There were many reasons why we were deemed to be crazy: Firstly, using a freezing technique in the South African climate seemed highly uneconomical and impractical; trying to recover water and salt from brine seemed noble but unfeasible and inventing a new technology that would need to be adopted by the famously conservative mining industry was just plain foolish.
The problem
Industrial brines are currently disposed of in waste ponds, which are both very expensive to build and unsustainable in the long term, as new ponds must be built once the existing ones are filled. They also require large areas of land set aside for the purpose, and risk contaminating groundwater if the lining integrity is compromised. Alternatively, brines can be treated using evaporative crystallisation, which uses heat to transform them into a purified distillate and a mixed salt waste. The disadvantages of this method are the high-energy costs for heating and the mixed composition of the salt waste, which usually necessitates additional treatment or disposal in a landfill.
A novel solution
Eutectic Freeze Crystallisation (EFC) is a novel potential treatment process for these brines, which involves cooling the contaminated solution down to the freezing point, at which point the water crystallises out as ice and the contaminants crystallise out as pure, usable salts. Because the ice is less dense than water, it floats and the salt, being denser, sinks. Thus, EFC is effectively a simultaneous purification and separation process that takes place in one vessel.
The beauty of EFC is that the energy required to separate the water as ice (involving a phase change from liquid to solid) is one-sixth of that required to separate it by evaporation (liquid to gas). What’s more, pure salt can be recovered, because each has its own unique temperature at which it will crystallise out in that particular mixture.
Other advantages of EFC include that it is operated at low temperatures and is, therefore, safer than evaporative crystallisation; for the same reason, simple materials of construction can be used; corrosion at low temperatures is minimal; design for liquid-solid systems is simpler (than for gas/liquid) and no added chemicals are required.
We started this research in our labs in 2007, when the only other work that had been done on the EFC process had been carried out in the Netherlands at the Delft University of Technology. They had progressed very far but their equipment was too complicated to be attractive for local mining applications. We started off by borrowing an “ice making machine” from PAM Refrigeration, a company that manufactures machinery for freezing seawater on board fishing vessels. We broke it in the very first experiment that we tried on a brine solution in the lab!
After that, we gradually developed and designed our own crystallisers and developed processes to conduct feasibility studies on different brines. We have shown that EFC is applicable to mining brines, fracking brines, textile wastes, phenolic brines, gold brines, nickel laterite brines, power station brines and acid mine drainage treatment brines. Our feasibility studies involve analysing a sample of the waste, using a thermodynamic modelling technique that predicts at what temperature the ice will crystallise out of the solution, and at what temperature each of the salts will crystallise out. We can also identify each salt and whether or not it will be feasible to recover it.
For example, we have shown that it is possible to recover both calcium sulphate and sodium sulphate from a brine that had been produced by a reverse osmosis plant treating coal-mining wastewater. Calcium sulphate, more commonly known as gypsum, can be used to make ceiling boards, drywall and plaster but it is also used as a fertiliser in agriculture and as a filler in dentistry and orthopaedic surgery. Sodium sulphate is a key ingredient of soaps and detergents but it is used in the production of paper, glass, textiles and a variety of other materials too.
Industrial driver—no brine waste
Following 12 years of research and development into mine water treatment technologies, the most cost-effective solution was a three-stage reverse osmosis plant capable of producing drinking water with a process efficiency or recovery of > 99%. This was realised when the 30 Ml/day eMalahleni Water Reclamation Plant was constructed by Anglo American and started operation in October 2007.
However, even though the process produced less than 1% of brine, the Life Cycle Cost of constructing and disposing of brine into triple plastic-lined evaporation lagoons as per the National Environment Waste Management Act (NEMWA) was estimated at over R300 million over a 20-year period.
So, the need to find a final solution for the brine was recognised. Evaporative Crystallisation, although a proven technology, is very expensive and therefore, an alternative, more cost-effective solution was needed.
Ten years of research
Over the last 10 years, 10 Masters and PhD students and 36 Honours students have graduated and contributed to the project. We have also published 14 internationally peer-reviewed journal publications on EFC, a patent and 18 international conference presentations. In addition, the research has been supported by local and international funders, with more than R21 million worth of funding.
Collaboration
One of the greatest strengths of this research project has been the collaboration between the university research group (the Crystallization and Precipitation Research Unit) and our industrial partners. This has enabled the project to progress beyond purely academic research and to be implemented in the field.
The Coaltech Research Association was one of the first industrially funded research entities that was willing to make the investment required to scale up the lab-based research into a real-world treatment facility. This resulted in the consulting company, Prentec, being commissioned to design and build a full-scale demonstration plant of 28 000L/day at the Optimum Colliery near Middelburg, which became operational in May 2016.
In April of 2017, Eskom commissioned a 2000L/day pilot plant (designed and built by Proxa) at its Research and Innovation Centre at Rosherville. The intention was to test whether the technology would be suitable for recycling water used in the electricity-generation process, and also for treating acid mine drainage so that it can be used at power stations. Eskom consumes almost 300 billion litres of water per year and EFC could provide a very attractive way of reducing its water footprint. At the same time, it would address one of South Africa’s most pressing environmental problems.
Students and staff from the Crystallisation and Precipitation Research Unit went up for the commissioning and spent a week on the plant. We were involved in troubleshooting and problem solving, and also in identifying other issues for which we need to consider developing new research projects.
However, the most significant development for EFC in South Africa has been the commissioning of the Eutectic Freeze Crystallization plant at Glencore’s Tweefontein Water Treatment Plant. This EFC plant has a treatment capacity of 750 000L/day and commissioning started in quarter 3 of 2017. Continuous operation is expected in quarter 4 of 2017.
Our collaboration in the Tweefontein project has been mostly with the company Prentec, led by Adrian Viljoen and Peter Günther. They have been longstanding collaborators with UCT and EFC research. The commissioning of this plant represents a breakthrough for EFC research and implementation. With its successful trials, we have shown that it is possible to implement EFC on a commercial scale, in a continuous process, with real, multicomponent brines.
More complex than it sounds
Although EFC sounds like a very simple process, it is fairly complicated to design and operate effectively, since each of the elements has its own complexity.
For example, the tendency of ice to form layers of ice scale on the sides of the crystalliser is a very real problem in implementing the technology. One of the major focus areas of our research has been into the causes and mechanisms of ice scaling. We have investigated how the crystalliser hydrodynamics such as crystalliser design, scraper speed and scraper and stirrer design affect ice scaling. We have also investigated ice scaling on a much more fundamental level and have shown that different types of ice (eg. dendritic ice and layer ice) have very different scaling tendencies. One of our more interesting fundamental findings is that water, just before it freezes, spontaneously structures itself whilst still in the liquid phase, and then forms solid ice. The current wisdom is that liquid water is completely unstructured, so this finding challenges the current beliefs.
We have also shown that the heat released when the ice crystallises (which is a challenge for the EFC process, since it tends to melt the newly crystallised ice) is mainly released into the surrounding liquid, and not into the solid, as was previously thought.
Another challenge for us has been moving from batch studies to developing a continuous EFC process that can be scaled from a 2L lab-scale crystalliser to an industrial-scale crystalliser with a volume of 25000L or more.
A continuous EFC process produces two solids simultaneously (ice and salt). However, the rate of ice production is 10 to 20 times more than that of salt. This presents a challenge in the continuous design, as the ice removal system must be able to remove ice from the top of the crystalliser at 10 to 20 times the rate that the salt removal system removes salt at the bottom of the crystalliser. In addition, to effectively remove both ice and salt, the ice and salt crystals need to be large enough to be mechanically separated. Small ice crystals will turn to “slush” and will entrain the mother liquor, leading to impure ice. Small salt crystals will become caught up in the ice crystals due to the stirring motion and will not be able to sink to the bottom of the crystalliser.
Another challenge is that mining brines often contain multiple components, so we needed to understand how to recover multiple components from the various brines. Antiscalants are often used in water treatment processes and these are concentrated into the brines. Therefore, understanding the effect of antiscalants on the EFC process was a very important part of the overall project.
These and many other interesting questions have been solved in the development of the process. Now that there is a functioning full-scale application, we look forward to new and interesting research problems on the road to optimising the process.
What is particularly exciting is that this technology has the potential to become standard in the mining industry. It will enable mines and other industries to change the way that they deal with wastewater. It also has the potential to change the way that we view waste. Instead of considering it as a liability, it has the potential to be a resource.
Professor Alison Lewis
