AN ECONOMIC AND TECHNICAL EVALUATION OF REGIONAL TREATMENT OPTIONS FOR POINT SOURCE GOLD MINE EFFLUENTS ENTERING THE VAAL BARRAGE CATCHMENT
FINAL REPORT 2000
Report No 800/1/00

EXECUTIVE SUMMARY

This report is divided into five chapters and three appendices. The first chapter gives an introduction to the project and an outline of relevant background information. The second chapter discusses the sequence of events that lead to the development of the proposed desalination strategies, while chapter three gives the methodology adopted for the comparison of the strategies. Chapter four presents the proposed conceptual desalination strategies together with budget costs and direct savings to users resulting from their implementation. In chapter five the specific and general conclusions and recommendations are discussed and recommendations for further research are proposed. The content of the five chapters is summarised below.

INTRODUCTION AND BACKGROUND

Recent studies have indicated that four gold mines (Grootvlei, DRD, WAGM and ERPM) are contributing as much as 26% of the salt load entering the Vaal Barrage by way of their point source discharges. It has further been estimated that this salt load is contained in only 5% of the flow.

In August 1995 an environmental impact assessment commissioned by the Department of Minerals and Energy concluded that the proposed quantity and quality of mining effluent to be discharged from Grootvlei to the Blesbokspruit would have a limited detrimental effect on the relevant receiving water bodies. Some months later a red slime was observed in the Blesbokspruit. This caused wide-spread concern that the Ramsar certified wetland within the Blesbokspruit was under threat from the introduction of polluted mine water. A number of studies were commissioned to assess the extent of the damage and to recommend methods of minimising the impact.

It was the conclusion of the Grootvlei Joint Venture Committee, that three possible future scenarios needed to be considered with regard to whether de-watering should be allowed to continue or not.

The first scenario considered, was that of ceasing the pumping at Grootvlei immediately, which would imply immediate closure of the mine. It was concluded that the implementation of this scenario would be detrimental to the regional economy and would merely defer the problem until the water table had risen sufficiently for decant to take place at Nigel.

The second scenario considered, was that of allowing pumping to continue (and hence mining) while ensuring that the pumped effluent was clarified and the high iron concentration was reduced to acceptable levels prior to discharge. This scenario was considered to be an acceptable short-term solution because it was concluded that the Blesbokspruit wetland would be able to withstand the effects of a saline discharge for approximately two years without undue negative impact.

The third scenario considered was that of allowing pumping and mining to continue but with the proviso that the pumped effluent be desalinated. It was recognised that this option would be costly and hence the second option was implemented immediately and desalination was deferred until a cost-effective method of desalination could be selected.

Against this backdrop and current water treatment pilot plant trials at Grootvlei this project was commissioned, with the following objectives:

1. Propose conceptual strategies for the cost-effective management/treatment of point source mining effluents emanating from Grootvlei, DRD, WAGM (Randfontein No 4 shaft) and ERPM.
2. Confirm the current contribution of these point source discharges to the total salt load entering the Vaal Barrage and ascertain the downstream effects of reducing the salt load upstream.
3. Estimate the costs to users of water from the Vaal river between the Vaal dam and Balkfontein, which can be ascribed to the point source discharges, for purposes of comparison with the costs of management/treatment.

BASIS FOR THE DEVELOPMENT OF STRATEGIES

In order to develop conceptual strategies for the management/treatment of mining effluents, it was necessary to gather sufficient information to determine the extent of the problem and the locations that required the most urgent attention. This was completed in four phases.

In the first phase, estimates of water qualities and quantities were compiled with the assistance of the four mines. In the second phase, the salt load emanating from each of the mines was compared. It was immediately obvious that Grootvlei contributed a disproportionately large load (approximately 80%) compared to the other three mines. Data was also gathered on the water qualities and quantities in the river system between the Vaal dam and the Vaal barrage. These data indicated that the four mines contributed approximately 35% of the total salt load entering the Vaal Barrage and in approximately 6% of the flow. The 9% increase in the salt load and 1% increase in flow, compared to studies prior to the end of 1996, are due to the increased discharge rate from Grootvlei since 1996.

In phase three it was necessary to devise a method of comparing various treatment processes in an equitable manner, which would cater for the different water recoveries and salt rejections of the processes. It was decided that each process would be compared in a side- stream configuration, the product water from which, would be re-blended with the main stream to conform to a pre-determined TDS concentration. Budget costs were then estimated for each process to provide an indication of the most cost-effective processes. Savings to users downstream were estimated to provide an indication of the benefits that would result from implementing of the different treatment processes. However, income derived from the sale of potable water or by-products was not included in these benefits. The annual expenses associated with desalination were calculated to be up to 8 times greater than the savings accruing to downstream users in an average 5-year river flow cycle. In a dry 5-year cycle, annual expenses were found to be up to 6 times greater than savings accruing to downstream users. From this finding it was clear that any income which could be derived from a particular process through the sale of water or by-products, should be included in the cost calculation to offset the annual expenses and obtain a more realistic figure. The "side-stream" method of comparing the processes therefore needed to be modified since the product water was to be sold and no longer re-blended with the main stream.

In phase four estimates were, therefore, made of the quantities of by-products and/or potable water (if applicable) which could be produced by each process and from these, the potential income from the sale of these by-products was estimated. This laid the foundation from which conceptual strategies could be formulated.

METHODOLOGY TO ASSESS THE STRENGTHS AND WEAKNESSES OF DIFFERENT STRATEGIES

In order to assess the strengths and weaknesses of the various strategies it was important to quantify the overall benefits or costs of each strategy.

Direct costs of treatment were estimated by subtracting the annual operating costs (which include repayment of capital) from the annual income derived from the sale of by-products. The effect of a given price in reducing the salt load entering the Vaal Barrage was then estimated by use of a water and salt balance model (Aquabat®). The model was set up in order to estimate the TDS concentrations at a number of different abstraction points along the Vaal river. A change in the salt load upstream could then be modelled to produce an estimate of the change in salt load at different abstraction points.

The estimated TDS concentrations at the major abstraction points allowed for user savings to be estimated using a salinity costing model specifically adapted for the purpose. According to the type of abstractor, (Eskom, Rand Water etc.) a reduction in salt load was thus converted into an annual monetary saving. The report provides details on how the treatment costs (e.g. rate of real interest rate and payback period used for the calculation of annual capital repayment) were estimated and how the Aquabat and salinity cost models were set up and used.

This methodology enabled not only a comparison of the costs of different strategies but just as importantly, a comparison of the effect of different treatment strategies.

The comparison generates three main outcomes:

1. Treatment is profitable and even without addition to the savings to users downstream; an overall benefit arises.
2. Treatment is not profitable but the savings to users downstream offset this annual loss and yield an overall benefit.
3. Treatment is not profitable and the savings to users downstream do not offset this annual loss and hence there is no overall benefit.

The development and assessment of the strategies were based on the following main assumptions:

1. Desalination processes that have been tested on a pilot scale are technically feasible for use as part of a water management strategy in this study.
2. Specific prices of by-products were assumed. Changes in these prices will have an affect on the costs determined in this study. Uncertainty exists regarding the quality of the elemental sulphur produced by the Biosulphate process it is thus not included as a by-product.
3. The capital required for initial construction of a process is paid back over a period of 20 years at a real interest rate of 8%.
4. Enhanced spray evaporation was assumed as the brine treatment option for those processes that do not include brine treatment.
5. Potable water is sold at 200 c/m³.
6. The additional levy of 90 c/m³ on subsurface abstraction would be varied and was thus not factored into the economic calculations.
7. The mix of activities within the various economic sectors is similar for the Middle Vaal- and the Vaal Barrage catchment (for which the salinity costing model was developed).
8. The ingress of water from Blesbokspruit into the Grootvlei mine workings is in the order of 50% of the volume pumped out of Grootvlei. This assumption is based on experience at the mine and should be investigated properly before any strategies that rely on this assumption are implemented.

The study and models developed for this study have the following limitations:

1. The costs determined and used in the study are budget costs. These costs are furthermore dependent on the quantity and quality of the water to be used.
2. The study focuses on the Vaal Barrage catchment. Additional economic benefits in the Middle Vaal, due to lower salinity in the Vaal River, were not included.
3. Future development and changes in catchment management that will affect the Vaal Barrage catchment were not considered in the study.
4. The minimum accuracy of the water and salt balance was set as 70%. The actual balances are in the range of 80% -90% accuracy.
5. The water and salt balance model is a static model calibrated for a 5-year moving average. The model is more accurate for average to dry 5-year periods and should not be used for very wet 5-year periods, unless it is re-calibrated for high rainfall scenarios.
6. The user cost model has a margin of error, e.g. an overestimation of 35% of salinity will result in costs being overestimated by 44% (see Table 3.2).

CONCEPTUAL STRATEGIES FOR REDUCING THE SALT LOAD TO THE VAAL BARRAGE

The eight conceptual strategies that were selected, are aimed at providing practical alternatives for reducing the salt load emanating from the four mines. Some of these strategies can be used in combination with others and cover alternatives that do not necessarily involve desalination.

Strategy 1 involves the desalination of 130 Ml/d of effluent (at a TDS of 4 156 mg/l) discharged from Grootvlei Mine. The costs and benefits of eight different desalination processes are given, namely: AQUA-K, ASTROP, Biosulphate, EDR, GYP-CIX, RO, SAVMIN and SPARRO.

For strategy 2 costs and benefits are estimated for the combined treatment of the l6 Ml/d of effluent produced at East Rand Propriety Mines (ERPM) and the 130 Ml/d produced at Grootvlei. The ERPM effluent is piped to Grootvlei and the capital and operating costs of the pipeline are included.

Strategy 3 deals with the combined treatment of the 7 Ml/d of effluent produced at Durban Roodepoort Deep Mine (DRD) with the 40 Ml/d of effluent produced at Western Areas gold Mine (WAGM). The DRD effluent is assumed to be piped to WAGM.

In keeping with the idea of regional treatment, strategy 4 was set up to provide for the combined treatment of the effluents from all four mines at Grootvlei. The costs of reticulating the effluents from DRD, WAGM and ERPM are included.

Strategy 5 explores the concept of "brine bleeding". It is sometimes possible to take advantage of the higher and low salinity flows experienced in summer for disposal of a portion of the brine produced by a desalination plant. The applicability of this method of reducing treatment costs is assessed in the context of the effluent discharged from Grootvlei to the Blesbokspruit.

It has been suggested that up to 50% of the water currently being de-watered from Grootvlei originates from the Blesbokspruit. Strategy 6 explores the possibility of constructing a 65 Ml/d open by-pass channel parallel to the Blesbokspruit to reduce the ingress of water into the Grootvlei workings. It is assumed that the construction of the channel reduces the quantity of water entering Grootvlei by 50% and hence only 65 out of a possible 130 Ml/d is required to be de-watered. The 65 Ml/d which is de-watered is assumed to be desalinated and the channel is designed to convey 65 Ml/d during summer and no flow during winter.

Strategy 7 also involves a by-pass channel but in this case it is designed to convey 130 Ml/d in summer (of which 65 Ml/d is de-watered from Grootvlei and the remaining 65 Ml/d originates from above Grootvlei in the Blesbokspruit) and 65 Ml/d in winter (this is de- watered from Grootvlei). This strategy is essentially the same as strategy six, although desalination is not assumed to take place.

For strategy 8 the flow being de-watered from Grootvlei be pumped to the Kafferspruit, which drains northwards to the Klipspruit and the Bronkhorstspruit dam. The capital and operating costs of the pipeline are estimated for flows of 35, 65 and 130 Ml/d. The savings to users of water from the Vaal system are included but the additional costs to users in the Bronkhorstspruit system are not included. The total costs and benefits of this strategy can therefore not be compared directly with the other seven strategies.

The strategies listed above were compared for an average 5-year hydrological cycle (calculated over the last 100 years) and a dry 5-year hydrological cycle. The reason for this was that a dry 5-year cycle produces a worst case scenario since the effect of dilution is minimal.

CONCLUSIONS, RECOMMENDATIONS AND FURTHER RESEARCH

The costs that were determined for the various strategies and the ranking of the strategies are presented in Tables E.1 -E.3, and summarised below.

It was concluded that during an average 5-year cycle the most promising strategy in terms of total benefit was strategy 7 (installation of 130 Ml/d channel with no desalination). The next most promising strategy was strategy 6 (installation of 65 Ml/d channel with desalination), followed by strategy 3 (desalination of combined flow from DRD and WAGM), 1 (desalination of Grootvlei only) and 2(desalination of combined flow from ERPM and Grootvlei). User savings were only estimated for abstractions below the Vaal dam up to and including the Vaal Barrage. The reason for this is that the inclusion or exclusion of user savings below the Barrage would not change the rank already established but would in all cases increase the benefits. In order to remain conservative therefore, the user savings below the Barrage were not included.

Table E.l: Summary of Total Costs

* Strategy 8 is not taken into consideration, as it necessitates an additional study, which is outside the scope of this project

Table E.2: Strategies ranked according to Total cost/benefit for Average 5 year cycle

Table E.3: Strategies ranked according to Total cost/benefit for Dry 5-year cycle

For the dry 5-year cycle the most promising strategy was strategy 6 (installation of 65 Ml/d channel with desalination). The next best was strategy 7 (installation of 130 Ml/d channel with no desalination) followed by strategies 1 (desalination of Grootvlei only), 2 (desalination of combined flow from ERPM and Grootvlei) and 3 (desalination of combined flow from DRD and W AGM).

Apart from the specific conclusions given above a number of general conclusions were also made and these are listed in point form below.

1. The desalination of typical calcium sulphate scaling mining effluents with a TDS exceeding 4 000 mg/l indicates that in order to minimise annual costs, the production and sale of high value by-products is essential (potable water can be considered to be a by-product).
2. The production and sale of potable water can usually offset the costs of treating a calcium sulphate scaling effluent with a TDS of below 700 mg/l.
3. The cost of treating the 130 MI/d effluent from Grootvlei is approximately equivalent to reticulating that same flow a distance of between 70 and 380 km depending on the process selected. With further process refinements and innovative financing, these figures can be reduced considerably.
4. The "treatment" of brine is usually expensive whatever method is used and hence the process philosophy adopted should as far as possible incorporate brine treatment into the main process in a synergistic manner to minimise transfer of the problem.
5. The reduction of ingress into a mine is an extremely cost-effective method of both minimising de-watering and minimising the quantity of water that requires desalination. If desalination is self-sustainable with regard to funding then prevention of ingress becomes less important. In the case of the Blesbokspruit it is recommended that a project be set up to determine the quantity of seepage which ultimately enters the Grootvlei workings.
6. In order to plan any desalination strategy it is essential to continually gather relevant information regarding water flows and qualities. The procurement of a complete data set is usually more time consuming than the implementation of the final solution.
7. The highest cost burden of combating salinity is currently being carried by the household sector and not by industry as might be expected.
8. The "user savings" are economic costs: they take into account the costs borne by those other than the polluter. As such, they can be used to form a basis for internalising externalities. Some economic instruments that could be of use in this regard, are discussed.
9. The "polluter pays principle" is based on the internalisation of externalities and therefore is central to the equitable resolution of pollution costs currently being borne by the end user .

RECOMMENDATIONS

From the conclusions and recommendations of this study the following topics for further research are proposed:

1. An evaluation of the feasibility of bypassing the Blesbokspruit wetland with a channel to minimise the ingress of water into the Grootvlei mine workings. This study will also have to determine the quantity, quality and location of the water ingress into the Grootvlei mine workings from the Blesbokspruit.
2. The future water management of the Blesbokspruit should be defined to optimise and integrate the implementation of strategies from this study and proposed future development within the Blesbokspruit catchment.
3. A study within the Vaal Barrage catchment is required to confirm and refine the salinity cost model used in this study.
4. The most feasible economic strategy to encourage the internalisation of externalities should be investigated before it is implemented.