Quantitative Risk-Based Approach for Improving Water Quality

Jul 28, 2011 - ACS eBooks; C&EN Global Enterprise .... Research and analysis of mine water management has led to the conclusion that the use of ... St...
0 downloads 0 Views 2MB Size
Download PDF
ARTICLE pubs.acs.org/est

Quantitative Risk-Based Approach for Improving Water Quality Management in Mining Wenying Liu,* Chris J. Moran, and Sue Vink Centre for Water in the Minerals Industry, Sustainable Minerals Institute, The University of Queensland, Brisbane, Queensland, 4072, Australia ABSTRACT: The potential environmental threats posed by freshwater withdrawal and mine water discharge are some of the main drivers for the mining industry to improve water management. The use of multiple sources of water supply and introducing water reuse into the mine site water system have been part of the operating philosophies employed by the mining industry to realize these improvements. However, a barrier to implementation of such good water management practices is concomitant water quality variation and the resulting impacts on the efficiency of mineral separation processes, and an increased environmental consequence of noncompliant discharge events. There is an increasing appreciation that conservative water management practices, production efficiency, and environmental consequences are intimately linked through the site water system. It is therefore essential to consider water management decisions and their impacts as an integrated system as opposed to dealing with each impact separately. This paper proposes an approach that could assist mine sites to manage water quality issues in a systematic manner at the system level. This approach can quantitatively forecast the risk related with water quality and evaluate the effectiveness of management strategies in mitigating the risk by quantifying implications for production and hence economic viability.

’ INTRODUCTION Mining activities can negatively affect local environment and water resources through both freshwater withdrawal and pollution from mine water discharge. In terms of access to freshwater for production purposes, the industry as a whole is operating under a general global context of widespread water scarcity,1,2 and limited freshwater availability from the natural environment in the long run.3,4 With respect to mine water pollution, three types of mine discharge exemplify the potential environmental threats posed by the industry: acid and metalliferous mine drainage (AMD) due to oxidation of metal sulfides (often pyrite), cyanide release, and saline water discharge. Unmitigated AMD can have adverse effects on water quality and ecosystem health.5 8 The environmental threats of cyanide (reagent added for extracting gold from low-grade ore) release from precious metal mining operations are receiving increased attention because of potential toxicity to aquatic life.9 12 Discharge of saline water into the local catchments degrades the aquatic habitat and can be toxic to aquatic organisms.13 Release of salts can also compromise downstream water uses such as drinking water and irrigation.14 These environmental challenges are among the factors motivating the mining industry to save freshwater and minimize mine water discharge into local streams. Research and analysis of mine water management has led to the conclusion that the use of multiple sources of water supply and introducing water reuse into r 2011 American Chemical Society

the site water system are effective and viable water management practices.15 Water reuse is common in minerals extraction and concentration activities, e.g., utilization of saline water in coal mining,16 and reusing water in gold cyanidation17 and sulphide ore flotation.18 However, as water reuse efficiency increases on a site there is an increased tendency for water quality to vary in response to different physical and physicochemical processes occurring in different parts of the water system. For example, dissolved constituents can be concentrated due to evaporation and diluted due to rainfall or runoff which may vary seasonally. In some cases water quality variation can bring fluctuations in minerals beneficiation efficiency, particularly in froth flotation, a process for separating valuable minerals from waste gangue by taking advantage of the differences in surface hydrophobicity.19 21 Even if minerals separation efficiency is not affected, changes in water quality can also dictate environmental consequences of noncompliant discharge events.22 Overall, it is clear that the interactions among water management practices that conserve water, efficiency of minerals separation processes, and environmental consequences are complex. Generally, water quality impacts are controlled separately in the local area where Received: June 1, 2011 Accepted: July 28, 2011 Revised: July 26, 2011 Published: July 28, 2011 7459

dx.doi.org/10.1021/es201876c | Environ. Sci. Technol. 2011, 45, 7459–7464

Environmental Science & Technology the impacts occur in a reactive manner, for example, changing flotation reagent regimes in response to water quality variation to maintain mineral recovery efficiency, and end-of-pipe dilution for discharge control. Given that water management practices, efficiency of minerals separation processes, and environmental consequences are intimately linked through the mine water system,23 there is an opportunity to develop an integrated approach that takes into account the mine water system’s connectivity to assist mine sites to manage water quality. In this paper, an approach is proposed that deals with complexity and allows specific situations to be considered in a generic manner. The approach combines risk quantification and

ARTICLE

system-wide risk control options. The latter allows opportunities to be identified from the entire mine water system as opposed to separately dealing with each impact in its local area. The approach is demonstrated in a case study of a coal mine in Central Queensland, Australia. Quantification of the risk posed to the coal washing from water conductivity variation is used to assess the value of a simple mitigation strategy outside the processing plant. The successful implementation of this mitigation strategy in production ensures that the environmental consequences of both freshwater withdrawal and mine water discharge are minimized. The form of the risk-based approach is considered to be generic and should be applicable to mine sites operating within different contexts.

’ RISK-BASED APPROACH

Figure 1. Framework of the risk-based approach.

General Framework. The general framework of the riskbased approach, as shown in Figure 1, involves two steps. Step 1 is to quantify the risk related to water quality variation by combining the frequency of water quality variation (likelihood) and the magnitude of the impact resulting from water quality variation (consequence). In addition, this step can diagnose the reasons why water quality varies. Step 2 is to effectively control the risk by changing water quality variation frequency and impact magnitude through developing systematic water quality management and reagents scenarios in light of the diagnosed reasons. Risk Quantification. In this step, the aim is to predict and account for the dynamic behavior of a key water quality variable in terms of the component or process affected. The simplest way of achieving this is to access mine site monitoring data and

Figure 2. Schematic of the framework of the risk-based approach: (A) time series of an example water quality variable; (B) frequency plot (solid line) and new frequency plot of the water quality variable after implementing proposed water quality management scenarios (dashed line); (C) impact magnitude: quantitative relationship between process response and the water quality variable, linear (9) or quadratic (0); (D) risk plot: solid line and single shading represent loss of opportunity; dashed line and cross-hatch region represent new risk plot and reduction of loss of opportunity after implementing proposed water quality management scenarios; black area is theoretical limit of risk mitigation. 7460

dx.doi.org/10.1021/es201876c |Environ. Sci. Technol. 2011, 45, 7459–7464

Environmental Science & Technology observe the time series (Figure 2A), which shows how the concentration of the variable changes over time. The time series is transformed into a frequency distribution (Figure 2B), which quantifies the time percentage for which the variable falls within a particular concentration interval. This transformation renders the time series amenable for combination with the impact magnitude information (Figure 2C). The frequency plot may be expected to exhibit different forms depending on the time period considered. In general, the time period of interest should be the period with water quality variations which could bring significant fluctuations to the performance of the process under investigation. In terms of flotation operations, key water quality variables are not always monitored with the desired time interval or at the locations desired for optimal analysis. In such cases, constructing a mass balance model of the mine site water system is a way to estimate the time series from available information. Empirical approaches could also be used but are more limited in their value. Quantification of the impact magnitude is generally determined experimentally. The response of the process under investigation, e.g., flotation recovery and grade, is tested across a range of concentrations of the water quality variable. The response function can thereby be fitted empirically so that the impact of variation in the variable within the experimental range can be forecast. Figure 2C shows two possible hypothetical response functions in mineral flotation operation. The linear function represents a positive impact of the variable on the process studied. The quadratic function represents an impact that varies depending on the concentration of the variable, changing from a positive impact at low concentrations to a negative impact at high concentrations. The response function may be expected to exhibit different forms to different water quality variables. Apart from quantifying the correlation between the response and the water quality variable, the experimental results can also demonstrate the extent to which water quality management scenarios should be implemented, i.e., whether there is sufficient magnitude of the impact to warrant action. The key innovation in this work is the combination of the water quality variation frequency and the impact magnitude function to quantify the risk related with water quality variation. That is, the frequency with which the process is impacted can be plotted as a function of the magnitude of the impact (Figure 2D). More specifically, the risk plot describes the time percentage for which the response of the process exceeds a certain value. In this risk plot, the shaded area represents the lost opportunity. The dark area in the top right corner represents the theoretical limit of performance. Risk Control. The aim of risk control is to minimize the shaded area (Figure 2) where the risk is loss of opportunity. Generally, the approaches employed to deal with the impacts resulting from water quality variation are applied locally in an isolated fashion, e.g., changing flotation reagents for maintaining recovery efficiency or end-of-pipe dilution for discharge control. These approaches, referred to as internal solutions by reference to the local area, may seem to be efficient ways to handle the impacts. However, the site water system consists of different water supplies and the resulting impacts of water quality are intimately linked through the site water system. Water quality from these water supplies can be affected by many chemical and physicochemical processes and interactions with the climate system. As a consequence, water quality is dynamic, associated with all of the feedbacks and interactions involved in the site water system.

ARTICLE

Figure 3. Simplified view of the water system at the study site.

Dealing with the impacts on a case by case basis by internal solutions has limitations. Lack of systems thinking has the potential to cause unintended impacts when managing one impact. Given that the site water system acts as a complex system, an understanding of the properties of different water streams and the connectivity of the whole water system are required. By understanding these properties, in some cases, relatively simple water management practices, termed herein external solutions, can be applied to manage the impacts related to water quality variation. To determine the best risk mitigation strategies, the shaded area should be managed by the best combination of internal and external solutions. A key advantage of the risk plot is that costs for control have a quantitative base of comparison with the costs (or revenue foregone) associated with lack of control. By implementing water quality management scenarios, the frequency plot of the variable can be altered by means of moving to a preferred concentration and/or decreasing the variance of the plot (Figure 2B). The shaded area in the risk plot can be quantitatively reduced from the single shading to the cross-hatched region (Figure 2D), showing that part of the risk is controlled by implementing water quality management strategies.

’ DATA SOURCES Study Site. Water quality data used in the risk-based approach are from a coal mine located in the Bowen Basin of Central Queensland, Australia. The mine predominantly extracted coal from underground collieries and also processed coal from some adjacent operations. A simplified view of the site water system is shown in Figure 3. There is a large central worked water store which is used to collect and distribute water around the site. Worked water, or nonfresh water,24 in the central water store is made up of saline water pumped from underground mines, runoff from spoil and roads, and water recovered from the tailings facility. Water quantity and quality in the central water store is also affected by climate conditions, such as high annual evaporation and seasonal rainfall events. Around 95% of water used for coal washing in the coal handling and preparation plant (CHPP) is sourced from the central water store with the remaining 5% being freshwater makeup. The blender is a virtual representation of site infrastructure used to demonstrate that worked water might need to be diluted with freshwater or treated 7461

dx.doi.org/10.1021/es201876c |Environ. Sci. Technol. 2011, 45, 7459–7464

Environmental Science & Technology

ARTICLE

Figure 4. Risk-based approach for the study site: (A) time series of surface water conductivity of the central water store between January and June 2008; daily rainfall is plotted for reference; (B) histogram (bars) and frequency plot derived from the time series (solid line), and new frequency plot after proposed water quality management (dashed line); (C) coal combustible recovery as a function of conductivity of synthetic saline water; (D) risk plot before (solid line) and after proposed water quality management (dashed line).

water to meet a specific concentration requirement of different operations, especially the CHPP. Previous work has indicated that the use of water with high salinity in coal washing provides benefits by improving recovery with reduced reagent use.25 However, a barrier for the site to reusing saline water in the CHPP is the water conductivity variation, which results in fluctuations in the performance of the CHPP.22 The key issue to be addressed is how to provide water with consistent conductivity to the CHPP in combination with appropriate reagent additions to maintain a steady and satisfactory coal recovery. One option is to dilute worked water with freshwater to a consistent but lower salinity together with higher reagent dosage to achieve recovery equivalent to a higher salinity. There are two problems with this option. (a) It will increase freshwater withdrawal externally to dilute the worked water to a constant salinity. The resulting requirement for larger water stores is expensive. Without this, there is an increase in the likelihood of saline water discharge because of insufficient space to accommodate this freshwater. (b) Higher reagent consumption means more costs for the mine site and more residual reagents in water, which would in turn increase the potential environmental damage if this water is discharged into local streams. In this work, we will assess a simple strategy for providing water with a consistent but higher salinity to the CHPP. Water Quality Data. Salinity in the central water store was continuously monitored between October 2006 and 2008 using a Hydrolab sonde.22 The study period included the wet season of 2007 2008 when the region experienced several heavy rainfall events. The sampling time was chosen to capture water quality changes in the water column of the central water store. Reagent use during the wet season was also monitored. The reagents used

in coal flotation were diesel as a collector and methyl isobutyl carbinol (MIBC) as a frother. Coal Flotation Experimental Data. Coal flotation experimental results for the site were obtained from previously published work.25 In that work, synthetic saline water was prepared by mixing a number of pure salts. Batch flotation experiments with synthetic saline water were conducted under controlled laboratory conditions to investigate the impact of water conductivity on coal flotation performance.

’ RESULTS AND DISCUSSION Risk Quantification. Water conductivity variation in surface water of the central water store between January and June 2008, when the region experienced several heavy rainfall events, is shown in Figure 4A. Surface water conductivity was significantly affected by rainfall, with decreased conductivity from 13 to 6 mS/ cm by early March following each of the major rainfall events in January and February. Over the following period, conductivity gradually increased to around 12 mS/cm. The time series was transformed into a histogram and a frequency distribution was fitted (Figure 4B), which was approximated by a normal distribution (μ = 10.4 mS/cm, σ = 1.5). The regression plot of combustible recovery (a measure of coal flotation recovery) versus water conductivity shows that combustible recovery increased linearly as water conductivity increased (Figure 4C). This means that the use of salty water in coal flotation has a positive influence on the process performance. This phenomenon supports the implementation of saline water reuse. The risk in relation to water conductivity variation was quantified by coupling the frequency plot of conductivity and flotation experimental results, which is represented by the solid 7462

dx.doi.org/10.1021/es201876c |Environ. Sci. Technol. 2011, 45, 7459–7464

Environmental Science & Technology

Figure 5. Weekly reagent use January June 2008 at the study site. Water conductivity was plotted for comparison.

Figure 6. Vertical profiles of water conductivity monitored in the central water store at the study site.

line in Figure 4D. The risk plot shows the time percentage of combustible recovery exceeding a certain value. For example, the point (58, 80) in the risk plot means that 80% of the time the combustible recovery exceeded 58%. The aim of the mine site is to realize higher recovery more often. Risk Control. The internal solution that the mine site adopted was to adjust flotation chemistry by adding flotation reagents including collector and frother. Figure 5 shows changes in use of flotation reagents between January and June 2008. The increase in both frother and collector usage failed to maintain flotation yields. Although there were some changes in the quality of ore feed, the primary factor affecting flotation was considered to be the variation in water conductivity due to rainfall events. At the time, the CHPP operators were unaware of this variation and were responding reactively to poor recovery by changing reagent use more-or-less haphazardly. This mine has a relatively simple water circuit because the water management system is based around a central water store. Figure 6 shows vertical profiles of water conductivity measured in the central water store. It can be seen that there was no vertical stratification of conductivity in the store prior to February 2008. The profile taken after the heavy rainfall in January and February 2008 shows that while surface water conductivity has decreased, water quality below around 6 m remained constant. Subsequent data, measured in April and June 2008, shows that the halocline

ARTICLE

was deepening and surface water conductivity was increasing until the pit was almost fully mixed in October 2008. Providing relatively consistent water quality for the CHPP, even during extraordinary rainfall events, is possible by pumping water from below the halocline. Figure 4B shows the frequency distribution of water conductivity if the pump intake had been lowered to 6 m. The mean value of conductivity was elevated from 10.4 to 12.0 mS/cm and the standard deviation was decreased from 1.5 to 0.5, i.e., higher and more consistent conductivity. Accordingly, flotation performance can be improved by pumping water from depth. This is shown quantitatively in the risk plot (Figure 4D), which indicates, for example, 80% of the time the combustible recovery exceeded 70% compared to 58% prior to implementing the proposed water quality management strategy. This means that the risk posed to production by water conductivity variation can be significantly reduced by making this simple change. The successful provision of water with consistent conductivity to the CHPP by pumping water from a depth of 6 m ensures that freshwater withdrawal for production purposes is reduced and that the potential environmental threat posed by flotation reagents and saline water discharge is attenuated. The risk-based approach provides a framework for considering water quality management, production, and environmental consequences in a systematic and quantitative manner, which was demonstrated by a single water-quality variable case. Other similar cases also exist in which ore variation may dominate, e. g., the use of seawater in almost all the metal-producing processes at Esperanza copper mine in Chile.26 It may be a challenging task to deal with a process with more than one water quality variable involved where nonlinear interactions occur. The main challenge will be to quantify the impact magnitude relationship, which may be multidimensional rather than a simple function as in this case study.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]; phone: +61 7 3346 4027; fax: +61 7 3346 4045.

’ ACKNOWLEDGMENT Funding for this study was provided by Australian Research Council linkage project (LP0883872) “Impact of recycled and low quality process water on sustainable mineral beneficiation practices”, which is part of AMIRA project P260E. Water quality data collection was funded by Australian Research Council linkage project (LP0667780) “Understanding salt dynamics to facilitate water reuse on coal mine sites”. ’ REFERENCES (1) Seckler, D.; Amarasinghe, U.; Molden, D.; Silva, R. d.; Barker, R. World Water Demand and Supply, 1990 to 2025: Scenarios and Issues; Research Report 19; International Water Management Institute: Colombo, Sri Lanka, 1998; http://pdf.usaid.gov/pdf_docs/PNACC575. pdf. (2) V€or€ osmarty, C. J.; Green, P.; Salisbury, J.; Lammers, R. B. Global water resources: vulnerability from climate change and population growth. Science 2000, 289, 284–288. (3) Ridoutt, B. G.; Pfister, S. Reducing humanity’s water footprint. Environ. Sci. Technol. 2010, 44, 6019–6021. 7463

dx.doi.org/10.1021/es201876c |Environ. Sci. Technol. 2011, 45, 7459–7464

Environmental Science & Technology (4) Peters, N. E.; Meybeck, M. Water quality degradation effects on freshwater availability: Impacts of human activities. Water Int. 2000, 25, 185–193. (5) Gray, N. F. Environmental impact and remediation of acid mine drainage: A management problem. Environ. Geol. (Heidelberg, Ger.) 1997, 30, 62–71. (6) Van Damme, P. A.; Hamel, C.; Ayala, A.; Bervoets, L. Macroinvertebrate community response to acid mine drainage in rivers of the High Andes (Bolivia). Environ. Pollut. (Oxford, U. K.) 2008, 156, 1061–1068. (7) Filipek, L. H.; Nordstrom, D. K.; Ficklin, W. H. Interaction of acid mine drainage with waters and sediments of West Squaw Creek in the West Shasta Mining District, California. Environ. Sci. Technol. 1987, 21, 388–396. (8) MacCausland, A.; McTammany, M. E. The impact of episodic coal mine drainage pollution on benthic macroinvertebrates in streams in the Anthracite region of Pennsylvania. Environ. Pollut. (Oxford, U. K.) 2007, 149, 216–226. (9) Eisler, R. Cyanide Hazards to Fish, Wildlife, and Invertebrates: A Synoptic Review; Biological Report 85(1.23); U.S. Fish and Wildlife Service, Patuxent Wildlife Research Center: Laurel, MD, 1991; http:// www.infomine.com/publications/docs/Eisler1991.pdf. (10) Hanawa, M.; Harris, L.; Graham, M.; Farrell, A. P.; BendellYoung, L. I. Effects of cyanide exposure on Dascyllus aruanus, a tropical marine fish species: Lethality, anaesthesia and physiological effects. Aquarium Sci. Conserv. 1998, 2, 21–34. (11) Johnson, C. A.; Grimes, D. J.; Leinz, R. W.; Rye, R. O. Cyanide speciation at four gold leach operations undergoing remediation. Environ. Sci. Technol. 2008, 42, 1038–1044. (12) Johnson, C. A.; Leinz, R. W.; Grimes, D. J.; Rye, R. O. Photochemical changes in cyanide speciation in drainage from a precious metal ore heap. Environ. Sci. Technol. 2002, 36, 840–845. (13) Chapman, P. M.; Bailey, H.; Canaria, E. Toxicity of total dissolved solids associated with two mine effluents to chironomid larvae and early life stages of rainbow trout. Environ. Toxicol. Chem. 2000, 19, 210–214. (14) Kefford, B. J. The effect of saline water disposal: Implications for monitoring programs and management. Environ. Monit. Assess. 2000, 63, 313–327. (15) C^ote, C.; Moran, C.; Hedemann, C. Evaluating the costs and benefits of salt management strategies at mine sites using a systems model. Mine Water Environ. 2007, 26, 229–236. (16) Turek, M.; Dydo, P.; Surma, A. Zero discharge utilization of saline waters from “Wesola” coal-mine. Desalination 2005, 185, 275–280. (17) Bahrami, A.; Hosseini, M.; Razmi, K. An investigation on reusing process water in gold cyanidation. Mine Water Environ. 2007, 26, 191–194. (18) Rao, S. R.; Finch, J. A. A review of water re-use in flotation. Miner. Eng. 1989, 2, 65–85. (19) Celik, M. S.; Somasundaran, P. The effect of multivalent ions on the flotation of coal. Sep. Sci. Technol. 1986, 21, 393–402. (20) Fornasiero, D.; Ralston, J. Effect of surface oxide/hydroxide products on the collectorless flotation of copper-activated sphalerite. Int. J. Miner. Process. 2006, 78, 231–237. (21) Yang, G. C. C.; Markuszewski, R.; Wheelock, T. D. Oil agglomeration of coal in inorganic salt solutions. Int. J. Coal Prep. Util. 1988, 5, 133–146. (22) Vink, S.; Moran, C. J.; Golding, S. D.; Baublys, K.; Nanjappa, V. Understanding mine site water and salt dynamics to support integrated water quality and quantity management. Trans. Inst. Min. Metall., Sect. A 2009, 118, 185–192. (23) C^ote, C. M.; Moran, C. J.; Hedemann, C. J.; Koch, C. Systems modelling for effective mine water management. Environ. Modell. Software 2010, 25, 1664–1671. (24) Barrett, D.; Moran, C.; Cote, C. A method for estimating the potential trading of worked water among multiple mines. Mine Water Environ. 2010, 29 (2), 92–98.

ARTICLE

(25) Ofori, P.; Firth, B.; Franks, G.; Nguyen, A.; Jameson, G. Impact of Saline Water on Coal Flotation; C13051; Australian Coal Association: Australia, 2005; http://www.acarp.com.au/cart.aspx. (26) Fox Business Website. CESCO: Chile Antofagasta’s Esperanza gets creative with seawater; http://www.foxbusiness.com/industries/ 2011/04/06/cesco-chile-antofagastas-esperanza-gets-creative-seawater/.

7464

dx.doi.org/10.1021/es201876c |Environ. Sci. Technol. 2011, 45, 7459–7464