Loadshedding also affected Megawatt Park, Eskom head offices at Sunninghill in Sandton. Picture: FREDDY MAVUNDA
Loadshedding also affected Megawatt Park, Eskom head offices at Sunninghill in Sandton. Picture: FREDDY MAVUNDA

Separating generation and transmission is another disastrously bad policy decision that should be reversed before any further damage is inflicted on Eskom and additional unnecessary costs  are imposed on the South African electricity consumer.

Understanding why this is the case requires some knowledge of how a power supply system operates and how the costs of managing the system are minimised.

Turbines driving electricity generators produce power within a range that varies between 100% and about 60% of rated capacity. In general, their most efficient operating point is about 80% of rated capacity.

The job of men and women operating a power system is to balance supply and demand to ensure that all the generators connected to the system operate within this range at all times. They do this by monitoring the system frequency. This must be held as closely as possible to 50 cycles per second, implying that the rotation of all the turbines must be synchronised and held at 3,000rpm.

When turbines are overloaded they slow down. When they are too lightly loaded they speed up. As long as all the machines are supplying power within their operating ranges the speed of rotation is held within narrow, pre-specified, limits by electro-mechanical devices and the system operators take no action.

An imbalance between supply and demand causes the system frequency to fall below or drift above its limits. When this happens, or when the operators anticipate it will happen in the near future, generating sets held in reserve must be activated or one or more generators supplying the system at that time need to be disconnected. Now the operators have to decide which generators should be brought into operation or which removed from the system.

In making their decision they take a number of considerations into account.

The first is the variable cost of operating each generating set. Those with lower variable costs are favoured over those with higher variable costs. This is known as the “merit order”. But, the effective position of a station in the merit order at any time depends on where it is located within the transmission network and the geographic locations of the centres of power demand at that time. This is because energy losses depend on the distances over which energy has to be transmitted. These losses need to be taken into account in the calculation of variable costs.

The second is the urgency of a requirement for additional capacity and the length of time that it will be needed. If there is plenty of time, then a cheap-to-run but slow-starting coal-fired station will be ordered into service. If additional power is needed immediately, or if needed for a short time only then more expensive gas turbines might be used instead.

The third important consideration is the issue of voltage stability. Unstable conditions occur when the geographic pattern of demands is not matched with appropriately located generation. Again, this might affect which generator is ordered into or removed from service, regardless is its position in the merit order.

These issues are well understood by both power system planners and transmission system planners who have to work in very close co-operation with one another on all new projects.

Separation of transmission from generation would introduce a number of complications. In this case, decisions on when to generate and when to disconnect from the system would be taken by the operators at each individual power station. They would do this on the basis of prices offered in some artificially contrived market for power.

This arrangement has several serious drawbacks. Despite efforts going back nearly three decades, no free market for bulk power has yet been devised that would ensure that existing and future generators would receive incomes that would cover their costs of operation, let alone provide a return on invested capital. This is known as the “missing money problem” and is attributable, ultimately, to the fact that electricity cannot be stored in bulk, and demand for it does not, in general, respond to changes in price.

Power system planners would work in isolation and base their project studies on the objective of optimising the return on investment based on these artificial prices, rather than the economic efficiency of the system as a whole. The result would be a chronic shortage of generating capacity in the medium and long term.

Independent power producers (IPPs) would, presumably, also be given access to this market. But IPPs in SA are hopelessly uncompetitive at present. If they had to operate in free-market conditions all would be bankrupted within months. Special arrangements for them to receive subsidies, either directly or indirectly through the structure of the market or via direct government payments, would be probably be declared unlawful in terms of the Competition Act (restrictive practices).

Each Eskom power station would probably have to be placed into a separate entity in terms of the same act (abuse of a dominant position). How current Eskom liabilities would be allocated in that event is a question with no easy answer.

Participants would be placed in positions where they all had conflicting interests and every reason to withhold vital information from the market and from the system controller. The predictable result would be sharply increased costs caused by a dramatic increase in the complication of managing the system as a whole

There is, therefore, no hope that this arrangement would produce either a practical or an economically efficient outcome. In fact, the fiasco that is the electricity market in the eastern states of Australia was caused by precisely this policy decision. The result has been absurdly expensive prices and an industry that exists in a permanent state of crisis.

Fortunately, there is a better solution for Eskom and for SA.

• Joubert, formerly corporate economic consultant at Eskom and now with Econometrix, is an engineer with a 20 year career at Eskom.