The Use of Computer Simulation for the Design of a Bioheap Leach for Sphalerite

Transcription

1 The Use of Computer Simulation for the Design of a Bioheap Leach for Sphalerite Todd J. Harvey GeoBiotics, LLC W. Alameda Pkwy, Suite 101 Lakewood, CO, USA Ph: Fax: Tharvey@geobiotics.com Jean Joubert, Willem Van Der Merwe Kumba Resources (PTY) Ltd PO Box 9229 Pretoria West, Gauteng 0001 South Africa Ph:+27 (012) Fax:+27 (012) Jean.Joubert@kumbaresources.com Willem.VDMerwe@kumbaresources.com ABSTRACT GeoBiotics, LLC and Kumba Resources (Pty) Ltd are jointly developing the GEOCOAT heap bioleaching technology and downstream processing technology for recovery of zinc from tailings at the Rosh Pinah Mine in Namibia. This paper outlines the development of a computer simulation for the bioleaching of sphalerite and the downstream processing. Computer modeling has shortened the product development cycle by allowing many flowsheet scenarios to be analyzed negating much of the expensive and time consuming laboratory testing. A model was developed using the Limn system which is an Excel Add-In flowsheet drawing and solution engine. Given that most engineers are familiar with Excel this software is extremely easy to use and has less hidden calculations making model verification much easier. Keywords: simulation, modeling, biooxidation, sphalerite INTRODUCTION The development of new technology for the extraction of metals is an extremely complicated process generally involving years of laboratory testing, pilot scale proving and hopefully commercial application. Many processes never manage to pass the laboratory stage, even fewer make it to pilot testing and rarely, is a new process introduced commercially. For the extraction of zinc from sphalerite only two commercial processes currently exist namely roast-leach-electrowin (RLE) and pressure hydrometallurgy. Pyrometallurgical extraction methods have been in existence since the 1800 s. Hydrometallurgical methods, although developed in the 1950 s, did not see wide scale commercial applications until the 1980 s. GeoBiotics, LLC and Kumba Resources (Pty) Ltd are jointly developing the GEOCOAT heap bioleaching and downstream processing technology to recover zinc from tailings at the Rosh Pinah Mine. The GEOCOAT process involves the coating of concentrates onto a suitable substrate, usually barren rock, then stacking the coated material in a conventional heap fashion. The heap is irrigated with acidic solutions containing iron and nutrients while low pressure ambient air is applied at the heap base. Biological activity oxidizes the sulphide minerals. In the case of sphalerite concentrates, the oxidation product is a solution carrying with it the solubilized zinc which is then extracted from the solution using one of several existing technologies. Several papers have been published outlining the GEOCOAT technology in more detail, Harvey, Holder, and Stanek (2002), Harvey, Holder, and Stanek (2002), Harvey and Potter, (1999). Given the time generally required to develop new process routes, particularly processes such as biooxidation where a single test can take several months, the only method feasible to shorten the product development cycle is computer modeling. Computer modeling of the GEOCOAT process flowsheet has allowed the project team to quickly identify potential process pitfalls and modify laboratory experiments to test the hypothesis. In this manner laboratory testing is optimized and full scale plant simulations can be created with much less data. Of particular

2 importance for this system is the modeling of solution flows. The bioleaching of sphalerite is a relatively straightforward exercise but the winning of the zinc from the pregnant leach solution (PLS) is much more complicated. The presence of a variety of impurities can cause significant problems. Elements such as fluorine, chlorine, manganese, cobalt, copper and iron need to be closely monitored particularly in circulating solutions. The use of computer simulation allows for these deleterious elements to be monitored in a closed loop system reducing the need for full scale lock-cycle laboratory or pilot testing. This paper outlines the development of a computer simulation for the bioleaching of sphalerite and the downstream processing. A model was developed using the Limn system which is an Excel Add-In flowsheet drawing and solution engine. Given that most engineers are familiar with Excel this software is extremely easy to use and has less hidden calculations making model verification much easier. FLOWSHEET TOOLS One of the keys to a good flowsheet design and thus, a successful plant operation, is good information. Good information manifests itself in a variety of ways in the metallurgical industry but the most important being sample representativeness and the ability to optimize the unit processes through testwork. Unfortunately several factors work against the metallurgist in obtaining good information; the cost of extensive testwork can be prohibitive and the costs of the required samples even more so. The pressure is on the metallurgist to maximize the information obtained from testwork while minimizing the amount of sample consumed. Several things can aid the metallurgist in these endeavors such as proper experimental design, sufficient experience, and computer modeling. Modeling is not a replacement for experience but it can aid in the development of the proper experimental design for unit operation testwork. Models should always be validated and the user must always know the boundary conditions. Having said that, process models can drastically reduce the number of laboratory tests required and can simplify the what if scenario testing. There are a variety of process tools available for modeling metallurgical processes, most of which with a very steep learning curve. The authors choose to use the Limn system for a variety of reasons. Limn is based on Excel, a tool which the vast majority of engineers know intimately, a variety of program Wizards make constructing a flowsheet very straight forward and finally the calculations are not hidden. Excel on its own is increasingly being utilized for everything from metallurgical data analysis, reagent databases, DCS data analysis, and more and more engineers are employing it to model their plant flowsheets. Excel on its own, however, can be cumbersome to utilize. It is a very powerful program but tracking all of the links and formulas can lead to built-in mistakes. Additionally, it is a very poor drawing tool for graphical representation of flowsheets. Limn automates much of the process thus reducing the chances for errors. Limn: The Flowsheet Processor for Microsoft Excel, uses the drawn flowsheet metaphor as the basis of a tool to provide the structure necessary to make efficient use of the Excel spreadsheet in flowsheet solution or simulation tasks. In Limn: The Flowsheet Processor, the structure of the process - the process connectivity - is derived directly from the user drawn process flowsheet. Discrete, uniform spreadsheet cell ranges are used to define process stream information, and discrete spreadsheet cell blocks, with defined input and output ranges, are used to define process units. Automated procedures (or "wizards") are provided to assist in setting up these stream data ranges and process unit model blocks. Since all of the calculation power of the spreadsheet is still available within the unit model block, a range of models, from the very simple, to the extremely complicated can be implemented in this environment. Wiseman, (2000) Many papers have been published outlining the use of simulation/modeling for the design of mineral/chemical flowsheets, Wills (1986), Merks (1999), Cleary (2001), Bailey, Patel and Kumar (2001), Abilov and Zeybek, (2000), Lynch and Morrison, (1999). Most of these papers have shown that the chosen software worked well to achieve the task at hand. The authors of this paper believe that the Limn system, with its commonly used Excel interface and its ready made Wizards, is one of the easiest systems to learn and use and has many other advantages over many of the much more expensive black box systems. The balance of this paper is dedicated

3 to illustrating how this software was employed to solve a unique metallurgical system and the advantages of modeling in process development. PROCESS DESIGN The process under development involves the extraction of zinc from a sphalerite concentrate using the GEOCOAT biooxidation technology, Harvey, Afewu and Van Der Merwe (2002). The solution from the heap biooxidation is processed via conventional downstream processing routes. The downstream process consists of solution purification, solvent extraction and electrowinning. Zinc downstream process optimization is notoriously difficult in that there are multiple purification circuits (iron, copper, cadmium, cobalt, etc.), the use of DEPHA for solvent extraction does not have high extractions, and electrowinning is sensitive to low levels of impurities and fairly inefficient. The result is many recycle streams that have to be carefully balanced and controlled. The GEOCOAT Process The GEOCOAT process incorporates elements of two successful and commercially proven technologies: heap leaching and biooxidation. Zinc-bearing sulphide minerals are concentrated by flotation and thickened. The resulting slurry is thinly coated onto crushed, screened support rock, stacked on a lined pad, and allowed to biooxidize. Coating is accomplished by spraying the concentrate slurry onto the support rock as it discharges from the end of a stacking conveyor onto the biooxidation heap. The coating solids density is highly dependent on the slurry viscosity and densities of 50-65% have been successfully coated at scale. The support rock is relatively uniformly sized, in the range of 6 to 30 millimeters in diameter and the concentrate coating is relatively thin, less than one millimeter in thickness. The weight ratio of support rock to concentrate is in the range of 5:1 to 10:1. The hydrophobic nature of the concentrate assists in the formation of a coating on the support rock. No binding agents are required. The concentrate naturally adheres to the support rock and does not wash out of the heap during solution application or during heavy rainstorms. Depending on the desired temperature of operation, the heap is inoculated with naturally occurring sulphideoxidizing bacteria, such as the mesophiles; Thiobacillus ferrooxidans, Thiobacillus thiooxidans, Leptospirillum ferrooxidans or moderate thermophiles; Acidothiobacillus caldus, Sulfobacillus thermosulfidooxidans, and the extreme thermophiles; Acidianus brierleyi, Acidianus infernus, Metallosphaera sedula, Sulfolobus acidocaldarious, Sulfolobus shibatae and Sulfolobus metallicus. Nutrients are added to the heap via recirculating solutions. As biooxidation progresses, the sulphides in the concentrate are oxidized and the solubilized zinc, iron and sulphate are carried from the heap by the recirculating solution. A portion of the solution stream is bled from the circuit for purification and metal recovery. The relatively uniform size of the support rock leads to large interstitial spaces within the heap and subsequently a low resistance to air and liquid flows. Sufficient air for biooxidation and heat removal is supplied to the heap by low-pressure blowers through a system of perforated pipes laid in the drain rock below the base of the heap. After biooxidation, additional lifts may be placed on the pad or the coated rock may be unloaded from the pad and the oxidized concentrate removed by trommeling or wet screening if precious metal recovery is warranted. The concentrate residue would be neutralized and then subjected to conventional precious metal recovery methods. The support can be recycled or, in the case of low grade sulphide ore, a portion can be bled out for disposal and replaced with fresh zinc bearing ore. The simplicity of the GEOCOAT process offers lower operating costs and capital costs compared to other oxidation processes. Lower operating costs are achieved by reducing energy consumption, reducing maintenance requirements and lowering manpower costs. The very simple unit operations of conveying material and solutions in the heap bioleaching configuration substantially reduces capital costs. The GEOCOAT technology has been proven in a wide variety of column tests (10cm to 1m diameter) and has been successfully

4 demonstrated in two field trials. A GEOCOAT commercial operation in early operation for a refractory gold application is scheduled for Downstream Processing The key to downstream treatment is the maximization of the zinc tenor while limiting the level of impurities. Purification of the PLS must take place to remove the contained iron, copper, cobalt, cadmium, manganese, magnesium, chlorine and fluorine. The removal or iron takes place via the addition of lime. The solution is then filtered and passed into a further metal removal stage where copper and cobalt are removed via zinc powder. Preacidification of the concentrate feed allows for the removal of a large portion of the magnesium with minimal zinc losses. A sufficient bleed stream is maintained to keep the Mg levels below 10 gpl with the balance of the solution returned to the circuit. The circuit is complicated by the variety of recycle streams and bleed streams that must all be balanced to maintain the concentrations of impurities at acceptable levels while maximizing the zinc tenor to solvent extraction. Furthermore, if the zinc tenor can be increased sufficiently, solvent extraction can be bypassed completely and direct electrowinning employed. This also depends on the levels of chlorine and fluorine in the PLS. The model developed with Limn reflects the basic circuit design and allows for many scenarios to be developed and analyzed for both the biooxidation and downstream processes. Each scenario, once optimized, is supported by laboratory testwork and the results tested against the model. MODEL DEVELOPMENT The current model evolved over a period of several months. Each iteration of the model added a new level of complexity to the system but also allowed a new level of analysis. These revisions where the result of both model refinement and process testwork. Figure 1 shows the current process model. The current model simulates both the bioleach and downstream process and has the following features included: 1. The ability to track all metals in solution throughout the process and follow tenor build up of elements in the system. 2. Simulates the use of one or two solution ponds, PLS (pregnant leach solution) and ILS (intermediate leach solution). 3. Simulates the bioleach reactions and acid recycle to provide a net acid balance. 4. Allows for the variation of solution bleeds to the following unit operations: a. Purification iron, copper, cobalt, cadmium removal, b. Bleed treatment (basic zinc sulphate precipitation), c. Preacidification (can either precipitate basic zinc sulphate or not in the process). 5. Tracks gypsum precipitation in all the processes. 6. Provides a complete balance to allow equipment sizing and material of construction selection. The current model uses multiple cells to produce the desired heap effluents. Since the GEOCOAT heap is a continuous process with material being added and removed constantly, the solution tenors from each heap section will be different. Instead of using an average solution tenor, the models flexibility was enhanced by breaking the heap into sectional cells each of which represent a portion of the heap at a different lifecycle. Additionally, the use of these cells splits the PLS solution into a variety of streams that can now be directed to either a PLS pond or an ILS pond, further enhancing the models flexibility and better simulating the true heap design. Based on model results the production heap will have the ability to route solutions to either pond and will be broken down into 5 cells. The use of multiple ponds allows the zinc tenor of the PLS to be maximized. The current model revisions are focusing on utilizing the data from the model to automatically develop capital and operating costs for each scenario. The model will now become intimately linked to the project financials thus allowing better project optimization.

5 L S LS LS GeoBiotics, LLC Intermediate solution 90 Pregnant solution Air H2SO4 Conc Residue Sec 1 Sec 2 Sec 3 Sec 4 Sec 5 Air Recycle H2O Intermediate Pond 92 Pregnant Solution Pond On recycle H2SO4 Concentrate 81 Recycle Raffinate H2SO4 Stripped Organic Rafinate split Recycle Raffinate 59 Acidification 60 Concentrate Feed To Heap EW Zn Metal Adv Electrolyte SX strip Loaded Organic SX load Purified Zn Solution SX Reagents CaO Mg bleed 91 Effluent treatment S/L Mg, Mn Solution Bleed Cu, Co, Cd Free Solution Effl bleed Cu, Co, Cd PPT S/L 70 H2O Pur bleed 68 Fe Free Solution Purification Zn Powder Bleed Fe precipitation 63 S/L FE PPT H2O CaO Basic Zinc Sulphate Figure 1 Zinc Bioleach Model

6 Many metallurgical flowsheets are developed in isolation from the overall project economics. Frequently, process flowsheets are developed and optimized to a local maxima (typically maximum recovery) as a result of the complexity and cost of performing multiple flowsheet scenarios in conjunction with the development of the overall project economics for each case. The use of a comprehensive model allows scenarios to be tested quickly and cost structures developed immediately. Obviously, scaled cost factors are employed which lack the resolution of detailed estimates but they do provide enough detail to quickly determine the best overall process option from a total project standpoint. FLOWSHEET DEVELOPMENT IMPLICATIONS The application of computer models has been shown to allow the rapid analysis of process scenarios. It has been found that linkage to the overall project economics has the ability to maximize not only the process economics but also the project economics. There are other equally important gains to be made from process modeling. The ability to test the validity of various process scenarios allows the amount of laboratory work to be minimized while maximizing the results. This has significant cost advantage both in terms of testwork and sample consumption. It also results in significant time savings. Being able to more rapidly determine the final process goals reduces the time to develop the final process flowsheet. From the standpoint of GeoBiotics, who are in the business of selling biooxidation technologies, reducing the time to market for biooxidation technologies is a significant and vital achievement. Rapid development reduces testing and thus cost, increases speed to market and thus market opportunities. A typical product life cycle is shown in Figure 2. From this figure it is evident that if s 1 and 2 can be reduced (conceptualization and development) then s 3 and 4 (acceptance and market share growth) can proceed sooner, thus, increasing income opportunities. Additionally, the ability of competing technologies to erode market share is greatly decreased, increasing the length of 5. The overall effect is to not only increase revenue streams earlier but to also extend them longer. From the standpoint of users of biooxidation technologies such as Kumba Resources, being able to bring projects on stream earlier and at reduced costs has obvious advantages. CONCLUSIONS One example of how computer modeling has been employed to aid in the development of a new process has been shown. From this example it can be seen that modeling can result in many benefits: the process can be quickly optimized to a global maxima ensuring maximum overall project economics, the costs of testwork can be reduced, valuable samples can be conserved and the project can progress more rapidly. From a product development standpoint, the advantages are obvious: Regardless of the project in question, the goal is the same: to get the best possible product to market ahead of the competition in order to extend the product's life cycle and increase profit. To be successful with both product quality and fast time to market, iterate, iterate, iterate at every stage of the design process. Build strong multidisciplinary teams, and use every rapid prototyping and tooling trick in the book to cut the time it takes to get these iterations into as many hands as possible. Evans, (1997)

7 Market +ve Conceptualization: Technology Development Product Development Market Introduction Market Acceptance and Growth Normal Market Cycle Income Time -ve Introduction Of Competing Technologies Figure 2 Typical Product Life Cycle

8 REFERENCES 1. Harvey, J.T., Holder, N. and Stanek, T., 2002, Thermophilic Bioleaching Of Chalcopyrite Concentrates with GEOCOAT Technology, Proceedings ALTA 2002 Nickel/Cobalt 8 Copper 7, Perth, Australia 2. Harvey, J.T., Holder, N. and Stanek, T., 2002, Thermophilic Bioheap Leaching Of Chalcopyrite Concentrates, European Journal Of Mineral Processing and Environmental Protection, Special Issue Biotreatment and Biosorption, Vol. 2, No Harvey, J.T. and Potter, G., 1999, GEOCOAT At Ashanti Goldfields Obuasi Operations, Proceedings Randol 99, Denver CO 4. Wiseman, DM., 2000, Spreadsheet Based Tools For Practical Flotation Data Analysis, Presented at Flotation 2000, Minerals Engineering International, Adelaide Australia 5. Wills, B.A., 1986, Complex Circuit Mass Balancing - A Simple, Practical, Sensitivity Analysis Method, International Journal of Mineral Processing, Vol. 16, pp Merks, J.W., 1999, Process Simulation With Spreadsheet Software, Minerals & Metallurgical Processing, Vol. 16, No. 2, pp Cleary, W., 2001, Recent Advances In DEM Modelling Of Tumbling Mills, Minerals Engineering, Vol.14, No Bailey, C., Patel, M., Kumar, S., 2001, Computational Modelling-A Key Component In Materials Processing, Trans. Institution Mining & Metallurgy, Sec. C, Vol Abilov, A., and Zeybek, Z., 2000, Use Of Neural Network For Modeling Of Non-Linear Process Integration Technology In Chemical Engineering, Chemical Engineering and Processing, Vol. 39, No Lynch, A.J., and Morrison, R.D., 1999, Simulation In Mineral Processing History, Present Status And Possibilities, Journal South African IMM, Vol. 99, No Harvey, T.J., Afewu, K., and Van Der Merwe, W., 2002, The Development Of The Geobiotics GEOCOAT Biooxidation Technology For The Treatment Of Sphalerite At Kumba Resource s Rosh Pinah Mine, Proceedings BioHydromet 02, Cape Town, South Africa 12. Evans, B., 1997, Accelerating The Product Development Cycle, Medical Device and Diagnostic Industry Magazine, Sept., pp.80