Abstract: Away back in 1953 few people in the world, let alone South Africa, knew or had heard about membrane desalination, but there was an increasing awareness that electrodialysis had considerable potential for the desalination of brackish water.In South Africa the development of the new gold fields in the northern Orange Free State and the problems posed by the presence of excessive volumes of very saline mine waters stimulated interest in desalination and the CSIR* in collaboration with the mining industry became involved in the development of the electrodialysis process. By 1959 the largest brackish desalination plant in the world had been built and commissioned. South Africans were thus in the forefront of this technology, even to the extent of making the required membranes locally.Our historical review of membrane development and the applications of membrane technology in Southern Africa encompasses both pressure- and voltage-driven processes. Examples of the pressure processes are microfiltration, ultrafiltration and charged membrane ultrafiltration or nanofiltration, and finally reverse osmosis with fixed and dynamically formed membranes. The voltage-drive processes considered are electrodialysis and electrodialysis reversal.

Abstract: The South African mining industry discharges relatively small quantities of mine service water to the environment, but these effluents contribute substantially to the salt load of the receiving waters. The poor quality of service water also has significant cost implications on the mining operations. Of the two main types of mine service water encountered in the gold mining industry, the so-called calcium sulphate scaling types is found in the majority of cases. Preliminary testwork on this type of water using membrane desalination processes revealed that only the seeded reverse osmosis type of process showed promise. To overcome certain process problems and high operating costs with this system, a novel membrane desalination technique incorporating seeded technology, called the SPARRO (slurry precipitation and recycle reverse osmosis) process, was developed. The novel features of the new process included; a lower linear slurry velocity in the membrane tubes, a lower seed slurry concentration, a dual pumping arrangement to a tapered membrane stack, a smaller reactor and a modified seed crystal and brine blow-down system. Evaluation of the SPARRO process and its novel features, over a five-year period, confirmed its technical viability for desalinating calcium sulphate-scaling mine water. The electrical power consumption of the process was approximately half that of previous designs, significantly improving its efficiency. Membrane performance was evaluated and was generally unsatisfactory with both fouling and hydrolysis dominating at times, although operating conditions for the membranes were not always ideal. The precise cause(s) for the membrane degradation was not established, but a mechanism for fouling (based upon the presence of turbidity in the mine water) and a hypothesis fora possible cause of hydrolysis (alluding to the presence of radionuclides in the mine water) were proposed. Product water from the SPARRO process has an estimated gross unit cost (including capital costs) of 383 c/m(3) (1994).