How much water will we need to keep the lights on?

A recently completed study by the Water Research Commission (WRC) and the University of Pretoria (UP) into the future water needs of the power generation sector highlights how a lack of technical knowledge is causing the country to delve in the dark.

That there are challenges in South Africa to meet the country’s electricity demand is no secret. Currently, our main electricity supplier, Eskom, is intermittently applying load-shedding in an effort to balance electricity supply and demand. In a nutshell, load-shedding means that Eskom is rotating and shedding the load in a planned and controlled manner to avoid the collapse of the electricity supply grid, which would have disastrous outcomes. Should this happen, it could take up to a month to restore power to the entire country.

Yet, while much attention is given to the availability of electricity, one of the most integral resources necessary to generate it is not being talked about – water. This is an important conversation to have, especially considering how scarce a resource water naturally is in South Africa.

While South Africa’s electricity generation activities only account for 6% to 8% of the country’s total freshwater resources used, power generation plants are mostly located within moderately to severely strained water management areas – close to coal reserves rather than water supplies. Seeing as conservative estimates would have it that South Africa is staring a 234 gigalitres shortfall of water in the face by 2025, water will have to be used more efficiently across all sectors in order to preserve the country’s water security.

“The field of water supply shortage is a hidden crisis. We already have shortages of water in areas of South Africa, but we have not seen many studies investigating the expected demand for, and supply of water,” says Prof Anastassios Pouris, director of the Institute for Technological
Innovation at UP. Pouris lead the WRC study that aimed to make a long-term forecast of water usage for electricity generation in South Africa until 2030. “It’s been forecast that by 2030 water demand nationally and globally will exceed supply,” says WRC Research Manager, Dr Jo Burgess, explaining the rationale behind the study. “We need to close the gap before then.”

The study team used their findings to propose water saving measures for power generation. In order to do this, they first aimed to forecast water usage patterns associated with coal-based electricity generation as well as water consumption factors. Secondly, they assessed scenarios of water usage patterns based on cooling technology and power plant type, in particular wet-cooled and dry-cooled power plants. Necessary historical data for the project was readily supplied by Eskom. While the study found that there are ways in which Eskom’s water use can be curbed substantially in future, the application of promising technologies are scuppered by a lack of research.

Eskom’s power supply setup
As much as 90% of the electricity that Eskom generates is fired by coal. There are 13 coal power plants spread across the country, of which ten are base-load power plants. These operate during normal demand. In order to meet growing energy demands, three power plants that were previously mothballed were brought back to the grid in recent years. Camden, Komati and Grootvlei, the so-called return-to-service (RTS) power plants, are used in conjunction with the base-load power stations during times of peak demand. Water is used for a number of processes during power generation, such as operating flow gas desulphurisation devices, ash handling, wastewater treatment and wash water. The most water is, however, used for cooling the thermos-electric power plants and this has, together with the choice of fuel technology, by far the biggest impact on the overall water supply needed as well as the ecological health of surface water bodies where it is subtracted from.

Of Eskom’s ten base load plants, eight use wet recirculation cooling technologies, which applies cooling water through condenser tubes with steam on the outside. The temperature variations between the water and steam cause condensation. The warm water in the condenser is collected in the cooling tower where an upward draft of air removes the heat. The cooled water is then recirculated to the condenser.

A major drawback of this technique is the water lost through evaporation when the warm cooling water comes in contact directly with air. Two of the base load plants (Kendal and Majuba) use dry cooling approaches. The dry cooling technology uses air instead of water in the heat exchange mechanism to cool down high temperature steam.

The RTS plants use wet recirculation techniques while Grootvlei also incorporate dry cooling technologies. These plants are water guzzlers, but their impact is exacerbated by their locations in the severely constrained water management areas of the Olifants and the Inkomati. While they satisfy energy requirements they adversely affect water needs.

The yet-to-be-completed Medupi and Kusile power plants will use direct-dry technology, drastically reducing water use.

Yet, all technologies have both advantages and disadvantages. For example, while dry cooling systems are more water-efficient, they are more expensive to build and the plant is less efficient. In the case of Medupi and Kusile however, they will be made more efficient from a system thermal efficiency perspective, because of the use of supercritical boilers.

“In summary,” said Pouris, “The need for water use forecasting in the topic of energy-water nexus, is essential in South Africa.”