Recently, somebody was discussing minibus taxis and how badly they drive.

It was mentioned that, in fact, they drive very well: they pass through the narrowest of gaps between cars, they glide down pavements. They can reverse at high speed through rush hour traffic. They can embark and disembark passengers while the taxi is still moving. They can move across three lanes and stop in less time than it takes you to read this sentence. Then somebody said the hallowed words: “They are doing a job. Nothing that you can do or say will stop them from doing it.”

The same has now occurred in the power generation industry. But first a story of times past. Leading up to the first free elections, in 1994, there was a period of waiting to see what would happen on all fronts, particularly economic. As a result, industry slowed down and the demand for electricity from State-owned utility Eskom was reduced. The Eskom management had changed, as had government. Thus, the blind led the blind into a deal whereby Eskom agreed to sell a significant chunk of its power capability to an aluminium smelting company. We engineers realised that this would leave Eskom without sufficient power generation by 2007. Since it takes about four years to build a power station, we looked out for an announcement of a new power station. None came. Finally, it was whispered: there would be a new power station. But the power station was not built in time and Eskom ran out of power and had to institute load-shedding.

Now government has said that the next round of power agreements with renewable-energy independent power producers will be signed next month. If construction proceeds, it will bring enough wind power into the grid to the point where the grid will become unstable. This is not hard to follow (even for Natal graduates). Say, the power demand is 32 000 MW and 26 000 MW is being supplied by coal and nuclear, 2 000 MW by solar, 4 000 MW by wind and the balance by hydro. Let us say, for example, the wind is as it was for Friday September 8, 2017. The wind was steady until about 15:00, when it dropped to nearly zero at 20:00. So, with this loading, if Friday happens again, Eskom will have to supply the 6 000 MW that is no longer supplied by solar or wind power. This will not happen, so the power system collapses.

But you cannot stop wind power construction. Too many palms have been greased and too may overseas trips paid for. You cannot ban taxis and you cannot stop wind power. What has to happen is that Eskom and the wind power suppliers have to come to an agreement. The limiting conditions are these: (a) wind power disappears suddenly (within a few hours) and unexpectedly; (b) Eskom coal- fired power stations cannot cope with sudden load changes of large magnitude in a short time; (c) Eskom can supply power of some magnitude in a short time using gas turbines, but these cost about four times more to run than a coal-fired power station (Eskom pays for this); and (d) running a gas turbine or coal-fired power station at less than full load is costly and inefficient. Currently, the gas turbines installed are about 1 900 MW.

The possible solutions to the problem of stopping the grid from collapsing are these: (a) Eskom be allowed to accept generation (or not) from wind turbine operators as and when they want it; and (b) refit some Eskom coal-fired power stations to more easily cope with sudden load changes.

The best solution is to build more gas turbine power. Then, by agreement, (a) either the wind turbine operators must give Eskom three hours’ notice if they are about to supply power or to disconnect, or (b) Eskom may start and run gas turbines given sudden loss of wind power, with costs for the wind turbine operator account (aka as ‘the accounts nightmare’). Or something. But we cannot just go on building wind turbines, the power output of which is variable – like an unregulated taxi industry. It just will not work.