Design Considerations for Underground RO Treatment

1 Underground RO Potable Water

The supply of potable water to underground mining operations poses unique challenges for each mining site. Employees who work underground on the mining face at the bank stations, technicians and engineers alike require access to potable water.  The National Norms and Standards for Domestic Water and Sanitation (2017) prescribes 15 to 20 litres per person per day, of which the average person will drink 5 litres per day. Some mines initially invested in the infrastructure to distribute potable water from the surface to the underground workings. This requires a significant upfront capital investment, and the static head in the shaft results in extreme pressures prone to leaks that result in water losses at a high cost. Another solution is to manage the logistics behind the distribution of drinking water bowsers at the various points of use. Considering the high level of logistics required for the successful operation of an underground mine, the additional requirements of potable water distribution can hamper production figures. These options are also further complicated because the actual mining takes place far from the mining shaft, which either requires piping or transporting the bowsers to the point of use.

Advances in water treatment technologies make the option of point-of-use water treatment increasingly feasible. Of these technologies, reverse osmosis is emerging as the most cost-effective treatment for potable quality. It has a very high rejection of salts, up to 99%, and does not contribute any additional salt to the water reticulation system. It achieves the separation of salts via the membranes for a pure permeate stream and a high salt load brine stream, and as long as there is a sustainable solution for brine disposal, this technology will be feasible. This has been implemented at various mines in South Africa with various levels of success, and this article will discuss technical details that need consideration when evaluating whether this is a solution at a specific mine.

2 Design Considerations for Underground RO Treatment: Typical Mine Water Characteristics

My water is unique to each location and depends on the bedrock’s geology, the ore body, water management strategies, and mining styles. With this in mind, it is clear that no silver bullet solution can be implemented across the mines. Analysis and understanding of the water quality that needs treatment play a pivotal role in selecting the appropriate treatment technology.

These qualities are merely examples of the references listed. They can vary widely on-site, with some mines having up to 4 or 5 different water sources in a single mine, each with unique characteristics. Successful water management will keep contaminated sources apart from the uncontaminated groundwater and recycle or reuse the contaminated water as far as possible. Where water is exposed to ore with high sulfidic concentrations, there is the potential for oxidation of the sulfide components to sulphates, which generates acid and can result in the liberation of further metals. These streams can have a pH as low as 3.0, with metal concentrations of Fe and Mn exceeding 100 mg/L and SO4 above 2000mg/L.  Highly impacted waters such as these should not be considered for treatment as they require significant pretreatment before being suitable for membrane processes. However, if such a source can be treated through pH correction and settling, available at some deeper gold mines, this water becomes viable for treatment.

2.1 Scaling Risks

Groundwater from dolomitic compartments is characterized by high hardness (Caused by Calcium and Magnesium) and carbonate content, which poses a scaling risk towards membranes. The use of antiscalant allows membranes to achieve high recoveries of hard water by increasing the calcium concentrations above the saturation levels. However, ensuring the correct dosage is applied is critical to extending membrane lifespan. Saturations levels for gypsum are 2 531mg/L and 1 500mg/L for limestone, respectively, in pure water, which can be exceeded with the use of antiscalant.  When concentrations are exceeded above the antiscalant capacity, there is the potential for scale formation to occur on the membrane surface. Precipitation on the membrane surface reduces the effective surface area, resulting in lower permeate flow rates and increases in feed pressure. Depending on the conditions, such as pH and the anions ratio between CO₃^2- or SO₄^2-  scale, it can be calcium carbonate or calcium sulphate.  Calcium carbonate can be readily dissolved using an acid such as citric acid under normal conditions; however, calcium sulphate requires a more rigorous alkaline rinsing to dissolve the scale.

2.2 Monovalent Elements

Typical mining minerals do not contribute large amounts of monovalent ions to the salt load of the mine water. However, some mining blasting methods contribute significant amounts of ammonium and nitrates to the mine water. Where the mine water is recycled, the ammonium values increase to 50 to 100 mg/L, whereas the SANS 241 limit is 1.5mg/L, which requires very high rejection to achieve potable quality. Because of the pH dependence, equilibrium between ammonia gas (NH₃) at high pH and ammonium ions at lower pH may require pH adjustments before treatment. This is because, in the gaseous phase, there will be little to no rejection of the ammonia. However, the RO membranes are very efficient at rejecting the ammonium ions.

2.3 Suspended Solids

Typical mine service water reticulation systems run a recycling system that cools water on the surface, distributes water to the underground workings for use, collects the water and treats it through mechanical settling with pH correcting and pumps it back to the surface for cooling and reuse. Using the cooled water in equipment, machinery and on the mining face results in contamination with super fine solids, which may have sulfidic content and contribute to the acidification of the water. Therefore, the settler installation for underground mines is a critical process, and even with optimal performance, the settler overflow contains between 5-10 mg/L of the superfine solids. These solids will be pumped with the mine water and reported to the feed inlet for any membrane installation. Mechanical filtration will be required to prevent the membranes from blocking. When there is a malfunction of the settlers, it is possible that the solid content can spike up to 500-800mg/L, which can instantly block a cartridge filter installation, so having a robust primary filtration step is vital.

3 Design Considerations for Underground RO Treatment: Membrane Treatment Process

With the requirement of being a point-of-use treatment solution, the entire membrane treatment solution must be skid-based and mobile to allow transportation between mining sites where potable water is needed.

Various alterations and improvements to the treatment process flow diagram provide unique solutions for the specific conditions for which they are designed. Some mining clients require very high mobility and state that the unit must function without an electricity supply. Other clients require a more permanent centralized solution with a focus on rejection and disinfection, which would require booster pumps and continuous UV disinfection.

3.1 Primary Filtration

An underground treatment facility requires a robust primary filtration solution due to the high variability of the suspended solids load in mine water. Some solutions only have cartridge filters installed as a filtration step, with successively smaller micron ratings to combat blockages. Unfortunately, the typical suspended solids slug experienced underground will block the primary cartridge and prevent any treatment. The unit will only be operational once the cartridge has been sourced and replaced, which could increase costs and result in long downtimes.

Active filtration, such as media filtration or Ultra-filtration membranes, is a preferable solution with higher capital costs; however, it will result in longer running times. If a high suspended solids slug is experienced, the active filter unit can switch to backwash mode, rinse the solids to the waste stream and resume operation. Depending on the unit’s requirement, this can be an automated (high cost, high efficiency) or manual (lower cost, operator dependent) process. Having an experienced service provider responsible for monitoring and maintaining these units is crucial to resolving blockages and investigating the potential causes thereof. As the first treatment process for potable water production, any downtime of this step automatically switches off the entire plant; hence, no potable water is available. This is an important factor to consider when evaluating the costs of various suppliers of these treatment units.

3.2 Membrane Filtration

The membrane treatment step is the constraining factor when considering the treatment capacity of mobile underground RO units. Units that are typically advertised as mobile can be handled manually and have a treatment range from 500L/day to 15,000L/day. This is because the RO membranes have a fixed surface area, and therefore, a maximum number of membranes can be fitted on a skid-based treatment plant that is still considered a “Mobile Treatment Plant.”

Reverse Osmosis rejection is a function of the osmotic pressure of the feed water and the additional pressure required to overcome this pressure and produce low salt load permeate water. High concentrations of monovalent ions require higher feed pressure to the RO membranes to maintain high salt rejections. Specifically, in the case of high concentrations of chloride, nitrate or ammonia, an inter-stage booster pump may be required before the RO membranes to increase the pressure to >15 bar. This is an example of power being required to operate the treatment unit, limiting the areas in which the unit can be used.

The membrane treatment process is the most sensitive step in producing potable water, as any changes in the operating conditions can affect the potable water quality, potentially damage the membranes or reduce potable water availability. Monitoring instruments can be used to protect the membrane treatment stage of which the most critical parameters to monitor are:

• Membrane differential pressure
• Permeate flux
• Permeate conductivity

Only one of these monitoring parameters is typically chosen to minimise costs on the underground treatment units, and the relevant instrumentation is installed. Depending on the mine and feedwater conditions, the technology supplier should recommend which is the most useful parameter to monitor. However, just having the monitoring in place means nothing if the measurements are not recorded and interpreted, which again highlights the importance of having an experienced technology supplier on board for the operation and maintenance of these units.

3.3 Product Water Disinfection

RO membrane permeate requires disinfection as a precautionary method to ensure no risk of microbiological contamination of the permeate. The preferred disinfection method is using UV lamps, which destroy any microbiological organisms present and prevent any growth that should enter the system. If the permeate is pumped to a distribution tank, it is suggested the water is circulated through a UV filter to prevent any contamination. Where there are long distribution lines underground with intermittent tanks, UV filters would also be required at these points, as UV does not provide a residual to keep the water disinfected downstream of the UV filter.

Due to the intermittent use of these units, dosing sodium hypochlorite is not a recommended solution. Dosing would require a complicated automated dosing system to ensure there is no overdose when consumption is low, whereas UV cannot overdose. Both these solutions require an energy source, limiting the unit’s flexibility to sites with a power source available.

Where the unit is required to operate without a power source, the high pressure of the service water is used as the driving force for the RO purification. Disinfection is possible without a power source, using Metallic Silver deposited ceramic filters. These filters have proven highly efficient at E.Coli destruction (K.N. Jackson & J.A. Smith, 2018); however, the disinfection efficiency reduces with continuous use. This can be mitigated by regeneration with nanosilver solutions or by replacing the cartridge filter, the recommended precaution to reduce risk. There is no physical indication to determine the disinfection efficiency over time, highlighting the importance of a competent and experienced monitoring and maintenance technology supplier. This technology does not provide a disinfectant residual and presents the same risks of contamination if an extended distribution network is utilized towards the water users.

4 Design Considerations for Underground RO Treatment: Treatment Viability Evaluation

Considering the various process considerations that need to be considered for an underground potable water treatment unit, it should not be seen as a silver bullet that can be implemented without the necessary feed water analysis and underground conditions. It is a unique technology solution that provides new underground potable water supply options as an alternative to historical supply from the surface. Working with an experienced technology expert will enable the correct solution to be designed cost-effectively, ensuring that underground mine workers have a safe, reliable source of potable water. Potable water quality monitoring is a crucial step in ensuring the water meets the required standards, which comes at an added cost that is usually overlooked. Using this technology to treat underground natural ingress water can provide a sustainable source of underground potable water for mining operations.

More about EPCM.

5 References

*B. Hutton, I. Kahan, T. Naidu & P. Gunther, 2009, “Operating and maintenance experience at the Emalahleni water reclamation plant”, International Mine Water Conference 2009.

Department of Water and Sanitation, 2017, “National Norms and Standards for Domestic Water and Sanitation Services”, Government Gazette No. 982, September 2017.

* GR Pretorius, 2019, “Sustainable mining through innovative water management, treatment technologies and remediation strategies to address current and historical environmental liabilities”, WISSYM 2019 4^th International Mining Symposium.

*K.A. Slatter et al., “Water Management in Anglo Platinum process operations: Effects of water quality on process operations”, Anglo Platinum Research and Development.

2018. Jackson & J.A. Smith, 2018, “A new method for the Deposition of Metallic Silver on Porous Ceramic Water Filters”, Journal of Nanotechnology, Volume 2018.

* SJW. Skinner, 2017, “Tracking nitrates sources at a Platinum Mine – Putting the Puzzle together”,  IMWA 2017, Mine Water and Circular Economy.