Eskom's dim understanding of the market
One thing most South Africans share is anxiety about the electricity supply this winter. Maintenance of the infrastructure, particularly generators, cannot be delayed any longer. Failures in the operations of other state-owned enterprises (SAA, SABC, Telkom, Metrorail) can be limited (don't fly or fly BA, don't watch TV or watch DStv, don't use telecoms or find another supplier, don't travel or go by taxi), but few of us can avoid the effects of a failure in our electricity supply.
Eskom, with revenue of more than R110-billion in 2012, is South Africa's fourth-biggest company by revenue alone.
It does almost everything in electricity – from generation and transmission to retail and distribution. Such a vertically integrated colossus is increasingly rare in electricity utilities; elsewhere in the world they are disaggregated and subject to competitive pressure. This monopoly presents obvious risks: if Eskom fails, the country risks losing its electricity supply.
The use of markets in the electricity sector is not about being a free-market fundamentalist. Many countries, such as those in Scandinavia, use markets as useful price signals to impose discipline on the allocation of scarce resources. What is clear from Eskom's most recent application for tariff increases is that it has no idea about how markets work or how to use them to its advantage.
Telkom, too, sought to avoid market discipline to its great disadvantage: Telkom's total value is now less than 7% of Vodacom's. Telkom never really understood what services the market wanted and how much consumers would pay for them. Vodacom did.
The result is that Eskom, despite being run independently of government, and regulated by an independent regulator, the National Energy Regulator (Nersa), fumbles from one poor decision to another. It seems to be at the beginning of a "death spiral" it will need to pull out of soon.
Whatever happens to Eskom, we will all be on the hook. So we had better understand what is going on. Some hard decisions on Eskom's long-term future are becoming urgent, particularly since Nersa limited Eskom's tariff hike to 8% year on year, rather than the 16% it wanted.
But the immediate problem is getting through this winter and the next. Eskom has to execute previously delayed maintenance over these months – maintenance ordinarily done in summer. Faults at one of the two 900MW generators at Koeberg and a 1 000MW reduction in power from Cahora Bassa in Mozambique as a result of floods mean that Eskom's reserve margin capacity is well below the recommended 10%.
Eskom has avoided uncontrolled blackouts by running its diesel-driven open-cycle gas turbines at maximum capacity and engaging in demand management by interrupting supply to large customers such as BHP Billiton. It is also paying energy-intensive ferrochrome smelters to shut down – and urging consumers to switch off everything possible.
Thus Eskom pays the enormous cost of diesel fuel as well as paying to shut down the smelters, while losing the electricity-sales revenue needed to meet these costs. Eskom's subsidy of solar water heaters, heat pumps and the like helps to reduce aggregate demand, but doesn't make sense for Eskom as a company.
The problem is this: Eskom's current generation capacity of about 37 000GW is just sufficient to generate enough electricity for South Africa's annual electricity demand of about 210 000GWh; it is insufficient to meet demand at peak periods.
Eskom either sells electricity directly to the end consumers or it is sold through a municipality. Although municipalities have more customers, Eskom sells more electricity. Eskom also sells about 35MWh per customer, municipalities about 21MWh. Eskom's industrial customers are larger and use more electricity than the smaller businesses that get their electricity from municipalities. Conversely, although Eskom and municipalities have a similar number of residential customers, municipalities sell much more electricity to this sector.
Neither Eskom nor the municipalities has any idea about use patterns of individual customers. All they have are the results of electricity metering on electricity already used. And this is a key to dealing with our problems (see graphic). Demand peaks at the end of the working day, between 6pm and 9pm. There is an increase of about 2 000MW in usage in the winter months over summer months. The peak demand for this winter is forecast at 36.8GW – but this is an average for the peak period. The "peak within the peak" could go as high as 38GW.
Most of Eskom's generating capacity comes from coal-fired power stations, which have to run 24 hours a day. The electricity generated by this string of power stations must be consumed as it is generated: storage is limited to pumping water uphill in hydroelectric schemes.
Any business leader looking at the curve shown in the graphic, say a telecommunications or airline executive, would know immediately what to do: the problem is less a generation problem than it is a daytime pricing problem or a yield-management issue. Variable pricing can flatten the curve. Clearly, the source of the two spikes are domestic users and the relatively "high-use" users are a big contributor to the peak spikes. Based on the All Media and Products Survey, those electricity users are likely to be the very people reading this newspaper.
For now, we are confronted with endless appeals for significant behavioural change so we can "beat the peak" by switching off nonessential appliances. These appeals seldom work. Each high-end consumer probably thinks that he or she, individually, does not make much of a difference. This is known as "the tragedy of the commons".
If Eskom was able to price electricity aggressively – charging, say, 30c/kWH (Eskom's base cost of generating electricity) from 11pm to 4am and, say, R2.50/kWH (the cost of generating electricity with diesel generators) during peak times – consumption habits of this category of users would change quickly. The queues of expensive cars outside petrol stations before a petrol price rise show the logic of this strategy.
It would require the installation of smart meters, very few of which are installed now, at selected customers. Eskom has just launched a flexi-tariff option but this is practically unknown to the consumer. The City of Tshwane is reported to be readying itself for the roll-out of its own smart meters, but only late next year. Customers will resist smart metering if they see it as another way for Eskom to gouge more money out of them. Instead, they must see that there is a benefit for them and a way to reduce their electricity bill.
Finding these customers ought not to be hard: if you spent more than R1500 a month on electricity in the winter of 2012 you qualify. A roll-out of 500000 or so smart meters with aggressive variable pricing would reduce the peak consumption of high-end consumers by just enough to see us all through this winter. Such a programme should have preceded all the others.
Eskom still needs to learn that the market can be your friend.
Dirk de Vos runs QED Solutions and consults in the renewable energy and ICT sectors