done by Jambor (2003) on tailings samples from the near by Cochenour mine indicated that the importance of iron oxides as a source of arsenic is greatly underestimated by X-Ray analysis using the SEM. He found that most of the iron oxide grains contained greater than 0.3 wt% when analyzed with the more sensitive microprobe. Some of these same grains did not show the presence of 60 arsenic using the SEM. He concluded that the proportion of As-bearing iron oxide particles is much higher than was est imated from his X-Ray analyses. He stated that the arsenic contained within roaster oxides is highly susceptible to mobil ization due to the porous texture of the roaster oxides and the sorbed association of the arsenic. He also stated that although weathering of arsenopyrite will continue to contribute arsenic to the surrounding pore waters over the long term, the roaster oxides will likely be responsible for contributing the bulk of the arsenic over the short term. The roaster tailings at the Red Lake mine are similar in nature to the roaster tailings at the Cochenour mine, therefore it can be inferred that the same sort of mechanism is occurring and will continue to occur at the Red Lake mine. In summary, material located around RLM-3, and RLM-6 is derived from the roasting process. A relatively large fraction of the arsenic in these locations (more so at RLM-6) is associated with iron oxide material and is and will continue to be mobile, especially under oxygen deficient condit ions (i.e. saturated condit ions). According to the f indings of Jambor (2003) the amount of arsenic bearing iron oxides determined through X-Ray analysis using the SEM is likely significantly less than what is actually present. This means that there is a high probability that the source of readily mobile arsenic is much larger than it appears through the results of the SEM work. The material sampled at RLM-2 does not appear to contain roasted material, however deeper down there is likely roaster material present. The material sampled at RLM-2 contains significant sulphides and the top few feet, which are unsaturated, have undergone substantial oxidation over the years. During the post deposit ional oxidation processes, some of the arsenic has been released from arsenopyrite and has become associated with iron oxide material, also formed during oxidation. The arsenic associated with the iron oxide material is likely to be much more mobile under reducing condit ions than the arsenic contained within arsenopyrite. 61 RLM-1, RLM-4, and RLM-5 contain essentially fresh tailings that have remained saturated. The vast majority of arsenic in these samples is contained within arsenopyrite that will continue to be stable as long as conditions remain reducing. A small amount of arsenic was found to be associated with a "junky" precipitate containing many species in the RLM-4 and RLM-5 samples. This material was likely formed during the milling process as a result of the addition of ferric iron. The stability of this material is unknown, however this material is likely to be a larger contributor of dissolved arsenic than arsenopyrite if the post depositional storage conditions for the fresh tailings remain saturated The material in the Primary and Secondary pond sediment was so fine that it was difficult to adequately characterize it. Past work performed on the pond sediment (Lorax, 2001) and analytical trends in the Secondary Pond water indicate that much of the arsenic in the Secondary Pond is associated with readily mobile iron oxide material. It is likely that this is the case in the Primary Pond as well. Under oxic conditions arsenic that is present in the pond water will naturally co-precipitate with iron that is also present in the water. In addition, arsenic and iron will reprecipitate from groundwater that is advecting and diffusing up into the ponds. Porewater profiles collected by Lorax (2001) indicate that arsenic and iron are reprecipitated in the Secondary Pond when the porewater from the pond sediments encounters the oxic interfacial layer. When conditions change (i.e. depleted oxygen), these precipitates readily dissolve releasing arsenic into the water column. The sequential extractions and XANES work performed on these samples give better insight into the composition of the precipitates. 4.3.4 Sequential Extractions A five-step sequential extraction procedure, slightly modified from Keon (2001), designed specifically for arsenic, was conducted on 4 samples in duplicate (RLM-2-1, RLM-5, RLM-6-1, and Secondary Pond Sediments). The fractions in order were: lonically Bound + Pore Water; Strongly Adsorbed; Coprecipitated with acid 62 volatile sulphides (AVS), Manganese oxides, and very amorphous iron oxyhydroxides; Coprecipitated with amorphous iron oxyhydroxides; Coprecipitated with crystalline iron oxyhydroxides; Residual. Samples were sent to SGS Chemex Environmental in Vancouver for low level arsenic and iron analysis using ICP - MS. The detection limits for aqueous arsenic and iron were 0.1 pg/L and 10 u,g/L, respectively. Based on these limits, and a sediment to extractant ratio of 0.4 g to 40 ml, the detection limits for extractable arsenic and iron in each step were less than 1 mg As/kg sediment and 2 mg Fe/kg sediment. Total arsenic and iron concentrations in each of the four sediment types tested were determined via near total four acid digestion at SGS Chemex in Vancouver, the results are shown in Table 3. The percent solids of each of the frozen samples was determined and used to calculate the dry equivalent mass of sample used in each of the extractions. Using the dry equivalent mass and the total arsenic and iron concentrations for each sample, the total mass of arsenic and iron, potentially available for extraction, was calculated. The data is tabulated in Table 3. Complete results can be found in Appendix IV. Tab le 3: Sequential Extraction Data IL) Location Average % Solids mass (wet) g mass (dry) g Total [As] mg/kg Total As mg Total [Fe] mg/kg Total Fe mg 1 2" Pond Sed 51.4 1.04 0.53 2670 1.426 71300 38.090 2 2 a Pond Sed 1.02 0.52 1.399 37.358 3 RLM-5 82.4 0.54 0.44 2015 0.896 74500 33.138 •i RLM-5 0.55 0.45 0.913 33.751 5 RLM-2-1 79.1 0.58 0.46 3440 1.578 89700 41.137 ti RLM-2-1 0.6 0.47 1.632 42.556 7 RLM-6-1 80.7 0.45 0.36 2550 0.926 85800 31.168 8 RLM-6-1 0.46 0.37 0.947 31.861 63 Reagent blanks were assayed and for the most part were found to contain negligible amounts of arsenic and iron. The Step 2 reagent (Nah^PCU) contained 91.8 u.g/L of arsenic. Solution assays that would have been affected by this contamination (Step 2b solutions of RLM-5) were adjusted by subtracting this value from the concentration reported for each of the solutions. The overall affect of this small amount of contamination was minor. The mass of arsenic in each of the solution samples obtained during the extraction process was calculated using the appropriate solution assay value and the measured volume of solution in each sample. The calculated mass of arsenic in each sample was divided by the dry equivalent mass of sample used in the extraction, to obtain a value in mg As/kg of sediment (Table 3). The calculated mass of arsenic in each sample was also divided by the total amount of arsenic in each sample (from Table 3) to obtain percent of total arsenic removed values. Repetit ions and water washes within each extraction step were added together to obtain the total amount of arsenic extracted in each step. The average was taken for the duplicates of each sediment type to obtain the values in Tables 4 and 5. Appendix IV contains all of the calculated data. Tab le 4: Percent of Total Arsenic Removed in Each Extraction Step Location Average °'c Total Arsenic Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 6 50 7 1 2 35 RLM-5 4 12 0 0 0 83 RLM-2-1 1 22 18 10 0 49 RLM-6-1 1 38 27 13 0 20 64 Tab le 5: Amount of Arsenic Removed in Each Extraction Step Location Average mg As/Kg dry sediment Step 1 Step 2 Step 3 Step 4 ! Step 5 Residual Secondary Pond Sediment 161 1334 176 20 53 926 RLM-5 86 233 9 1 1 1681 RLM-2-1 19 754 605 360 3 1700 RLM-6-1 36 978 686 338 12 499 The values shown in Tables 4 and 5 are depicted graphically in Figures 23 and 24. Secondary Pond Sediment A small fraction (6%) of the arsenic was ionically bound/exchangeable (including the aqueous fraction contained in the pore water of the sample) in the Secondary Pond Sediment. The majority of arsenic in this sample was strongly adsorbed (Step 2), 5 0 % or 1334 mg/kg was removed during the second extraction step. Only seven percent of the arsenic was coprecipitated with acid volatile sulphides (AVS), carbonates, Mn oxides, and very amorphous Fe oxyhydroxides (Step 3). Minimal amounts of arsenic were coprecipitated with amorphous Fe oxyhydroxides (Step 4) and crystalline Fe oxyhydroxides (Step 5). The residual fraction of the arsenic was calculated to be 3 5 % and was likely associated with sulphides. In summary, approximate 6 5 % (1744 mg As/kg Sediment) of the arsenic contained in the Secondary Pond Sediment is likely to be fairly mobile under the changing redox conditions that exist in the Secondary Pond. RLM-5 The vast majority of the arsenic in the RLM-5 samples (End of Pipe discharge) was not mobil ized by the extraction procedure. A small fraction (4%) of the arsenic was ionically bound/exchangeable (including the aqueous fraction contained in the pore water of the sample) and 12% was strongly adsorbed (Step 2). Negligible amounts of arsenic were removed during Steps 3, 4, and 5. The 65 remainder of the arsenic, 8 4 % (1685 mg As/kg sediment), was calculated to be the residual fraction and was likely associated with sulphides. In summary, only 16%, or 330 mg As/kg sediment of the arsenic contained in End of Pipe tailings (RLM-5) is likely to be fairly easily mobil ized, indicating that the vast majority of the arsenic in these tailings should remain stable as long as saturated condit ions are maintained. RLM-2-1 A small fraction (1%) of the arsenic was ionically bound/exchangeable (including the aqueous fraction contained in the pore water of the sample) in the RLM-2-1 sediment. Twenty two percent (754 mg As/kg sediment) of the arsenic was strongly adsorbed and was extracted in Step 2. A relatively significant portion of the arsenic (18%) was coprecipitated with acid volatile sulphides (AVS), carbonates, Mn oxides, and very amorphous Fe oxyhydroxides (Step 3). An additional 10% of the arsenic was coprecipitated with amorphous Fe oxyhydroxides (Step 4) and a negligible amount was coprecipitated with crystalline Fe oxyhydroxides (Step 5). In summary, approximately 2 3 % of the arsenic in this sample is adsorbed, 2 8 % is associated with iron oxyhydroxides, and the remaining 4 9 % (1700 mg As/kg sediment) is likely associated with sulphides. Under the currently unsaturated condition of these tailings the sulphide portion will continue to oxidize, releasing arsenic from sulphides only to be immobil ized by sorption on iron oxides phases. RLM-6-1 A small fraction (1%) of the arsenic was ionically bound/exchangeable (including the aqueous fraction contained in the pore water of the sample) in the RLM-6-1 sediment. The largest fraction of the arsenic (38%) was strongly adsorbed and Was extracted in Step 2. A significant portion of the arsenic (27%) was coprecipitated with acid volatile sulphides (AVS), carbonates, Mn oxides, and 66 very amorphous Fe oxyhydroxides (Step 3). In addition, 13% of the arsenic was coprecipitated with amorphous Fe oxyhydroxides (Step 4). A negligible amount was coprecipitated with crystalline Fe oxyhydroxides (Step 5). In summary, approximately 3 9 % of the arsenic in this sample was adsorbed, 4 0 % was associated with iron oxyhydroxides, and the remaining 2 0 % (2051 mg As/kg sediment) was likely associated with sulphides. Only 2 0 % of the arsenic contained in this sample was calculated to be remaining after the extraction was complete, significantly less than all the remaining samples indicating that the large majority of the arsenic in RLM-6-1 sediment is likely to be fairly easily mobil ized. 67 Figure 23: Percent of Total Arsenic Removed in Each Extraction Step Secondary Pond Sediment RLM-5 RLM-2-1 RLM-6-1 Figure 24: mg As/kg Sediment Removed in Each Extraction Step 1800 1600 H Secondary Pond Sediment RLM-5 RLM-2-1 RLM-6-1 68 Similar calculations were made for iron to determine the percent and amount of iron removed in each extraction step for each sample. This was done for comparison purposes, as arsenic and iron concentrat ions are usually l inked. Tables 6 and 7 show the percent and amount of iron removed in each step, respectively, and Figure 25 depicts the percent or iron removed. Tab le 6: Percent of Total Iron Removed in Each Extraction Step Location Average % Total Iron Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 0 8 26 2 7 57 RLM-5 0 8 23 6 5 58 RLM-2-1 0 4 26 20 4 46 RLM-6-1 0 8 18 23 5 46 Tab le 7: Amount of Iron Removed in Each Extraction Step Location Average mg Fe/Kg dry sediment Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 192 5482 18674 1695 4694 40563 RLM-5 29 5921 17163 4622 3741 43024 RLM-2-1 133 3682 23419 18103 3380 40984 RLM-6-1 32 6648 15854 19529 4636 39101 The reagents used in the extraction procedure were selected specifically for arsenic extraction, therefore not all of the steps are relevant to iron (namely steps 1 and 2). Steps 3, 4, and 5 can be used to give and indication of the fraction of iron associated with AVS/very amorphous iron oxyhydroxides, amorphous oxyhydroxides, and crystalline oxyhydroxides. The amount of iron extracted in Step 3 (AVS/very amorphous iron oxyhydroxides) ranged from 18 - 26%, with RLM-6-1 containing the least amount. RLM-2-1 and RLM-6-1 contained significantly more amorphous iron oxyhydroxides (20% and 23%, respectively) than the other two samples. All the samples contained relatively small amounts of crystalline iron oxyhydroxides (4 - 7%). The Step 4 and 5 iron and arsenic 69 results correspond well, confirming that Secondary Pond Sediment and RLM-5 contain much less amorphous iron oxyhydroxides than RLM-2-1 and RLM-6-1 . F igure 25: Percent of Total Iron Removed in Each Extraction Step 7 0 T —r- — r - —i 1 60 ] Secondary Pond Sediment RLM-5 RLM-2-1 The sequential extraction results support the SEM results fairly well. Both of the RLM-2-1 and RLM-6-1 samples were found to contain arsenic bearing iron oxyhydroxide material in the S E M work, with much more being found in the RLM-6-1 sample. The sequential extractions indicate that 5 1 % of the arsenic in the RLM-2-1 sample and 8 0 % of the arsenic in the RLM-6-1 sample is sorbed or precipitated with iron oxyhydroxides. An adequate SEM analysis of the Secondary Pond Sediment could not be done due to the fine grained nature of the sample. Historical t rends in Secondary Pond water chemistry however, indicate that a large fraction of the arsenic in the sediment must be associated with a readily mobile iron oxyhydroxide phase. The sequential extraction results indicate that over 6 5 % of the arsenic in the Secondary Pond Sediment is sorbed or precipitated with an iron oxyhydroxide phase, with more than 5 0 % of the arsenic being sorbed. 70 SEM analysis of RLM-5 tailings indicated that the vast majority of the arsenic was associated with arsenopyrite. A small fraction of the arsenic was found to be associated with a "junky" precipitate. The sequential extraction data supports this result as more than 8 3 % of the arsenic reported to the residual phase (likely associated with sulphides) while the remaining 16% was ionically bound or adsorbed. 4.3.5 Synchrotron X-Ray Absorption Spectroscopy The Canadian Light Source prepared a report summarizing the results of the XANES data collected at the National Synchrotron Light Source located at Brookhaven National Laboratories, Upton, New York. The bending magnet beam line X11A (Navel Research Laboratory-Synchrotron Radiation Consort ium) was used. The following is a summary of the results reported in Kotzer (2003). Three arsenic oxidation states were found in the samples: As" 1 (as in arsenopyrite), A s 3 + (as in arsenic trioxide), and A s 5 + (as in iron arsenate). An arsenopyrite sample from the high grade zone of the Red Lake Mine was provided for use as a reference material. Figure 26 shows the XANES spectra for each of the three arsenic oxidation states found in the samples. 71 Figure 26 : Arsenic K-edge XANES spectra of three different model compounds with different oxidation states. FeAsS — i < 1 > 1 < 1— 11860 11870 11880 11890 Photon Energy (eV) Table 8 summarizes the results of the least squares fitting for all of the samples. Tab le 8: Semi-quantitative Arsenic Distribution (+/- 1 0 % ) . Calculated from linear least-squares fitting of the As K-edge Synchrotron XANES spectra. Location % Total Arsenic As(-1) As(lll) As(V) RLM-1 89 0 11 RLM-2-1 31 11 58 RLM-2-2 85 8 7 RLM-2-3 88 8 4 RLM-3-1 28 4 68 RLM-3-2 22 5 73 RLM-3-3 93 2 5 RLM-4 78 7 15 RLM-5 77 9 14 RLM-6-1 20 16 65 RLV.-6-2 63 10 28 RLM-7 92 0 8 Secondary Pond Sediment 25 40 35 Primary Pond Sediment 50 11 40 Primary Pond Backhoe 84 7 9 72 RLM-2 W h o l e rock a n a l y s i s a n d S E M resu l t s i n d i c a t e d tha t t h e ta i l i ngs l o c a t e d at R L M - 2 c o n t a i n e d a la rge a m o u n t of s u l p h i d e ma te r i a l a n d w e r e d e p o s i t e d at t i m e w h e n r o a s t i n g a n d s u l p h i d e c o n c e n t r a t i o n p r a c t i c e s h a d c e a s e d . T h e S E M a n d R ie tve ld resu l ts i n d i c a t e d tha t s ign i f i can t pos t d e p o s i t i o n a l o x i d a t i o n of t h e s u l p h i d e s h a d o c c u r r e d in t h e u n s a t u r a t e d layer o f t h e ta i l i ngs . T h e s e q u e n t i a l ex t rac t i on resu l t s a l s o i n d i c a t e d tha t a la rge po r t i on of t h e a r s e n i c in R L M - 2 - 1 w a s s o r b e d o r a s s o c i a t e d w i t h i ron o x y h y d r o x i d e p h a s e s . T h e X A N E S resu l t s c o r r e s p o n d we l l w i t h t h e s e o b s e r v a t i o n s . R L M - 2 - 1 h a d o n l y 3 1 % of t h e a r s e n i c in t h e A s ( - 1 ) o x i d a t i o n s ta te , w i t h 5 8 % p r e s e n t a s A s ( V ) . R L M - 2 - 2 a n d R L M - 2 - 3 w e r e s im i la r in c o m p o s i t i o n to e a c h o the r , w i th m o r e t h a n 8 0 % of t h e a r s e n i c p r e s e n t a s A s ( - 1 ) , w i t h t h e r e m a i n i n g po r t i on p r e s e n t a s a c o m b i n a t i o n of A s ( l l l ) a n d A s ( V ) . T h e s e resu l ts i nd ica te tha t s ign i f i can t s u l p h i d e o x i d a t i o n h a s on l y o c c u r r e d in t h e u p p e r m o s t l ayer of t h e ta i l i ngs in t h e v ic in i ty of R L M - 2 . F i g u r e s 2 7 , 2 8 , 2 9 , a n d 3 0 s h o w t h e X A N E S s p e c t r a fo r t h e R L M - 2 s a m p l e s . Figure 27: Arsenic K-edge XANES spectra of RLM-2 series samples —i 1 1 •——r 1 < 1— 11850 11860 11870 11880 11890 Photon Energy (eV) Figure 28: Fitted Arsenic K-edge spectrum of RLM 2-1 Figure 29: Fitted Arsenic K-edge spectrum of RLM 2-2 F igure 30: Fitted Arsenic K-edge spectrum of RLM 2-3 Fit — i 1 1 1 1 1 1— 11860 11870 11880 11890 Photon Energy (eV) 75 FILMS Historical information and SEM results indicate that the material in the vicinity of RLM-3 consists of roaster tailings, with numerous particles of arsenic bearing iron oxide detected via the SEM in both the RLM-3-1 (0 - 1 ft below ground surface) and RLM-3-2 (1.5 - 2.5 ft below ground surface) samples. The XANES result for RLM-3-1 and RLM-3-2 are similar with only 2 8 % of the arsenic in the form of As(-1) in RLM-3-1 and 2 2 % in RLM-3-2. A small amount (< 5%) of the arsenic existed in the form of As(l l l ) , with the remainder present as As(V). These results correspond well with a material that has been derived from a roasting process. During the roasting process the material is subjected to extremely Oxidizing condit ions, therefore it is expected that the majority of the arsenic should be present in the As(V) form in a roasted material. The XANES results, therefore, support the theory that the material sampled in RLM-3-1 and RLM-3-2 is roaster-derived. RLM-3-3 (3 - 3.5 ft below ground surface) had a much different XANES spectra than the other two RLM-3 samples. Most of the arsenic (93%) was present in the form of As(-1), with 5% present as As(V) and 2 % as As( l l l ) . This result would seem to indicate that the deeper tailings in the vicinity of RLM-3 were not roaster-derived. However, historical reports indicate that at the point in t ime when tailings began to be discharged near RLM-3 the roaster was in operation. As well, the low sulphur content of the RLM-3-3 sample also indicates that the material has been roasted. A possible explanation for the vast difference in arsenic speciation between the deep and shallow tailings at RLM-3 is that the unstable, readily mobile oxidized arsenic species in the tailings could have dissolved into the pore water in the saturated zone (RLM-3). The total arsenic concentration at RLM-3-3 was 2660 mg/kg, while the concentration at RLM-3-2 was 3210 mg/kg. The difference in concentration between the two samples can be reasonably explained by the 76 above theory (oxidized arsenic species dissolving into pore water). However the difference in concentrations in not great enough to entirely explain the difference in the fraction of arsenic seen as As(-1) in the two samples. It is possible that some of the oxidized arsenic species have been reduced to As(-1) and have formed secondary sulphide mineral species. RLM-6 Historical information and SEM results indicate that the tailings deposited in the area of RLM-6 were roaster-derived. Only 20% of the arsenic at RLM-6-1 was present in the form of As(-1), with 16% present as As(lll) and 65% present as As(V). The XANES spectra for RLM-6-2 was much different than for RLM-6-1 with 63% of the arsenic in the form of As(-1). As in RLM-3, the total arsenic concentration in the saturated zone (RLM-6-2) was less than in the unsaturated zone (RLM-6-1), with total concentrations of 2050 mg/kg and 2630 mg/kg, respectively. The difference in total arsenic concentration and the difference in the distribution of arsenic species indicates that a fraction of the oxidized arsenic has dissolved into the pore water in the saturated zone. Elevated arsenic concentrations have been observed in the pore water in the vicinity of RLM-6, therefore this is plausible. The difference in concentrations however, only accounts for a small fraction (about 5%) in the difference in the fraction of As(-1). A possible explanation for the remaining difference is that oxidized arsenic species have undergone post depositional reduction and may have formed secondary sulphide minerals. Fresh Tailings Samples RLM-1, RLM-4, RLM-5, and RLM-7 consisted of fresh or relatively new tailings. The arsenic in RLM-7 tailings was predominately in the As(-1) oxidation state (arsenopyrite), with only 8% of the arsenic in the As(V) state and no arsenic in the As(lll) state. The small amount of As(V) was likely produced as a result of 77 oxidation of arsenopyrite during the milling process. RLM-5 consists of RLM-7 tailings that have gone through a cyanide detox process, have had ferric iron added to them, and passed through a sulphide flotation circuit that removes a large portion of the sulphide material from the tailings. The proportion of oxidized arsenic in the RLM-5 tailings was higher than in RLM-7 tailings, with 9% in the form of As(l l l) and 14% in the form of As(V). RLM-4 had a similar composit ion to RLM-5, as was expected as RLM-4 tailings were located just downstream of the End of Pipe discharge (RLM-5). RLM-1 tailings were more similar in composit ion to RLM-7 than RLM-4 or RLM-5. RLM-1 contained no As(l l l ) , and 8 9 % of the arsenic was in the form of As(-1). These results all compare well with what was seen via the SEM, and via sequential extractions in the case of RLM-5. Pond Sediments The Primary Pond sediment contained 5 0 % As(-1), 1 1 % As(l l l ) , and 4 0 % As(V). The vast majority of the oxidized arsenic species were likely formed via the natural precipitation of arsenic oxyhydroxides formed from dissolved arsenic and iron species in the water column. The As(-1) fraction is a result of the tailings that are lining the bottom the Primary Pond. The Primary Pond Backhoe sample, which is a sample of the deeper tai l ings/sediment in the primary pond had a similar composit ion to RLM-2-3. This result makes sense, as the tailings in the Primary Pond were likely to have been the same as RLM-2-3 when they were deposited. Since the tailings are saturated there is limited opportunity for the sulphide (arsenopyrite) fraction to undergo oxidation. However, the total arsenic concentration in the Primary Pond Backhoe sample (2150 mg/kg) was significantly less than in the RLM-2-3 sample (4100 mg/kg), in addition the sulphur content of the Primary Pond Backhoe 78 sample is much lower than in the RLM-2-3 sample indicating that the tailings in the Primary Pond were deposited at a time when sulphide concentration (i.e. sulphide flotation was a part of the milling process) was occurring. The Secondary Pond sediment contained 25% As(-1), 40% As(lll), and 35% As(V). Past work performed on the pond sediment (Martin, 1996) and analytical trends in the Secondary Pond water indicate that much of the arsenic in the Secondary Pond is associated with readily mobile iron oxide material. Under oxic conditions arsenic that is present in the pond water will naturally co-precipitate with iron that is also present in the water. In addition, arsenic and iron will reprecipitate from groundwater that is advecting and diffusing up into the ponds. Porewater profiles collected by Lorax (2001) indicate that arsenic and iron are reprecipitated in the Secondary Pond when the porewater from the pond sediments encounters the oxic interfacial layer. When conditions change (i.e. depleted oxygen), these precipitates readily dissolve releasing arsenic into the water column. These results correspond well with the sequential extraction data for Secondary Pond Sediments. The aqueous concentration of arsenic in the Secondary Pond is seen to cycle seasonally with higher concentrations appearing in the warmer summer months when biologically activity limits the oxygen content of the pond bottom waters resulting in the reductive dissolution of oxidized arsenic species. 4.4 C o n c l u s i o n s Historical tailings samples (RLM-2, RLM-3, and RLM-6) exhibited significantly higher arsenic concentrations than the samples containing newer tailings (RLM-1, RLM-4, and RLM-5). The high sulphur content in the RLM-2 samples, ranging from 1.86 - 2.63 %, indicates that these tailings were produced during a period of time when roasting and concentration of the sulphide portion of the ore had ceased. The unsaturated old tailings (RLM-2-1, RLM-3-1, and RLM-6-1) contain gypsum, an indicator that sulphide oxidation has occurred. 79 The tailings sampled at RLM-2 do not appear to contain roasted material, however deeper down there is likely roaster material present. The material sampled at RLM-2 contains significant sulphides and being unsaturated, the top few feet have undergone significant oxidation over the years. RLM-2-1 had only 3 1 % of the arsenic in the As(-1) oxidation state, with 5 8 % present as As(V). From the sequential extractions it was found that approximately 2 3 % of the arsenic in RLM-2-1 is adsorbed, and 2 8 % is associated with iron oxyhydroxides. RLM-2-2 and RLM-2-3 were similar in composit ion to each other, with more than 8 0 % of the arsenic present as As(-1), with the remaining portion present as a combination of As(l l l ) and As(V). Based on the above information, the surface of this deposit has undergone significant oxidation, resulting in the majority of the arsenic present having undergone post depositional transformation from arsenopyrite to and oxidized form. Deeper in the deposit the majority of arsenic has not undergone oxidation and exists as arsenopyrite. The unoxidized material will be most stable under a water cover. It is likely, however, that roaster tailings are present below the depth of material sampled in this study (based on historical tailings deposition information and observed elevated groundwater arsenic concentrat ions in the area). Based on the results of the roaster material sampled at RLM-3 and RLM-6, the majority of arsenic present in this material is likely in an oxidized form, and will be most stable under oxidizing conditions. The roasted material, however, is already contained within the saturated zone of this deposit (severely elevated groundwater arsenic concentrations have been observed in the vicinity of the RLM-6 sampling location), therefore flooding the surface of the deposit will not alter redox condit ions and will likely not affect the release of arsenic from this material. In summary, f looding of the dry tailings beach in the Primary Pond is not likely to result in a net increase in dissolved arsenic concentrations. The material located around RLM-3, and RLM-6 was produced f rom the roasting process. A large fraction of the arsenic at these locations (more so at RLM-6) is 80 associated with iron oxide material and is and will continue to be mobile under the oxygen deficient condit ions (i.e. saturated) that exist in the tailings deposits (extremely elevated arsenic concentrat ions in the vicinity of RLM-6 have been observed). It would be difficult to maintain aerobic condit ions in these subaerial tailings deposits due to the fine grained nature of the tailings, the existence of perched water tables, and the vast amount of biological activity (marsh) growing on top of the tailings. Less than 3 0 % of the arsenic in RLM-3-1 and RLM-3-2 was in the form of As(-1). A small amount (< 5%) of the arsenic existed in the form of As(l l l ) , with the remainder present as As(V). Most of the arsenic in the saturated zone in the area of RLM-3 (RLM-3-3) was present in the form of As(-1), with 5% present as As(V) and 2 % as As(l l l ) , suggesting that the oxidized forms of arsenic present in the unsaturated zone have been mobil ized and transformed to As(-1) in the saturated zone. Only 2 0 % of the arsenic contained in RLM-6-1 the form of As(-1), with 16% present as As(l l l ) and 6 5 % present as As(V). According to the sequential extractions, 3 9 % of the arsenic in RLM-6-1 was adsorbed while 4 0 % was coprecipitated with iron oxyhydroxides. The saturated tailings in the area of RLM-6 (RLM-6-2) contained 6 3 % of the arsenic in the form of As(-1) suggesting that a considerable portion of the oxidized arsenic has been mobil ized, and transformed into As(-1). RLM-1 , RLM-4, and RLM-5 contain essentially fresh tailings that have remained saturated. The vast majority of arsenic in these samples is contained within arsenopyrite that should continue to be stable as long as condit ions remain saturated. A small amount of arsenic was found to be associated with a " j unky precipitate, containing many species, in the RLM-4 and RLM-5 samples. 81 Based on the sequential extraction data, only 16% of the arsenic contained in End of Pipe tailings (RLM-5) is likely to be fairly easily mobil ized under the existing saturated conditions. The XANES spectra indicate that 2 3 % of the arsenic in RLM-5 tailings is in the oxidized form. All of the solids identification results indicate that the vast majority of the arsenic in the fresh tailings produced at the mine site should remain stable under saturated conditions. RLM-4 tailings were similar in composit ion to RLM-5 tailings. RLM-1 tailings contained no As( l l l ) , and 8 9 % of the arsenic was in the form of As(-1). The Primary Pond sediment contained 5 0 % As(-1), 1 1 % As(l l l ) , and 4 0 % As(V). The material f rom the Primary Pond (Primary Pond Backhoe sample), contained significantly more arsenic in the As(-1) form. Approximately 6 5 % of the arsenic contained in the Secondary Pond Sediment is likely to be fairly mobile under the changing redox condit ions that exist in the Secondary Pond. The sequential extraction results indicate that over 6 5 % of the arsenic in the Secondary Pond Sediment is sorbed or precipitated with an iron oxyhydroxide phase, with more than 5 0 % of the arsenic being sorbed. The Secondary Pond sediment contained 2 5 % As(-1), 4 0 % As(l l l ) , and 3 5 % As(V). 82 5.0 In situ and Laboratory Experiments A series of experiments were conducted to investigate the behaviour of arsenic in several tailings types under various conditions. The objective of the experiments was to see if arsenic could be stabilized under reducing conditions, ideally in the form of arsenopyrite or arsenian pyrite. In general, arsenic can be stabilized in two ways: • Form a ferric oxide solid phase that will adsorb As 5 + and keep it under oxidizing conditions • Form an arsenic sulphide phase (e.g. arsenopyrite or arsenian pyrite) and keep it under reducing conditions Unfortunately, it is likely not possible to maintain oxidizing conditions throughout a tailings deposit in the natural environment in the long term (even if the tailings are not covered with water). Tailings are fine grained in nature and perched water tables often develop in the deposits. In addition, traditional, non-engineered tailings ponds (i.e. unlined ponds formed in natural valleys and creek beds) are in contact with biological activity. Tailings pond waters are often rich in nutrients (as a result of the reagents used in the mining and milling processes) that promote biological activity, which can bring about reducing conditions. Maintaining oxidizing conditions throughout the tailings deposits at the Red Lake Mine would likely be impossible, especially considering that some of the tailings deposits are covered with water. Field and laboratory experiments were conducted to investigate whether or not it is possible to stabilize arsenic through the precipitation of an arsenic sulphide phase. 83 5.1 In s i t u C o l u m n E x p e r i m e n t The Primary and Secondary Ponds were constructed on top of historical tailings. The surface sediments in the Primary and Secondary Ponds contain a large fraction of arsenic associated with iron oxyhydroxides. Seasonal increases in aqueous arsenic concentration are seen, predominately in the Secondary Pond water, with increases occurring in the summer and winter. The water covering the pond sediments limits oxygen transport into the tailings (remedy for acid rock drainage is to cover tailings with water as dissolved oxygen has a much lower diffusivity in water than in air). Limiting the flux of oxygen into the tailings helps to bring about reducing condit ions, which has been shown to lead to the mobil ization of arsenic from oxidized tailings. Bacteria play a major role in the development of reducing conditions. As described previously, bacteria use dissolved organic carbon as an electron donor to reduce various chemical species for energy. The Secondary Pond has become biologically active, due to the low concentrat ions of cyanide and dissolved metals present, the elevated concentrat ions of nitrogen and phosphorous present and the addition of Beaver Pond water through the Diversion Ditch. The Diversion Ditch water has introduced a fish population into the Secondary Pond that has resulted in additional biological growth. The biological growth sinks to the bottom of the pond once it dies and provides a source of organic carbon. With an ample source of organic carbon and limited oxygen flux, biologically mediated reducing condit ions can develop. As a result of the reducing conditions, arsenic is released from the sediment. The exact mechanism of release of arsenic from iron oxide phases is not completely understood. A combination of reductive dissolution of the iron oxide ( F e 3 + being reduced to F e 2 + resulting in the solubilization of the iron oxide phase and the release of sorbed arsenic) and the direct reduction of arsenate to arsenite (arsenite sorbtion to iron oxide phases at near neutral pH is far less than arsenate sorption (Pierce and Moore, 1982)) is thought to occur. Another result 84 of the reducing conditions is the formation of sulphide via the reduction of sulphate. Any sulphide (S 2") formed (as a result of sulphate reduction) would be expected to combine with reduced dissolved metal species, such as F e 2 + , to form low -solubility sulphide species (eventually pyrite). Little evidence of arsenic removal as an arsenic sulphide species has been documented (Smedley and Kinniburgh, 2002), and it is unclear as to why high dissolved arsenic concentrat ions are observed under reducing condit ions in the presence of sulphide. There are numerous studies that document the increase in aqueous arsenic concentrat ions following the development of anaerobic condit ions in sediments containing arsenic bearing iron oxyhydroxides. These studies include: Deuel and Swoboda (1972), McGeehan and Naylor (1994), Azcue and Nriagu (1995), McCreadie et al. (2000), and Martin and Pedersen (2002). McCreadie et al. (2000) saw increased arsenic concentrat ions in a sulphate reducing zone of the Campbel l Mine tailings impoundment. Meng et al. (2003) indicate that biotic reductions can convert arsenic and sulphide into arsenian pyrite, although there is limited evidence of this occurring in natural systems. Martin and Pedersen (2002) report that in the deeper sediments of Balmer Lake arsenic is consumed as an authigenic sulphide phase. In the shallow sediment, arsenic is released to the surface water due to seasonal anoxia that develops in the near surface pond sediments (Martin and Pedersen, 2002). It is speculated that condit ions may not be reducing enough in some situations to cause the formation of an arsenic sulphide phase, and or not enough sulphide is available (not enough sulphate available to be reduced to sulphide) to precipitate all of the arsenic (plus other dissolved metal species). 85 Pyrite formation in low temperature sedimentary environments has been studied intensively, however there is only limited information available on arsenopyrite or arsenian pyrite formation in low temperature sedimentary environments. 5.1.1 Methods A relatively inexpensive method for conducting an in situ experiment to investigate the geochemistry behind the dissolution of arsenic from the tailings located at the bottom of the tailings ponds was developed. A low-cost l imnocoral was developed by sinking pieces of 8 inch (20.3 cm) diameter pipe that extended above the surface of the pond into the tailings sediment, resulting in the isolation of a column of pond water and underlying sediments. Potential experimental locations were investigated in the Secondary Pond, however no feasible spot was located. An access road needed to be built to the experiment location so that a backhoe could sink the pieces of pipe into place. The logical location for the experiment was off the Primary Dam. Erosion of the Primary Dam, which provides access to the pond, caused the nearby tailings to be covered with sand and gravel, making them inappropriate for the study. Other locations in the pond were too deep, did not contain tailings, or would require too long of an access road to be built. It was decided that the experiment could be much more easily conducted in the Primary Pond as the water was shallower and only a short access road would be needed. On July 3 r d , 2002 four columns of water were isolated in the Primary Pond. The columns consisted of 10-foot lengths of 8-inch diameter f ibreglass pipe and were located about 20 feet upstream of the Primary Dam. The pieces of pipe were taken to the correct position in a boat, and held in place while the backhoe operator pushed the pipes into the tailings with the bucket of the hoe. The pipes were sunk approximately 3 feet into the tailings, in about 5 feet of water, leaving 2 feet of pipe to stick up above the water surface. 86 Figure 31 depicts an installed column. One column was used as an unaltered control, and different treatments were added to the other three columns. Thin tubing outfitted with a filter at one end (landscape cloth was used as a filter) and weighted down with fishing weights was used as sampling ports in the columns. One length of tubing was placed in each column prior to the addition of the treatment layer (bottom sample), and another length of tubing was placed in the columns after the treatment layer had been added (middle sample), enabling sampling at the tailings/treatment layer interface and the treatment layer/water interface. Aqueous samples could also be taken from the surface of the columns using a sampling stick or a pump (surface sample). Figure 31: Diagram of an installed column Sampling Locations Tailings 87 Column #1 (C1) was chosen as the control column. Approximately 150 dry grams of peat were added to Column #2 (C2), and approximately 150 dry grams of peat and 200 g of sulphate (in the form of gypsum from crushed drywall) were added to Column #3 (C3). The treatment added to Column #4 (C4) consisted of approximately 100 dry grams of peat, 200 g of sulphate (in the form of gypsum from crushed drywall), and 5 kg of zero valent iron (in the form of fine iron filings). Peat was used as a source of organic carbon, gypsum was added to ensure that sufficient sulphate was present to be reduced to sulphide, and zero valent iron was chosen as a strong reductant. The treatments were added on July 4 t h , 2002. Prior to adding the treatments the depth of water was measured and the pH, temperature, specific conductance, dissolved oxygen, and oxidation/reduction potential were measured using a Hydrolab probe. The Hydrolab probe was calibrated using pH 4 and 7 buffers, a 1000 mv specific conductance standard, and a Thermo Orion ORP standard. Hydrolab readings were taken again on July 6 t h, July 8 t h, and July 10 t h, and then once per week until August 27 t h. Aqueous samples were taken from the top and bottom of the columns on July 4 t h , and the pond surface prior to treatment additions to the columns. Aqueous samples from the top and bottom of the columns were tested for dissolved metals, including arsenic, and surface water samples were also tested for chloride, nitrate and sulphate. Samples were taken again from all locations within the columns and the pond surface for a complete analysis suite (including dissolved metals, ammonia, chloride, dissolved organic carbon, nitrate, phosphate, and sulphate) on July 26 t h, August 22 n d , and September 18 t h. On August 9 t h samples were taken for dissolved metals analysis only. Samples were obtained by connecting the thin tubing installed in the columns to a peristaltic pump. All samples were analyzed by Envirotest Laboratories Inc. located in Thunder Bay, Ontario (a CAEAL certified laboratory). Samples for dissolved metals and dissolved organic carbon analysis were field filtered using 0.45 urn syringe filters. Metals analysis was done by ICP-OES, anion analysis was done was by ion chromatrography, ammonia analysis was 88 done by colourimetry, and dissolved organic carbon analysis was done by the method APHA 5310 B. 5.1.2 C o l u m n E x p e r i m e n t Resu l t s Complete results from the column experiments can be found in Appendix V. The pH in the columns prior to the addition of the treatment layers (July 4 t h , 2002) was approximately 7.7. After the treatments had been added the pH in C1 - C4 was 7.7, 5.5, 5.7, and 6.0, respectively (measurements taken on July 6 t h , 2002). The results from each sampling port (bottom, middle, and surface) will be discussed separately. Figure 32 shows the dissolved arsenic concentration at the bottom sampling port (tail ings/treatment layer interface) for C1 - C4. The arsenic concentration is normalized to the initial value in each column (i.e. concentration expressed as the ratio of arsenic concentration divided by the initial arsenic concentration). The average concentration of dissolved arsenic at the bottom location in the control column (C1) was 1.08 mg/L, and was relatively constant throughout the duration of the experiment. The arsenic concentration in C2 remained relatively constant around a value of 1.18 mg/L until September 1 8 t h when the concentration increased to 1.72 mg/L. As can be seen from Figure 32, the arsenic concentration at the bottom sampling location in C3 increased to a maximum value of 3.33 mg/L on August 2 2 n d (nearly three t imes the initial value) and showed a slight decrease to 3.04 mg/L on September 2 1 s t . In C4 the arsenic concentration dropped below the detection limit (<0.02) prior to the first sampling t ime (July 26 t h ) and was not detectable throughout the remainder of the experiment. 89 Figure 32: Dissolved Arsenic Concentrat ion at the Bottom Sampling Port Bottom (Top of tailings layer) 03-Jul 13-Jul 23-Jul 02-Aug 12-Aug 22-Aug 01-Sep 11-Sep 21-Sep - • - C I ( C O N ) - » - C 2 ( P e a t ) - A - C3 (Pea t -Gypsum) - x - C4 (Pea t -Gypsum-Fe) The average dissolved organic carbon concentration was elevated above the control concentration of 11 mg/L to 21 mg/L in C2 and 33 mg/L in C3, due to the addition of peat. The DOC concentration in C4 was only slightly elevated and averaged 13 mg/L. The average sulphate concentration in C3 and C4 was 1650 and 1573 mg/L, respectively which, due to the addition of gypsum, was significantly higher than the dissolved sulphate concentration of 440 mg/L in C 1 . The addition of gypsum to C3 and C4, in the form of crushed up drywall, resulted in an increase in the concentration of dissolved strontium. The concentration of dissolved Sr in C3 and C4 averaged approximately 3.1 mg/L while the Sr concentration in the control column was 0.797. Nitrate is a redox sensitive species and can give and indication of the redox status of the water. Nitrate is one of the first species to be depleted in the development of reducing condit ions, therefore if condit ions were reducing at the bottom of the columns no nitrate would be seen. The concentration of nitrate was below detection (<0.03 mg/L) at the bottom sampling port in C3 and C4 90 indicating that at least some what reducing condit ions had developed. The nitrate concentration in CT and C2 was 6.31 mg/L and 4.27 mg/L, respectively, indicating that conditions were still oxidizing in these columns. Figure 33 shows the concentration of dissolved iron at the C3-bottom sampling location. As can be seen f rom the figure the concentration of iron increases in a similar manner as the concentration of arsenic in C3, strongly suggesting that an arsenic containing iron solid phase was being dissolved. The iron concentration at C3 increased from 0.584 mg/L to 11.4 mg/L. The presence of increasing dissolved iron concentration in C3 indicates that condit ions are reducing enough for ferric iron to be reduced to ferrous iron. In addit ion, during sampling on September 18 t h , a distinct hydrogen sulphide gas smell was noticed, indicating that sulphate was being reduced to sulphide. The iron concentration in C2 averaged 0.458 mg/L while the control iron concentration equaled 0.02 mg/L. The iron concentration in C4 was only slightly elevated above the control concentration (0.07 mg/L). It is believed that the redox status in C4 was strongly reducing as a result of the addition of zero valent iron. Under strongly reducing condit ions, arsenic-iron oxides are reductively dissolved and iron and arsenic may be reprecipitated as a reduced solid phase (possible containing sulphide). It is possible that this process was occurring in C4. It is also possible that the arsenic and iron were precipitated on the surface of the zero valent iron in the form of an iron arsenate species. Su and Puis (2001) found zero valent iron to be effective at removing both arsenate and arsenite from solution and also was found to degrade nitrate. According to Oblonsky et al. (2000) zerovalent iron corrodes in solution forming products such as magnetite and maghemite on the Fe° surface. Su and Puis (2001) report that zero valent iron removes arsenic from solution via adsorption of the arsenic onto corrosion products present on the surface of the fillings, they described the adsorption with as a first order reaction. 91 Concentrations of dissolved manganese were slightly elevated over the control concentration in all of the columns. The concentration of cadmium increased from 0.013-0.041 mg/L in C2 and from 0.014 - 0.075 mg/L in C3. Figure 33: Concentration of Dissolved Iron at the C3 Bottom Sampling Port C3(Peat-Gypsum)-Bottom 03-Jul 23-Jul 12-Aug 01-Sep 21-Sep Figure 34 shows the dissolved arsenic concentration at the middle sampling port (treatment layer/water interface) for C1 - C4. The arsenic concentration is normalized to the initial value in each column (i.e. concentration expressed as the ratio of arsenic concentration divided by the initial arsenic concentration). There was no "middle" sample location in the control column, as a treatment layer was not added, therefore the middle sample concentrations in C2-C4 are compared with the surface concentrations in C1. The average concentration of dissolved arsenic at the surface location in the control column (C1) was 1.08 mg/L, and decreased slightly as the experiment progressed. The arsenic concentration in C2 averaged 0.925, initially decreasing to a minimum value of 0.68 mg/L on August 22 n d , then increasing to slightly greater than the starting value. In C3 the concentration of arsenic decreased to a minimum value of 0.49 mg/L on August 9 t h, then began to increase up to a value of 1.17 mg/L. Up until August 22 n d , the treatment layer was able to stop the 92 released arsenic, from the tailings in C3, from migrating up into the overlaying water column. In C4 the arsenic concentration dropped below the detection limit (<0.02) prior to the first sampling t ime (July 26 t h ) and was not detectable throughout the remainder of the experiment. F igure 34: Concentrat ion of Dissolved Arsenic at the Middle Sampling Port Middle (Water - Treatment Layer interface) 03-Jul 13-Jul 23-Jul 02-Aug 12-Aug 22-Aug 01-Sep 11-Sep 21-Sep - • - C I - ( C O N ) - r a - C 2 ( P e a t ) - ± - C 3 ( P e a t - G y p s u m ) C 4 ( P e a t - G y p s u m - F e ) The average DOC concentrations at the middle sampling port were elevated in C2 and C3 to 25 mg/L and 20 mg/L, respectively over the control value of 10 mg/L. The average sulphate concentration in C3 and C4 was 1076 mg/L, which was lower than the bottom sulphate concentration in C3 and C4 but still significantly higher than the control sulphate concentration of 440 mg/L. Strontium concentrations were also elevated at in C3 and C4 at the middle sample location, averaging 2.48 mg/L and 1.81 mg/L, respectively, while the control concentration averaged 0.796 mg/L. 93 Significant concentrations of nitrate were present in all of the middle sampling locations averaging 6.4, 2.4, 4.0, and 4.3 mg/L in C1 - C4 respectively. The presence of nitrate may indicate that condit ions were oxidizing at the treatment layer/water interface in all of the columns, or that kinetic controls prevented the reduction of nitrate. The iron concentrat ion was elevated in C2 averaging 1.11 mg/L while the control iron concentrat ion equaled 0.014 mg/L. The iron concentration was also elevated in C3 and C4, averaging 0.41 and 0.26 mg/L, respectively (the presence of ferrous iron indicates reducing condit ions). Concentrat ions of dissolved manganese were slightly elevated over the control concentration in all of the columns, while the concentration of cadmium was approximately equal to the control concentration in all the columns except C4 in which all the cadmium had been depleted. Figure 35 shows the dissolved arsenic concentration at the surface of the columns. The arsenic concentrat ion is normalized to the initial value in each column. As can be seen from the figure, the arsenic concentrat ions in C1 - C3 were nearly constant and similar to each other throughout the duration of the experiment. In C4 the arsenic concentration dropped below the detection limit (<0.02) prior to the first sampling t ime (July 26 t h ) and was not detectable throughout the remainder of the experiment. 94 Figure 35: Concentration of Dissolved Arsenic at the Surface of the Columns Surface (Top of water column) 1.2 i 03-Jul 13-Jul 23-Jul 02-Aug 12-Aug 22-Aug 01-Sep 11-Sep 21-Sep - » - C 1 ( C O N ) -B - C 2 ( P e a t ) - A - C 3 ( P e a t - G y p s u m ) H * - C 4 ( P e a t - G y p s u m - F e ) The average DOC concentrat ions at the surface of C2 and C3 averaged 18 mg/L and 17 mg/L, respectively, which was elevated over the control value of 10 mg/L. The DOC concentration at the surface of C4 was slightly elevated over the control value, and averaged 12 mg/L. The average sulphate concentrat ions in C3 and C4 were 1053 mg/L and 1083 mg/L, respectively, still significantly higher than the control sulphate concentrat ion of 440 mg/L. Strontium concentrations were also elevated in C3 and C4 at the surface sample location, averaging 1.35 mg/L and 1.80 mg/L, respectively, while the control concentration average 0.796 mg/L. Significant concentrations of nitrate were present in all of the surface water samples from the columns averaging concentrations of 6.4, 5.8, 5.1 and 4.1 mg/L in C1 - C4 respectively. The presence of nitrate indicates that condit ions were oxidizing at the surface of all of the columns. This is expected as the surface of the water is exposed to atmospheric oxygen. Concentrat ions of dissolved manganese were slightly elevated over the control concentration in all of the 95 columns, while the concentration of cadmium was approximately equal to the control concentration in all the columns except C4 in which all the cadmium had been depleted. In summary, the addition of a strong reductant (iron fillings) was effective in reducing the dissolved arsenic concentrat ion, and prevented the release of additional arsenic into the water column. The exact mechanism that resulted in the removal of arsenic from the water column in C4 (iron fillings column) is unclear. The arsenic may have been adsorbed onto an oxidized iron corrosion product on the surface of the fillings or may have been incorporated into a reduced iron phase (that may or may not contain sulphide). In column C4, arsenic was likely removed from solution through precipitation/adsorption with an oxidized iron phase at the treatment/ layer interface and upwards through the water column. The combination of the addition of organic carbon and sulphate (C3) resulted in the development of reducing condit ions at the tail ings/treatment layer interface leading to the reductive dissolution of oxidized iron/arsenic phases and the development of high concentrat ions of dissolved arsenic and iron. Al though the redox potential was obviously low enough to bring about the reduction of iron, it is unclear if significant sulphate reduction occurred (hydrogen sulphide gas odour was observed during sampling on some occasions). These results allow one to conclude that released arsenic and iron are not being effectively removed through the precipitation of iron/arsenic/sulphide phases, in the C3 (peat and sulphate treatment). It is not clear what these results imply for the behaviour of arsenic in the primary pond. The isolated water column was narrow, which limited mixing. In reality the pond waters are usually well mixed which could result in a reduced effectiveness of the iron fillings treatment. In addit ion, the sides of the columns create an artificial media for precipitates to bind too that would not be present in the pond. A larger scale field experiment would be required to confirm the effectiveness of the iron filling treatment. 96 To better understand the mechanisms responsible for arsenic release and removal, a laboratory experiment was designed to augment the results of the insitu column testing. 5.2 Designed Laboratory Experiments A series of experiments were conducted to investigate how the manipulation of various factors influenced the stability of solid phase arsenic bearing species. The objective of the experiments was to determine how arsenic concentrations in solution would be affected by adjusting factors that were suspected to have an influence on arsenic stability in tail ings samples from the Red Lake Mine. Six factors were identified from research papers and from field observations at the site. The factors selected were: concentration of dissolved organic carbon, concentration of sulphate, concentration of dissolved oxygen, presence of zero-valent iron, tailings source, and presence of elemental sulphur. The experimental method for the designed laboratory experiments was based on the methods of Reynolds et al. (1999), McGeehan (1996), Dowdle et al. (1996), Rochette et al. (1998), Rochette et al. (2000), Guo et al. (1997) and Rittle et al. (1995). Dowdle et al. (1996) identify lactate as the most effective source of metabolically available organic carbon. Rochette et al. (1998) synthesized various arsenate minerals, subjected them to reducing condit ions (by f looding them in the presence of soil containing organic carbon), and determined the relative solubility of the substances. They found that under reducing condit ions scorodite ( F e A s 0 4 * 2 H 2 0 ) was the most soluble mineral. The iron arsenate underwent reductive dissolution releasing As(l l l ) to solution and solid phases (Rochette et al., 1998). Rochette et al. (2000) studied the effect of aqueous sulphide on arsenate minerals. It was found that the presence of sulphide brought about the rapid reduction of arsenate (more so at low pH) and lead to the formation of dissolved arsenic - sulphide complexes that 97 persisted for days (Rochette et al., 2000). The formation of orpiment only occurred at high S:As ratios. Rittle et al. (1995) explored whether arsenopyrite could be formed in the Milltown Reservoir sediments by enhancing bacterial sulphate reduction (additional sulphate and organic carbon were added to the sediments in laboratory experiments). It was found that both sulphate and organic carbon amendments were required in order for arsenic to be removed as a sulphide phase (some arsenopyrite with a stoichiometry of approximately 1:1:1 was detected through SEM analysis (Rittle et al., 1995). Reynolds et al. (1999) also detected arsenopyrite formation in their experiments involving the f looding of soils amended with organic carbon. As described earlier, organic carbon is an energy source for microorganisms. The presence of sufficient organic carbon can result in the development of reducing conditions. Mildly reducing conditions are known to cause certain arsenic bearing mineral phases (namely iron oxyhydroxides) to dissolve, resulting in the release of arsenic into solution. In addit ion, at pH < 6, arsenate (As(V)) is much more readily adsorbed onto mineral surfaces than arsenite (As(l l l)). Under reducing condit ions, sorbed arsenate will begin to be reduced to arsenite, resulting in the release of previously sorbed arsenic into solution. Under more strongly reducing conditions, it is possible that sulphide minerals such as pyrite, arsenopyrite, and arsenian pyrite will form. Arsenic could potentially be sequestered into these stable sulphide mineral phases if condit ions were appropriate. A high concentration of dissolved organic carbon, in the form of lactate (a readily available form of organic carbon) was used in the experiments (there was some additional organic carbon initially present in the tailings samples and pond water used in the experiments). For sulphide minerals to form there needs to be an ample source of sulphur present that can be easily reduced to sulphide. A high concentration of sulphur, 98 in the form of sulphate was used in the experiments to ensure that sulphide mineral formation would not be limited by a lack of sulphur. The atmosphere that the experiments were conducted under was manipulated in order to mimic oxygen sufficient, and oxygen deficient conditions that may develop in the tailings ponds. Zero valent iron was added with the intention that it would act as a strong reductant, and would bring about strongly reducing condit ions in the experiments, potentially resulting in the formation of sulphide minerals. Iron fillings were added in the in situ column experiments and as a result all arsenic was removed from solution and the arsenic release from the sediments ceased. By using iron fillings in the laboratory experiments, the mechanism by which arsenic was removed could be observed. The tailings source is an important factor in studying the release of arsenic f rom the tailings. Various types of tailings are located around the mine site, 4 were selected for the experiment. RLM-5 was selected because it represents the tailings that are currently being produced at the mine site. RLM-7 was selected in order to observe the difference between the final tail ings currently produced, and the tailings produced prior to the Detox circuit and ferric addition. RLM-2 was selected as it represents tailings that are currently unsaturated in the Primary Pond that may be f looded in the near future. In addition, it was believed at the time that the RLM-2 tailings may have been roaster-derived. Secondary Pond sediments were selected as it has been shown that a large amount of arsenic is released from the sediment every year. Elemental sulphur was added based on research conducted into sedimentary pyrite formation. All of the literature reviewed stated that in order to get pyrite to form it was necessary to add elemental sulphur or polysulphides. If only sulphide 99 was added, iron monosulphides would form but would not be converted into pyrite. 5.2.1 Methods The design of experiment approach was used to evaluate the factors described above, in the most efficient and defensible way. Designed experiments allow for the testing of several variables at one time while being able to determine which factors are significant. By using the design of experiment approach, the individual effects of a factor as well as the combined effects of that factor with other factors, can be determined in the most efficient manor. One of the classes of experimental designs is the two level factorial design. Two level factorial designs, known as 2 n , develop a linear equation that relates some response to various factors. Y = f ( X 1 , X 2 , X n ) + e Where Y is the response, Xj is the factor, n is the number of factors and e is the error term. For three factors the equation would be: Y = ao + a i X i + a 2 X 2 + 83X3 + a 4 X i X 2 + 35X1X3, aeX 2 X 3 + a y X i X 2 X 3 + e Where a.\ is a constant and the other variables have the same meaning as above. By conducting the appropriate tests, the constants can be calculated to yield a model that gives an estimate of the response. By performing a statistical analysis of the results the significant factors and interactions can be determined. This type of experimentation is referred to as a two-level factorial design because each factor is only tested at two levels, low and high. In a full two-level factorial design all possible combinat ions of levels are run to determine the above constants. For 3 factors the number of combinations is 2 3 = 8. If the number of factors is large it is usually desirable to conduct only a fractional two-level 100 factorial design, such as a half or quarter fraction. In a quarter fraction factorial design 2 n " 2 runs are performed. In fractional designs not all of the interaction effects can be determined as there is not enough information, however, higher level interactions are rarely significant, and the most valuable information can still be determined from the reduced number of experimental runs performed. Two, 6 factor, 14 fraction, two level factorial experiments were run simultaneously. If the full fraction designed experiment was conducted this would result in 64 runs. By doing a VA fraction, only 16 runs for each experiment (32 in total) had to be conducted. Table 9 shows the design matrix for the experiment. The factor levels are coded as e i t h e r - 1 (low level) or +1 (high level). For example, run 1 would contain DOC at the high level, Sulphate at the low level (no addition), would have a low level atmosphere (sealed with rubber stopper), low level zero valent iron (no addition), tailings at the low level (in this case RLM-5 tailings), and elemental sulphur at the high level. 101 Tab le 9: Design Matrix for Laboratory Experiments Trial DOC S04 Atm Fe(0) Tailings S(0) Number A B C 0 E F 1 1 -1 -1 -1 -1 1 2 -1 -1 -1 -1 1 -1 3 -1 1 1 -1 1 -1 4 -1 1 1 1 -1 -1 '••"•5 -1 -1 1 -1 -1 1 6 -1 1 -1 1 1 1 7 1 -1 1 1 -1 -1 8 1 -1 1 -1 1 -1 9 1 -1 -1 1 1 1 10 1 1 -1 -1 1 -1 11 1 1 1 1 1 1 12 -1 -1 -1 1 -1 -1 13 1 1 -1 1 -1 -1 14 1 1 1 -1 -1 1 15 -1 -1 1 1 1 1 16 -1 1 -1 -1 -1 1 17 -1 -1 1 -1 -1 1 18 1 -1 -1 1 1 1 19 1 -1 1 -1 1 -1 20 -1 -1 -1 1 -1 -1 21 -1 1 -1 1 1 1 22 -1 1 -1 -1 -1 1 23 1 -1 -1 -1 -1 1 24 -1 1 1 -1 1 -1 25 1 -1 1 1 -1 -1 26 -1 1 1 1 -1 -1 27 1 1 1 1 1 1 28 1 1 -1 -1 1 -1 29 1 1 1 -1 -1 1 30 -1 -1 1 1 1 1 31 1 1 -1 1 -1 -1 32 -1 -1 -1 -1 1 -1 Each run of the experiment was conducted in 250 ml Erlenmeyer flasks. A solution of approximately 2 g of sodium lactate syrup in 100 ml of water was created. High level DOC runs received 1 ml of this solution at the start of the experiment (approximate addition of 20 mg/L DOC). For sulphate, an assumption was made that the Secondary Pond water used in the experiment contained about 250 mg/L sulphate. In order to achieve 1000 mg/L of sulphate in the high sulphate runs, 0.25 g of sodium sulphate was added. Zero Valent iron 102 was first washed with distilled water to remove the fine iron dust, then 15 g (wet) or iron fillings was added to the high level flasks. Low level atmosphere condit ions were created by capping flasks with Teflon lined rubber stoppers. Flasks were only opened in an anaerobic chamber to minimize oxygen influx into the flasks. High level atmosphere conditions were created by leaving the f lasks open to the atmosphere throughout the duration of the experiment (flasks were loosely capped with a sponge stopper in an attempt to minimize evaporation). In the first set of experiments (runs 1 - 1 6 ) RLM-5 represented the low level tail ings source, while RLM-7 represented the high level. In the second set of experiments (runs 17 - 32) RLM-2 represented the low level tail ings source, while Secondary Pond Sediments represented the high level. One gram of elemental sulphur was added too each of the high level f lasks. Fifty grams (dry equivalent) of each tailings type was added to the appropriate flask and secondary pond water was added to make the total amount of solution equal to 225 ml (taking into account the water contained within the wet solids). The appropriate amount of each reagent was added to the flasks. The flasks were stirred and shaken to thoroughly mix the ingredients. The pH in each flask was adjusted to 7.0 using environmental grade HCI. The flasks were capped (rubber stoppers for low level atmosphere, foam stoppers for high level atmosphere) and placed on a shaker table. Buckets of water, and additional f lasks full of water were placed inside the shaking table unit in an attempt to minimize evaporation from the foam capped flasks. Figure 36 shows a digital image of the flask set up on the shaker table. 103 Figure 36: Designed Experiment Laboratory Setup Photograph Initial samples were taken on Day 1, with additional samples being taken after 7 days, 21 days, and finally after 42 days. During the day 7 sampling, the pH was adjusted to 6.5 and an additional 1 ml of 2 g/100 ml sodium lactate solution was added to the high level DOC flasks. During the day 21 sampling, the pH was also readjusted to 6.5. At the final sampling t ime (day 42), solid samples were also taken from the f lasks. The solid samples were immediately f rozen after being removed f rom the flasks. 5.2.2 Results Distinct visual changes had occurred in the flasks by day 2 1 . In the first set of samples, f lasks #6, #9, # 1 1 , and #15 contained black material on the bottom of the flasks with silvery shiny spots present. In addit ion, an orange coating had appeared on the upper portion of the glass flask. All of these f lasks had both iron fillings and elemental sulphur present, and were constructed with RLM-7 tail ings. Flasks #4, #7, #12, and #13 had a darker shading present on the bottom of the 104 f lasks. This darker shading was believed to be just the iron fillings, as all four of these flasks had iron filing added to them. Figure 37 shows flask #6, depicting the orange coating on the glass walls of the f lask and the bottom black layer with silvery shiny spots. F igure 37: Flask #6 after Day 21 In the second set of samples, f lasks #18, #27, and #30 contained black material, with f lasks #18 and #27 also containing spots of a silvery shinny substance. All three flasks contained iron fillings and elemental sulphur and were constructed with Secondary Pond sediments. Flasks #18 and #27 contained DOC while flask #30 did not. Flask #21 also contained iron fillings, and elemental sulphur, however black material did not appear to have formed in this flask. In all f lasks containing Secondary Pond Sediment, distinctive layering of the solids was present and bubbles were entrapped throughout the sediment layer. The black material is believed to be iron monosulphides. The shiny spots may have been arsenopyrite. 105 A r s e n i c a n d i ron c o n c e n t r a t i o n s w e r e m e a s u r e d u s i n g a g r a p h i t e f u r n a c e a t o m i c a d s o r p t i o n s p e c t r o p h o t o m e t e r . S a m p l e s w e r e d i l u ted u s i n g 1 % e n v i r o n m e n t a l g r a d e nitr ic a c i d . B l a n k s a n d s t a n d a r d s w e r e a l so p r e p a r e d u s i n g 1 % e n v i r o n m e n t a l g r a d e ni t r ic a c i d . T h e d e t e c t i o n l imi ts fo r a r s e n i c a n d i ron w e r e 0 .05 m g / L a n d 0 .01 m g / L , respec t i ve l y . A r s e n i c c o n c e n t r a t i o n s w e r e m e a s u r e d in al l t h e s a m p l e s , w h e r e a s i ron c o n c e n t r a t i o n s w e r e m e a s u r e d o n l y in t h e D a y 1 s a m p l e s a n d in t h e D a y 4 2 s a m p l e s ( d u e to t i m e c o n s t r a i n t s a n d m a c h i n e ava i lab i l i t y ) . D i s s o l v e d o r g a n i c c a r b o n c o n c e n t r a t i o n s w e r e m e a s u r e d u s i n g a C a r l o E r b a N A - 1 5 0 0 E l e m e n t a l A n a l y z e r in t h e D a y 1 a n d D a y 4 2 s a m p l e s . S u l p h a t e a n d n i t ra te c o n c e n t r a t i o n s w e r e m e a s u r e d in t h e D a y 1 a n d D a y 4 2 s a m p l e s a s w e l l , u s i n g Ion c h r o m a t o g r a p h y . T h e a v e r a g e s u l p h a t e c o n c e n t r a t i o n fo r e a c h ta i l i ngs t y p e w a s d e t e r m i n e d fo r f l a s k s t h a t h a d s u l p h a t e a d d e d to t h e m a n d fo r f l a s k s t h a t d id not , t h e v a l u e s in m g / L a re s h o w n in T a b l e 10 . C o m p l e t e s u l p h a t e resu l t s c a n b e f o u n d in A p p e n d i x V I . T h e c o n c e n t r a t i o n of s u l p h a t e in t h e S e c o n d a r y P o n d w a t e r u s e d in t h e e x p e r i m e n t w a s 2 7 6 m g / L . Tab le 10: A v e r a g e S u l p h a t e C o n c e n t r a t i o n ( m g / L ) fo r E a c h T a i l i n g s T y p e Tailings Type Dnv - Day 42 No Sulphate Sulphate No Sulphate Sulphate RLM-5 640 1925 676 2550 RLM-7 682 2189 1149 1778 RLM-2 1679 2550 2536 3591 Secondary Pond 633 2168 1759 3320 T h e a v e r a g e d i s s o l v e d o r g a n i c c a r b o n c o n c e n t r a t i o n f o r e a c h ta i l i ngs t y p e w a s d e t e r m i n e d f o r f l a s k s t h a t h a d lac ta te a d d e d to t h e m a n d f o r f l a s k s tha t d id not . T h e v a l u e s a r e s h o w n in T a b l e 11 in m g / L , c o m p l e t e resu l t s c a n b e f o u n d in A p p e n d i x V I . T h e c o n c e n t r a t i o n of d i s s o l v e d o r g a n i c c a r b o n in t h e S e c o n d a r y P o n d w a t e r u s e d in t h e e x p e r i m e n t w a s 17 m g / L . 106 Tab le 1 1 : Average Dissolved Organic Carbon Concentrat ion (mg/L) for Each Tail ings Type Tailings Type Day 1 Day 42 No DOC DOC No DOC DOC RLM-5 11.7 18.9 8.4 7.7 RLM-7 11.3 20.1 6.3 5.7 RLM-2 8.8 21.3 8.9 13.2 Secondary Pond 23.5 34.7 4.8 7.9 The concentration of arsenic in the Secondary Pond water used to construct the experimental f lasks was 2.7 mg/L. Arsenic results were graphed for each tailings type and are briefly analyzed below. Complete arsenic and iron results can be found in Appendix VI . Secondary Pond Sediment Flask #19 was omitted from the analysis as an unusually high dissolved organic carbon concentration was observed, which resulted in non-representative condit ions occurring in the flask. The abnormally high DOC concentration may have been due to error in lactate addition or due to some anomalous growth. Figure 38 shows the change in arsenic concentration for each of the f lasks containing Secondary Pond Sediment. Initial arsenic concentrations (Day 1) were highest in f lasks #32 and #24, both did not contain iron fillings or sulphate. Lowest initial arsenic concentrations were seen in f lasks #21 and #27, each of which contained both iron fillings and sulphate. After day 21 the arsenic concentration in all the flasks was less than 0.5 mg/L with the lowest values seen in #18, #27, and # 2 1 . These flasks all contained both elemental sulphur and iron fillings. After 42 days arsenic concentrat ions were found to have significantly increased in f lasks #18 and #30, and all f lasks except for #27 and #21 had , arsenic concentrat ions greater than 0.5 mg/L. Flasks #27 and #21 maintained below 0.5 mg/L arsenic concentrations. Both of these flasks contained sulphate, 107 elemental sulphur and iron filings. In order to maintain arsenic concentrations below 0.5 mg/L it appears as if (based on this preliminary assessment) elemental sulphur and iron, as well as sulphate are needed. Figure 38: Arsenic Results for Flasks Containing Secondary Pond Sediment 12 1 0 • v., X • E 6 (0 < 1 0 1 5 2 0 2 5 Time (days) 3 0 3 5 4 0 4 5 • • - • # 3 2 - - • - • # 2 8 * # 1 9 - x - # 2 4 - - * - # 1 8 - - • - - # 2 1 — + — # 3 0 • # 2 7 RLM-2 All flasks contained less than 0.5 mg/L arsenic throughout the duration of the experiment. Initially, the highest arsenic concentrations were seen in flasks that did not contain iron or sulphate (#17 and #23), followed by flasks that contained sulphate but no iron (#29 and #22), flasks that contained iron but no sulphate (#25 and #20), and finally the lowest initial concentrations were seen in flasks containing both iron and sulphate (#26 and #31). After Day 21 the arsenic concentration in all of the flasks had dropped below 0.3 mg/L, and at the end of the experiment the concentration in all flasks was less than 0.2 mg/L. Figure 39 108 shows the change in arsenic concentration over t ime. With these tailings it does not seem to matter what is done to them, arsenic concentrations still remain low, even the f lasks containing organic carbon addition and low oxygen environments did not produce significantly elevated arsenic concentrations. These tailings appear to have a capacity to remove arsenic from solution as even after 1 day the aqueous arsenic concentration was reduced from 2.7 mg/L (concentration in Secondary Pond water added to flask at start of experiment) to below 0.5 mg/L. F igure 39: Arsenic Results for Flasks Containing RLM-2 Tail ings 0.45 -| '• 0.05 ^ 0 -I 1 , I 1 1 1 1 1 0 5 10 15 20 25 30 35 40 45 Time (days) -••--#23 --«--#22 —*~~~#17 —*—#29 ---*--#20 •-••-#31 —+—#25 #26 RLM-5 The highest initial concentrations were seen in flasks that did not contain iron fillings (#1 , #14, #16, #5). It should be noted that the high concentration seen in flask #1 initially (20 mg/L) is likely due to error as all other concentrations were less than 3 mg/L. Lowest initial concentrations were seen in flasks containing 109 both iron and sulphate. With the exception of f lasks #12 and #16, an increase in arsenic concentration was seen at Day 7 in all f lasks, fol lowed by a decrease for the remainder of the experiment. At the end of the experimentation period all f lasks contained less than 0.4 mg/L arsenic. In general, the lowest final concentrat ions were seen in f lasks with an oxygen rich atmosphere. Figure 40 shows the change in arsenic concentrat ion over t ime . F igure 40: Arsenic Results for Flasks Containing RLM-5 Tail ings 10 15 20 25 Time (days) 30 35 40 45 - • - -#1 - • • - - # 1 6 * #5 —*—#14 - - « - - # 1 2 -#13 —•— #7 -#4 RLM-7 Of the four tail ings types, the highest concentrat ions were seen in the RLM-7 flasks. This result is understandable as the RLM-7 tailings sample contained the highest concentration of total arsenic. The highest initial concentrat ions occurred in flasks #2 and #3 which both did not contain iron. The lowest initial concentrat ions were seen in f lasks #6 and # 1 1 , both of which contained sulphate, 110 and iron. In some flasks concentrat ions were seen to rise after Day 7 and then drop, while in other flasks concentrat ions just dropped from their initial value. By the third week the concentration of arsenic in flasks #6 and #9 had dropped below 0.1 mg/L. At the end of the experiment the arsenic concentrat ions in flasks #6 and #9 were below detection, and in flask #11 the concentration was 0.02 mg/L. Significantly higher concentrat ions (greater than 0.5 mg/L) were seen in the remainder of the flasks. Flasks #6, #9, and #11 all contained both elemental iron and sulphur, while f lasks #6 and #9 had an oxygen depleted environment. Figure 41 shows the arsenic concentrations over t ime. F igure 4 1 : Arsenic Results for Flasks Containing RLM-7 Tailings "8> \ V -^ '"•'•••:^^_lfr^^^^ | 1 1 1 r . ! 1 = ••- - - I • 1 0 5 10 15 20 25 30 35 40 45 Time (Days) - - • - - # 2 • • • - # 1 0 #8 — * ^ # 3 ---*--#9 - - • - • # 6 — + — # 1 5 #11 The experimental data was analyzed using the Design - Expert software for experimental Design, version 6.0.1.0, created by Stat-Ease Inc. Data and experimental parameters are inputted into the software, a regression is I l l preformed, significant factors determined, and a model is fitted to the data. The first step involves viewing a half normal probability plot. Data points on the plot represent factors and interaction terms. A line is fitted through the points closest to the origin (the points that produced the smallest effect), and any points not falling on the line are initially considered to be important. Next an analysis of variance table (ANOVA) is created for the selected factors. F values for the model and for each factor/interaction are shown, as well as the sum of squares for each term. F values indicate the importance of the terms, the larger the F value the more likely that the term is significant. Probability greater than F (Prob > F) values are also shown, these values signify the percentage chance that an F value that large could occur due to noise. Prob > F values less than 0.05 indicate that model terms are significant. Ideally the selected model will account for the majority of the total sum of squares (i.e. the residual values will be close to zero). The closer the residual sum of squares is to zero the better the model was able to fit the experimental data. Another check of the model fit is the predicted R squared value (how good the model predicts the response) compared to the adjusted R squared value (a measure of the amount of variation about the mean explained by the model). These two values should be within 0.2 of each other. At this point any terms initially selected that are determined to be insignificant can be removed and the model re-regressed. Once the final model has been determined the program gives an equation for the model with calculated constant values. Actual values can be compared to calculated values as an additional check of the adequacy of the model to fit the data. Next the normal probability plot of the studentized residuals is v iewed to check for normality of the results. In order to check for constant error the studentized residuals are graphed versus the predicted values. The outlier T graph can be viewed to check for any outlier points. If all of these checks are okay then the model can be concluded to be a 112 good fit of the data. Model graphs are created so that the impact of each factor and interaction can be easily seen. Set 1 - RLM-5 and RLM-7 The Day 42 data from Set 1 (RLM-7 represented by factor E at the +1 level, RLM-5 represented by factor E at the - 1 level) was analyzed using the Design Ease software. The statistically significant factors and interactions determined by the software were D (iron filings), E (tailings type), F (elemental sulphur), AD (the interaction between DOC and iron fil ings), and AF (the interaction between DOC and elemental sulphur). The A N O V A table, diagnostic plots, and model graphs can be found in Appendix Vl l . The predicted model in terms of coded factors (concentrations represented by +1 or - 1 ) is: Arsenic = 0.74 - 0.51 D + 0.49E - 0.5F - 0.35AD - 0.34AF The presence of iron filings (D) and elemental sulphur (F) both resulted in decreased arsenic concentrations for both tailings types. Lower arsenic concentrat ions were seen with RLM-5 tailings (factor E at a - 1 level), the positive constant associated with E in the above equation accounts for this. Both interaction terms involve DOC. Figures 42 and 43 show the effect of the interaction terms AD and AF, respectively. From Figure 42 it can be seen that when no iron is present (D = - 1 , the black line) lower DOC concentrat ions are correlated with lower arsenic concentrat ions. In essence, when iron filling are not present, lower arsenic concentrations were seen in the f lasks in the f lasks that did not have DOC added. When iron is present (D = + 1 , the red line), however, the opposite is true. In essence, when iron fillings are present, lower arsenic concentrat ions were seen in the flaks that had DOC added. Therefore when elemental iron is present the addition of DOC results in lower arsenic concentrat ions, however when elemental iron is not present the addition of DOC results in an increase in arsenic concentrations. This trend makes sense as 113 when elemental iron is present reducing conditions are induced and the presence of DOC will encourage these reducing condit ions (potentially leading to microbially mediated production of arsenopyrite). However when elemental iron is not present, the addition of DOC will lead to the onset of mildly reducing conditions, possibly leading to reductive dissolution of arsenic bearing iron oxyhydroxides. If DOC is present, the addition of iron filings will help to reduce the arsenic concentration. F igure 42: Set 1 - Day 42 Interaction Graph for AD Interact ion G r a p h DESIGN-EXPERT Plot Arsenic A: DOC D: Fe(0) • D--1.000 A D+ 1.000 Actual Factors B: S 0 4 = 0.00 C: ATM = 0.00 E: Tail ings = 0.00 F: S(0) = 0.00 -0.53535 A: DOC The interaction term AF had the same behaviour pattern as AD. When no elemental sulphur was present (F = - 1 , the black line), lower arsenic concentrat ions were seen with lower DOC concentrations. W h e n elemental sulphur was present (F = + 1 , the red line), lower arsenic concentrat ions were seen with higher DOC concentrations. 114 Figure 43: Set 1 - Day 42 Interaction Graph for AF Interact ion G r a p h DESIGN-EXPERT Plot Arsenic X = A: DOC Y = F: S(0) • F- -1.000 A F+ 1.000 Actual Factors B: S 0 4 = 0.00 C: ATM = 0.00 D: Fe(0) = 0.00 E: Tail ings = 0.00 -0.51535 A: DOC According to the above equation, the lowest arsenic concentrations will be achieved when iron filings, elemental sulphur, and DOC are present (i.e. D = + 1 , F = + 1 , A = +1), and when the tailings type is RLM-5 (E = -1). The majority of the arsenic present in both of the tailings types used in Set 1 was in the form of arsenopyrite. Arsenopyrite is most stable under reducing conditions, therefore it is logical that the presence of a strong reductant leads to the lowest arsenic concentrations. Overall, significantly lower concentrat ions were seen for RLM-5 than for RLM-7. The experiment demonstrated that under all condit ions the aqueous arsenic concentrations after 42 days for RLM-5 tailings were less than 0.5 mg/L. RLM-7 tailings will never likely be discharged to the tailings pond as this would result in a large loss of recoverable gold. The experiment showed that in order to obtain aqueous arsenic concentrations less that 0.5 mg/L in water overlying RLM-7 tailings it is necessary to maintain reducing conditions. Earlier t ime data (i.e. Day 21 and Day 7) were also analyzed using the software. For Day 21 data the software identified C, D, E, F, and CD as the significant 115 factors. The presence of C (atmosphere) had a significant impact on RLM-7 tailings (E = +1) and only a minimal impact on RLM-5 (E = -1) tailings. The CE interaction graph shown in Figure 44 depicts this trend. Arsenic concentrations were significantly higher when C = +1 (oxygen present) and E = +1 (RLM-7 tailings), then when C = -1 (oxygen depleted atmosphere). This result demonstrates the importance of maintaining reducing condit ions when tailings with a high arsenopyrite content are present. C was not found to be a significant factor in the Day 42 data, meaning that any additional arsenic that was mobil ized due to presence of an oxygen rich atmosphere in early days of the experiment, was later removed from solution via a precipitation or adsorption process (either with a sulphide or an oxide phase). The predicted model for the Day 21 data, in terms of coded factors, is: Arsenic = 2.93 + 0.98C - 1.07D + 1.48E - 1.38F + 0.89CE F igure 4 4 : Set 1 - Day 21 Interaction Graph for CE 9.68 H DESIGN-EXPERT Plot Arsenic X = C:ATM Y = E: Tailings • E- -1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 B: S04 = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 o c S> 4.8615-2.45225-E+C 0.043 -In teract ion G r a p h E: T a i l i n g s 0.50 1.00 C: ATM For Day 7 data the software identified the same factors as for the Day 21 data, with the addition of B (sulphate). Higher sulphate concentrations produced lower 116 arsenic concentrations for both tailings types. The predicted model for the Day 7 data, in terms of coded factors, is: Arsenic = 5.61 - 0.85B + 1.75C - 1.54D + 3.74E - 1.44F + 1.23CE The presence of sulphate was not found to be statistically significant in the later t ime data. Either the effect of sulphate was small in comparison to the other effects, or sulphate had no effect in the long term. Set 2 - RLM-2 and Secondary Pond Sediment The Day 42 data from Set 2 (Secondary Pond Sediment represented by factor E at the +1 level, RLM-2 represented by factor E at the - 1 level) was analyzed using the Design Ease software. The statistically significant factors and interactions determined by the software were B (sulphate), C (atmosphere), E (tailings type), AC (the interaction between DOC and sulphate), AD (the interaction between DOC and iron filings), AF (the interaction between DOC and elemental sulphur), BD (the interaction between sulphate and iron filings), BE (the interaction between sulphate and tailings type), and CE (the interaction between atmosphere and tailings type). The ANOVA table, diagnostic plots, and model graphs can be found in Appendix VII. The predicted model in terms of coded factors (concentrations represented by +1 or - 1 ) is: Arsenic = 0.84 - 0.58B + 0.44C +0.72E - 0 . 1 7 A C - 0.47AD - 0.45 AF - 0.17BD - 0.6BE + 0.44CE The presence of sulphate (B) resulted in a decrease in arsenic concentrat ion, while the presence of an oxygen rich atmosphere (C = +1) resulted in an increase in arsenic concentration. Both of these factors, however, were involved in an interaction with tailings type. Tail ings type had a major impact on the arsenic concentration, with much lower concentrations being seen with RLM-2 (E 117 = -1) tailings. The BE and CE interaction terms show that factors B and C are only significant for the Secondary Pond Sediment (E = +1). Figures 45 and 46 show the behaviour of the BE and CE interaction terms. For E = +1 (Secondary Pond Sediment), the arsenic concentration decreases significantly with increasing sulphate concentration, and the arsenic concentration is significantly higher with a +1 atmosphere (oxygen rich). Figure 45: Set 2 - Day 42 Interaction Graph for BE Interact ion G r a p h DESIGN-EXPERT Plot Arsenic X = B: S 0 4 Y = E: Tail ings • E- -1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 C: ATM = 0.00 D. Fe(0) = 0.00 F: S(0) = 0.00 -0.0275028 B: S04 Figure 46: Set 2 - Day 42 Interaction Graph for CE DESIGN-EXPERT Plot Arsenic X = C : A T M Y = E: Tail ings • E- -1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 B: S 0 4 = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 In teract ion G r a p h -0.0115028 - f -1.00 -0.50 0.00 0.50 1.00 C: ATM 119 The interaction terms AC, AD, AF, and BD have an impact on both tailings types. For AC, when C = -1 (oxygen depleted atmosphere), the arsenic concentration increases slightly when DOC is present, however when C = + 1 , the arsenic concentration decreases slightly when DOC is present. For the AD term, the arsenic concentration decreases when iron filings are present as the DOC concentration increases. When iron filings are not present, the arsenic concentration is lowest with lower DOC concentrations. The same situation exists for the A F term as described for the AD term. The effect of the BD term is much more significant for the Secondary Pond Sediments than for RLM-2 tailings. For the Secondary Pond Sediments, the arsenic concentrat ion decreases as the sulphate increases (more so when iron filings are present). For RLM-2 tailings, when iron filings are not present the arsenic concentrat ion increases slightly as the sulphate concentration increases. W h e n iron is present the arsenic concentration decreases slightly as the sulphate concentrat ion increases. The graphs for the interaction discussed above can be found in Appendix VII . By manipulating the model factors it is possible to predict how the arsenic concentration will be affected under certain conditions. According to the model, for RLM-2 tailings, the lowest possible arsenic concentrations are seen when all the factors are at the low level, in essence, an oxygen depleted environment with no additional sulphate, DOC, elemental sulphur or iron filing present (the impact of iron filings and elemental sulphur on the concentration is negligible). If the situation arises where DOC will be present under a depleted oxygen environment, and all the other factors remain at the - 1 level, the arsenic concentration increases dramatically, f rom near zero to about 1 mg/L (predicted by the model). If iron and elemental sulphur are added (now at +1 level) the arsenic concentration is seen to drop back down to near zero. The addition of sulphate decreases the arsenic concentration slightly, but only when iron filings are present, otherwise the addition of sulphate increases the arsenic 120 concentration. These results are intuitive, and coincide with the theory of microbially mediated arsenic dissolution. They show that the addition of iron filings can significantly reduce arsenic concentrations. For the opposite situation for RLM-2 (i.e. under oxygen rich condit ions, C = +1), the lowest arsenic concentrations are seen when all other factors are at the low level (the impact of sulphate, iron filings and elemental sulphur is minor). If DOC is added (A = +1), the arsenic concentration, as predicted by the model, increases from near zero to about 0.7 mg/L. By adding iron filings and elemental sulphur the arsenic concentration is returned to near zero. The addition of sulphate has the same result as under oxygen depleted condit ions. The case of the Secondary Pond Sediments is quite different than for RLM-2. The lowest arsenic concentrations are seen when all factors are at the low level except for sulphate (B = +1). The addition of sulphate dramatically decreases the arsenic concentration from approximately 0.6 mg/L, down to near 0 mg/L. Without sulphate the addition of DOC drives the arsenic concentration up to greater than 2.5 mg/L, by adding sulphate the concentration is seen to drop to about 0.75 mg/L. By adding iron and elemental sulphur the concentration can be dropped further to near zero, however without the sulphate the arsenic concentration is greater than 1 mg/L. In this situation it is likely that the sulphate is precipitating with calcium in the water to form gypsum and arsenic is adsorbing and/or coprecipitating with the gypsum. Unfortunately, arsenic - calcium precipitates are known to have limited stability. It was surprising that lower concentrations were seen under an oxygen depleted environment than under an oxygen rich environment, as the release of arsenic from Secondary Pond Sediments due to reductive dissolution of arsenic bearing iron oxyhydroxides is believed to contribute a significant source of arsenic to the Secondary Pond waters. 121 Under oxygen rich conditions, and with all other factors at the low level, the predicted arsenic concentration in water overlying the Secondary Pond Sediments in the experiment is about 2.7 mg/L (compared with 0.6 mg/L under oxygen depleted conditions, with all other factors the same). A portion of the difference in the concentration can be attributed to evaporation that may have occurred in the flasks open to the atmosphere. When DOC is added, the arsenic concentrat ion cl imbs to over 4 mg/L. The addition of sulphate drops the concentrat ion down to about 2.2 mg/L, and by adding iron filings and elemental sulphur the concentration can be dropped further to near zero. In order to confirm the mechanism of arsenic removal in the experiments it is necessary to identify the black/shiny material formed. An attempt was made to mineralogically identify this, however it was not possible to confirm its composit ion for several reasons. The fine grained nature of the material and the presence of iron filings and pre-existing arsenic mineral phases in the tailings samples, it made it impossible to determine the composit ion of this material. The black material was also seen to readily oxidize upon exposure to air (changed from black to brownish-orange). The formation of arsenic bearing sulphide phases (e.g. mono sulphides, arsenopyrite, arsenical pyrite) could therefore not be mineralogically confirmed. It is however likely that the black/shiny material was a sulphide phase. 5.3 Conclusions The addition of iron fillings was effective in reducing the dissolved arsenic concentrat ion, in both the in situ and laboratory experiments. In addit ion, the release of arsenic from the sediments into the water column was prevented in the in situ experiments. It is unclear as to the exact mechanism that resulted in the removal of arsenic from the water column, however it is speculated that arsenic became associated with an iron sulphide phase. In the laboratory experiments, black sediments were seen to form in some of the flasks containing iron filings. 122 These black sediments were likely iron monosulphides. The combination of the addition of organic carbon and sulphate in the column experiments resulted in the reductive dissolution of oxidized iron/arsenic phases and the development of high concentrations of dissolved arsenic and iron. In this case, condit ions were not appropriate to bring about arsenic removal via the formation of sulphide species. In the laboratory experiments, the black sediment only formed in flasks containing zero valent iron. It appears as if reducing condit ions are necessary to bring about the formation of iron sulphide species. Under all of the conditions tested in the laboratory experiments, RLM-5 tailings did not produce aqueous arsenic concentrations greater than 0.5 mg/L after 42 days. For the RLM-7 tailings the experiment showed that in order to obtain aqueous arsenic concentrat ions less that 0.5 mg/L it is necessary to maintain reducing conditions. All f lasks contained less than 0.5 mg/L of arsenic throughout the duration of the experiments for RLM-2 tailings. The model does show however that if DOC is present under oxygen deficient conditions, the arsenic concentration will increase significantly. According to the model, if iron filings and elemental sulphur are added the arsenic concentration will drop back down to near zero. For the Secondary Pond Sediments, sulphate was found to have a major impact on the aqueous arsenic concentrat ion. The addition of sulphate resulted in a dramatic decrease in the arsenic concentrat ion. The model showed that if oxygen deficient conditions exist and DOC is introduced into the system (with all other factors at the low level), the arsenic concentration will increase up to greater than 2.5 mg/L. The addition of sulphate, iron filings and elemental sulphur returns the arsenic concentration to near zero. Lower arsenic concentrations were seen with an oxygen deficient atmosphere than with an oxygen rich atmosphere. 123 6.0 Conclusions Arsenic has always been a problem at G1 (discharge from the Secondary Pond), with the concentration exceeding 0.5 mg/L since 1991 except for a brief period during the shutdown and early start up of the new mill. Measures need to be taken to reduce the concentration of arsenic at the discharge of the Secondary Pond if Balmer Lake is to be removed from the Tail ings Management facility. Natural degradation has not been successful in reducing the concentration of total arsenic in the effluent from the Secondary Pond to within MISA (and MMER) standards. Historical tailings samples (RLM-2, RLM-3, and RLM-6) exhibited significantly higher total arsenic concentrations than the samples containing newer tailings (RLM-1 , RLM-4, and RLM-5). The high sulphur content in the RLM-2 samples, ranging from 1.86 - 2.63 %, indicates that these tailings were produced during a period of t ime when roasting and concentration of the sulphide portion of the ore had ceased. The surface of the unsaturated old tailings (RLM-2-1, RLM-3-1 , and RLM-6-1) contains gypsum, an indicator that sulphide oxidation has occurred. The tailings sampled at RLM-2 (downstream of SD#2 on dry tailings beach in Primary Pond) do not appear to contain roasted material, however deeper down there is likely roaster material present (elevated arsenic levels have been observed in the vicinity of RLM-2). The material sampled at RLM-2 contains substantial sulphides and the top few feet have undergone significant oxidation over the years due to the unsaturated condit ions present. RLM-2-1 had only 3 1 % of the arsenic in the As(-1) oxidations state, with 5 8 % present as As(V). From the sequential extractions it was found that approximately 2 3 % of the arsenic in RLM-2-1 is adsorbed, and 2 8 % is associated with iron oxyhydroxides. RLM-2-2 and RLM-2-3 were similar in composit ion to each other, with more than 8 0 % of the arsenic present as As(-1), with the remaining portion present as a combinat ion of As(l l l ) and As(V). Results from the designed laboratory 124 experiments for RLM-2-1 material indicate that under all condit ions tested the aqueous arsenic concentration remained below 0.5 mg/L for the duration of the experiments. The model produced from the experiments for this material shows however, that if dissolved organic carbon is present under oxygen deficient condit ions, the arsenic concentration will increase significantly. According to the model, if iron filings and elemental sulphur are added the aqueous arsenic concentrat ion will drop back down to near zero. Based on all of the data for tail ings located around RLM-2 (i.e. dry tailing beach in Primary Pond), the f looding of this tailings deposit is not likely to result in a significant increase in aqueous arsenic concentrations. The material located around RLM-3 (old tailings on south side of access road), and RLM-6 (revegetated tailings east of Balmer Creek) has been subjected to roasting. A large fraction of the arsenic at these locations (more so at RLM-6) is associated with iron oxide material and is and will continue to be mobile, especially under oxygen deficient condit ions (i.e. saturated condit ions). Severely elevated arsenic levels in the groundwater surrounding RLM-6 support this conclusion. Less than 3 0 % of the arsenic in RLM-3-1 and RLM-3-2 was in the form of As(-1). A small amount (< 5%) of the arsenic existed in the form of As( l l l ) , with the remainder present as As(V). Most of the arsenic in the saturated zone in the area of RLM-3 (RLM-3-3) was present in the form of As(-1), with 5% present as As(V) and 2 % as As(l l l ) , suggesting that the oxidized forms of arsenic present in the unsaturated zone have been mobil ized and transformed to As(-1) in the saturated zone. Based on the above results, reducing infiltration into the deposit and maintaining unsaturated condit ions will minimize the mobilization of arsenic. The majority of the arsenic present in the saturated zone is in the As(-1) form and should remain stable as long as saturated conditions exist. 125 Only 2 0 % of the arsenic contained in RLM-6-1 was in the form of As(-1), with 16% present as As(l l l ) and 6 5 % present as As(V). According to the sequential extractions, 3 9 % of the arsenic in RLM-6-1 was adsorbed while 4 0 % was coprecipitated with iron oxyhydroxides. The saturated tailings in the area of RLM-6 (RLM-6-2) contained 6 3 % of the arsenic in the form of As(-1) suggesting that a considerable portion of the oxidized arsenic has been mobil ized, and transformed into As(-1). The saturated tailings still contain a significant amount of arsenic in oxidized forms that are likely to continue to be mobil ized under the saturated condit ions that exist in the deposit. Minimizing infiltration and lowering the water table will help to reduce mobilization of arsenic in the deposit. RLM-1 (downstream of SD#1), RLM-4 (upstream of SD#1), and RLM-5 (End of Pipe) contain essentially fresh tailings that have remained saturated. The vast majority of arsenic in these samples is contained within arsenopyrite that will continue to be stable as long as conditions remain saturated. A small amount of arsenic was found to be associated with an amorphous and spongy textured precipitate that containing many species, in the RLM-4 and RLM-5 samples. Sequential extraction data for the End of Pipe tailings indicate that only 16% (or 330 mg As/kg sediment) of the arsenic present is likely to be easily mobil ized. The XANES spectra indicate that 2 3 % of the arsenic in RLM-5 tailings is in the oxidized form. All of the solids identification results indicate that the vast majority of the arsenic in the fresh tailings produced at the mine site should remain stable under saturated conditions. Under all of the condit ions tested in the laboratory experiments, RLM-5 tailings did not produce aqueous arsenic concentrations greater than 0.5 mg/L after 42 days. RLM-4 tailings were similar in composit ion to RLM-5 tailings. RLM-1 tailings contained no As(l l l ) , and 8 9 % of the arsenic was in the form of As(-1). For the RLM-7 tailings (CIP tail ings), containing high concentrations of arsenopyrite, the laboratory experiments showed that in order to obtain aqueous arsenic concentrat ions less that 0.5 mg/L it is necessary to maintain reducing conditions. 126 The Primary Pond sediment contained 5 0 % As(-1), 1 1 % As(l l l ) , and 4 0 % As(V). The material f rom the Primary Pond (Primary Pond Backhoe sample), contained significantly more arsenic in the As(-1) form. A large fraction of the arsenic contained in the primary pond sediment will become mobil ized if conditions become reducing. Approximately 6 5 % of the arsenic contained in the Secondary Pond Sediment is likely to be fairly mobile under the changing redox condit ions that exist in the Secondary Pond. The sequential extraction results indicate that over 6 5 % of the arsenic in the Secondary Pond Sediment is sorbed or precipitated with an iron oxyhydroxide phase, with more than 5 0 % of the arsenic being sorbed. The Secondary Pond sediment contained 2 5 % As(-1), 4 0 % As(l l l ) , and 3 5 % As(V). The laboratory experiments conducted on Secondary Pond Sediments indicate that dissolved sulphate has a major impact on the aqueous arsenic concentrat ion. The addition of sulphate resulted in a dramatic decrease in the aqueous arsenic concentration. The model also showed that if oxygen deficient condit ions exist and dissolved organic carbon is introduced into the system (with all other factors at the low level), the arsenic concentration will increase up to greater than 2.5 mg/L. The addition of sulphate, iron filings and elemental sulphur returns the arsenic concentration to near zero. The addition of iron fillings was effective in reducing the dissolved arsenic concentrat ion, in both the in situ and laboratory experiments. In addition, the release of arsenic from the sediments into the water column was prevented in the in situ experiments. It is unclear as to the exact mechanism that resulted in the removal of arsenic from the water column, however it is speculated that arsenic became associated with an iron sulphide phase. In the laboratory experiments, black sediments were seen to form in some of the f lasks containing iron filings. These black sediments were likely iron monosulphides. 127 In the column experiments, the addition of organic carbon and sulphate resulted in the reductive dissolution of oxidized iron/arsenic phases and the development of high concentrations of dissolved arsenic and iron. In this case, condit ions were not appropriate to bring about arsenic removal via the formation of sulphide species. In the laboratory experiments, the black sediment only formed in flasks containing zero valent iron. It appears as if strongly reducing condit ions are necessary to bring about the formation of iron sulphide species. As was found at the neighboring Campbel l Mine (McCreadie et al., 2000) the source of high concentrations of dissolved arsenic in groundwater at the Red Lake Mine is likely from the reductive dissolution of arsenic bearing iron oxyhydroxides found in the roaster derived tailings. Field and laboratory experiments confirmed that the creation of semi reducing condit ions, via the addition of organic carbon, resulted in increased aqueous arsenic and iron concentrations, supporting the reductive dissolution theory. Field and laboratory experiments also showed that the addition of a strong reductant resulted in significantly lowered aqueous arsenic concentrations, even in the presence of high concentrations of dissolved organic carbon. It is believed that under strongly reducing condit ions arsenic contained within a relatively unstable iron oxyhyroxide phase can be stabilized as a sulphide phase. 128 7.0 Recommendations In order to effectively manage arsenic bearing solids it is necessary to fully understand the speciation of the arsenic present as well as the mechanisms with which the arsenic is associated with the solids. The solids identification work conducted in this study proved to be useful in understanding the nature of the arsenic associated with the various types of tailings present at the Red Lake Mine. Any new arsenic bearing solids that are produced should be adequately characterized in order to determine the optimal storage condit ions and to evaluate the long term stability of the solids. XANES and or sequential extractions are required to determine the amount of arsenic in each oxidation state. SEM analysis provides a qualitative and visual understanding of the arsenic present in the solids, while whole rock analysis is a simple test to determine the total amount of arsenic present. The combination of these methods will provide the required insight into the material. A further step of laboratory and/or field experimentation can be conducted in order to better evaluate the long term stability of arsenic bearing solids. In the case of solids bearing arsenic stabilized as an oxidized form (e.g.ferric arsenate precipitates, arsenic sorbed to iron oxyhydroxides, etc.) field and/or laboratory experiments should be designed to examine what happens to the solids and to the aqueous arsenic concentrat ions under an oxygen deficient environment (e.g. solids covered with water in the presence of organic carbon). In the case of solids bearing arsenic stabilized as a reduced from (i.e. arsenic associated with mono sulphides, arsenopyrite, arsenical pyrite, etc.) field and/or laboratory experiments should be designed to examine what happens to the solids and how the pore water arsenic concentrat ions change under an oxygen rich environment (e.g. solids exposed to the atmosphere). The key questions to answer in evaluating the production and storage of arsenic bearing solid wastes are: 129 • What arsenic bearing mineral phases are present? • How much of each arsenic bearing mineral phase is present? • How stable are the arsenic bearing mineral phases under changing redox condit ions? • What happens to aqueous arsenic concentrations under changing redox condit ions? • Can optimal redox condit ions for the arsenic bearing mineral phases present be maintained? Once these questions have been answered a decision can be made on whether the arsenic bearing minerals phases should be produced (if this is a choice) and how the solids should be stored. It is important to store arsenic bearing solids under appropriate condit ions so that arsenic mobilization is minimized. Solids containing reduced arsenic species (i.e. arsenopyrite, As(-1)) will be most stable if stored under saturated/reducing condit ions where contact with oxygen is minimized. Material containing oxidized arsenic species (As(V) and As(l l l)) will be most stable if stored under unsaturated/oxidizing condit ions, where contact with oxygen is maximized. Often the nature of these solids makes it difficult to maintain unsaturated conditions, and if located in a natural setting (where organic carbon and biological activity are present) the onset of reducing condit ions is a realistic concern. At the Red Lake Mine, fresh tailings (RLM-1 , RLM-4, and RLM-5) contain arsenic predominately associated with arsenopyrite, therefore the fresh tailings should be stored under saturated condit ions to minimize oxidation. The fresh tailings, however, contain a low concentration of sulphides due to the concentration of sulphides in the milling circuit. Arsenic release from unsaturated fresh tailings, therefore, should be minimal. 130 Historical tailings located around RLM-3 (south of the access road) and RLM-6 (revegetated tailings on east side of Balmer Creek) contain a significant amount of oxidized arsenic species in the unsaturated zone. Mobilization of arsenic from these solids will be minimized by reducing infiltration and maintaining unsaturated conditions (capping the deposits with an appropriate engineered cover system). In addit ion, during reclamation efforts the use of fertilizer should be avoided in order to limit the supply of nutrients to the system. The tailings located around RLM -2 (i.e. dry tailing beach in Primary Pond) can be f looded (i.e. result of raising dams) with little increased arsenic mobilization expected. The changing redox condit ions in the Secondary Pond (and Primary Pond) and the nature of the arsenic species that are present in the pond sediment indicate that arsenic will continue to be released from the sediment over the long term. Amendments to the pond sediments should be further investigated to determine their applicability. For example, Lorax (2001) suggests placing a layer of fresh tailings over the pond sediments to act as a diffusion barrier. In this study it was determined that the fresh tailings contain arsenic predominately in the As(-1) state, as a result the arsenic in these tailings is likely to remain if placed under the water cover of the Secondary Pond. An investigation (i.e. water balance) should be conducted to determine the significance of mobil ized arsenic from the pond sediments in the over all arsenic loading to the pond. If arsenic loading f rom the pond sediments is found to be significant, a feasibility and effectiveness study on the use of fresh tailings as a diffusion barrier should be conducted. The addition of iron fillings proved to be successful in reducing aqueous arsenic concentrat ions in both the field and laboratory experiments, however the mechanisms by which the arsenic was being removed are not understood. Addit ional experiments are required in order to gain an understanding of the mechanism of arsenic removal in the presence of iron fillings. The experiments 131 must be structured in a manner allowing the produced solids to be mineralogically identified. For example, laboratory experiments could be constructed in which the tailings are separated f rom the treatment medium by a layer of silica sand. Ideally, newly generated mineral phases would form in the silica sand layer and could be extracted for analysis (in this way the generated solids would not be contaminated with pre existing arsenic phases from the tailings solids or with iron fillings). It is likely that XANES (using a Synchrotron light source) and/or sequential extractions would be required to identify the generated solids. The analysis should be conducted on wet solids under an oxygen deficient atmosphere. In order for the formation of arsenic bearing sulphide phases to be a practical solution for mining operations it is necessary for the solids produced to be stable. The stability of the generated solids would therefore have to be evaluated. The use of iron fillings in a field setting may not be feasible due to the cost associated with the iron fillings. The addition of organic carbon in the laboratory and field experiments showed that the production of semi reducing conditions results in increased arsenic concentrations, therefore efforts should be made to reduce the influx of organic carbon (i.e. reduce nutrient loading) into tailings deposits containing arsenic in oxidized forms. Strongly reducing condit ions are required in order to cause arsenic to be removed from solution. 132 8.0 References Azcue, J.M., Nriagu, J.O., 1995. Impact of abandoned mine tailings on the arsenic concentrations in Moira Lake, Ontario. J. Geochem. Explor. 52, 8 1 -89. Barr, J. 1983. Industrial Abatement Section, Northwestern Region, Ministry of the Environment, Report and Recommendat ions of Provincial Officer. Berner, R.A. 1970. Sedimentary pyrite formation. American Journal of Science 268, 1-23. Benning, L.G., Wilkin, R.T, and Barnes, H.L. 1999. Reaction pathways in the Fe-S system below 100°C. Chemical Geology 167, 25 -51 . Buykx, S.E.J., Bleijenberg, M., van den Hoop, M.A.G.T, and Gustav Loch, J.P. 2000. The effects of oxidation and acidification on the speciation of heavy metals in sulfide-rich freshwater sediments using a sequential extraction procedure. J. Environ. Monit. 2, 23-27. Bancroft, G.M. and Hallin, E.L. 2002. The Canadian Light Source: Progress, and opportunit ies for earth, environmental, and material science applications. Mineralogical Association of Canada, Short Course Series Volume 30 -Synchrotron Radiation, pp 25-42. Deuel, L.E., Swoboda, A.R., 1972. Arsenic solubility in a reduced environment. Soil Sci. Soc. A m . Proc. 36, 276-278. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Carlosn, A.R., and Ankley, G.T. (1992). Acid volatile sulfide predicts the acute toxicity of cadmium and nickel in sediments. Environmental Science and Technology 26, 96-101 . Dowdle, P.R., A .M. Laverman, and R.S. Oremland. 1996. Bacterial dissimilatory reduction of arsenic (V) to arsenic (III) in anoxic sediments. Applied Environmental Microbiology 62, no. 5: 1664-1669 Furukawa, Y. and Barnes, H.L. 1995. Reactions forming pyrite from precipitated amorphous ferrous sulfate. A C S Symposium Series 612, American Chemical Society, Washington, D.C., pp. 194-205. Guo, T.Z. DeLaune, R.D., Patrick, W.H. , 1997. The influence of sediment redox chemistry on chemically active forms of arsenic, cadmium, chromium, and zinc in estuarine sediment. Environ. Internat. 23, 305-316. 133 Gleyzes C., Tellier, S., Sabrier, R., and Astruc, M. 2 0 0 1 . Arsenic characterization in industrial soils by chemical extractions. Environmental Technology, Vol. 22. pp 27-38. Huerta-Diaz, M.A. and Morse, J.W. 1992. The pyritization of trace metals in anoxic marine sediments. Geochimica et Cosmochimica Acta 56, 2 6 8 1 -2702. Huerta-Diaz, M.A., A. Tessier, and R. Carignan. 1998. Geochemistry of trace elements associated with reduced sulfur in freshwater sediments. Applied Geochemistry 13: 213-233. Jambor, J.L. 2003. Arsenic in Solids from the Tailings Containment Facility at the Cochenour Mine, Ontario. Leslie Investments Ltd. Research and Consult ing. Jain, A., K.P. Raven and R.H. Loeppert. 1999. Arsenite and arsenate adsorption of ferrihydrite: Surface charge reduction and net OH release stoichiometry. Environmental Science and Technology, v.33, pp 1179-1184. Jiang, D.T. 2002. X-ray absorption fine structure spectroscopy. Mineralogical Association of Canada, Short Course Series Volume 30 - Synchrotron Radiation, pp 65-98. Kotzer, T. 2003. Use of X-ray absorption spectroscopy for the determination of arsenic speciation in gold-ore mine tailings. Canadian Light Source report prepared for Goldcorp Inc. Keon, N.E., Swartz, C.H., Brabander, D.J., Harvey, C , Hemond, H.F., 2 0 0 1 . Validation of an arsenic sequential extraction method for evaluating mobility in sediments. Environ. Sci. Technol . 35, 2778-2784. Krause, E., and V.A. Ettel. 1988. Solubility and stability of scorodite, FeAsO 4 . 2H 2 0 : New data and further discussion. American Mineralogist 73, no. 7-8: 850-854. Krause, E. and Ettel, V.A. 1985. Ferric arsenate compounds: Are they environmental ly safe? Solubilities of basic ferric arsenates. Canadian Institute of Mining, Metallurgy and Petroleum, Montreal. 5/1-5/20. Krause, E. and Ettel, V.A. 1987. Solubilites and stabilities of ferric arsenates. Pergamon Press, New York, NY, U.S.A, pp. 195-210. Krause, E., and V.A. Ettel. 1989. Solubilities of ferric arsenate compounds. Hydrometallurgy, vol. 22, pp. 311-337. 134 Loeppert, R.H., Jain, A., El-Haleem, M.A., and Biswas, B.K. 2003. Quantity and speciation of arsenic in soils by chemical extraction. Biogeochemistry of Environmentally Important Trace Elements, A C S Symposium Series: 835, pp 42-56. Lombi, E., Sletten, R.S., and Wenzel , W.W. 1999. Sequential ly extracted arsenic from different size fractions of contaminated soils. Water, Air, and Soil Pollution 124: 319-332. Lorax 2 0 0 1 . Goldcorp - Red Lake Mine Tailings Management Plan. Internal report prepared by Lorax Environmental Services. McCreadie, H., Blowes, D.W., Ptacek, C.J., and Jambor, J.L. 2000. Influence of Reduction Reactions and Solid-Phase Composit ion of Arsenic. Environmental Science and Technology: 34, 3159-3166. McGeehan, S.L., 1996. Arsenic sorption and redox reactions: Relevance to transport and remediation. J. Enviro. Sci. Health Part A-Environ. Sci. Engineer. Toxic Hazard. Subst. Cotnrol 3 1 , 2319-2336. McGeehan, S.L., and Naylor, D.V. 1994. Sorption and redox transformation of arsenite and arsenate in two f looded soils. Soil Sci. Soc. Am. J. 58, 337-342. Martin, A.J . 1996. The geochemistry of Fe, Mn, Ni, Cu, Zn, and As in the water column, sediments, and porewaters in a seasonally anoxic lake. MSc thesis, Department of Earth and Ocean Sciences, University of British Columbia. Martin, A.J. , and Pedersen, T.F. 2002. Seasonal and Interannual Mobility of Arsenic in a Lake Impacted by Metal Mining. Environmental Science and Technology: 36, 1516-1523. Meng X., Jing, C. and Korfiatis, G.P. 2003. A review of redox transformation of arsenic in aquatic environments. Biogeochemistry of Environmentally Important Trace Elements, ACS Symposium Series: 835, 70-83. Moldovan, B.J., Hang, D.T., and Hendry, J. 2003. Mineralogical Characterization of Arsenic in Uranium Mine Tailings Precipitated from Iron-Rich Hydrometallurgical Solutions. Environmental Science and Technology: 37, 873-879. Moore, J.N., W . H . Ficklin, and C. Johns. 1988. Partit ioning of arsenic and metals in reducing sulfidic sediments. Environmental Science and Technology 22: 432-437. 135 Morse, J.W. 1999. Sulfides in sandy sediments: New insights on the reactions responsible for sedimentary pyrite formation. Aquatic Geochemistry, 5: 75-85. Morse, J.W. and Wang, Q. 1997. Pyrite formation under conditions approximating those in anoxic sediments: II. Influence of precursor iron minerals and organic matter. Marine Chemistry 57, 187-193. Morse, J.W. and Luther, G.W. 1999. Chemical influences on trace metal-sulfide interactions in anoxic sediments. Geochimica et Cosmochimica Acta Vol. 63, No. 19/20, pp. 3373-3378. Morse, J.W. 1994. Interactions of trace metals with authigenic sulfide minerals: Implications for their bioavailability. Marine Chemistry 46, 1-6. Neal, A.L., Techkarnjanaruk, S., Dohnalkova, A., McCready, D., Peyton, B.M., and Gessey, G.G. 2 0 0 1 . Iron sulfides and sulfur species produced at hematite surfaces in the presence of sulfate-reducing bacteria. Geochimica et Cosmochimica Acta Vol. 65, No. 2, pp. 223-235. Oblonsky, L.J., Ryan, M.P., and Isaacs, H.S. 2000. Corrosion Science 42, 229-2 4 1 . Postma, D. and Jakobsen, R. 1996. Redox zonation: Equil ibrium constraints on the Fe( l l l ) /S0 4 - reduct ion interface. Geochimica et Cosmochimica Acta, v. 60, pp 3169-3175. Raudseep, M., and E. Pani. 2003. Application of Rietveld Analysis to Environmental Mineralogy. Mineralogical Associat ion of Canada Short Course 30 "Environmental Mineralogy". Reynolds, J.G., Naylor, D.V., Fendorf, S.W., 1999. Arsenic sorption in phosphate-amended soils during f looding and susbsequent aeration. Soil Sci. Soc. A m . J. 63, 1149-1156. Rickard, D.T. 1995. Kinetics of FeS precipitation: Part 1. Compet ing reaction mechanisms. Geochimica et Cosmochimica Acta, v. 59, pp. 4367-4379 Rittle, K.A., J.I. Drever, and P.J.S. Colberg. 1995. Precipitation of arsenic during bacterial sulfate reduction. Geomicrobiology Journal 13, no. 1: 1-11. Riveros, P.A., Dutrizac, J.E., and Spencer, P. 2 0 0 1 . Arsenic disposal practices in the metallurgical industry. Canadian Metallurgical Quarterly, Vol. 40, No. 4, pp 395-420. 136 Rochette, E.A., Li, G.C., Fendorf, S.E., 1998. Stability of arsenate minerals in soil under biotically generated reducing conditions. Soil Sci. Soc. Am. J. 62, 1530-1537. Rochette, E.A., Bostick, B.C., Li, G., and Fendorf, S. 2000. Kinetics of arsenate reduction by dissolved sulfide. Environ. Sci. Technol. 34, 4714-4720. Schoonen, M.A.A. and Barnes, H.L. 1991. Reactions forming pyrite: I. Nucleation of F e S 2 below 100°C. Geochimica et Cosmochimica Acta 55:1495-1504. Sham, T.K. 2002. Synchrotron Radiation: An overview. Mineralogical Association of Canada, Short Course Series Volume 30 - Synchrotron Radiation, pp 1-24. Smedley, P.L., and Kinniburgh, D.G. 2002. A Review of the Source, Behaviour and Distribution of Arsenic in Natural Waters. Applied Geochemistry: 17, 517-568. S tumm, W. and Morgan, J.J. 1995. Aquatic Chemistry: Chemical Equilibria and Rates in Natural Waters. Wiley-lnterscience, New York. Su, C , and Puis, R.W. 2 0 0 1 . Arsenate and Arsenite Removal by Zerovalent Iron: Effects of Phosphate, Silicate, Carbonate, Borate, Sulfate, Chromate, Molybate, and Nitrate, Relative to Chloride. Environ. Sci. Technol : 35, 4562-4568. Sveinson, M.G. 1956. New Dickenson Mines Limited. Internal memo dated September 17 t h , 1956. Wei , Dawei and Osseo-Asare, K. 1996. Particulate pyrite formation by the Fe 3 7HS" reaction in aqueous solutions: effects of solution composit ion. Colloids and Surfaces A: Physiochemical and Engineering Aspects 118, 5 1 -6 1 . Wenzel , W.W., Kirchbaumer, N., Prohaska, T., Stingeder, G., Lombi, E., and Adriano, D.C. 2001 . Arsenic fractionation in soils using an improved sequential extraction procedure. Analyt ica Chimica Acta: 436, 309-323. Wilkin, R.T., and Barnes, H.L. 1996. Pyrite formation by reactions of iron monosulf ides with dissolved inorganic and organic sulfur species. Geochimica et Cosmochimica Acta 60 (21), 4167-4179. Zhang, S., Wang , S. and Shan, X. 2 0 0 1 . Effect of sample pre-treatment upon the metal speciation in sediments by a sequential extraction procedure. Chemical Speciaiton and Bioavailability: 13(3), pp 69-74. Appendix I: Historical Surface Water Data 138 Date PP G1 L2 As CN As Fe CN As Fe CN Jan-91 1.69 60 1.88 64 0.79 0.34 Feb-91 1.92 59 1.03 0.3 Mar-91 1.61 38 1.14 0.36 Apr-91 1.1 22 1.28 17.25 0.41 1.13 May-91 1.26 15.2 1.62 2.55 0.89 0.71 Jun-91 0.82 3.75 1.02 1.38 0.78 0.31 Jul-91 1.42 7.76 0.81 0.62 0.79 0.11 Aug-91 0.91 1.2 0.5 0.81 0.33 0.06 Sep-91 1.24 10.31 0.86 1.08 2.54 0.95 0.56 0.08 Oct-91 1.16 86.9 1.06 3.65 86.9 0.826 0.7 0.12 Nov-91 2.49 76.1 1.54 11.9 0.746 0.5 0.12 Dec-91 2.57 59.8 2.51 21.2 44.3 0.816 0.61 0.33 Jan-92 1.416 34.8 0.69 0.58 0.25 Feb-92 1.56 44.5 0.71 0.57 0.213 Mar-92 3.61 82 0.489 0.66 0.239 Apr-92 1.24 56.4 1.02 5.79 21.9 0.77 0.76 0.82 May-92 1.6 3.41 1.07 6.57 3.05 0.23 0.9 1.09 Jun-92 0.72 2 0.63 2.23 0.98 0.35 0.52 0.24 Jul-92 1.36 6.76 0.6 1.06 0.45 0.31 0.3 0.15 Aug-92 1.63 2.16 1.26 1.05 1.5 0.293 0.23 0.068 Sep-92 6.69 3.89 2.92 8.18 1.28 0.301 0.6 0.1 Oct-92 3.31 3.75 1.95 6 0.6 0.273 0.74 0.12 Nov-92 Dec-92 1.94 39 Jan-93 0.435 34.7 0.28 1.161 0.67 Feb-93 0.15 30.9 0.13 0.7 0.429 Mar-93 1.1 71.7 1.78 8.64 8.8 0.329 0.89 0.464 Apr-93 1.1 17.1 0.76 2.12 9.94 0.37 2.01 0.826 May-93 1.63 6.55 1.31 4.38 4.84 0.234 0.98 0.501 Jun-93 1.22 10.2 1.02 0.424 1.6 0.154 1.01 0.19 Jul-93 1.22 4.9 1.14 1.52 0.29 0.19 0.28 0.06 Aug-93 0.698 7.33 0.517 2.17 0.47 0.19 0.242 0.052 Sep-93 0.824 4.37 0.859 3.31 0.876 0.21 0.38 0.06 Oct-93 1.64 10.3 0.78 3.06 1.74 0.21 0.38 0.06 Nov-93 1.38 16.4 0.935 3.25 0.22 0.589 0.265 Dec-93 1.17 39.6 0.99 1.016 3.17 0.32 0.84 0.74 Jan-94 0.435 34.7 0.24 1.509 0.25 Feb-94 0.937 2.759 0.045 Mar-94 0.354 0.605 0.63 Apr-94 1.14 19.56 0.69 1.54 10.65 0.148 1.09 0.033 May-94 1.34 12.6 0.8 2.11 8.06 0.107 1.864 0.13 Jun-94 0.67 8.17 0.78 1.56 0.67 0.176 0.48 0.093 Jul-94 0.9 7.88 1.42 1.15 0.45 0.16 0.33 0.087 Aug-94 0.9 7.88 1.62 2.96 0.321 0.175 0.14 0.066 Sep-94 1.58 8.42 1.29 1.66 0.445 0.183 0.158 0.08 Oct-94 Nov-94 0.59 3.74 1.6 Dec-94 0.98 7.03 0.496 1.97 4.29 0.21 0.611 0.31 Jan-95 1.11 31.3 0.26 0.83 0.66 139 Date PP G 1 L2 As CN As Fe CN As Fe CN Feb-95 2.29 25.5 0.28 0.83 0.83 Mar-95 1.39 73.4 0.86 0.696 0.34 0.041 0.794 0.746 Apr-95 1.25 28.8 1.14 0.955 11.8 0.232 0.918 0.543 May-95 1.08 3.99 0.893 3.13 2.48 0.25 1.01 0.11 Jun-95 1.53 3.5 0.805 2.47 2 0.24 1.02 0.05 Jul-95 1.59 3.4 1.16 0.45 0.22 4.23 0.05 Aug-95 2.68 8.75 1.08 6.98 0.35 0.18 2.75 0.05 Sep-95 2.04 4.88 1.3 0.818 0.22 0.264 0.395 0.06 Oct-95 1.72 6.6 1.26 0.881 0.8 0.302 0.37 0.078 Nov-95 1.58 14 1.05 1.56 2.52 0.321 0.304 0.235 Dec-95 3.23 23.4 1.87 2.09 6.9 0.396 0.31 0.56 Jan-96 1.61 26.9 3.42 17.07 24.1 0.364 0.406 0.239 Feb-96 2.31 11.15 0.42 0.53 0.73 Mar-96 1.12 43.2 0.063 0.648 0.028 Apr-96 1.98 48.7 0.332 3.72 0.107 0.734 0.026 May-96 2 2.89 0.782 7.51 • 5.38 0.44 0.79 0.14 Jun-96 0.296 10 0.796 2.22 1.6 0.22 0.5 0.09 Jul-96 0.607 0.052 0.821 0.2 0.042 0.22 0.17 0.06 Aug-96 0.257 0.047 0.641 0.349 0.046 0.22 0.14 0.06 Sep-96 0.244 0.063 0.641 0.349 0.046 0.22 0.23 0.06 Oct-96 0.247 0.034 0.658 1.01 0.082 0.27 0.27 0.07 Nov-96 0.201 0.11 0.196 2.1 0.088 0.23 0.36 0.08 Dec-96 0.18 0.151 0.236 2.99 0.046 0.19 0.31 0.1 Jan-97 0.175 0.142 0.316 0.714 0.083 0.19 0.31 0.1 Feb-97 0.264 0.939 0.28 0.68 0.15 Mar-97 0.228 0.266 0.279 0.111 Apr-97 0.215 2.588 0.31 0.71 0.16 May-97 0.278 0.049 0.278 0.024 0.15 0.98 0.05 Jun-97 0.209 0.03 0.228 0.017 0.2 0.389 0.048 Jul-97 0.282 0.024 0.774 0.374 0.034 0.274 0.183 0.039 Aug-97 2.28 0.111 0.864 0.274 0.025 0.42 0.15 0.04 Sep-97 0.434 0.022 1.79 1 0.1 0.45 0.48 0.04 Oct-97 0.318 0.03 0.469 1.45 0.029 0.37 0.25 0.04 Nov-97 0.23 0.067 0.19 0.786 0.017 0.31 0.26 0.06 Dec-97 0.24 0.106 0.355 0.69 0.022 0.27 0.27 0.09 Jan-98 0.327 0.094 2.015 1.63 0.062 0.316 0.419 0.086 Feb-98 0.31 0.66 0.08 Mar-98 0.25 0.158 0.687 0.683 0.108 0.423 0.541 0.07 Apr-98 0.327 0.037 0.183 0.823 0.045 0.174 0.662 0.035 May-98 0.315 0.025 0.194 0.837 0.019 0.261 0.378 0.043 Jun-98 0.145 0.021 0.204 0.318 0.015 0.316 0.196 0.04 Jul-98 • 0.593 0.026 0.553 0.177 0.024 0.446 0.148 0.053 Aug-98 1.75 0.023 3.24 0.367 0.04 0.732 0.193 0.052 Sep-98 4.05 0.084 1.28 0.171 0.026 0.664 0.18 0.064 Oct-98 0.44 0.029 0.311 0.755 0.068 0.328 0.422 0.041 Nov-98 0.368 0.079 0.841 0.546 0.086 0.454 0.154 0.102 Dec-98 0.454 0.077 1.24 0.338 0.018 0.592 0.199 0.106 Jan-99 1.67 0.370 0.028 0.704 0.229 0.095 Feb-99 12.9 1.950 3.05 0.683 0.037 0.690 0.163 0.106 140 Date PP G1 L2 As CN As Fe CN As Fe CN Mar-99 0.551 0.341 0.104 Apr-99 0.321 0.068 1.60 0.602 0.047 0.243 0.236 0.038 May-99 0.457 0.052 0.29 0.389 0.019 0.333 0.367 0.073 Jun-99 0.444 0.040 0.44 0.365 0.019 0.442 0.238 0.059 Jul-99 0.566 0.032 0.81 0.271 0.008 0.530 0.247 0.031 Aug-99 0.884 0.040 1.56 0.242 0.034 0.556 0.385 0.046 Sep-99 0.797 0.047 2.25 0.367 0.030 0.534 0.136 0.062 Oct-99 0.453 0.020 0.83 0.490 0.030 0.416 0.244 0.069 Nov-99 0.316 0.043 0.37 0.666 0.012 0.327 0.300 0.081 Dec-99 0.332 0.072 0.36 1.200 0.016 0.328 0.181 0.077 Jan-00 0.333 0.207 0.075 Feb-00 0.559 0.413 0.336 0.230 0.076 Mar-00 0.53 0.225 0.260 0.620 0.015 0.341 0.441 0.071 Apr-00 0.278 0.035 0.210 0.421 0.017 0.175 0.601 0.032 May-00 0.333 0.029 0.341 0.385 0.016 0.254 0.392 0.045 Jun-00 0.328 0.064 0.251 0.341 0.015 0.303 0.308 0.032 Jul-00 0.455 0.047 0.586 0.399 0.022 0.368 0.386 0.036 Aug-00 0.659 0.181 0.851 0.329 0.030 0.381 0.343 0.052 Sep-00 0.798 0.065 0.467 0.387 0.036 0.344 0.304 0.034 Oct-00 1.15 0.111 0.354 0.189 0.040 0.291 0.156 0.052 Nov-00 0.717 0.593 0.322 0.622 0.061 0.240 0.125 0.053 Dec-00 0.931 1.03 0.292 0.653 0.065 0.213 0.252 0.050 Jan-01 1.8 0.736 0.403 0.058 0.223 0.681 0.059 Feb-01 0.348 0.512 0.078 0.225 0.798 0.047 Mar-01 2.25 1.01 0.313 0.409 0.069 0.36 0.787 0.05 Apr-01 1.9 3.58 0.386 0.555 0.262 0.216 0.845 0.043 May-01 3.18 10.5 0.389 0.530 0.104 0.238 0.719 0.035 Jun-01 2.16 0.176 0.376 0.363 0.062 0.255 0.323 0.061 Jul-01 1.24 0.541 0.740 0.213 0.092 0.385 0.266 0.038 Aug-01 2.23 0.646 1.450 0.136 0.099 0.63 0.483 0.033 Sep-01 3.8 7.49 1.285 0.158 0.084 0.569 0.583 0.029 Oct-01 1.97 6.24 1.066 0.137 0.088 0.447 0.239 0.065 Nov-01 1.31 7.57 0.886 0.181 0.099 0.37 0.123 0.545 Dec-01 0.893 0.239 0.097 0.362 0.126 0.044 Jan-02 1.64 2.66 0.81 0.26 0.117 0.576 0.499 0.042 Feb-02 1.12 1.38 0.84 0.358 0.121 0.42 0.579 0.046 Mar-02 0.92 0.316 0.24 0.38 0.683 0.048 Apr-02 0.515 0.371 0.177 0.236 1.17 0.023 May-02 0.98 0.166 0.562 0.442 0.083 0.208 0.598 0.015 Jun-02 1.12 2.68 0.602 0.428 0.022 0.314 0.363 0.019 Jul-02 2.09 0.025 1.24 0.23 0.011 0.56 0.312 0.006 Aug-02 3 0.072 1.3 0.191 0.044 0.648 0.274 0.013 Sep-02 2.3 0.182 1 0.208 0.044 0.512 0.232 0.021 Oct-02 1.82 0.188 0.77 0.393 0.132 0.014 141 Primary Pond Data Date TSS PH Total CN Cu Ni Zn As Fe Ammonia 05-Jan-93 16 8 34.7 3.59 2.71 1.98 0.44 4.52 7.53 08-Feb-93 8.4 7.2 30.9 3.42 2.69 0.4 0.15 01-Mar-93 5.2 9.7 71.7 3.65 3.25 7.19 1.1 06-Apr-93 4 9.1 17.1 2.62 2.04 1.68 1.1 4.1 4.84 04-May-93 100.4 8.7 6.55 0.81 0.87 0.35 1.63 1.14 17-May-93 10.5 1.01 1.56 0.92 2.15 26-May-93 10.2 1.2 1.68 0.84 1.28 03-Jun-93 22.8 8.4 10.2 2.15 1.99 0.62 1.22 10-Jun-93 7.32 3.08 2.04 0.18 1.42 7.3 15-Jun-93 11.08 3.9 1.63 0.2 1.32 4.6 23-Jun-93 8.03 4.97 2.23 0.12 1.48 7 29-Jun-93 3.95 2.87 1.29 0.13 1.18 7.2 05-Jul-93 120.5 12 4.9 2.45 0.89 0.17 1.22 10.4 3 13-Jul-93 5.46 2.26 1.06 0.19 0.62 3.5 20-Jul-93 11.8 4.04 1.36 0.04 0.61 4.08 29-Jul-93 1.28 0.32 0.16 0.11 0.9 1.86 05-Aug-93 4.51 0.4 0.15 0.2 0.57 1.52 11-Aug-93 107.5 12.7 7.33 2.46 0.64 0.14 0.7 3.56 20-Aug-93 7.32 2.41 0.84 0.13 1.5 15.1 3.02 23-Aug-93 9.5 3.47 1.17 0.14 1.05 3.86 01-Sep-93 3.59 1.51 0.5 0.17 0.86 3.82 08-Sep-93 20.1 5.94 1.61 0.61 1.21 3.65 13-Sep-93 87 7.85 4.37 2.26 0.63 0.16 0.82 3.1 22-Sep-93 11.44 3.9 1.2 0.64 1.24 3.53 29-Sep-93 7.44 3.31 1.12 0.57 0.78 3.57 04-Oct-93 241 7.56 10.3 4.47 1.35 0.68 1.64 23.1 3.08 13-Oct-93 12.7 5.11 1.49 0.63 0.72 3.33 10-Jan-94 1.6 9.98 1.09 0.32 0.11 0.03 0.37 0.37 0.79 12-Apr-94 81.2 8.28 19.56 4.81 1.89 1.54 1.14 7.75 3.78 02-May-94 189 12.6 5.27 1.67 0.55 1.34 9.5 2.73 06-Jun-94 20.4 7.78 8.17 3.32 1.56 0.19 0.67 4.44 5.53 13-Jul-94 33.2 8.07 7.88 2.07 1.37 0.219 0.9 3.55 6.72 04-Aug-94 17.6 7.46 4.88 1.9 1.36 0.9 1.61 2.32 7.63 06-Sep-94 19.2 7.58 8.42 2.69 1.68 0.275 1.58 2.43 4.49 03-Oct-94 19.2 7.93 5.48 2.21 1.79 0.382 1.58 1.77 6.77 07-Nov-94 25.6 7.67 5.83 1.89 0.796 0.641 0.54 .3.12 2.07 05-Dec-94 13.6 8.34 7.03 4.83 1.84 2.7 0.98 2.94 3.41 11-Jan-95 8.8 7.69 31.3 9.22 3 2.11 1.11 2.4 6.49 08-Feb-95 23.5 7.52 25.5 6.215 2.6 1.69 2.29 3.8 6.27 09-Mar-95 4.4 9.02 73.4 5.096 3.842 15.01 1.39 10.1 13-Apr-95 14 8.56 28.8 2.26 1.98 3.52 1.25 3.45 4.52 17-May-95 160 7.68 3.99 0.689 0.57 0.29 1.08 11.47 1.5 02-Jun-95 115 11.36 3.5 0.547 0.462 0.4 1.53 7.12 2.81 05-Jul-95 24.5 7.75 3.4 0.468 0.784 0.36 1.59 3.75 4.39 11-Aug-95 84 7.23 8.75 0.708 1.84 1.017 2.68 15.39 3.68 11-Sep-95 64 6.83 4.88 0.834 1.41 0.34 2.04 4.13 02-Oct-95 12 6.79 6.6 1 1.91 0.67 1.72 6.35 08-Nov-95 56 7.34 14 1.45 1.42 1.94 1.58 4.72 4.34 04-Dec-95 26 7.1 23.4 3.31 1.94 2.27 3.23 5.79 142 Date TSS PH Total CN Cu NI Zn As Fe Ammonia 08-Jan-96 17.6 26.9 2.12 2.18 3.68 6.38 05-Feb-96 121 11.15 2.89 2.46 0.204 2.31 9.29 11-Mar-96 12.8 6.29 43.2 5.05 4.7 2.2 1.12 1.83 13.8 08-Apr-96 18.2 7.99 48.7 5.91 4.26 9 1.98 3.84 9.83 06-May-96 2.89 0.16 0.408 0.44 2 0.614 03-Jun-96 53.5 10 1.51 0.892 0.307 0.296 2.3 23-Jul-96 22.4 0.052 0.055 0.049 0.013 0.607 0.264 20-Aug-96 4 0.047 0.027 0.034 0.005 0.257 0.406 0.14 04-Sep-96 7 8 0.063 0.027 0.036 0.004 0.244 0.12 07-Oct-96 14 0.034 0.023 0.028 0.009 0.247 0.084 15-Nov-96 7 7.26 0.11 0.036 0.074 0.019 0.201 1.38 0.61 09-Dec-96 11 0.151 0.014 0.073 0.013 0.18 0.89 Jan-97 17 0.142 0.023 0.083 0.015 0.175 1.5 Feb-97 27 7.37 0.939 0.051 0.11 0.014 0.264 0.96 Mar-97 15 7.77 0.266 0.019 0.07 0.011 0.228 0.83 Apr-97 9 7.78 2.588 0.107 0.11 0.011 0.215 1.4 May-97 34 7.58 0.049 0.022 0.037 0.015 0.278 0.35 Jun-97 5 8.26 0.03 0.016 0.026 0.004 0.209 0.469 0.06 Jul-97 6 7.96 0.024 0.018 0.02 0.012 0.282 0.38 0.14 Aug-97 28 8.07 0.111 0.091 0.04 0.036 2.28 1.34 0.16 Sep-97 5 8.02 0.022 0.016 0.011 0.002 0.434 0.08 Oct-97 11 7.76 0.03 0.014 0.021 0.014 0.318 0.12 Nov-97 3 8.08 0.067 0.014 0.019 0.007 0.23 0.491 0.83 Dec-97 1 7.28 0.106 0.017 0.023 0.022 0.24 0.491 1.9 Jan-98 5 6.94 0.094 0.009 0.045 0.014 0.327 1.19 1.83 Mar-98 6 7.69 0.158 0.016 0.052 0.018 0.25 1.02 1.04 Apr-98 34 7.68 0.037 0.019 0.006 0.029 0.327 2.85 0.31 May-98 7 7.9 0.025 0.017 0.022 0.01 0.315 0.359 <0.05 Jun-98 5 8.03 0.021 0.011 0.018 0.003 0.145 0.296 0.05 Jul-98 6 7.81 0.026 0.009 0.011 0.006 0.593 0.368 0.12 Aug-98 2 7.84 0.023 0.008 0.006 0.007 1.75 0.263 0.08 Sep-98 28 8.07 0.084 0.059 0.035 0.07 4.05 0.804 <0.05 Oct-98 2 6.82 0.029 0.003 <0.002 0.002 0.44 0.141 <0.05 Nov-98 3 7.89 0.079 0.01 0.003 0.004 0.368 0.153 <0.05 Dec-98 3 7.65 0.077 0.012 0.036 0.009 0.454 0.158 0.34 Feb-99 88.00 7.47 1.950 <0.005 0.405 0.05 12.9 0.669 5.4 Apr-99 7 7.68 0.068 0.013 0.048 0.018 0.321 0.801 0.99 May-99 8 7.95 0.052 0.014 0.033 0.014 0.457 0.706 0.58 Jun-99 <2 8.07 0.040 0.015 0.027 0.007 0.444 0.562 0.17 Jul-99 6 8.02 0.032 0.016 <0.005 0.009 0.566 0.236 0.27 Aug-99 6 8.05 0.040 0.021 <0.02 0.006 0.884 0.282 0.34 Sep-99 3<T 7.79 0.047 <0.02 <0.02 0.003 0.797 • 0.089 0.78 Oct-99 6 7.81 0.020 <0.02 0.031 0.007 0.453 0.308 0.86 Nov-99 9 7.84 0.043 <0.02 <0.02 0.004 0.316 0.838 1.59 Dec-99 2<T 7.56 0.072 <0.02 <0.02 0.018 0.332 0.269 1.15 Feb-00 65 7.46 0.413 0.066 <0.02 0.068 0.559 1.52 3.99 Mar-00 4<T 7.49 0.225 0.046 0.041 0.029 0.53 0.912 6.38 Apr-00 12 7.7 0.035 <0.02 <0.02 0.011 0.278 0.729 1.09 May-00 11 8.21 0.029 <0.02 <0.02 0.01 0.333 0.445 0.85 Jun-00 5 7.71 0.064 0.024 <0.02 0.023 0.328 0.731 2.77 Jul-00 2<T 7.99 0.047 <0.02 <0.02 0.012 0.455 0.243 0.57 143 Date TSS PH Total CN Cu Ni Zn As Fe Ammonia Aug-00 2<T 8.33 0.181 <0.02 <0.02 0.047 0.659 0.157 0.29 Sep-00 17 7.66 0.065 0.026 <0.02 0.01 0.798 1.34 7.27 Oct-00 8 8.06 0.111 0.04 0.028 0.045 1.15 0.871 17 Nov-00 8.1 0.593 0.333 0.054 0.006 0.717 <0.005 16.5 Dec-00 7.7 1.03 0.236 0.033 0.011 0.931 <0.005 23.1 May-01 <2 7.6 1.010 0.118 0.019 0.020 1.16 0.770 20.3 Jun-01 8 0.080 0.035 0.017 0.008 1.01 0.188 14.9 Jul-01 <2 8 0.149 0.045 0.025 0.027 1.34 0.243 29.7 Aug-01 4 8.0 0.177 0.024 0.015 0.011 2.14 0.085 33.4 Sep-01 2 7.8 0.527 0.330 0.021 0.012 2.5 0.535 33.4 Oct-01 6 7.6 0.210 0.156 0.011 0.002 2.84 0.353 35.7 Nov-01 6 7.8 0.430 0.130 0.019 0.013 1.17 0.803 24.6 Dec-01 3 8.7 0.847 0.352 0.020 <0.002 1.77 0.433 39.8 Jan-02 6 7.8 2.660 0.053 0.009 0.011 1.64 1.190 52.0 Feb-02 15 8.1 1.380 0.215 0.041 0.009 1.12 0.765 37.2 May-02 11 7.9 0.166 0.102 0.011 0.007 0.98 0.415 14.0 Jun-02 6 7.8 2.680 0.074 0.016 0.026 1.12 0.191 19.8 Jul-02 3 7.4 0.025 0.036 0.013 <0.006 2.09 0.083 15.9 Aug-02 2 7.7 0.072 0.027 0.015 <0.006 3.00 0.063 24.5 Sep-02 3 7.4 0.182 0.023 0.010 <0.006 2.30 0.092 23.5 Oct-02 <2 8.0 0.188 0.029 0.010 <0.006 1.82 0.063 26.8 Nov-02 2 8.7 0.250 0.042 0.013 <0.006 1.32 0.120 31.7 Dec-02 4 8.6 0.313 0.038 0.011 <0.006 1.13 0.127 37.4 144 Secondary Pond Data Date TSS PH Total CN Cu Ni Zn As Fe Ammonia 03-Mar-93 8.4 9.1 68.8 5.12 4.48 6.6 1.78 8.64 06-Apr-93 4.8 8.5 9.94 3.56 1.65 0.85 0.76 2.12 5.56 23-Apr-93 7.06 0.72 0.88 0.41 1.09 29-Apr-93 5.26 1.64 1.23 0.34 1.08 04-May-93 43.6 8.6 4.84 1.1 1.06 0.22 1.31 4.38 4.06 17-May-93 2.24 0.37 0.88 0.19 1.4 4.78 26-May-93 2.6 0.45 1.01 0.25 0.97 03-Jun-93 15.2 8.1 1.6 0.36 0.91 0.18 1.02 0.42 10-Jun-93 0.25 0.64 0.98 0.12 1.26 10.8 15-Jun-93 0.85 0.83 0.96 0.11 1.32 11.3 23-Jun-93 0.61 1.04 0.96 0.09 1.44 11.9 29-Jun-93 2.7 1.36 0.96 0.1 1.37 11 05-Jul-93 109 12.1 0.29 1.31 0.85 0.14 1.14 1.52 11.6 13-Jul-93 0.87 1.04 0.71 0.1 0.84 10.7 20-Jul-93 1.15 1.16 1.03 0.08 0.93 10.82 29-Jul-93 0.75 0.82 0.46 0.09 0.77 8.71 05-Aug-93 ' 0.38 0.28 . 0.17 0.06 0.62 7.36 09-Aug-93 67.5 12.4 0.47 0.46 0.17 0.06 0.52 2.17 5.67 20-Aug-93 0.8 0.42 0.25 0.04 0.86 4.51 23-Aug-93 0.61 0.55 0.28 0.03 0.93 6.03 01-Sep-93 0.86 0.76 0.38 0.06 0.89 6.13 08-Sep-93 0.97 0.8 0.35 0.07 0.91 6.45 13-Sep-93 18.5 7.76 0.88 0.84 0.38 0.08 0.86 3.31 6.97 22-Sep-93 0.95 1.15 0.47 0.15 0.83 6.73 29-Sep-93 1.3 1.37 0.56 0.19 0.69 6.88 04-Oct-93 20.8 7.37 1.74 1.54 0.61 0.23 0.78 3.06 6.49 13-Oct-93 2.13 1.79 0.72 0.27 0.66 6.32 12-Apr-94 17.6 8.02 10.65 5.62 1.61 0.88 0.69 1.54 3.69 02-May-94 13:2 8.6 8.06 3.75 1.29 0.36 0.8 2.11 3.36 06-Jun-94 12.8 7.64 0.67 2.13 1.17 0.09 0.78 1.55 8.02 12-Jul-94 16 7.92 0.45 0.833 0.657 0.068 1.42 1.15 13.8 12-Aug-94 10 7.61 0.44 0.575 0.597 0.062 1.62 2.96 14.5 06-Sep-94 12.4 7.49 0.445 0.747 0.709 0.07 1.29 1.66 12.8 03-Oct-94 8.4 7.66 0.368 0.767 0.847 0.118 0.78 0.546 10.4 07-Nov-94 41 7.41 1.6 0.99 0.63 0.28 0.59 3.74 3.87 05-Dec-94 16.4 7.41 4.29 2.05 0.83 0.44 0.5 1.97 4.77 09-Mar-95 4.4 7.3 0.34 1.688 0.848 0.162 0.86 0.696 8.38 10-Apr-95 3.6 7.57 11.8 2.7 1.47 1.04 1.14 0.955 4.09 18-May-95 31.6 7.66 2.48 1.08 0.993 0.37 0.893 3.13 5.02 07-Jun-95 41 11.21 2 0.228 0.241 0.082 0.805 2.47 4.21 05-Jul-95 4.8 7.13 0.45 0.182 0.333 0.062 1.16 1.16 2.96 11-Aug-95 10.4 6.47 0.35 0.17 0.407 0.037 1.08 6.98 3.68 11-Sep-95 12.4 6.26 0.22 0.218 0.527 0.054 1.3 0.818 2.5 02-Oct-95 10.4 6.47 0.8 0.359 0.765 0.164 1.26 0.881 3.08 08-Nov-95 10 6.2 2.52 0.49 0.95 0.262 1.05 1.56 3.24 04-Dec-95 16 6.55 6.9 1.54 1.19 0.621 1.94 2.09 4.85 08-Jan-96 51.2 24.1 2.09 1.94 2.17 3.42 17.07 8.39 30-Apr-96 6 6.18 3.72 1.21 0.621 0.553 0.332 1.59 06-May-96 112 5.38 0.945 0.65 0.305 0.782 7.51 1.87 145 Date TSS PH Total CN Cu Ni Zn As Fe Ammonia 03-Jun-96 21.2 1.6 0.598 0.44 0.136 0.796 2.22 3.95 23-Jul-96 7.2 0.042 0.088 0.205 0.013 0.821 0.2 2.17. 20-Aug-96 9 0.046 0.047 0.122 0.012 0.641 0.349 0.23 04-Sep-96 13 8 0.087 0.057 0.091 0.016 0.516 0.749 0.13 07-Oct-96 18 0.082 0.063 0.068 0.024 0.658 1.01 0.052 15-Nov-96 7 6.84 0.088 0.055 0.039 0.083 0.196 2.1 0.2 09-Dec-96 32 0.046 0.036 0.046 . 0.037 0.236 2.99 0.17 Jan-97 2.73 0.083 0.022 0.067 0.016 0.316 0.714 1 May-97 11 7.45 0.024 0.008 0.017 0.008 0.278 0.21 Jun-97 17 7.43 0.017 0.039 0.032 0.019 0.228 <0.05 Jul-97 20 7.88 0.034 0.042 0.031 0.006 0.774 0.374 0.07 Aug-97 5 8.09 0.025 0.017 0.022 0.023 0.864 ..274 0.1 Sep-97 32 8.12 0.1 0.062 0.017 0.026 1.79 1 0.07 Oct-97 16 7.58 0.029 0.029 0.024 0.018 0.469 1.45 <0.05 Nov-97 4 7.53 0.017 0.013 <0.002 0.019 0.19 0.786 0.12 Dec-97 2 7.48 0.022 0.046 0.035 0.022 0.355 0.69 0.1 Jan-98 5 7.06 0.062 0.041 0.064 0.037 2.015 1.63 0.4 Mar-98 3 7.88 0.108 0.02 0.023 0.018 0.687 0.683 0.96 Apr-98 8 7.29 0.045 0.004 0.008 0.026 0.183 0.823 0.24 May-98 9 7.49 0.019 0.026 0.029 0.016 0.194 0.837 <0.05 Jun-98 3 7.85 0.015 0.021 0.011 0.006 0.204 0.318 <0.05 Jul-98 4 7.88 0.024 0.028 0.01 0.002 0.553 0.177 0.05 Aug-98 8 7.82 0.04 0.038 0.005 0.011 3.24 0.367 0.06 Sep-98 3 7.97 0.026 0.013 0.02 0.041 1.28 0.171 <0.05 Oct-98 9 7.58 0.068 0.056 0.031 0.019 0.311 0.755 0.06 Nov-98 4 7.85 0.086 0.029 0.018 0.015 0.841 0.546 <0.05 Dec-98 4 7.66 0.018 0.031 0.024 0.013 1.24 0.338 0.05 Jan-99 <2 7.43 0.028 0.043 0.040 0.065 1.67 0.370 0.22 Feb-99 4 7.30 0.037 0.082 0.073 0.045 3.05 0.683 0.50 Apr-99 3 7.23 0.047 0.013 0.028 0.014 1.60 0.602 0.94 May-99 6 7.84 0.019 0.013 0.022 0.023 0.29 0.389 0.28 Jun-99 6 8.48 0.019 0.023 0.023 0.024 0.44 0.365 0.09 Jul-99 10 8.01 0.008 0.032 0.041 0.006 0.81 0.271 0.06 Aug-99 3 7.73 0.034 0.036 0.034 0.006 1.56 0.242 0.19 Sep-99 6 7.76 0.030 0.041 <0.02 0.008 2.25 0.367 0.15 Oct-99 29 7.80 0.030 <0.02 <0.02 <0.003 0.83 0.490 0.10 Nov-99 5 7.63 0.012 0.020 <0.02 <0.003 0.37 0.666 0.17 Dec-99 4<T 7.73 0.016 <0.02 <0.02 0.007 0.36 1.200 0.23 Mar-00 <5 7.38 0.015 <0.02 <0.02 0.010 0.260 0.620 0.24 Apr-00 7.00 7.55 0.017 <0.02 <0.02 <0.008 0.210 0.421 0.24 May-00 7.79 7.58 0.016 <0.02 <0.02 <0.011 0.341 0.385 0.18 Jun-00 6.84 7.47 0.015 <0.021 <0.025 0.013 0.251 0.341 0.09 Jul-00 3<T 7.22 0.022 <0.02 <0.023 <0.012 0.586 0.399 0.07 Aug-00 7.00 7.64 0.030 <0.02 <0.024 0.050 0.851 0.329 0.19 Sep-00 6.67 7.46 0.036 <0.016 <0.014 0.016 0.467 0.387 1.09 Oct-00 7.00 7.52 0.040 0.016 <0.01 <0.032 0.354 0.189 1.93 Nov-00 9.33 7.52 0.061 0.048 0.019 0.026 0.322 0.622 4.35 Dec-00 6.00 7.54 0.065 0.049 0.019 0.009 0.292 0.653 6.22 Jan-01 4<T 7.4 0.058 0.050 0.030 0.008 4.340 0.403 5.81 146 Date TSS PH Total CN Cu Ni Zn As Fe Ammonia Feb-01 6 7.4 0.078 0.060 0.020 0.016 0.348 0.512 6.27 Mar-01 <2 7.6 0.069 0.050 0.020 0.017 0.313 0.409 6.61 Apr-01 7 7.4 0.262 0.050 0.020 0.011 0.386 0.555 7.09 May-01 <5 7.7 0.104 0.043 0.017 0.027 0.389 0.530 10.50 Jun-01 <4 7.7 0.062 0.024 0.012 0.013 0.376 0.363 6.56 Jul-01 <4 7.9 0.092 0.019 0.016 0.013 0.740 0.213 4.50 Aug-01 4 7.9 0.099 0.017 0.019 0.008 1.450 0.136 4.50 Sep-01 <3 7.4 0.084 0.023 0.018 0.009 1.285 0.158 6.62 Oct-01 <2 7.7 0.088 0.031 0.014 <0.007 1.066 0.137 8.69 Nov-01 <5 7.9 0.099 0.044 0.013 <0.004 0.886 0.181 10.08 Dec-01 6 8.0 0.097 0.048 0.011 <0.004 0.893 0.239 11.55 Jan-02 4 7.9 0.117 0.055 0.017 0.010 0.810 0.260 14.0 Feb-02 7 7.6 0.127 0.049 0.016 0.019 0.840 0.358 14.5 Mar-02 <2 7.8 0.240 0.047 0.018 <0.006 0.920 0.316 15.2 Apr-02 8 7.4 0.177 0.039 0.012 <0.007 0.515 0.371 10.10 May-02 7 7.8 0.083 0.035 0.010 <0.007 0.562 0.442 8.74 Jun-02 12 8.0 0.022 0.028 0.010 <0.006 0.602 0.428 8.28 Jul-02 <2 7.7 0.011 0.016 0.009 <0.006 1.239 0.230 3.32 Aug-02 3 7.8 0.044 0.012 0.009 <0.006 1.30 0.191 4.10 Sep-02 5 7.5 0.044 0.012 0.009 <0.006 1.00 0.208 4.47 Oct-02 4 8.0 0.032 0.016 0.015 <0.006 1.000 4.06 5.18 Nov-02 3 7.9 0.009 0.017 0.015 0.009 0.77 0.109 7.71 Dec-02 2 7.45 0.034 0.015 0.016 0.012 0.81 0.118 12.1 Appendix II: Whole Rock Analysis Results CO 3 CD 131 CO cn cn U J cn CD OD i n cn m cn co cn CO < CO CO 3PI [Q's: £3 ko ko i o :a O CD > co 1 TD Ci — 3 CD m i ft .X t o ko 3'.: XI '.Is. CO CO Ln 'fir. c o l CD CO CD OD NJ cn cn o i 33 S io In la) 1 S i 3D I -S O i > so (Cf. .3 m o w K J CD cn CD cn cn CO' CD 6 , 73 r— TD IXII CO CD CD CD P CO 3 73 r-TD TD GD 4*. 00 CO :o. o: < co co O CD ~0 • B) 3 3 , o | U 2 > 71 a r-0) X 3 CD tf) o E V) 30 <D (A C : V) CO o CO CO o o CO •a = 3 t = f P KO^ 3? 0) cr a> 1 71 <t> a |— CD m CD 3 CD w o a tn 73 CD (A C (/) Appendix III: Rietveld Refinement Reports 150 Quantitative Phase Analysis of 14 samples using the Rietveld Method and X-ray Powder Diffraction Data. Attention: Desiree Meilleur Mati Raudsepp, Ph.D. Elisabetta Pani, Ph.D. Dept. of Earth & Ocean Sciences 6339 Stores Road The University of British Columbia Vancouver, BC V6T1Z4 April 23, 2003 151 EXPERIMENTAL METHODS The particle size of the fourteen samples from Cochenour Mine (CM series) and from Red Lake Mine (RLM series) was further reduced to the optimum grain-size range for X-ray analysis (<5 pm) by grinding under ethanol in a vibratory McCrone Micronising Mill (McCrone Scientific Ltd., London, UK) for 6 minutes. Fine grain-size is an important factor in reducing micro-absorption contrast between phases. Samples were pressed from the bottom of an aluminum sample holder against a ground glass slide; the cavity in the holder measures 43 x 24 x 1.5 mm. The textured surface of the glass minimizes preferred orientation of anisotropic grains in the part of the powder that is pressed against the glass. Step-scan X-ray powder-diffraction data were collected over a range 3-70°2G with CuKa radiation on a standard Siemens (Broker) D5000 Bragg-Brentano diffractometer equipped with a diffracted-beam graphite monochromator crystal, 2 mm (1°) divergence and antiscatter slits, 0.6 mm receiving slit and incident-beam Soller slit. The long sample holder used (43 mm) ensured that the area irradiated by the X-ray beam under these conditions was completely contained within the sample. The long fine-focus Cu X-ray tube was operated at 40 kV and 40 mA, using a take-off angle of 6°. X-ray powder-diffraction data were refined with Rietveld Topas 2.0 (Broker AXS) running on a Pentium III 1000 MHz personal computer. RESULTS AND DISCUSSION The X-ray diffractograms were analyzed using the International Centre for Diffraction Database PDF2 Data Sets 1-49 plus 70-86 using Search-Match software by Siemens (Broker). The results of quantitative phase analysis by Rietveld refinement are given in Table 1. Rietveld refinement plots are given in Figures 1-14. •c Pi I A cj u C/D. O v o m m o v n AO MO o as " O " C N CN r n o 00 v d I ^ d o c T o C N m r n - r n -C N O C N r n o O o o o o o 152 C N ' C N C N o o o © o C N i V O i s VO I a I OS C N ^3" O N VO tN m v© O N v d o o C N o -Ov-- o o -m ' - T O -O N C N o C N C N C N - C N -r n - C N -- C N -</-) CN* - v o -- T N -o o o © ' -f-O -tN-O O O O O o o o o o r n •si-rs V O C N -o~ -m-ON p i -rn C N m o t - - v o -o - O v -C N -CN-O -CN-O O O O O C N i m • m s <? CN C N C N m C O m VO m C N C N - C N -C N - o o -C N -CN-VO r n -r-C N C N C N m - M S -C D o ~Or-o -CN-- t -o o " o - m -o - C N -O o o o o o o CN I s CJ C N . — 4 IS CJ o C N m CN C N r n C N o " VO OS - C N -- C t -ON - o o -C N - T N -C N - C N -m CN m m C N m -sj-CN" - O r -O C N - m -C N - o r O o -Or--1X3-O -t30~ o - o o -o © o o o o © ca o o C3 o u •4—» •c o o cc) e co ft- JJL cu _4) CM C o cd a 6 cd ^ 3 U O C2_ o 153 <^ Top as - C:\Topas 2-1 FJral\1estfiJes\Go!d£ot-p zi Aprit2Q(ra\GoIdcotpRlM2^ I-PRI [LO 1 <y>fqt Wtw f * l a u n c h T j P * Wndow Hefc D £ 0 1 fl 3 ] r<> i\ H ^ U t U oMc<xpHLM21RAW j j f "31 4.400 4.200-4,000-3.800 3;600J 3,400-3200 3000 2^00 2 POO 2.400-2,20a 2JD00-1JB0O-1j600-1.400-1200-1J0OO 800 600 400 -200J -40D-| -600 -800 -1JD0O -1200 -1.400 -1JE00--1,800 Quati G y p s u m Pyntiotio Gmnente AcNfiOtr Stole 43.46 % 9.06% 3.87% £77 <4 2S6S 3 5 6 • » 14jD7% 081% : . 3 s % 2 X 0 J 4 * in i ; 'il iV i ! rn f i 3 0 3 2 3 4 3 B 3 8 4 0 42 44 4 6 4 8 5 0 S 2 S 4 58 58 62 64 66 68 70 Figure 8: . Rietveld refinement plot for sample RLM-2-1 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 154 i j > Topas - C:\Topas2-1 F inar iTest f l les iGoldcw-p Z2 ApriE003> G o l d L o t p R L M .PRO [ G o k k o r p R l M - AW] AW 7]! 3,400-3^00 3 POO 2 j H » -2J6O0 2.400 2.200-2 pOO 1JMO-J 1JB00 1.400-1.200-IflOO-800 600-| 400 20&f 0 fasm Quartz DotemXe B i o U e l M CfeTochliin; 1 Grunerte Cafcie Afeefio&yrte Mttptffis A t v i s s r e G o e t h i e 32.08% 2 1 5 4 % 7.94's 6 5 9 % 2 9 6 % 0.21 -!i 0.79 % 15.41 % 0.13 S -400-SOD--\PCD-Vu. ( 1 i l i I II II I I III I I H i l l i I ! I r S in si | l I | i i i II my la i i ni I i I I I I p I M M l I I I . I I ISI I I ! I I I I i 11 i I I I I I I I H I I IIII JIB i^ ynmy yjiuiijii i|iniiiijiiyii i | i i i p y i i i i i ^ » i i i | ii«l|i«yieii |[id|yyjiift»^fl jfljy 'S, Figure 9: . Rietveld refinement plot for sample RLM-5 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 155 ^VTopas - C:\Topas 2-] final\Testfites\GoWcorp 22 April20B3\GolctcotpRLM6- 1.PRO -1 'fii^ fStfi PJ?*"-: TAJJ*.". T S T ^ A C ^ 5,000 4,800 4,400-4,200 4J0O0-3jS00-3J60&1 3.400J 3,200-: 3JOOO^ 2,800-2:600 2,400-2,200-ipao 1*00 1,400-1J00-; 1JOOO| 8O0-; 600^ 400-j - 1 Setose Dctorste Gypsun EMU1M Magr*'jte tAisc*jvile2Mi Trcoivgc Pyrrhotae 43S1 % 3.16% 10.15% 1JJ1 ' * 3 B 1 % 4 5 8 % 1 0 . 4 4 * jJ .?2% S71 % '. .16% 074% 067-5. -20O -400--£00--800 -1J0OO--1,200--1,400--1,600 •1jB00 : -2J0Oo4 II I i B I i l i ' i ' i i M "II ii |' ii 'VU iWi iJWIi l l iVi i j ' i i ' i l i i N ' i r ' t i i t i i in i i l . ' i l j I I | I I! I I I II 6 8 10 12 14 16 18 20 22 24 26 30 32 34 3E 38 Figure 10:. Rietveld refinement plot for sample RLM-6-1 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 156 T o p a s - C : \ T o p a s 2-1 F ina lYJest f l !es \Gokfcorp 22 Aprn2003\Gi>!dcofpRl .M-6- ; c . 5200-i spoa-4000-4.600 4.400-4200-40OO : 3.600 3,600-3,400-i 3200-3000-2000-2000-2.400 2 2 0 0 2000-1000 1000-1.400 1200-1000^ 800-600-2 8 7 % 14.90% SderiSe Dotomle BM *1M Muscovite 2M1 i.54 -X ircserjtt y.>4 Tiraf lc , 117-=, Pyntwtte 058 % Calcte 0 0 0 % 6 58? 1 3 S S -400 -600 -800 -1000--1200 -1,4004 -10004 i i 6 8 10 12 14 16 18 20 22 24 28 28 30 32 34 38 38 40 42 44 46 48 50 52 S4 Figure 11:. Rietveld refinement plot for sample RLM-6-2 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 157 s * Topas - C : \ T o p a s 2-1 f i r w h l e s t M e ' ; - G a H c o m Z2 Apri l2003\KoIcfcorpRI>f7. fit- Si fi. a at ' " ' M i . 3 i i' ? •Mf, « £ . T I . H 3,200-3,000-2000 2 £ 0 0 2,100-2200-2000 1000-1000-1.400 1 2 0 0 1000-800-| GOO 400 200-f 0 -200-t^oo -600-] -€00 -1000-Quartz Coter ie autelM C&SCClasreS Grunerte AlVJ?5iP3 G M W C Pynhatte 3 2 2 5 % 1 8 3 8 % 5.30-X 3.02 'ia :• :i --. 5 0 8 % 2.40% 1.24% 0.S6 - ; , 15.34% 0.47% 2.16% l i V . / ' I . I , I I II II I I I I I I I . Il I I I I. M l l III II 1 II 111, II I I III II £  II JlJll^lglj! Figure 12: . Rietveld refinement plot for sample RLM-7 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 158 l o p a s - C : \ T o p a s 2-1 f lnal\7i;5tfi!es\Go!dcoiu 22 Apr i12003 \ fo lcko tpKLMS e d ortd.PRO [ C o k k o r p R L ^ S e c P o n d . f t A * ] M M 3,200-3.000H 2.800 2.600-) 2,400^ 2,2O0 ; 2J0O0; 1^00; 1,600; 1,400; 1.200J 1.000 600-600-400 Quartz 3 7 3 1 % Ooionie 11.25% B M i a l M 3 3 2 % dsjccli i jre a 1 7 5 s % Aflhdte 2 61 --. Grunerte 3 5 8 % ArsenoPVrSfi 0 . 1 7 ^ MasneUs G.70 •» ,l III 'niiini)inii'i^ iiiwu'>Vlwirwwj i i I I ; i | iifi I I | I I I ni I I I n«n,i|U)i|iif I iiiiinBiii ifiiw Jim iijiju ji.mi'mi'i 10 12 14 16 18 20 22 24 28 28 30 40 42 44 46 4 8 5 0 S 2 S 4 S 6 5 8 6 0 6 2 6 4 6 6 6 8 7 0 Figure 13: . Rietveld refinement plot for sample RLM-Secondary Pond (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 159 an '.-Mai "3r 3 * 2,900 2.BOCH 2.700-2,6004, 2,500; 2,400^ 2,300-2,200-1000 1.700^ 1,600 I S O O } 1,400-1000-900--3004, •400; -500-i' i Quartz Bo»e1M A V T l r s g i Gmnerte C a U a A i s e n o s y r S i Oosttte 3528% 1123% 10.66% 1S.11 'A 5.04% 1.36% 0J1 % 0.54 1£.S?-% 0.43% n,,rjy',|,r^"lV I I I II II 1 I Bl II Ul II I I I I 8 11 IIII II I I ifpiiin ii nn'iiarni iifniitf i vi la l rn i nurini mini tra oajnt fatt ri"iT Uu rataiia JIB , ' ni, m i 1 mn I I 11 1 1 1 iii r I I I I , mi I I si m I,II yipjijliiiyi ^ .iianuniii Ejiiiiin^aiiiiflin^ra^ 6 8 10 12 14 16 18 20 22 24 26 28 30 32 36 38 40 42 44 48 48 50 5 2 5 4 5 8 5 8 Figure 14: . Rietveld refinement plot for sample RLM-Primary Pond (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 160 Quantitative Phase Analysis of 9 samples using the Rietveld Method and X-ray Powder Diffraction Data. Attention: Desiree Meilleur Goldcorp Inc.- Red Lake Mine Balmertown ON POV1CO Mati Raudsepp, Ph.D. Elisabetta Pani, Ph.D. Dept. of Earth & Ocean Sciences 6339 Stores Road The University of British Columbia Vancouver, BC V6T1Z4 May 26, 2003 161 EXPERIMENTAL METHODS The particle size of the nine samples RLM-1, RLM-2-3, RLM-3-1, RLM-3-3, RLM-4, #3, #5, #24 and #29 was further reduced to the optimum grain-size range for X-ray analysis (<5 pm) by grinding under ethanol in a vibratory McCrone Micronising Mill (McCrone Scientific Ltd., London, UK) for 6 minutes. Step-scan X-ray powder-diffraction data were collected over a range 3-70°29 with CuKcc radiation on a standard Siemens (Bruker) D5000 Bragg-Brentano diffractometer equipped with a diffracted-beam graphite monochromator crystal, 2 mm (1°) divergence and antiscatter slits, 0.6 mm receiving slit and incident-beam Soller slit. The long fine-focus Cu X-ray tube was operated at 40 kV and 40 mA, using a take-off angle of 6°. X-ray powder-diffraction data were refined with Rietveld Topas 2.0 (Bruker AXS) running on a Pentium III 1000 MHz personal computer. RESULTS AND DISCUSSION The X-ray diffractograms were analyzed using the International Centre for Diffraction Database PDF2 Data Sets 1-49 plus 70-86 using Search-Match software by Siemens (Bruker). The results of quantitative phase analysis by Rietveld refinement are given in Table 1. Rietveld refinement plots are given in Figures 1-9. 162 o /) 3 c« I r -T M " \ 0 X ' I I r * I v e i — • I ON o -5 o " T M -O N C M tN =tfc CO 1 i co C N CO C M C O vq co' co o C M O N oo v© C O VO CM" i n 00 C O od C M CM' vq <n" ON CM C O CM" CO -co-vd C M - C O -CM" C O VO CO C M C O od C M C M ON ON vo o C M N CO CU CO CO »—i o o 03 O • i-H *> O o CO S •c O 3 J2_ CU -*—» *o r—I ca o • T M -u a, a •MS-C M CU -»—» o xi -Ov -Ov •CM-o " tM-- C M -CM o o o o o o o o o o o o o o o o o o o o o o " o xi CU o I2_ 13 o 163 T o p a s - C : \ T o p a s 2-1 F i n a l ' a e s l f i l e s \ G o W c D r p M a y Z 0 0 3 ' . G o W c f ( L M - J . 44, A C T S i* s t SPA 4 < --54 I T -jaT*l -Y-.i.-M£ 4,800 4^00 4.400-4300 4,000 3B00-3,400 3.200 2 ^ 2j00fr 1,600-IjSOO-1,400-1300 1,000 BO0-] Quartz 3G£5 % Dotoml!! 24.14% B o t t e l M 3 3 9 % 3.46 ' i Grunerte 654 % fcoyf>*lse 0.76 Antenna 15.09% A c S w S e 6 3 9 '=» Ouc4? 0 78 -r. -200--400--600--800--1J00O--120O -1.400-•i II I I i i » II un u i i in on HI leiiilimiiu B i i i i i i i i m iiiiiiaiiiiiiiiiiimiiiiJiiiiii nnnnnnii si mis insuiaii'uiiii! smia^  I I I II i in, i i i i 11 y H 'tin iiiiiiii IIIIJIII 8 1 1 1 n , i' ! 1 1 * " t i 'T 1 1 1 1 1111 , s m ^ ' i ' ? 1 ! 1 " 1 rafliiiiipiH^i y 11 III! Mil 5 6 5 0 6 0 6 2 6 4 6 6 6 8 7 0 Figure 1: . Rietveld refinement plot for sample R L M - 1 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 164 Iopas - L : \ r o p a s 2-1 l m a K I c s t f l l e r - . G o k t e C r t p M a v 2 u 0 3 l G o l d c R l M 2 - 3 . P R O - [ G i r f d c P , l M 2 - - . R A W ] • "iiiti-i.''ti" '^ ^>-vM Q r • Figure 2 : . Rietveld refinement plot for sample RLM-2-3 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 165 ^ T o p a s - C : \ T o p a s 2-1 F ina l \7est f i les - .GoldcorpMayZ0u3\Gt i !dcRl>13-3 .PRO. f G o M c a i K a -3r Quartz 49.18% Ddomle 9 0 8 % Seta* stf G.-?t -X-C & e c S i s r e l 14.76% Grunerie 356 % Gypsusi 0.76 % TiemcSe 1.58*& i 1 i i 1 i i i 1 i 1 1 i 1 i II 1 i V i " 1 i i i i ' H \ I ' I" i 4 6 8 10 12 14 16 1B 20 22 24 26 2 8 3 0 3 2 3 4 3 S 38 40 42 44 46 48 S 0 5 2 5 4 5 S 5 8 6 0 6 2 6 4 6 6 6 8 7 0 6000 5500 5000 4500 4000 3500-3000 2500 2,000 1500-1000 -1000-Figure 3: . Rietveld refinement plot for sample RLM-3-1 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 166 V T o p a s - C : \ T o p a s 2-1 f ina lMes t f i l es ' iGo idcofpMayZOOS^GordcRLMS-Sa .PRO - [ C o l c k R U M 3 - . 3 a . R A W ] . 5,600 5,4004, S 2 W 5000^ 4.8O0: 4000-; 4.40OJ 4230 4000 3000 30004 3,400 32304 3 0 0 0 20004 2000 2,400-22004 20Ot>J 10001 1000-1,4004 I2ar>j 1000 600 Quartz DolomSo 5 0 5 4 % 11.93% <Sreci*.Tel =.82% Grunerite 333 % Magnefa 141 % ArKMsS-* 1i6^*s> pyrrhoSo 2 JO i AGinoKe 0.S6 'i* S W r i r 1 OS l l l i l i l i l l H I i S I W i l l i nis nm ii an g u p psiiiiiii'innniiiffiNigiliiiii niiii!miiiii3ii(M|i|iiai4UTOi;iii«ii| in upi HUM \ I I i ymamt i l ia i i i i i i^iij i i \ la l i i i j i ni<|ajHiBrj Figure 4:. Rietveld refinement plot for sample RLM-3-3 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 167 Figure 5: . Rietveld refinement plot for sample RLM-4 (blue line - observed intensity at each step; red line - calculated pattern, solid grey line below - difference between observed and calculated intensities; vertical bars, positions of all Bragg reflections. Coloured lines are individual diffraction patterns of all phases. 168 Appendix IV: Sequential Extraction Data and Calculations 169 Table AIV-1 : Raw Sequential Extraction Data Sample Tare Total Liquid [As] Mass As Fraction As [Fe] Mass Fe Fraction Fe ml mg/L mg mg/L mg 1-1a 13.09 54.33 41.24 1.358 0.0560 0.0393 1.22 0.0503 0.0013 2-1 a 13.01 55.71 42.70 0.951 0.0406 0.0290 0.58 0.0248 0.0007 3-1 a 13.04 54.83 41.79 0.618 0.0258 0.0288 0 0.0000 0.0000 4-1 a 13.04 54.06 41.02 0.669 0.0274 0.0301 0 0.0000 0.0000 5-1 a 13.15 55.91 42.76 0.0847 0.0036 0.0023 0.23 0.0098 0.0002 6-1 a 13.00 54.05 41.05 0.0866 0.0036 0.0022 0.27 0.0111 0.0003 7-1 a 13.06 55.38 42.32 0.1475 0.0062 0.0067 0.05 0.0021 0.0001 8-1 a 13.04 60.04 47.00 0.1485 0.0070 0.0074 0 0.0000 0.0000 1-1b 12.98 55.44 42.46 0.832 0.0353 0.0248 1.72 0.0730 0.0019 2-1 b 13.13 55.17 42.04 0.594 0.0250 0.0179 0.95 0.0399 0.0011 3-1 b 13.03 54.47 41.44 0.239 0.0099 0.0111 0.14 0.0058 0.0002 4-1 b 13.01 55.26 42.25 0.291 0.0123 0.0135 0.25 0.0106 0.0003 5-1 b 13.01 54.61 41.60 0.099 0.0041 0.0026 0.92 0.0383 0.0009 6-1 b 13.00 54.97 41.97 0.124 0.0052 0.0032 1.29 0.0541 0.0013 7-1 b 13.05 56.14 43.09 0.1325 0.0057 0.0062 0.26 0.0112 0.0004 8-1 b 13.07 56.79 43.72 0.14 0.0061 0.0065 0.15 0.0066 0.0002 1-1w 13.14 24.17 11.03 0.76 0.0084 0.0059 1.07 0.0118 0.0003 2-1w 12.98 24.10 11.12 0.503 0.0056 0.0040 0.36 0.0040 0.0001 3-1w 13.03 22.89 9.86 0.0693 0.0007 0.0008 0.43 0.0042 0.0001 4-1 w 13.07 23.25 10.18 0.094 0.0010 0.0010 0.52 0.0053 0.0002 5-1 w 13.00 23.05 10.05 0.0449 0.0005 0.0003 0.38 0.0038 0.0001 6-1 w 13.00 23.19 10.19 0.0611 0.0006 0.0004 0.69 0.0070 0.0002 7-1 w 13.04 23.34 10.30 0.0675 0.0007 0.0008 0.24 0.0025 0.0001 8-1 w 13.06 23.09 10.03 0.0663 0.0007 0.0007 0.14 0.0014 0.0000 1-2a 13.00 56.09 43.09 16.48 0.7101 0.4979 47.69 2.0550 0.0539 2-2a 13.01 56.41 43.40 12.61 0.5473 0.3912 35.7 1.5494 0.0415 3-2a 13.00 58.09 45.09 1.94 0.0875 0.0976 38.52 1.7369 0.0524 4-2a 13.08 57.95 44.87 2.605 0.1169 0.1280 49 2.1986 0.0651 5-2a 13.03 57.24 44.21 5.969 0.2639 0.1673 18.56 0.8205 0.0199 6-2a 13.04 58.30 45.26 7.414 0.3356 0.2056 22.12 1.0012 0.0235 7-2a 13.09 56.25 43.16 5.477 0.2364 0.2552 22.48 0.9702 0.0311 8-2a 13.00 57.99 44.99 6.021 0.2709 0.2861 25.45 1.1450 0.0359 1-2b 13.08 57.81 44.73 1.861 0.0832 0.0584 26.83 1.2001 0.0315 2-2b 13.03 58.05 45.02 1.526 0.0687 0.0491 21.65 0.9747 0.0261 3-2b 13.06 58.10 45.04 0.0252 0.0011 0.0013 14.75 0.6643 0.0200 4-2b 13.08 49.91 36.83 0.0627 0.0023 0.0025 17.11 0.6302 0.0187 5-2b 13.02 57.62 44.60 1.116 0.0498 0.0315 17.47 0.7792 0.0189 6-2b 13.04 58.27 45.23 1.17 0.0529 0.0324 18.19 0.8227 0.0193 7-2b 13.01 59.21 46.20 2.127 0.0983 0.1061 21.99 1.0159 0.0326 8-2b 13.13 59.11 45.98 2.383 0.1096 0.1157 37.83 1.7394 0.0546 1-2w 12.96 22.79 9.83 0.1895 0.0019 0.0013 1.38 0.0136 0.0004 2-2w 13.02 23.15 10.13 0.216 0.0022 0.0016 1.45 0.0147 0.0004 3-2w 13.07 22.86 9.79 0.0102 0.0001 0.0001 0.75 0.0073 0.0002 4-2w 13.12 31.49 18.37 0.0725 0.0013 0.0015 4.49 0.0825 0.0024 5-2w 13.02 22.35 9.33 0.12 0.0011 0.0007 0.69 0.0064 0.0002 6-2w 13.04 22.86 9.82 0.1475 0.0014 0.0009 0.83 0.0082 0.0002 7-2w 13.09 23.24 10.15 0.1685 0.0017 0.0018 0.88 0.0089 0.0003 8-2w 13.05 23.67 10.62 0.226 0.0024 0.0025 1.27 0.0135 0.0004 170 Table AIV-1 (Cont): Raw Sequential Extraction Data Sample Tare | Total Liquid [As] Mass As Fraction As [Fei Mass Fe Fraction Fe ml mg/L mg mg/L mg 1-3a 13.03 55.28 42.25 1.935 0.0818 0.0573 208.86 8.8243 0.2317 2-3a 13.04 55.13 42.09 2.325 0.0979 0.0700 245.28 10.3238 0.2764 3-3a 13.01 65.35 52.34 0.0477 0.0025 0.0028 132.76 6.9487 0.2097 4-3a 12.94 53.22 40.28 0.0851 0.0034 0.0038 199.7 8.0439 0.2383 5-3a 13.09 54.14 41.05 5.769 0.2368 0.1501 237.23 9.7383 0.2367 6-3a 13.02 54.81 41.79 7.27 0.3038 0.1862 266.26 11.1270 0.2615 7-3a 13.11 34.46 21.35 4.103 0.0876 0.0946 129.82 2.7717 0.0889 8-3a 13.05 55.18 42.13 9.186 0.3870 0.4087 197.02 8.3005 0.2605 1-3w 13.12 23.30 10.18 0.284 0.0029 0.0020 27.68 0.2818 0.0074 2-3w 12.97 23.59 10.62 0.308 0.0033 0.0023 29.54 0.3137 0.0084 3-3w 13.12 23.35 10.23 0.034 0.0003 0.0004 18.64 0.1907 0.0058 4-3w 13.04 23.02 9.98 0.17 0.0017 0.0019 23.63 0.2358 0.0070 5-3w 13.07 23.29 10.22 1.095 0.0112 0.0071 50.27 0.5138 0.0125 6-3w 13.07 23.40 10.33 1.297 0.0134 0.0082 47.26 0.4882 0.0115 7-3w 13.02 23.32 10.30 1.03 0.0106 0.0115 23.12 0.2381 0.0076 8-3w 13.13 22.98 9.85 2.211 0.0218 0.0230 40.3 0.3970 0.0125 1-4a 13.09 55.01 41.92 0.208 0.0087 0.0061 18.83 0.7894 0.0207 2-4a 13.02 55.15 42.13 0.236 0.0099 0.0071 16.89 0.7116 0.0190 3-4a 12.97 54.63 41.66 0.0112 0.0005 0.0005 43.21 1.8001 0.0543 4-4a 13.02 55.43 42.41 0.0143 0.0006 0.0007 52.37 2.2210 0.0658 5-4a 12.99 55.50 42.51 3.833 0.1629 0.1033 191.02 8.1203 0.1974 6-4a 13.06 55.14 42.08 3.972 0.1671 0.1024 194.5 8.1846 0.1923 7-4a 13.08 54.82 41.74 2.081 0.0869 0.0938 130.2 5.4345 0.1744 8-4a 13.04 55.06 42.02 3.731 0.1568 0.1656 198.74 8.3511 0.2621 1-4w 13.06 23.15 10.09 0.139 0.0014 0.0010 16.11 0.1625 0.0043 2-4w 13.13 23.26 10.13 0.1455 0.0015 0.0011 12.92 0.1309 0.0035 3-4w 13.04 23.14 10.10 0.0023 0.0000 0.0000 5.95 0.0601 0.0018 4-4w 12.97 23.24 10.27 0.0037 0.0000 0.0000 7.06 0.0725 0.0021 4-5w 13.08 23.22 10.14 0.26 0.0026 0.0017 25.87 0.2623 0.0064 4-6w 13.01 23.09 10.08 0.279 0.0028 0.0017 31.83 0.3208 0.0075 4-7w 13.02 23.32 10.30 0.1945 0.0020 0.0022 24.23 0.2496 0.0080 4-8w 13.01 22.99 9.98 0.376 0.0038 0.0040 34.29 0.3422 0.0107 1-5a 13.08 55.78 42.70 0.587 0.0251 0.0176 48.81 2.0842 0.0547 2-5a 13.00 55.44 42.44 0.673 0.0286 0.0204 60 2.5464 0.0682 3-5a 13.00 55.44 42.44 0.012 0.0005 0.0006 34.3 1.4557 0.0439 4-5a 13.03 55.49 42.46 0.0138 0.0006 0.0006 40.21 1.7073 0.0506 5-5a 13.07 56.18 43.11 0.0245 0.0011 0.0007 31.65 1.3644 0.0332 6-5a 13.09 55.13 42.04 0.0342 0.0014 0.0009 38.74 1.6286 0.0383 7-5a 13.05 55.88 42.83 0.0854 0.0037 0.0039 34.34 1.4708 0.0472 8-5a 13.01 57.04 44.03 0.106 0.0047 0.0049 41.25 1.8162 0.0570 1-5w 13.11 23.48 10.37 0.1185 0.0012 0.0009 18.82 0.1952 0.0051 2-5w 13.15 23.90 10.75 0.086 0.0009 0.0007 12.72 0.1367 0.0037 3-5w 13.05 24.12 11.07 0.0093 0.0001 0.0001 8.39 0.0929 0.0028 4-5w 13.01 23.20 10.19 0.0093 0.0001 0.0001 10.28 0.1048 0.0031 5-5w 13.08 24.10 11.02 0.0147 0.0002 0.0001 7.98 0.0879 0.0021 6-5w 13.01 23.59 10.58 0.0144 0.0002 0.0001 7.16 0.0758 0.0018 7-5w 13.02 23.66 10.64 0.0281 0.0003 0.0003 5.49 0.0584 0.0019 8-5w 12.98 23.61 10.63 0.0377 0.0004 0.0004 5.97 0.0635 0.0020 171 Table AIV-2: Percent Solids Data for Sequential Extractions Sample Tare Total Wet Total Dry % Solids Average RLM-5 2.56 20.75 17.57 82.5 82.4 RLM-5 1.61 23.83 19.88 82.2 RLM-2a 2.12 27.9 22.45 78.9 79.1 RLM-2b 1.68 27.55 22.19 79.3 RLM-6a 1.83 21.21 17.47 80.7 80.7 RLM-6b 3.82 23.3 19.55 80.7 Secondary Pond a 1.44 18.69 10.35 51.7 51.4 Secondary Pond b 2.07 25.15 13.86 51.1 Table AIV-3: Summary Data for Sequential Extractions Location % Total Arsenic Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 6.99 55.75 5.93 0.71 1.84 28.77 Secondary Pond Sediment 5.09 44.19 7.23 0.82 2.11 40.57 RLM-5 4.06 9.90 0.32 0.05 0.07 85.60 RLM-5 4.46 13.20 0.56 0.07 0.07 81.63 RLM-2-1 0.52 19.95 15.72 10.50 0.08 53.23 RLM-2-1 0.57 23.89 19.44 10.41 0.10 45.58 RLM-6-1 1.37 36.31 10.60 9.59 0.43 41.70 RLM-6-1 1.45 40.43 43.17 16.95 0.54 -2.54 Location % Total Iron Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 0.35 8.58 23.91 2.50 5.98 58.67 Secondary Pond Sediment 0.18 6.80 28.47 2.26 7.18 ' 55.11 RLM-5 0.03 7.27 21.54 5.61 4.67 60.87 RLM-5 0.05 8.63 24.53 6.80 5.37 54.63 RLM-2-1 0.13 3.90 24.92 20.38 3.53 47.14 RLM-2-1 0.17 4.31 27.29 19.99 4.01 44.24 RLM-6-1 0.05 6.40 9.66 18.24 4.91 60.75 RLM-6-1 0.02 9.10 27.30 27.29 5.90 30.40 Location mg As/Kg dry sediment Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 187 1489 158 19 49 768 Secondary Pond Sediment 136 1180 193 22 56 1083 RLM-5 82 199 6 1 1 1725 RLM-5 90 266 11 1 2 1645 RLM-2-1 18 686 541 361 3 1831 RLM-2-1 20 822 669 358 3 1568 RLM-6-1 35 926 270 245 11 1063 RLM-6-1 37 1031 1101 432 14 -65 Location mg Fe/Kg dry sediment Step 1 Step 2 Step 3 Step 4 Step 5 Residual Secondary Pond Sediment 253 6118 17046 1782 4267 41835 Secondary Pond Sediment 131 4845 20303 1608 5121 39292 RLM-5 23 5415 16051 4182 3481 45348 RLM-5 35 6426 18276 5063 4000 40700 RLM-2-1 113 3502 22355 18278 3167 42285 RLM-2-1 152 3862 24483 17928 3593 39683 RLM-6-1 43 5492 8285 15647 4210 52122 RLM-6-1 21 7804 23422 23411 5062 26080 Appendix V: Column Experiment Results 173 July 4th Probe Readings (before addition) Location Depth Temp SPC DO PH ORP Water Level m Celcius mg/L units inches 0 19.66 1.441 4.63 7.72 102 Column 1 0.4 19.35 1.439 4.35 7.71 102 60.5 0.8 19.14 1.44 4.67 7.7 101 1.2 18.75 1.443 5.31 7.69 101 0 20.18 1.447 4.7 7.73 103 Column 2 0.4 19.64 1.444 4.56 7.73 101 65.0 0.8 19.18 1.443 4.49 7.73 99 1.4 18.5 1.445 4.66 7.73 92 0 21.33 1.444 5.12 7.73 100 Column 3 0.3 20.65 1.441 4.8 7.73 101 70.5 0.8 19.58 1.441 4.67 7.72 98 1.5 18.6 1.446 4.76 7.72 88 0 20 1.44 5.63 7.73 100 Column 4 0.4 19.51 1.44 5.52 7.73 99 63.0 0.8 19.19 1.439 5.5 7.73 97 1.4 18.62 1.438 6.01 7.72 92 >be Readings Location Depth T e m p SPC DO pH ORP Water Level m Celcius mg/L units inches 0 20.29 1.47 4.35 7.68 117 Column 1 0.4 19.64 1.47 4.17 7.65 99 59.0 0.8 19.07 1.47 4.45 7.64 88 1.3 18.63 1.47 4.78 7.64 82 0 22.96 1.359 3.93 5.35 186 Column 2 0.4 20.43 1.368 3.99 5.56 148 69.0 0.8 19.31 1.371 4.06 5.57 143 1.5 18.73 1.375 4.39 5.61 143 0 23.54 2.01 3.8 5.55 215 Column 3 0.4 22.43 2.02 3.8 5.63 197 75.5 •0.9 19.97 2.03 3.68 5.69 151 1.7 18.6 1.94 2.02 5.77 148 0 22.42 2.12 2.37 6 154 Column 4 0.4 20.74 2.12 2.3 5.99 69.0 0.8 19.45 2.13 2.61 5.98 108 1.5 18.82 2.13 2.54 6.02 110 174 July 8th Probe Readings Location Depth Temp SPC DO pH ORP Water Level m Celcius mg/L units inches 0 23.39 1.455 3.8 7.81 134 Column 1 0.3 23.09 1.454 4.26 7.8 132 58.5 0.8 22.74 1.455 3.88 7.79 129 1.3 21.59 1.461 3.39 7.71 128 0 25.74 1.375 2.56 5.46 216 Column 2 0.4 23.85 1.37 2.55 5.54 193 68.5 0.9 22.68 1.374 2.59 5.56 181 1.5 21.75 1.378 2.6 5.58 182 0 27.68 2.03 2.94 5.61 64 Column 3 0.4 24.33 2.03 2.82 5.69 235 74.5 1 22.99 2.03 2.95 5.71 244 1.6 21.67 2.03 2.91 5.76 231 0 26.02 2.14 1.92 6.11 -50 Column 4 0.3 23.54 2.13 1.62 6.06 122 68.5 0.8 22.88 2.13 1.76 6.06 109 1.4 21.91 2.14 1.9 6.12 97 July 10th Probe Readings Location Depth Temp SPC DO PH ORP Water Level m Celcius mg/L units inches 0 23.06 1.449 5.56 7.89 136 Column 1 0.4 22.52 1.448 5.32 7.87 139 58.5 0.9 22.08 1.451 4.97 7.84 144 1.3 21.93 1.453 5.48 7.79 164 0 24.85 1.384 3 5.38 220 Column 2 0.4 23.36 1.379 2.66 5.42 188 67.5 0.8 22.52 1.379 2.69 5.47 198 1.4 22.02 1.378 1.56 5.58 193 0 26.43 2.05 5.06 5.6 225 Column 3 0.4 23.63 2.05 5.09 5.59 178 73.0 1 22.47 2.06 4.91 5.59 175 1.6 22.11 2.06 2.52 5.64 170 0 24.48 2.14 1.99 5.99 137 Column 4 0.4 23.71 2.14 1.9 5.97 125 68.0 0.8 22.74 2.14 1.69 5.96 117 1.4 22.13 2.14 1.71 5.96 123 0 24.09 1.8 6.5 7.81 135 Pond 0.3 23.82 1.8 6.75 7.82 100 63.0 0.8 22.24 1.79 6.4 7.81 86 1.3 21.85 1.81 6.03 7.79 86 175 July 11th Sampling Location Water level Volume Removed ml Before inches After inches Column 1 58.5 56.0 2060 Column 2 67.5 64.0 2883 Column 3 73.0 69.5 2883 Column 4 68.0 64.0 3295 Water level Volume Evaporated ml Initial inches July 10th inches 60.5 58.5 1648 70.0 67.5 2060 76.5 73.0 2883 70.0 68.0 1648 July 18th Probe Readings Location Depth Temp SPC DO pH ORP Water Level m Celcius mg/L units inches 0 25.13 1.324 7.04 8.27 134 Column 1 0.4 25.05 1.323 6.73 8.27 134 55.0 0.8 24.97 1.322 7 8.27 133 1.2 24.67 1.324 6.84 8.24 133 0 22.03 1.286 4.7 5.71 184 Column 2 0.4 22.05 1.285 4.54 5.71 183 62.0 0.8 22.03 1.285 4.57 5.71 185 1.2 21.96 1.286 4.84 5.73 183 0 21.95 1.92 6.5 5.96 192 Column 3 0.4 21.92 1.92 6.1 5.96 194 65.5 0.9 21.88 1.92 6.31 5.97 197 1.4 21.86 1.92 6.42 5.97 189 0 25.1 1.96 3.49 6.21 -37 Column 4 0.4 25.01 1.96 3.68 6.21 -50 63.0 0.8 24.86 1.96 3.47 6.22 -59 1-2 24.51 1.97 3.79 6.3 -85 0 21.97 2.08 4.88 7.61 145 Pond 0.4 21.96 2.08 4.93 7.62 147 62.0 0.8 21.9 2.08 4.99 7.61 152 1.2 21.82 2.08 5.1 7.6 159 176 July 24th Probe Readings Location Depth Temp SPC DO pH ORP Water Level m Celcius mg/L units inches 0 21.62 1.325 5.54 7.98 79 Column 1 0.4 21.55 1.325 5.55 7.97 73 55.0 0.7 21.45 1.325 5.34 7.95 77 1.1 21.36 1.325 5 7.93 83 0 21.83 1.298 4.26 6.07 100 Column 2 0.4 21.68 1.297 5.00 6.07 105 61.0 0.8 21.61 1.297 4.17 6.15 104 1.1 21.46 1.296 4.76 6.33 100 0 21.77 1.95 5.85 6.39 118 Column 3 0.5 21.5 1.95 6.61 6.37 124 64.5 0.9 21.38 1.94 5.91 6.37 124 1.3 21.29 1.95 6.6 6.44 135 0 21.67 1.98 5.25 6.22 101 Column 4 0.4 21.54 1.98 4.95 6.6 105 62.0 0.8 21.43 1.97 4.48 6.61 105 1.2 21.31 1.98 5.06 6.65 107 0 22.02 2.36 6.24 7.76 91 Pond 0.4 21.98 2.36 6.26 7.76 90 62.0 0.9 21.9 2.36 6.15 7.76 90 1.3 21.84 2.36 6.04 7.75 90 July 26th Sampling Location Water level Volume Removed ml Before inches After inches Column 1 55.0 52.0 2472 Column 2 61.0 57.5 2883 Column 3 64.0 61.0 2472 Column 4 62.0 58.0 3295 Water level Volume Evaporated ml July 11th inches July 26th inches 56.0 55.0 824 64.0 61.0 2472 69.5 64.0 4531 64.0 62.0 1648 177 August 1st Probe Readings Location Depth Temp SPC DO pH ORP Water Level m Celcius mg/L units inches 0 18.11 1.332 4.52 7.62 118 Column 1 0.5 17.99 1.331 3.73 7.61 120 52.5 0.8 17.86 1.331 3.77 7.6 122 1.2 17.73 1.331 4.51 7.55 132 0 17.77 1.317 7.47 6.48 153 Column 2 0.4 17.68 1.316 6.75 6.53 137 57.0 0.8 17.63 1.315 6.9 6.69 136 1.2 17.54 1.313 7.04 6.96 134 0 17.87 1.96 8.16 6.95 170 Column 3 0.4 17.79 1.96 7.45 6.98 168 60.3 0.8 17.71 1.96 7.12 7.04 160 1.2 17.59 1.96 7.9 7.22 154 0 17.99 1.98 6.4 7 152 Column 4 0.4 17.87 1.99 6.43 6.99 170 57.5 0.8 17.83 1.98 6.42 6.98 171 1.1 17.68 1.99 6.35 6.93 169 0 18.12 2.4 8.21 7.79 162 Pond 0.4 18.09 2.4 8.21 7.8 162 0.8 18.07 2.4 8.29 7.79 163 1.1 18.06 2.4 8.3 7.74 160 Probe Readings Location Depth Temp SPC DO PH ORP Water Level m Celcius mg/L units inches 0 21 1.337 6.36 7.54 117 Column 1 0.4 20.46 1.336 5.41 7.51 119 52.0 0.7 20.13 1.338 5.61 7.51 122 1.1 19.84 1.336 5.75 7.53 124 0 21.1 1.323 6.85 6.4 123 Column 2 0.4 20.87 1.32 5.65 6.35 123 57.0 0.7 20.58 1.32 7.13 6.23 122 1.2 20.1 1.316 4.54 5.82 136 0 21.21 1.97 9.51 7.18 97 Column 3 0.4 21.02 1.97 8.61 7.15 98 60.0 0.7 20.64 1.97 8.61 7.15 98 1.2 20.18 1.97 8.55 7.05 110 0 21.23 1.99 5.87 7.18 97 Column 4 0.4 20.8 1.99 6.65 7.14 114 57.0 0.7 20.32 1.99 6.99 7.09 120 1.2 20.13 1.99 5.87 7.17 117 0 21.56 2.49 8.24 7.72 101 Pond 0.4 21.57 2.5 7.93 7.71 99 0.7 21.25 2.49 7.53 7.71 109 1.2 20.16 2.49 7.2 7.68 132 178 August 9th Sampling (just metals) Water level Volume Water level Volume Location Before After Removed July 26th Aug 9th Evaporated inches inches ml inches inches ml Column 1 52.0 51.5 412 52.0 52.0 0 Column 2 57.0 56.5 412 57.5 57.0 412 Column 3 60.0 59.5 412 61.0 60.0 824 Column 4 57.0 56.5 412 58.0 57.0 824 Probe Readings Location Depth Temp SPC DO pH ORP Water Level m Celcius mg/L units inches 0 19.87 1.27 2.82 7.35 113 Column 1 0.4 19.59 1.27 2.7 7.34 115 54.0 0.8 19.42 1.271 2.54 7.36 105 1.1 19.41 1.266 3.35 7.4 104 0 19.77 1.257 4.02 5.37 132 Column 2 0.3 19.56 1.256 3.73 5.38 128 57.5 0.7 19.43 1.256 3.94 5.46 123 1.1 19.2 1.256 4.22 5.68 111 0 20.1 1.88 4.64 7.11 117 Column 3 0.4 19.74 1.9 4.75 7.03 124 60.3 0.8 19.58 1.9 4.81 6.94 123 1.1 19.35 1.9 5.14 6.93 121 0 20.07 1.88 3.83 7.09 107 Column 4 0.4 19.83 1.92 3.96 6.98 112 59.0 0.8 19.61 1.93 3.76 6.95 113 1.1 19.42 1.94 3.95 7.01 116 0 20.2 2.21 5.26 7.49 102 Pond 0.4 19.96 2.2 4.89 7.48 104 0.8 19.68 2.21 4.54 7.47 107 1.2 19.55 2.21 5.02 7.45 111 179 August 22nd Sampling (just metals) Water level Volume Water level Volume Location Before After Removed Aug 9th Aug 22nd Evaporated inches inches ml inches inches ml Column 1 53.5 51.5 1648 51.5 53.5 -1648 Column 2 57.0 54.5 2060 56.5 57.0 -412 Column 3 59.5 57.0 2060 59.5 59.5 0 Column 4 57.5 56.0 1236 56.5 57.5 -824 Probe Readings Location Depth Temp SPC DO PH ORP Water Level m Celcius mg/L units inches 0 17.93 1.287 3.16 7.51 143 Column 1 0.4 17.89 1.286 3.15 7.46 149 51.8 0.8 17.73 1.287 3.66 7.39 164 1.2 17.63 1.286 3.61 7.21 179 0 17.94 1.276 3.29 4.91 194 Column 2 0.4 17.92 1.276 3.27 4.96 200 54.5 0.8 17.84 1.276 3.27 4.96 200 1.1 17.75 1.275 3.84 5.23 190 0 17.98 1.92 6.01 7.06 177 Column 3 0.4 17.98 1.92 5.29 7.02 175 57.0 0.8 17.94 1.93 5.81 6.95 168 1.1 17.76 1.93 6.35 6.91 166 0 17.96 1.94 4.61 7.2 131 Column 4 0.4 17.89 1.94 3.94 7.19 130 55.8 0.8 17.84 1.94 3.86 7.17 128 1 17.75 1.94 4.93 7.18 132 0 18.11 2.3 6.72 7.73 124 Pond 0.4 18.11 2.31 6.42 7.73 120 0.9 18.1 2.31 6.32 7.73 120 1.2 18.1 2.31 6.52 7.73 129 180 Sept. 17-18th Sampling Location Water level Volume Removed ml Before inches After inches Column 1 52.8 50.5 1854 Column 2 55.0 52.0 2472 Column 3 58.0 Column 4 57.0 54.0 2472 Water level Volume Evaporated ml Aug 22nd inches Sep 17th inches 51.5 52.8 -1071 54.5 55.0 -412 57.0 58.0 -824 56.0 57.0 -824 Column 1 (Control) I 18-Sep I C1-Top | 1.05 | 3.97 | in CM 7.79 | <0.04 | m CO <0.0001 | 0.025 | 0.14 | 0.0286 | <0.0002 | 0.026 | 0.06 | <0.0007 | 0.0211 | 0.031 0.01 | <0.006 | <0.006 | 0.805 <0.006 | <0.05 | <0.002 | 0.012 | 23.7 | 21.9 | 69.7 | 0.007 | 0.0014 | Column 1 (Control) I 18-Sep I C1 -Bottom 1.08 4.15 O) CM 7.75 <0.04 o <0.0001 0.026 0.14 0.0304 <0.0002 0.025 0.058 <0.0007 0.0188 0.042 0.012 <0.006 <0.006 CO d <0.006 <0.05 0.007 0.01 co 23.4 21.8 69.9 0.018 0.0063 Column 1 (Control) I 22-Aug C1-Top 0.98 11.4 CM 6.33 0.04 CM <0.0001 0.039 0.14 0.0288 <0.0002 0.009 0.059 <0.0007 0.024 0.025 0.012 <0.006 <0.006 0.765 <0.006 <0.05 <0.002 <0.006 m co 23.5 21.5 70.3 0.022 0.0035 Column 1 (Control) I 22-Aug C1-Bottom 0.99 5.44 co CM o 6.27 <0.04 CM t^" <0.0001 0.054 0.14 0.028 <0.0002 0.008 0.059 <0.0007 0.0192 0.024 0.011 <0.006 <0.006 0.77 <0.006 <0.05 <0.002 0.007 o CO CM 21.7 71.3 0.032 0.0169 Column 1 (Control) I 9-Aug C1-Top -<0.0001 0.065 0.16 0.0308 <0.0002 0.009 0.063 <0.0007 0.0283 0.027 0.015 <0.006 <0.006 0.826 0.011 <0.05 <0.002 0.009 24.6 CO CM in 0.027 0.0078 Column 1 (Control) I 9-Aug C1-Bottom 1.14 <0.0001 0.053 0.16 0.0312 0.0002 0.009 0.064 <0.0007 0.025 0.026 0.011 <0.006 <0.006 0.822 <0.006 <0.05 <0.002 <0.006 24.6 CO CM 74.5 0.042 0.0187 Column 1 (Control) I 26-Jul C1-Top 1.09 8.89 CM CO 4.96 <0.04 CO m <0.0001 0.083 0.14 0.0307 <0.0002 0.017 0.069 <0.0007 0.0317 0.025 0.013 <0.006 <0.006 0.817 <0.006 <0.05 <0.002 0.009 24.5 22.2 74.3 0.012 0.0038 Column 1 (Control) I 26-Jul fc o o m i n 1.07 8.53 co <0.04 m <0.0001 0.072 0.14 0.0319 <0.0002 0.016 0.065 <0.0007 0.0288 0.036 0.013 <0.006 <0.006 0.821 <0.006 <0.05 <0.002 <0.006 co 24.8 22.6 76.1 0.015 0.0111 Column 1 (Control) I 4-Jul (Before Addition) C1-Top 1.19 CM CO LO iri in <0.0001 0.726 0.14 0.0319 <0.0002 0.014 0.07 0.0026 0.0544 0.031 0.021 <0.006 <0.006 0.794 0.011 <0.05 <0.002 0.045 CO co 23.8 CO CM 73.3 0.726 0.124 Column 1 (Control) I 4-Jul (Before Addition) C1 -Bottom -<0.0001 0.482 0.14 0.0293 <0.0002 0.013 0.067 0.0014 0.0351 0.034 0.016 <0.006 <0.006 0.786 <0.006 <0.05 <0.002 0.019 r-co 23.9 22.5 73.1 0.44 0.107 Units mg/L mg/L mg/L mg/L | mg/L mg/L I mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Parameter m < Ammonia O DOC Nitrate Phosphate Sulphate O) < < CO CO m CD m TD O Co o o Mo iz QL c CO co i - H > c N Ca Mg Na CD LL Mn C2-Top | 1.21 | <0.0001 | 0.045 | 0.15 | 0.0523 | 0.0005 | 0.01 | 0.066 | 0.0009 | 0.0053 | 0.017 | 0.005 | <0.006 | <0.006 1 co CD 0.006 | <0.05 | <0.002 | 0.02 | co 25.4 I 26.4 | 77.7 I 0.037 | 0.126 | 9-Aug C2-Middle 0.73 <0.0001 0.078 CD 0.0333 0.0003 0.007 0.05 0.0007 0.0028 0.0007 0.003 <0.006 <0.006 0.497 <0.006 <0.05 <0.002 0.011 O) T — 25.8 23.5 76.3 cq 0.118 C2-Bottom 1.21 <0.0001 0.049 0.15 0.0582 <0.0002 0.009 0.064 <0.0007 0.0052 0.01 <0.002 <0.006 <0.006 0.599 <0.006 <0.05 <0.002 0.013 CO 25.2 26.1 78.3 0.046 0.125 >lumn 2 (Peat) C2-Top 1.18 14.5 CO co oo 5.49 <0.04 o co <0.0001 0.044 0.14 0.0546 <0.0002 0.017 0.072 <0.0007 0.0094 0.015 0.002 <0.006 <0.006 0.608 <0.006 <0.05 <0.002 0.033 LO •3-25.5 25.6 77.1 0.061 0.134 >lumn 2 (Peat) 26-Jul C2-Middle 1.03 14.8 00 CM co CM 3.16 0.09 LO CO <0.0001 0.057 0.12 0.0483 <0.0002 0.014 0.059 <0.0007 0.0027 0.019 <0.002 0.027 <0.006 0.546 <0.006 <0.05 <0.002 0.028 co 24.8 24.4 75.4 0.705 0.127 o C2-Bottom 1.36 O) <M CM CM 5.15 <0.04 oo r -<0.0001 0.045 0.16 0.473 <0.0002 0.02 0.069 <0.0007 0.0067 0.018 0.003 <0.006 <0.006 0.617 <0.006 <0.05 <0.002 0.017 <y> •o-26.6 26.7 81.3 0.274 0.149 ion) C2-Top 1.22 CM CO 5.43 <0.0001 0.448 0.14 0.0313 <0.0002 0.014 0.068 0.0017 0.0469 0.031 0.021 <0.006 <0.006 0.801 0.006 <0.05 <0.002 0.037 •*t r~~ 24.3 23.1 LO r-- 0.43 0.111 Before Addit C2-Middle 4-Jul ( C2-Bottom 1.23 <0.0001 0.33 0.14 0.0295 <0.0002 0.013 0.067 0.0008 0.0392 0.032 0.016 <0.006 <0.006 0.785 <0.006 <0.05 <0.002 0.021 CM c- 24.1 22.6 73.2 0.294 0.104 CO _ i _ i _ J _ l _ i _ l _ l _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ j _ i _ i _ i _ j _ i _ i _ j _ i Uni E mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ Parameter CO < Ammonia o DOC Nitrate Phosphate Sulphate < < CO Ba Be Cd Co o Cu Mo Iz Pb Sn CO i - i - > c N | Ca I Mg I Na 0 LL Mn C2-Top | 1.08 | 10.3 | CM CM o CM 6.59 | 0.26 | CO <0.001 | 0.076 | 0.13 | 0.0535 | <0.0002 | 0.025 | 0.058 | <0.0007 | 0.0046 | 0.017 | 0.004 | <0.006 | <0.006 | 0.557 | <0.008 | <0.05 | <0.001 | 0.034 | co 23.4 | 23.6 | 70.4 | 0.039 | 0.132 | 17-Oct C2-Middle 1.71 11.2 CM CD 5.82 <0.04 <0.001 0.051 0.12 0.0453 <0.0002 0.041 0.055 <0.0007 0.0033 0.006 0.004 0.151 <0.006 0.541 <0.008 <0.05 <0.001 0.032 CO 24.1 24.8 72.5 0.147 0.138 t o o ffl CN o 4.68 13.4 LO CM CM 4.14 <0.04 <0.001 0.035 d 0.0846 <0.0002 0.114 0.047 <0.0007 0.0071 0.012 0.003 0.025 <0.006 0.507 <0.008 <0.05 0.022 0.031 CM LO CN 24.8 72.7 0.454 0.14 C2-Top 1.14 10.5 CO co 7.01 <0.04 o r-<0.0001 0.069 0.14 0.0631 <0.0002 0.026 0.067 0.001 0.0044 0.012 0.004 0.022 <0.006 0.595 <0.006 <0.05 <0.002 0.033 CD CO 24.4 24.7 74.6 0.042 0.136 Column 2 18-Sep C2-Middle 1.26 13.9 CO CM 2.92 d <M <0.0001 0.042 d 0.0367 <0.0002 0.03 0.048 0.0007 0.0018 0.008 0.003 <0.006 <0.006 0.516 <0.006 <0.05 <0.002 0.027 CN 25.4 23.9 73.5 1.18 0.129 Column 2 C2-Bottorr 1.72 12.3 o -a- CM 4.49 <0.04 5-<0.0001 0.055 0.12 0.049 <0.0002 0.041 0.054 0.0008 0.0023 0.009 0.013 <0.006 <0.006 0.548 <0.006 <0.05 <0.002 0.031 CD CM LO CN 23.9 73.9 0.664 0.132 C2-Top 1.17 5.34 CO CO oo 4.85 <0.04 CO LO <0.0001 0.042 0.13 0.0603 <0.0002 0.01 0.064 <0.0007 0.0057 0.009 0.003 <0.006 <0.006 0.583 <0.006 <0.05 <0.002 0.027 CD CO 24.2 24.8 75.2 0.041 0.129 22-Aug C2-Middle 0.68 14.1 CM CO co CM 1.14 0.28 co CD CO <0.0001 0.059 0.09 0.0471 <0.0002 0.006 0.047 0.0009 0.0013 <0.006 <0.002 0.014 <0.006 0.482 <0.006 <0.05 <0.002 0.008 CD 24.9 23.1 75.6 1.26 0.116 C2-Bottom 0.96 CM CM CO CM 3.18 0.05 h-co <0.0001 0.042 0.11 0.0363 <0.0002 0.007 0.055 0.0009 0.0019 0.008 0.002 0.009 <0.006 0.531 <0.006 <0.05 <0.002 0.008 CO 24.4 24.6 74.7 0.769 0.129 Units mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Parameter As Ammonia O DOC Nitrate Phosphate Sulphate O) < < CO I Ba I Be Cd I Co o I Cu Mo z Pb Sn Sr i- > Zn Ca Mg Na Fe Mn C3-Top 1.27 | <0.0001 | 0.043 | 0.27 | 0.0712 <0.0002 | 0.01 | 0.066 | <0.0007 | 0.006 | 0.017 | 0.004 | <0.006 | <0.006 | 1.26 | <0.006 | <0.05 | <0.002 | 0.008 | CD co 26.8 | 31.2 | 80.6 | 0.03 | 0.141 | 9-Aug C3-Middle 0.49 <0.0001 0.036 0.14 0.0314 0.0003 0.004 0.043 0.0008 0.0018 0.026 0.003 <0.006 <0.006 3.46 <0.006 <0.05 <0.002 <0.006 CO 26.9 43.3 80.6 0.845 0.151 C3-Bottom 3.15 <0.0001 0.06 0.07 0.227 0.0003 0.024 0.028 <0.0007 <0.0005 0.024 0.011 <0.006 <0.006 3.63 0.008 <0.05 <0.002 <0.006 m CD 28.8 C\J 77.8 10.3 0.32 ypsum) C3-Top 1.21 13.7 CD CO 00 5.61 <0.04 1060 <0.0001 0.028 0.25 0.073 <0.0002 0.018 0.069 <0.0007 0.0083 0.026 <0.002 <0.006 <0.006 1.16 <0.006 <0.05 <0.002 0.012 oo co 26.3 29.9 79.5 0.035 0.149 3 (Peat + G< 26-Jul C3-Middle 0.78 15.9 00 C\J 2.42 <0.04 1120 <0.0001 0.028 0.15 0.0669 <0.0002 0.012 0.054 <0.0007 0.0014 0.022 <0.002 <0.006 <0.006 CD c\i <0.006 <0.05 <0.002 0.017 co CO 25.7 o ^t- 77.9 0.721 0.153 Column C3-Bottom 2.65 16.6 o co CD co <0.03 <0.04 1540 <0.0001 0.032 o" 0.115 <0.0002 0.04 0.041 <0.0007 0.0006 0.023 0.011 <0.006 <0.006 2.47 <0.006 <0.05 <0.002 <0.006 co CO CD 27.9 88.2 75.7 8.49 0.345 ion) C3-Top 1.18 CO co iri CD <0.0001 0.261 0.13 0.0298 <0.0002 0.013 0.068 0.0014 0.046 0.037 0.017 <0,006 <0.006 0.797 <0.006 <0.05 <0.002 0.017 co f -24.4 22.9 74.7 0.244 0.101 Before Add it C3-Middle 4-Jul ( C3-Bottom 1.16 <0.0001 0.595 0.14 0.0307 <0.0002 0.014 0.069 0.002 0.0374 0.028 0.021 <0.006 <0.006 0.787 0.008 <0.05 <0.002 0.011 co r-24.3 22.9 74.3 0.584 0.109 Units mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L | mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Parameter CO < Ammonia O DOC Nitrate Phosphate Sulphate < < m CO CQ CD CQ T3 O o O O o Mo iz Q. c C/D CO y— H > c N CO O Mg CO CD LL Mn C3-Top | 10.9 | (M O) 4.77 | <0.04 | 1010 | <0.001 | 0.021 | 0.22 | 0.0655 | <0.0002 | 0.025 | 0.056 | <0.0007 | 0.0036 | 0.012 | 0.004 | <0.006 | <0.006 | 1.77 | <0.008 | <0.05 | 0.002 | 0.018 | LO CM CO 25.1 | 29.6 | CO IV. 0.029 | 0.0518 | 17-Oct C3-Middle 10.9 O) CM 00 4.97 <0.04 1050 <0.001 0.019 0.21 0.0657 <0.0002 0.026 0.053 <0.0007 0.0047 0.016 0.007 0.049 <0.006 1.79 <0.008 <0.05 0.004 0.023 CO CM co 24.9 29.6 73.7 0.029 0.0518 C3-Bottom 5.75 16.2 co co CM <0.3 <0.4 1710 0.001 0.012 <0.05 0.0728 <0.0002 0.138 0.016 <0.0007 <0.0005 0.01 0.009 <0.006 <0.006 3.57 <0.008 <0.05 <0.001 0.018 IO co LO 27.1 83.5 CM 18.8 0.684 C3-Top 1.13 11.7 o LO 4.66 <0.04 1050 <0.0001 0.017 0.23 . 0.0715 <0.0002 0.027 0.06 0.0007 0.0035 0.018 0.005 <0.006 <0.006 1.67 <0.006 <0.05 <0.002 0.024 O) CM CO 25.9 o CO 76.2 0.011 0.023 Column 3 18-Sep C3-Middle 1.17 12.5 00 CO co 4.62 <0.04 1050 <0.0001 0.015 0.23 0.0745 <0.0002 0.025 0.059 <0.0007 0.0032 0.014 0.004 0.092 <0.006 . 1.68 <0.006 <0.05 0.005 0.016 co 00 CO 25.9 30.2 75.6 0.03 0.0246 Column 3 C3-Bottorr 3.04 18.7 o o CO 0.03 0.08 1520 0.0001 0.017 0.06 0.0426 <0.0002 0.075 0.023 <0.0007 <0.0005 0.007 0.006 <0.006 <0.006 3.12 <0.006 <0.05 <0.002 0.015 CM O) LO 28.2 61.4 74.4 11.4 0.364 C3-Top 1.15 13.3 co CO 00 4.97 0.07 1050 <0.0001 0.03 0.24 0.0668 <0.0002 0.01 0.062 <0.0007 0.0055 0.016 <0.002 <0.006 <0.006 1.32 <0.006 <0.05 <0.002 0.008 CO CO 26.3 29.9 78.8 0.043 0.102 22-Aug C3-Middle 0.96 13.3 a> CO 4.83 0.05 1060 <0.0001 0.024 0.23 0.0816 <0.0002 0.007 0.059 <0.0007 0.0032 0.015 <0.002 0.336 <0.006 1.87 <0.006 <0.05 <0.002 <0.006 O) CM co 25.9 30.8 77.7 0.053 0.11 C3-Bottom 3.33 00 LO CO o CO <0.03 <0.04 1890 <0.0001 0.033 <0.05 0.248 <0.0002 0.027 0.023 <0.0007 <0.0005 0.028 0.006 <0.006 <0.006 3.25 <0.006 <0.05 <0.002 <0.006 co CO 28.8 68.5 78.7 T— 0.374 Units mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Parameter As Ammonia o DOC Nitrate Phosphate Sulphate < < Ba Be Cd Co O Cu Mo z -O CL Sn CO i - H > c N Ca Mg Na cu LL Mn C4-Top | <0.02 | <0.0001 | 0.242 | 0.24 | 0.0581 | 0.0003 | <0.002 | 0.063 | 0.0037 | 0.0284 | 0.03 | 0.047 | 0.036 | <0.006 | 1.64 | <0.006 | <0.05 | <0.002 | 0.024 | CO m CO 25.3 | 26.3 | 74.8 | 8.19 | 0.679 | 9-Aug C4-Middle <0.02 <0.0001 0.037 0.29 0.0539 0.0004 <0.002 0.054 <0.0007 0.0068 0.032 0.017 0.017 <0.006 1.73 <0.006 <0.05 <0.002 0.01 CM CD co 25.6 27.2 76.1 0.816 0.636 CO a c C4-Bottom <0.02 <0.0001 0.067 0.69 0.0583 0.0002 <0.002 <0.003 <0.0007 <0.0005 0.065 <0.002 <0.006 <0.006 3.75 <0.006 <0.05 <0.002 <0.006 55 in 31.9 23.4 82.2 0.318 0.435 + Iron Filli C4-Top <0.02 13.3 tn co CO 4.88 0.04 1080 <0.0001 0.013 0.27 0.06 <0.0002 <0.002 0.063 <0.0007 0.0127 0.035 0.019 <0.006 <0.006 in <0.006 <0.05 <0.002 0.027 CM CO 25.6 26.2 77.2 0.489 0.668 t + Gypsum 26-Jul C4-Middle <0.02 13.2 CO CM T— 4.86 <0.04 1090 <0.0001 0.011 0.27 0.0607 <0.0002 <0.002 0.064 <0.0007 0.0068 0.036 0.02 <0.006 <0.006 1.52 <0.006 <0.05 <0.002 0.018 CM i ~ » CO 25.7 26.2 76.6 0.525 0.672 lumn 4 (Peal C4-Bottom <0.02 17.9 OJ m CO 0.06 <0.04 1550 <0.0001 0.01 0.64 0.0593 <0.0002 <0.002 0.006 <0.0007 <0.0005 0.074 0.002. <0.006 <0.006 3.13 <0.006 <0.05 <0.002 <0.006 oo o CD 33.6 26.3 84.6 0.043 0.491 Co ion) C4-Top 1.17 CM CO 5.37 CM •3-<0.0001 0.437 0.14 0.0293 <0.0002 0.013 0.07 0.0019 0.0443 0.025 0.017 <0.006 <0.006 0.786 <0.006 <0.05 <0.002 0.019 CM I"- CM 22.7 74.7 0.441 0.106 Before Addit C4-Middle 4-Jul ( C4-Bottom 1.21 <0.0001 0.096 0.14 0.0319 <0.0002 0.014 0.069 <0.0007 0.0268 0.028 0.016 <0.006 <0.006 0.826 <0.006 <0.05 <0.002 0.012 co r- in CM 23.3 76.1 0.052 0.0977 Units mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/Ll mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L mg/L Parameter CO < Ammonia O DOC Nitrate Phosphate Sulphate O J < < CQ CO CO CD CQ o o O L . o o Mo iz J D CL c CO CO l— > c N Ca Mg CC z <x> LL Mn C4-Top | <0.02 | 12.1 | cn CM 3.46 | <0.04 | 1080 | <0.0001 | 0.017 | 0.25 | 0.0533 | <0.0002 | <0.002 | 0.044 | <0.0007 | 0.0043 | 0.029 | 0.013 | 0.013 | <0.006 | 2.14 | <0.006 | <0.05 | 0.005 | 0.028 | cn LO CO 24.4 | 25.8 | 72.5 | 0.088 | 0.424 | 18-Sep C4-Middle <0.02 12.6 CO CM o 3.29 d 1080 <0.0001 0.014 0.24 0.0528 <0.0002 <0.002 0.046 <0.0007 0.0044 0.025 0.014 <0.006 <0.006 2.16 <0.006 <0.05 <0.002 0.024 LO LO CO 24.3 25.7 0.097 0.425 C E ti o +•» O CD • y <0.02 18.5 LO <0.03 <0.04 1400 <0.0001 0.013 0.47 0.0495 <0.0002 <0.002 0.004 0.0011 <0.0005 0.034 <0.002 <0.006 <0.006 2.86 <0.006 <0.05 <0.002 0.006 CM r-» LO 29.5 21.9 oo r» 0.048 0.603 Colu C4-Top <0.02 12.3 CO CM co 3.84 0.04 1090 <0.0001 0.018 0.25 0.0534 <0.0002 <0.002 0.05 <0.0007 0.0055 0.029 0.014 <0.006 <0.006 1.72 <0.006 <0.05 <0.002 <0.006 CM co 23.8 25.4 72.4 0.126 0.559 22-Aug C4-Middle <0.02 CM o CO CM 4.83 0.05 1060 <0.0001 0.02 0.26 0.0532 <0.0002 <0.002 0.052 <0.0007 0.0055 0.026 0.013 <0.006 <0.006 1.76 <0.006 <0.05 <0.002 0.011 CO 24.5 25.9 71.6 0.16 0.561 C4-Bottom <0.02 16.6 o LO CM <0.03 <0.04 1770 <0.0001 0.045 0.61 0.0528 <0.0002 <0.002 0.004 <0.0007 0.0005 0.06 <0.002 <0.006 <0.006 3.33 <0.006 <0.05 <0.002 <0.006 co co LO CO 23.4 79.3 0.119 0.42 (o _i _i _ l _1 _i _ i _ i _ i _ i _ i _j _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _ i _J i _ i _ i _ i _ i Uni mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ mg/ Parameter As Ammonia o DOC Nitrate Phosphate Sulphate CD < < CQ Ba Be Cd Co o Cu Mo iz Pb Sn CO i- h- > Zn i Ca I Mg | Na I Fe • Mn Appendix VI: Designed Laboratory Experiment Results O N 00 O E L O £M CM . Q CD T3 CD > .. O £ E 3 CD S = M (1) o I - > c o E E o o cu *-» 03 H Cfl o J O oi Q E (0 c L O <M C o CD T 3 CO E 0) c o 5 CD C CD o L O L O IE5 \°\ 3 o as O E lo 3 IS a •o <u > .. o 2 E 3 <B Jr CO 0) O I- > c o E E o o f t O S) P. E (0 c 2 f >- fc — 3 CO 0 ) 5 C M ao 3 cri ao C O CO CO ISI m co co i n cn C O CM CO O ) CO SI O) CO CO CO CM CO CO CO CO ISI Comments | no black, all grey I no black, all grey I no dark spots, grey brown throughout I Idarker shading on bottom, likely just FeO I |no dark spots, grey brown throughout I orange water, lots of black/shiny | darker shading on bottom, likely just FeO I |no dark spots, grey brown throughout | black/shiny I no black, all grey I black/shiny on bottom (some), and some orange spots on top I darker shading on bottom, likely just FeO I darker shading on bottom, likely just FeO I no dark spots, grey brown throughout I black/shiny on bottom I darker shading on bottom, likely just FeO | pure brown, no black | tonnes of black/shiny, layering and bubbles I no black, still colloidal, bubbles in solids | darker shading on bottom, likely just FeO | darker shading on bottom, likely just FeO, layering and bubbles | darker shading on bottom, likely just FeO I darker shading on bottom, likely just FeO I really murky water, no black, layering and bubbles I darker shading on bottom, likely just FeO I darker shading on bottom, likely just FeO I orange precip. On top of solids, layering with bubbles, then black/shiny | no black, layering and bubbles I pure brown, no black I green layer, brown layer with bubbles, black layer, grey layer | darker shading on bottom, likely just FeO I layering and bubbles, no black | [Sulphate] mg/L [DOC] mg/L o CO CO CO CO oS CO CO I 10.1 I OO CO CD CO I 12.5 | in CO r v CO m CD CO CO 0 0 OO 0 0 i — co CO in cri | 769.3 | O O cri cn CO I 11.3 I | 20.9 | CM CO i v I 10.8 | I 15.2 | m cri I 13.4 | co Iv' I 15.2 | [Fe] mg/L [As] mg/L | 2.03 | I 4.81 | I 7.72 | I 1.51 I | 0.29 | | 0.04 | I 2.64 | I 9.68 | | 0.05 | I 5.24 | I 1-50 | I 1-57 | I 1-32 | I 1-74 | | 6.20 ] | 0.50 | 1 0.12 | | 0.09 | | 4.64 | I 0.14 | I 0.12 | I 0.16 | I 0.19 | | 0.23 | I 0.27 | | 0.19 | | 0.09 | | 0.24 | I 0.27 | | 0.22 | I 0.11 | | 0.24 | Final Wt. | 358.52 | | 373.25 | | 357.52 | | 370.41 | | 354.53 | | 372.41 | | 354.94 | | 364.88 | I 346.87 | I 354.58 | | 354.84 | | 365.59 | | 363.58 | | 369.01 | | 354.35 | | 361.04 | Int. Wt. [ 366.96 | | 383.08 | | 368.03 | | 380.43 | | 364.54 | [ 382.76 | | 364.11 | | 375.07 | I 357.53 | I 365.01 | | 365.28 | | 373.9 | [ 373.06 | | 379.76 | | 364.31 | | 371.84 | X a. 1 7.51 | | 7.55 | I 7.71 | I 7.88 | i v I 8.12 | | 8.02 | I 8.11 | I 7.18 | I 7.93 | cn r v I 7.63 | cd I 8.16 | cn t v | 7.38 | I 7.31 | I 6.43 | | 6.59 | I 7.27 | | 6.48 | i v I 7.04 | | 7.06 | I 7.69 | I 7.25 | | 6.91 | I 6.72 | I 7.57 | | 6.84 | t v | 6.62 | Trial Number CM CO in oo cn o CM CO LO CO rv 00 cn o CM CM CM CM CO CM *r CM LO CM CO CM |v CM 00 CM cn CM o CO CO CM CO C N O N O E 3 CO Q •a o > .. o £ E 3 01 ra « 9 M CD O H > CU E E o o co _i It CO O oi a E •—• E x a :S 1 a J co x : co x x Z J X I T3 I Z CO _ CO z>~ E > c jz o _CC X I c CB CO _a> X X Z J X I g CD >> CO I E" o o X c o c 1 II a) CD CD > _cfl o •g < X < CO L _cu i X : -o •; X ( X r o JS x o c co" _g> x J D zi X CO _cz a> >. CO co co c i C O co d IS I Appendix VII: Design Ease Results 194 Set 1 - Day 40 Des ign Ease G r a p h s a n d A N O V A Tab le il p 73 Normal Plot of Res idua ls -1.88 -0.76 0.35 2.58 Studentized Residuals CO "3 •3 CO CD CC N C * CO 3.00-1.50-0.00--1.50--3.00-Res idua ls vs . P red ic ted -0.47 0.38 1.23 2.08 2.93 Predicted 1.50-Res idua ls vs . Run 195 cc S o.oo -N T5 cu •o 3 CO -1.50-1 I 1 1 I 1 1 I 1 1 1 1 1 I 1 4 7 10 13 16 4.24 H 2.31 -Run Number Out l ier T •5 O -1.56-I 1 1 I 1 1 I 10 13 16 Run Number 196 Pred ic ted vs . Ac tua l -0.47 0.68 1.84 3.00 4.16 Actual Box-Cox Plot for Power T r a n s f o r m s - 3 - 2 - 1 0 1 2 3 Lambda Interact ion G r a p h DESIGN-EXPERT Plot Arsenic : A: DOC : D: Fe(0) • D--1.000 A D+ 1.000 Actual Factors B: S04 = 0.00 C: ATM =0.00 E: Tailings = 0.00 F: S(0) = 0.00 -1.00 -0.50 0.00 0.50 1.00 A: DOC Interact ion G r a p h DESIGN-EXPERT Plot Arsenic A: DOC F: S(0) • F- -1.000 A F+ 1.000 Actual Factors B: S04 = 0.00 C: ATM =0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 -0.51535 H -1.00 -0.50 0.00 0.50 1.00 A: DOC O n e Fac to r Plot Warning! Factor invo lved in an interaction. 3.09231 - \ o c g 2.02461 < 0.956919 --0.110774 --1.00 -0.50 0.00 0.50 1.00 D: Fe(0) O n e Fac to r Plot 4 . 1 6 -3.09778 - \ E: Tailings O n e Fac to r Plot 4 16 J Warning! Facto r i nvo lved in an interaction. 3.09459 - \ o 'c g 2.02918-< 0.963763 -4 -0.101649 --1.00 -0.50 0.00 0.50 1.00 F: S(0) D E S I G N A r s e n i c E X P E R T Plot A: D O C B: S 0 4 C : A T M D : F e ( 0 ) E: Ta i l i ng s F: S(0) 5 H a If N o r m a l p lo t 0.00 0.26 0.77 1 .02 | E ffe ct| Response: Arsenic ANOVA for Selected Factorial Model Analysis of variance table [Partial sum of squares] Source Sum of DF Mean F Prob >F Squares Square Value Model 15.84 5 3.17 8.69 0.0021 significant D 4.19 1 4.19 11.49 0.0069 E 3.84 1 3.84 10.53 0.0088 F 4.04 1 4.04 11.08 0.0076 AD 1.95 1 1.95 5.35 0.0433 AF 1.83 1 1.83 5.02 0.0490 Residual 3.64 10 0.36 Cor Total 19.482 15 The Model F-value of 8.69 implies the model is significant. There is only a 0.21% chance that a "Model F-Value" this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case D, E, F, AD, AF are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve your model. Std. Dev. 0.6037 R-Squared 0.8129 Mean 0.7370 Adj R-Squared , 0.7194 CV. 81.9113 Pred R-Squared 0.5211 PRESS 9.3296 Adeq Precision 9.2025 The "Pred R-Squared" of 0.5211 is in reasonable agreement with the "Adj R-Squared" of 0.7194. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your ratio of 9.203 indicates an adequate signal. This model can be used to navigate the design space. Factor Coefficient Estimate Intercept 0.737 D-Fe(0) -0.512 E-Tailings 0.490 F-S(0) -0.502 AD -0.349 AF -0.338 DF Standard 95% Cl 95% Cl VIF Error Low High 0.1509 0.4007 1.0733 0.1509 -0.8478 -0.1752 1 0.1509 0.1534 0.8259 1 0.1509 -0.8386 -0.1661 1 0.1509 -0.6853 -0.0127 1 0.1509 -0.6744 -0.0019 1 201 Final Equation in Terms of Coded Factors: Arsenic = 0.737 -0.5115 * D 0.489625 * E -0.502375 *F -0.349 * A * D -0.338125 * A * F Diagnostics Case Statistics Standard Actual Predicted Residual Leverage Student Cook's Outlier Run Order Value Value Residual Distance t Order 1 0.712 1.553 -0.841 0.375 -1.7629 0.3108 -2.0146 11 2 0.36 0.267 0.093 0.375 0.1941 0.0038 0.1845 14 3 0.376 0.246 0.130 0.375 0.2732 0.0075 0.2601 13 4 4.16 2.928 1.232 0.375 2.5822 0.6668 4.244* 15 5 0.098 0.246 -0.148 0.375 -0.3093 0.0096 -0.2949 12 6 2.66 2.928 -0.268 0.375 -0.5608 0.0314 -0.5405 8 7 1.43 1.553 -0.123 0.375 -0.2585 0.0067 -0.2461 6 8 0.192 0.267 -0.075 0.375 -0.1579 0.0025 -0.1500 2 9 0.275 0.249 0.026 0.375 0.0542 0.0003 0.0514 4 10 0 -0.474 0.474 0.375 0.9940 0.0988 0.9933 1 11 0.003 0.900 -0.897 0.375 -1.8792 0.3532 -2.2167 7 12 0.218 0.227 -0.009 0.375 -0.0196 0.0000 -0.0186 16 13 0.832 0.900 -0.068 0.375 -0.1422 0.0020 -0.1351 9 14 0.254 0.227 0.027 0.375 0.0558 0.0003 0.0529 10 15 0.206 0.249 -0.043 0.375 -0.0904 0.0008 -0.0858 3 16 0.016 -0.474 0.490 0.375 1.0275 0.1056 1.0307 5 * Case(s) with |Outlier T| > 3.50 Proceed to Diagnostic Plots (the next icon in progression). Be sure to look at the: 1) Normal probability plot of the studentized residuals to check for normality of residuals. 2) Studentized residuals versus predicted values to check for constant error. 3) Outlier t versus run order to look for outliers, i.e., influential values. 4) Box-Cox plot for power transformations. If all the model statistics and diagnostic plots are OK, finish up with the Model Graphs icon. 202 Set 1 - Day 7 D e s i g n Ease G r a p h s a n d A N O V A Tab le Normal Plot of Res idua ls 2 0-"53 0.09 0.92 1.75 Studentized Residuals "53 S3 rr c CD X J 3 cn 0.00--1.50-^ -3.00-Res idua ls vs . Pred ic ted 0.40 4.33 8.27 12.21 Predicted 2 0 3 Res idua ls vs . Pred ic ted 3.00-0.00--1.50-^ -3.00 -0.40 4.33 8.27 12.21 16.15 Predicted Res idua ls vs . P red ic ted 3.00-oo •53 •3 "8 1.50-0.00- " 1 c cu TD a cn -1.50-4 -3.00-0.40 4.33 8.27 12.21 16.15 Predicted 204 Res idua ls vs . P red ic ted 3 . 0 0 -1 . 5 0 -•3 0 ) CD rr N C CD T3 CO - 3 . 0 0 -0 . 4 0 4 . 3 3 8 . 2 7 1 2 . 2 1 1 6 . 1 5 Predicted Res idua ls vs . P red ic ted 1 . 5 0 H 0 ) "S3 •a rr "8 o ° ° -c cu 73 3 CO - 1 . 5 0 -0 . 4 0 4 . 3 3 8 . 2 7 1 2 . 2 1 1 6 . 1 5 Predicted O n e Fac to r Plot DESIGN-EXPERT Plot Arsenic X = C:ATM Actual Factors A: DOC = 0.00 B: S04 = 0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 0.504 -1.00 -0.50 0.00 0.50 1.00 C: ATM DESIGN-EXPERT Plot Arsenic X = C: ATM Y= E: Tailings • E--1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 B:SO4=0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 15.32 0.401904 In teract ion G r a p h -1.00 -0.50 0.00 C: ATM O n e Factor Plot 15.32 DESIGN-EXPERT Plot Arsenic X= D:Fe(0) Actual Factors A: DOC = 0.00 B: S04 = 0.00 C: ATM = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 -1.00 -0.50 0.00 0.50 1.00 D: Fe(0) O n e Fac to r Plot DESIGN-EXPERT Plot Arsenic X = B: S04 Actual Factors A: DOC = 0.00 C: ATM = 0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 0.504 B: S04 O n e Factor Plot DESIGN-EXPERT Plot Arsenic X = E: Tailings Actual Factors A: DOC = 0.00 B: S04 = 0.00 C: ATM = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 15.32 - \ -1.00 -0.50 0.00 E: Tailings O n e Fac to r Plot DESIGN-EXPERT Plot Arsenic X = F: S(0) Actual Factors A: DOC = 0.00 B: S04 = 0.00 C: ATM = 0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 o 'tz CD S2 < 11.616 7.912 4.208 -i 0.504 -1.00 -0.50 0.00 0.50 1.00 F: S(0) |Ef fect | 209 Response: Arsenic ANOVA for Selected Factorial Model Analysis of variance table [Partial sum of squares] Source Sum of DF Mean F Prob >F Squares Square Value Model 379.37 6 63.23 45.47 < 0.0001 significant B 11.56 1 11.56 8.31 0.0181 C 49.14 1 49.14 35.34 0.0002 D 37.72 1 37.72 27.13 0.0006 E 223.80 1 223.80 160.96 < 0.0001 F 32.98 1 32.98 23.72 0.0009 CE 24.16 1 24.16 17.37 0.0024 Residual 12.51 9 1.39 Cor Total 391.88 15 The Model F-value of 45.47 implies the model is significant. There is only a 0.01% chance that a "Model F-Value" this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case B, C, D, E, F, CE are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve your model. Std. Dev. Mean CV. PRESS 1.1792 5.6088 21.0238 39.5506 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.9681 0.9468 0.8991 20.2013 The "Pred R-Squared" of 0.8991 is in reasonable agreement with the "Adj R-Squared" of 0.9468. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your ratio of 20.201 indicates an adequate signal. This model can be used to navigate the design space. Factor Coefficient Estimate Intercept 5.609 B-S04 -0.850 C-ATM 1.753 D-Fe(0) -1.536 E-Tailings 3.740 F-S(0) -1.436 CE 1.229 DF Standard 95% Cl 95% Cl Error Low High 0.2948 4.9419 6.2756 0.2948 -1.5169 -0.1831 0.2948 1.0856 2.4194 0.2948 -2.2024 -0.8686 0.2948 3.0731 4.4069 0.2948 -2.1026 -0.7689 0.2948 0.5619 1.8956 VIF 210 Final Equation in Terms of Coded Factors: Arsenic = 5.60875 -0.85 *B 1.7525 *C -1.5355 *D 3.74 * E -1.43575 *F 1.22875 * C * E Final Equation in Terms of Actual Factors: Arsenic = 5.60875 -0.85 *S04 1.7525 * ATM -1.5355 *Fe(0) 3.74 * Tailings -1.43575 *S(0) 1.22875 * ATM * Tailings Diagnostics Case Statistics Standard Actual Predicted Residual Leverage Student Cook's Outlier Run Order Value Value Residual Distance t Order 1 10.56 10.189 0.371 0.4375 0.4198 0.0196 0.3997 11 2 2.29 2.295 -0.005 0.4375 -0.0054 0.0000 -0.0051 14 3 0.504 0.595 -0.091 0.4375 -0.1026 0.0012 -0.0968 13 4 8.44 8.489 -0.049 0.4375 -0.0551 0.0003 -0.0520 15 5 2.04 3.342 -1.302 0.4375 -1.4725 0.2409 -1.5934 12 6 15.32 16.151 -0.831 0.4375 -0.9399 0.0982 -0.9332 8 7 14.96 14.451 0.509 0.4375 0.5753 0.0368 0.5526 6 8 3.04 1.642 1.398 0.4375 1.5805 0.2775 1.7531 2 9 1.69 2.095 -0.405 0.4375 -0.4582 0.0233 -0.4372 4 10 5.32 4.246 1.074 0.4375 1.2141 0.1638 1.2518 1 11 1.15 2.546 -1.396 0.4375 -1.5788 0.2770 -1.7505 7 12 0.896 0.395 0.501 0.4375 0.5662 0.0356 0.5436 16 13 11.76 10.209 1.551 0.4375 1.7541 0.3419 2.0385 9 14 2.69 3.143 -0.453 0.4375 -0.5119 0.0291 -0.4898 10 15 1.8 1.443 0.357 0.4375 0.4040 0.0181 0.3844 3 16 7.28 8.509 -1.229 0.4375 -1.3894 0.2145 -1.4780 5 Proceed to Diagnostic Plots (the next icon in progression). Be sure to look at the: 1) Normal probability plot of the studentized residuals to check for normality of residuals. 2) Studentized residuals versus predicted values to check for constant error. 3) Outlier t versus run order to look for outliers, i.e., influential values. 4) Box-Cox plot for power transformations. If all the model statistics and diagnostic plots are OK, finish up with the Model Graphs icon. 2 1 1 Set 1 - Day 14 D e s i g n Ease G r a p h s a n d A N O V A Tab le Normal Plot of Res idua ls 99-95^ i 00 CD mil tn IHM 11111 robe 70-50-| CO 30 4 20-1 z 10-i 5-= Studentized Residuals 3.00-1.50 H tn co •a CO a> rr •g o.oo-"E CD T3 a co -1.50-4 -3.00-0.08 Res idua ls vs . Pred ic ted 2.24 4.40 6.57 8.73 Predicted Res idua ls vs . Run 212 3.00-1.50 H SH co -3 cn CD or "S c CD T3 a CO -1.50--3.00 -- [ — T — I 1 I I | I I | 7 10 13 16 Run Number 3.50-Out l ier T 1.75 H •s o -1.75 — -3.50 -I 1 1 I 1 1 I 1 1 1 1 1 I 1 ' I 1 4 7 10 13 16 Run Number 213 Pred ic ted vs . Ac tua l "7 7.2 Actual Box-Cox Plot for Power T r a n s f o r m s Lambda DESIGN-EXPERT Plot Arsenic C: ATM E: Tailings • E- -1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 B: S04 = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 0.043 H In teract ion G r a p h C: ATM O n e Factor Plot DESIGN-EXPERT Plot Arsenic X=C:ATM Actual Factors A: DOC = 0.00 B: S04 = 0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 0.043 - \ -1.00 -0.50 0.00 0.50 1.00 C: ATM O n e Factor Plot DESIGN-EXPERT Plot Arsenic X=D:Fe(0) Actual Factors A: DOC = 0.00 B: S04 = 0.00 C: ATM = 0.00 E: Tailings = 0.00 F:S(0) = 0.00 9.68 H 7.27075 -o 'c CD < 4.8615-^ 2.45225-0.043 --1.00 -0.50 0.00 0.50 D: Fe(0) O n e Factor Plot DESIGN-EXPERT Plot Arsenic X= E: Tailings Actual Factors A: DOC = 0.00 B: S04 = 0.00 C: ATM = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 7.27075 -o c <u 4.8615-2.45225 - J 0.043 --1.00 -0.50 0.00 0.50 1.00 E: Tailings 216 | E f f e c t | 217 Response: Arsenic ANOVA for Selected Factorial Model Analysis of variance table [Partial sum of squares] Source Sum of DF Mean F Prob >F Squares Square Value Model 112.07 5 22.41 13.71 0.0003 significant C 15.45 1 15.45 9.45 0.0117 D 18.45 1 18.45 11.29 0.0072 E 34.92 1 34.92 21.37 0.0009 F 30.64 1 30.64 18.75 0.0015 CE 12.60 1 12.60 7.71 0.0195 Residual 16.34 10 1.63 CorTotal 128.410 15 The Model F-value of 13.71 implies the model is significant. There is only a 0.03% chance that a "Model F-Value" this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case C, D, E, F, CE are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve your model. Std. Dev. 1.2784 R-Squared 0.8727 Mean 2.9274 Adj R-Squared 0.8091 CV. 43.6702 Pred R-Squared 0.6742 PRESS 41.8395 Adeq Precision 11.0568 The "Pred R-Squared" of 0.6742 is in reasonable agreement with the "Adj R-Squared" of 0.8091. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your ratio of 11.057 indicates an adequate signal. This model can be used to navigate the design space. Factor Coefficient Estimate Intercept 2.927 C-ATM 0.983 D-Fe(0) -1.074 E-Tailings 1.477 F-S(0) -1.384 CE 0.888 CE 1.229 jp Standard Error 1 0.3196 1 0.3196 1 0.3196 1 0.3196 1 0.3196 1 0.3196 1 0.2948 95% Cl 95% Cl Low High 2.2153 3.6396 0.2706 1.6948 -1.7861 -0.3618 0.7652 2.1894 -2.0959 -0.6717 0.1754 1.5997 0.5619 1.8956 218 Final Equation in Terms of Coded Factors: Arsenic = 2.9274375 0.9826875 * C -1.073938 * D 1.4773125 * E -1.383813 * F 0.8875625 * C * E Final Equation in Terms of Actual Factors: Arsenic = 2.9274375 0.9826875 * ATM -1.073938 *Fe(0) 1.4773125 'Ta i l ings -1.383813 *S(0) 0.8875625 * ATM * Tailings Diagnostics Case Statistics Standard Actual Predicted Residual Leverage Student Cook's Outlier Run Order Value Value Residual Distance t Order 1 4.81 4.992 -0.182 0.375 -0.1803 0.0033 -0.1713 11 2 2.03 1.045 0.985 0.375 0.9745 0.0950 0.9718 14 3 0.5 1.045 -0.545 0.375 -0.5394 0.0291 -0.5193 13 4 5.24 4.992 0.248 0.375 0.2451 0.0060 0.2333 15 5 0.291 1.235 -0.944 0.375 -0.9344 0.0873 -0.9279 12 6 9.68 8.733 0.947 0.375 0.9372 0.0878 0.9310 8 7 7.72 8.733 -1.013 0.375 -1.0020 0.1004 -1.0023 6 8 1.74 1.235 0.505 0.375 0.4993 0.0249 0.4797 2 9 1.57 1.665 -0.095 0.375 -0.0939 0.0009 -0.0891 4 10 0.045 0.077 -0.032 0.375 -0.0314 0.0001 -0.0298 1 11 0.043 0.077 -0.034 0.375 -0.0334 0.0001 -0.0317 7 12 1.32 1.665 -0.345 0.375 -0.3412 0.0116 -0.3256 16 13 6.2 3.817 2.383 0.375 2.3576 0.5558 3.3559 9 14 2.64 1.855 0.785 0.375 0.7766 0.0603 0.7600 10 15 1.51 1.855 -0.345 0.375 -0.3415 0.0117 -0.3259 3 16 1.5 3.817 -2.317 0.375 -2.2928 0.5257 -3.1582 5 Proceed to Diagnostic Plots (the next icon in progression). Be sure to look at the: 1) Normal probability plot of the studentized residuals to check for normality of residuals. 2) Studentized residuals versus predicted values to check for constant error. 3) Outlier t versus run order to look for outliers, i.e., influential values. 4) Box-Cox plot for power transformations. If all the model statistics and diagnostic plots are OK, finish up with the Model Graphs icon. 219 Set 2 - Day 40 D e s i g n Ease G r a p h s a n d A N O V A Tab le JD O Normal Plot of Res idua ls -2.35 -1.17 0.00 2.35 Studentized Residuals 3.00-1.50-CO •3 CO CD cc "S 0-c.o-N "E CD "D C/D -1.50--3.00-Res idua ls vs . P red ic ted 0.00 1.23 2.45 3.67 4.90 Predicted Residuals vs. Run 220 3.00-1.50 H JO CO •a co CU DC "8 N C CD "O -2 CO - J — 1 — I — I I I I I I I 7 10 13 16 i ' 1 r 1 4 Run Number Outlier T 7.59-3.79 —I Z J o -ta B _ - 3 . 7 9 — r -7.59 H I 1 1 I 1 1 I 10 13 16 i i i i r Run Number 221 Pred ic ted vs . Ac tua l 5.12-3.84 H T3 CD TJ & CL 1.28 -J 0.00-0.00 1.28 2.56 3.84 5.12 Actual Box-Cox Plot for Power T r a n s f o r m s 10.15H 6.58 2 3.01 CD DC 4.56 Lambda In teract ion G r a p h DESIGN-EXPERT Plot Arsenic A: DOC C: ATM • C- -1.000 A C+ 1.000 Actual Factors B: S04 = 0.00 D: Fe(0) = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 0.055 0.50 1.00 A: DOC In teract ion G r a p h DESIGN-EXPERT Plot Arsenic A: DOC D: Fe(0) • D--1.000 A D+ 1.000 Actual Factors B: S04 = 0.00 C: ATM = 0.00 E: Tailings = 0.00 F: S(0) = 0.00 3.85375 H o 'c CD (2 < 2.5875 1.32125 - h 0.055 H -1.00 -0.50 0.00 0.50 1.00 A: DOC I n t e r a c t i o n G r a p h DESIGN-EXPERT Plot Arsenic A: DOC F: S(0) • F- -1.000 A F+ 1.000 Actual Factors 8: S 0 4 = 0.00 C: ATM = 0.00 D:Fe(0) = 0.00 E: Tail ings = 0.00 -1.00 -0.50 0.00 0.50 1.00 A: DOC I n t e r a c t i o n G r a p h 5.12 H DESIGN-EXPERT Plot Arsenic 3.8337 -X = B: S 0 4 Y = D : F e ( 0 ) • D- -1.000 A D+ 1.000 Actual Factors A: DOC = 0.00 C: ATM = 0.00 E: Tail ings = 0.00 F: S(0) = 0.00 o 'c g 2.54739-I < 1.26109-^ -0.0252128--1.00 -0.50 0.00 0.50 1.00 B : S 0 4 Interaction Graph •0.0275028 H B : S 0 4 Interaction Graph DESIGN-EXPERT Plot Arsenic C: ATM E: Tailings • E--1.000 A E+ 1.000 Actual Factors A: DOC = 0.00 B: S04 = 0.00 D: Fe(0) = 0.00 F: S(0) = 0.00 -1.00 -0.50 0.00 0.50 1.00 C: ATM O n e Fac to r Plot 5 1 2 J Warning! Factor involved in an interaction. o c <jj 2.5875 -< 1.32125-0.055 H i i i i : r~ -1.00 -0.50 0.00 0.50 1.00 B: S04 O n e Fac to r Plot 5 1 2 J Warning! Factor involved in an interaction. 3.85375 H o ' c 0j 2.5875 -< 1.32125-0.055 H ~1 1 1 1 I -1.00 -0.50 0.00 0.50 1.00 C: ATM O n e Fac to r Plot 0.0285261 H -1.00 -0.50 0.00 0.50 1.00 E: Tailings D E S I G N - E X P E R T Plot Arsenic A: D O C B: S 0 4 C: ATM D: Fe(0) E: Ta i l i ngs F : S ( 0 ) H a l f N o r m a l p l o t 1 0.00 | E f f e c t | 227 Response: Arsenic ANOVA for Selected Factorial Model Analysis of variance table [Partial sum of squares] Source Sum of DF Mean F Prob >F Squares Square Value Model 33.59 9 3.73 157.29 < 0.0001 B 5.46 1 5.46 230.15 < 0.0001 C 3.15 1 3.15 132.69 < 0.0001 E 8.32 1 8.32 350.76 < 0.0001 AC 0.46 1 0.46 19.22 0.0046 AD 3.51 1 3.51 148.06 < 0.0001 AF 3.31 1 3.31 139.50 < 0.0001 BD 0.460 1 0.46 19.39 0.0046 BE 5.783 1 5.78 243.74 < 0.0001 CE 3.134 1 3.13 132.09 < 0.0001 Residual 0.142 6 0.02 Cor Total 33.728 15 The Model F-value of 157.29 implies the model is significant. There is only a 0.01% chance that a "Model F-Value" this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are significant. In this case B, C, E, AC, AD, AF, BD, BE, CE are significant model terms. Values greater than 0.1000 indicate the model terms are not significant. If there are many insignificant model terms (not counting those required to support hierarchy), model reduction may improve your model. Std. Dev. Mean CV. PRESS 0.1540 0.8439 18.2513 1.0123 R-Squared Adj R-Squared Pred R-Squared Adeq Precision 0.9958 0.9894 0.9700 40.1923 The "Pred R-Squared" of 0.9700 is in reasonable agreement with the "Adj R-Squared" of 0.9894. "Adeq Precision" measures the signal to noise ratio. A ratio greater than 4 is desirable. Your ratio of 40.192 indicates an adequate signal. This model can be used to navigate the design space. Factor Coefficient Estimate Intercept 0.844 B-S04 -0.584 C-ATM 0.444 E-Tailings 0.721 AC -0.169 AD -0.469 AF -0.455 BD -0.170 BE -0.601 CE 0.443 DF Standard 95% Cl 95% Cl Error Low High 0.0385 0.7497 0.9382 0.0385 -0.6784 -0.4900 0.0385 0.3493 0.5378 0.0385 0.6270 0.8154 0.0385 -0.2630 -0.0746 0.0385 -0.5628 -0.3743 0.0385 -0.5490 -0.3606 0.0385 -0.2638 -0.0753 0.0385 -0.6954 -0.5070 0.0385 0.3483 0.5368 VIF 228 Final Equation in Terms of Coded Factors: Arsenic = 0.8439375 -0.584188 *F3 0.4435625 *C 0.7211875 *E -0.168813 * A* C -0.468563 * A* D -0.454813 * A* F -0.169563 * B * D -0.601188 * B * E 0.4425625 * C * E Diagnostics Case Statistics Standard Actual Predicted Residual Leverage Student Cook's Outlier Run Order Value Value Residual Distance t Order 1 0.632 0.603 0.029 0.625 0.3114 0.0162 0.2866 11 2 0.11 0.118 -0.008 0.625 -0.0822 0.0011 -0.0750 14 3 0.092 0.126 -0.034 0.625 -0.3578 0.0213 -0.3302 13 4 0.66 0.755 -0.095 0.625 -1.0112 0.1704 -1.0134 15 5 0.055 0.092 -0.037 0.625 -0.3949 0.0260 -0.3653 12 6 4 4.222 -0.222 0.625 -2.3496 0.9201 -7.590 * 8 7 0.584 0.681 -0.097 0.625 -1.0271 0.1758 -1.0327 6 8 0.144 0.155 -0.011 0.625 -0.1193 0.0024 -0.1090 2 9 0.127 0.119 0.008 0.625 0.0822 0.0011 0.0750 4 10 1.25 1.279 -0.029 0.625 -0.3114 0.0162 -0.2866 1 11 0.174 0.079 0.095 0.625 1.0112 0.1704 1.0134 7 12 0.158 0.124 0.034 0.625 0.3578 0.0213 0.3302 16 13 5.12 4.898 0.222 0.625 2.3496 0.9201 7.590 * 9 14 0.131 0.094 0.037 0.625 0.3949 0.0260 0.3653 10 15 0.165 0.154 0.011 0.625 0.1193 0.0024 0.1090 3 16 0.101 0.004 0.097 0.625 1.0271 0.1758 1.0327 5 * Case(s) with |Outlier T| > 3.50 Proceed to Diagnostic Plots (the next icon in progression). Be sure to look at the: 1) Normal probability plot of the studentized residuals to check for normality of residuals. 2) Studentized residuals versus predicted values to check for constant error. 3) Outlier t versus run order to look for outliers, i.e., influential values. 4) Box-Cox plot for power transformations. If all the model statistics and diagnostic plots are OK, finish up with the Model Graphs icon. 

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.831.1-0081126/manifest

Comment

Related Items