The Case for Karpowership in South Africa

Shining a new light on the deal

Since 2008, South Africa has been grappling with a progressively worsening electricity crisis. The country's state-owned power utility, Eskom, is debt ridden and struggling to meet demand, with rolling blackouts having become a regular occurrence. In an effort to address the crisis, the government has proposed using Karpowership to generate electricity.

Karpowership is a Turkish based company that uses powerships- floating power plants that use natural gas to generate electricity. They are a quick and easy way to add power to the grid. See below for a simple illustration:

The company won a bid to provide 1,220 megawatts to South Africa. Eskom has said that they need an additional 4,000 to 6,000 megawatts to end load shedding. So while the additional power from Karpowerships is not promised to end load shedding, it will reduce it by 20-30%. The public sentiment has mostly been negative. Interestingly, despite the frequent complaints about load shedding and its effects, there has been significant opposition to this potential solution. The main concerns raised have been 1. The cost and 2. environmental blocks. 

The apparent cost of the 20 year deal is R220 billion, which is dated in 2021. The Organisation Undoing Tax Abuse (OUTA) has recently made claims that it will cost R500 billion over the 20 years. The reasons given for their much higher estimate include a weakening rand and the price of gas increasing due to the Ukraine war. I have done some rough calculations below using July 2021 and May 2023 data:

It's unclear what numbers OUTA used to arrive at a 127% increase, as the figures used here do not align with such a projection.

I then wanted to contextualize this “cost” to South Africa. Instead of looking at the price per KWh that consumers would pay, I started by looking at how much load shedding is currently costing South Africa. Bloomberg reported that the South African Reserve Bank estimates load shedding costs the economy R899 million per day. This is based on the revised economic growth of the country given the impact of load shedding.

Cost of Load Shedding vs Cost of Karpowership Deal

If load shedding persists for 250 days a year for the next 20 years, the reserve bank estimates that it will cost South Africa R4.5 trillion. If the Karpowership deal is indeed able to reduce load shedding by 20-30%, it looks promising to enter the agreement even at the overinflated cost that OUTA is suggesting (i.e 20% of the cost of load shedding is greater than the cost of the Karpowership contract). Despite active media coverage on the cost of load shedding and the Karpowership deal, a direct cost comparison seems to be missing, even though such an analysis is relatively straightforward.

Instead of focusing on the cost per KWh to consumers, this analysis considers the broader economic impact, given the universal understanding of supply and demand dynamics. At the start of 2023 there was outrage over the NERSA suggested price increase of 18.65%. People couldn’t understand that electricity prices were going to increase even though they couldn't buy any. However, this makes perfect sense: when the supply of a good or service is limited, the price of the good or service will tend to increase. Even if the price consumers pay increases by 100%, it will at least give consumers the chance to decide to reduce their usage to save money. They will also have the ability to compare the cost of electricity from Eskom and the cost if they had their own alternative solution.

The next aspect that has been raised in this discussion is around the environment, with activists and the Department of Forestry, Fisheries, and the Environment opposing the ships. The following reasons were quoted:

“Environmental activists lodged complaints with the Department of Forestry, Fisheries, and the Environment (DFFE) about the power ships’ impact on fishing, local ecosystems, and greenhouse gas emissions.”

This opposition was particularly perplexing, considering that the power generated would come from gas.

Almost 70% of all electricity consumed in South Africa is done so by burning coal, with another c18% from burning oil. The EIA claims the following around CO2 emissions: 

“Burning natural gas for energy results in fewer emissions of nearly all types of air pollutants and carbon dioxide (CO2) than burning coal or petroleum products to produce an equal amount of energy. About 117 pounds of CO2 are produced per million British thermal units (MMBtu) equivalent of natural gas compared with more than 200 pounds of CO2 per MMBtu of coal and more than 160 pounds per MMBtu of distillate fuel oil.”

In other words, gas is around 42% and 27% cleaner than coal and oil respectively. So the environmental activists are against a cleaner energy source.

I know the ESG debate can be extremely nuanced, with multiple trade offs over varying timelines but at the end of the day people want electricity. Those that can afford it will install alternative renewable solutions or burn oil in generators, with the gap between the haves and have nots widening. If faced with this decision I would most likely conclude that a near immediate 20-30% decrease in load shedding, combined with a cleaner energy source, outweighs the negatives around fishing concerns and the delay of the long term switch to renewable energy.

That being said, it is clearly not the long term solution (and doesn't even fully solve load shedding). It does however buy the government some more time to restore current plants, such as Kusile, and find a longer term solution. It's positive to me that a temporary solution exists, and it happens to be cleaner too. This is not the only immediate solution out there either, but with 6,000 megawatts needed to end load shedding we will need more than one.

You might notice that I have not mentioned corruption or the ANC in the above. This is because I don’t see it as a relevant topic, i.e. it will occur either way. Any alternative to increasing electricity supply will be subject to some sort of corruption.