12 Feb 10 News South Africa

Wind: The cheapest solution for electricity generation in South Africa

Continuation of the theme that wind and other renewable power sources reduce the cost of providing an electricity service.

Firstly let me say that the debate sparked off by my last blog is most welcome. That different opinions exist is known. What is happening now is that they are being expressed in public and the people will be allowed to judge for themselves as to the merits of what each has to say. I will be in South Africa next week and would welcome a public debate with Dr. Kemm, the pro nuclear person.

South Africa makes most of its electricity from coal. The coal fired power stations are located in the area between Johannesburg and Swaziland. From here the power is transported to every load centre in South Africa. There are considerable electrical losses in the system. Losses in alternating current systems are dependent on voltage and distance travelled, and their exact calculation will also take into account the electrical loads along the way. Building a 150MW wind farm in Jeffrey’s Bay area would save more than 11 million kilo-watt hours each year in electrical losses between the coal fired stations and the South Coast. This is enough to power more than 4,000 residential households for 1 year. If the wind energy output is scaled up to 1,000 MW in the South (still less than 3% of the system) then the losses saved rises to 137 million kilo-watt-hours each year which could power more than 50,000 homes.

If we take the fines that will be applied to all coal generation in a world where we have agreement on reducing global warming we see a further value accruing to wind energy. Right now the price of a tonne of CO2 is €13. Every unit of coal fired power puts 0.8kgs of CO2 into the atmosphere. So each unit of wind generated electricity saves R 0.11. If the price of the fine went to €30 then the saving would be R0.25. It is confidently expected that this price for CO2 fines will be at the lower end of the range by the middle of this decade. There is a clear message here for the policy makers in South Africa. Investing in wind is the really effective hedge against fines for coal fired generation.

In our last blog we mentioned water use by coal fired power stations. 1.2 litres for every unit of electricity produced. In a water stressed country like South Africa, how much does fresh water cost? So far, the 167 local authorities charged with supplying water have got coverage of 87% of all households, a considerable achievement given the 65% coverage of 15 years ago. If water is limited, water used in electricity production competes with water that might be supplied to households. A 1,000 MW coal fired power station uses 8,935,000 cubic meters (m3) of water per annum. When it is considered that the FBW per household per month is 6 m3, it can be seen that the development of a coal fired power station competes with households for water. In fact to keep one 1,000 MW coal fired station going would in a constrained water situation deny the basic water quantity to 124,000 houses. How can one put a cost on this?

How will this situation change with global warming? Unfortunately only one way: for the worst.

The contention that wind is completely unpredictable runs contrary to experience elsewhere in the world. It also runs counter to what can be figured out from a study of wind patterns in South Africa.

An analysis of wind speed data taken over 10 years from 8 geographically dispersed locations in South Africa (Figure 1) shows that the power output from a portfolio of generators rarely differs by more 15% over 10 minute intervals. In fact 40% of the time there is no difference in power.

Figure 1:

SAmap_political(stations marked)

In addition, the daily power output from such a portfolio of generation is shown to be not only predictable but also follows the daily electricity load shape (Figure 2). Figures 3 and 4, are based on a 2025 scenario showing that wind can provide base-load generation up to 6,000 MW displacing coal generation and that during the day wind will replace expensive diesel powered peaking generation.

Figure 2:


Figure 3:


Figure 4:


These graphs tell the real story about the prospects for wind energy in South Africa. To summarise them in words is helpful:

· They are based on real data compiled over ten years. This data comes from real sites, and actual power curves from existing wind turbines are used.

· They show that wind power has a strong capacity factor. In environmental terms it reduces the amount of fossil fuels that have to be used to meet customer demand.

· Wind power does not have to be backed up with coal or nuclear.

· Wind power makes its biggest contribution during the day when the demand for electricity is most valuable.

· In a situation where open cycle gas turbines have to be used because of shortages of generation, wind energy costing R 1.25 per unit replaces gas fired electricity at R 2 to R 3 per unit

· These graphs show how, by dispersing wind energy generation across South Africa that the variation in an individual wind farm is smoothed out. The wind is always blowing somewhere.

Operators like predictability. Today, wind power can be forecast accurately to enable supply-demand balancing. Modern forecasting systems rely on advanced numerical forecasting computer models and real time production data. Accuracy of the order of 4-8% of power for a portfolio of wind farms can be achieved with a 24 hour ahead forecast.

This is not new technology. In 2008 wind power contributed 9% of overall power in Spain, a country with limited interconnection to other countries. Notably wind power peaked at 41% of total system power during this period. To manage the system REE, the system operator, requires wind farm operators to provide short-term forecasts. Deviation from these forecasts incurs penalties.

The UK has approximately 4 GW of installed wind capacity. Forecasts must be supplied by operators of wind farms with an installed capacity greater than 50 MW. Penalties are incurred by electricity suppliers for deviation from their overall forecast power for all generation types. This ensures that greater forecasting accuracy benefits the wind farm operator and grid operator. In some countries such as Australia the system operator has implemented their own forecasting system.

Forecasting enables scheduling of other power sources so that expensive peaking plant can be avoided. This ensures that the cheapest possible electricity is always on the grid.

We pointed out before that wind was the cheapest solution for electricity generation in South Africa. We reiterate our contention here. There is no contradiction between wind energy needing a period of support, to help overcome the capital cost, and being the cheapest source of generation. The support lasts only for a period, while the wind goes on being free forever. Once built, a wind farm will be there for 100 or more years and for the last 80 years electricity will be produced at the marginal cost of running a wind farm. That’s really cheap, native to South Africa and, free from pollution.

My next blog will contrast wind generation with nuclear.