Desalination for South Africa

The drought in the Western Cape has caused thought to be given to how the water supply can be increased. With available surface water being a finite resource there are basically only two possible solutions to effectively augmenting that supply, the repeated recycling of waste water or desalination. To these two can almost be added the more efficient use of water, especially reducing leakages from the system, as this increases supply for productive purposes. The Western Cape has been some way through the process of using water more efficiently, or simply using less water, but it is unlikely that this can be sustained indefinitely as using less water can have immediate economic or social consequences or come at a long term cost, especially in agriculture. Added to which the local residents would not tolerate restrictions indefinitely and that that population is in any case expanding all the time. Given the obviously precarious water availability position of the Western Cape it is surprising that so little progress has been made in installing desalination plants, especially given that these are exported to the Middle East from the Western Cape. The number of plants in the whole of South Africa is only just into double figures compared to 30,000 thousand or so substantial ones (>20,000 m3/day) worldwide.

Mention of desalination usually triggers thoughts of reverse osmosis but that is not the only commercial desalination process. I find Wikipedia a little confusing on this. It credits multi flash distillation on one of its pages as accounting for 60% of the world’s desalination but this appears to come from a very out of date report. The section on desalination on the other hand refers to reverse osmosis (RO) as the leading process in terms of both installed capacity and yearly growth. The picture is of such a plant in Barcelona and, to provide a perspective on size, I assume is the plant in that city which provides 200,000 m3 per day.

There are various other distillation processes apart from multi flash and other methods apart from RO or distillation, but not appearing to account for much of the world’s installed capacity. These range from solar evaporation (with condensation on a cooled surface) to freeze/thaw, in cold countries, where sea water is sprayed onto a cold pad and the subsequent ice meltwater is salt free (just as sea ice has a much lower salt concentration than the sea).

Reverse osmosis appears currently the way to go. Osmosis occurs widely in nature, being the passage of a solvent through a membrane (which is permeable in relation to the solvent but not the solute) such that the concentration of the solute is equal on both sides of the membrane. Thus a potato slice placed in a saline solution will shrink as water leaches into the solution in an effort to equalise the salt concentration. The principle of RO is thus to engineer the reverse situation whereby the membrane separates a body of solvent (pure water) from another body which is solute rich (very salty). The process requires around 4kWh of electrical energy per m3 of desalinated water to force pressurised water through a permeable membrane which does not allow salt to pass. A layman might easily refer to the process as filtering. In fact the process does normally involve passing the water through an initial nanofilter membrane which removes bulky ions such as magnesium and sulphate in addition to bacteria and large particles.

No small cost of the project is involved in sourcing the water from the sea which can either be by covered trench or tunnel. The former is cheaper but depends on the height above sea level of the plant, the need to protect the local environment and facilities, etc. The water must be drawn from beyond the surf zone, with intake and outflow tunnels sometimes extending over a kilometre out to sea, depending on the offshore seabed profile. The process clearly involves drawing a quantity of water from the sea and discharging a much smaller quantity but with the same salt load. This should not be allowed to harm the marine environment. The casual observer of the intake tunnel configuration is often surprised at their size, believing that the input water required for a given output could be obtained via a much smaller diameter tunnel. The snag is that the inflow water speed, in order to avoid an undesirable marine life influx, should not exceed 0.1m/second, and just as importantly 2.5 units of sea water is required to produce 1 unit of desalinated water.

These various considerations dictate the siting of reverse osmosis plants and can encourage co-location with power plants. The advantages can be significant. The power plant will already be drawing water from the sea and its cooling water output is typically 10°C above ambient which conveniently translates to approximately 6% less pressure needed in the RO process. Additionally, the highly saline RO output is partially diluted by the normally much higher output from the thermal plant. Surprisingly I have seen no reference to the highly saline output water being discharged into evaporations pans (on a personal note I also recall salt water being sprayed onto sand roads in the Namib Desert, apparently to stabilise the surface).

As with any other process there can be unforeseen downsides. For example, a survey in Israel found that 62% of school age children had an iodine deficiency (ie below the WHO adequacy range) which is highly correlated to the consumption of iodine depleted desalinated water even if causation has not been conclusively proved. On the other hand claims are made that desalinated water is more healthy than that from traditional dams.

Returning then to the South African situation, the above considerations appear to support RO desalination. The country is one of the more arid, has a long coastline with a number of its larger cities located there and seawater temperatures are cool to warm. Thermal power stations are generally located well inland except for the one nuclear plant but is there not a case to be made for purification of polluted mine water, assuming that is a possibility. With world output at very approximately 100 million m3 per day it does not seem a big ask to make greater inroads into Cape Town’s 500,000 m3 per day requirement. A recent temporary plant only produces 2,000 m3 per day (to be upgraded to 7,000). Costs are still high but technological advances continue and South Africa should take advantage of these. For example Science last month carried a report of how the current smooth nanofiltration membranes can be improved by incorporating Turing structures (named after Alan Turing) featuring complex patterns of blobs and striations, effectively a rougher surface presenting a much greater surface area and in turn a fourfold efficiency increase. Zhang Lin, of Zhejiang University in Hangzou, China, is now following up on his research to see if the same techniques can be applied to the RO membrane itself.

Sources and references: Wikipedia; SA Water Research Commission;;; The Economist.


One thought on “Desalination for South Africa

  1. Your very interesting and informative blog gave me a much better understanding of the financial costs and environmental considerations to be taken into account with an efficient desalination process. Thank you.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s