South Africa has had bouts of load shedding (localised blackouts) since 2006. They go away to periodically return for longer periods and more severely each time. The latest and most severe bout of load shedding which hit us late last year will be with us for a long time. There is no end in sight.

Load shedding is, of course, the chosen solution to the actual problem. Eskom’s ageing generation plants, which have not been adequately maintained or operated for many, many years, are breaking down. Newly commissioned plants (Medupi and Kusile) are not yet fully commissioned and poor construction practice means these new generators are even more unreliable than existing plants. 

Eskom’s generation division has two problems. One: taking generation capacity out of operation to undertake comprehensive maintenance and perhaps a complete overhaul (scheduled maintenance); and two: taking plants out of commission that break at random points (unscheduled maintenance). Anyone who has owned an old car knows that while scheduled maintenance is important (regular service intervals), it does not prevent unscheduled maintenance. Things, especially when they operate under heat and pressure, become less reliable with age. It’s unavoidable.

In any electricity system at any point in time, supply and demand must be in balance (charging and discharging storage/batteries gives a little latitude). Eskom has a problem with reliable and predictable supply, and load shedding like we have is the chosen route to cut demand. If demand was not cut, the whole system would collapse with consequences beyond our comprehension.

Still, load shedding, as we now have it, is very damaging. We can measure it by the costs of “unserved electricity” or how much not having electricity (especially if it is randomly unavailable) costs the economy. By some estimates used in our electricity planning, which tries to estimate how much reserve (not generally used) capacity should be provisioned, the cost of unserved electricity is R87.50/kWh.

A more recent study by the CSIR puts the figure at between R43.75-R87.50 per kWh not delivered. Energy expert Chris Yelland uses a much smaller figure of R20/kWh and mentions that this figure is in respect of planned, rather than unplanned, load shedding. Even at R20/kWh, with the almost 24-hour Stage 2 load shedding, the economy bleeds a billion rand a day. Over a year this is R365-billion. Given that the government, including local government, extracts around 27% of GDP in taxes, lost government revenue on this R365-billion/annum figure is nearly R100-billion or just under a quarter of Eskom’s total debt.

If load shedding, as we experience it, seems a very crude response to the problem, it is because it is. It is crude because our electricity system as a whole is a very crude and inflexible system.

Eskom, effectively our total electricity system, was built as a vertically integrated monopoly.  It owns almost all of South Africa’s generating capacity. Of the electricity that it generates, half is supplied via its high-voltage transmission grid to large mining and industrial customers, and the other half to distributors, the metros, cities and towns but also its own distribution division, to areas where Eskom, not the local government, is the supplier. The distributors step down the voltage until it connects at 220V to your DB board in your home or place of work.

While generating capacity is relatively new, in excess of what is needed, reliable and can supply electricity cheaply (excluding environmental costs), the system, crude as it is, works well enough. The rest of us don’t think about electricity too much. Instead, we go about developing society and an economy based on the assumption that we can get the electricity we need by paying for it largely on a per kWh basis.

None of these assumptions is valid but we are stuck with a crude and inflexible system that has become increasingly unreliable.

We all know how load shedding works. The electricity distributors have divided their distribution areas into zones. These zones correspond to the main switches in their distribution area. In response to a call (taken on a telephone) following an event from Eskom, various stages of load shedding are called. Each stage corresponds to 1,000MW that must be dropped. 

Eskom has its own arrangements with large industrial and mining customers, but distributors serving cities and towns have a certain set protocol. Simply drop their share of the load that must be dropped. Distributors do this by switching off zones and then switching on and switching other zones off according to a schedule. When your zone is switched off, it is not in the system, when it is switched on, it is back in the system – it’s binary.

Eskom, as a generator, only cares that enough load is dropped. On a purely technical level, it is indifferent as to how it is dropped.  The system works but is causing chaos. The price we pay for our electricity supply is much, much less than the costs we pay for not having electricity. This begs for a different approach.

Almost all of us, whether at home or in our workplaces could easily drop large amounts of our electricity demand at a point in time if we could retain just enough power to keep our essentials, like computers, some lights and especially all the traffic lights going. We would probably accept an arrangement where some percentage of our average demand could be guaranteed at a higher cost per kWh provided the cost of our total electricity consumption over a period remained unchanged. We can be sure of this as many businesses and households have resorted to buying PV systems and standby generators.

So, what could an alternative to the load shedding we now have look like? One should remember that we are not trying to solve the deep structural problems of our energy system. We need to eke out roughly 2,000MW to deal with mostly unscheduled load shedding. In addition, whatever intervention that is chosen should work in the short term but also provide long-term benefits so that the system becomes more flexible and robust.  Finally, each component of a worthwhile plan must stand on its own. If just some of it can be implemented it must still be worthwhile doing.

The obvious first step would be for Minerals and Energy Minister Gwede Mantashe to amend schedule 2 of the Electricity Regulation Act to allow for unlimited self-provision. The reluctance to do so is based on the fear of Eskom losing key customers. This is not a valid fear: any future in mining or industry will require a large utility to service most of this sector’s electricity requirements. Miners want to mine, industrialists want to manufacture/refine. Very few of them want to use their capital to produce electricity for their operations. Further, self-provisioning of even a small part of their requirements provides a useful price signal to Eskom and therefore what and how it should invest as well as the commercial terms (tariff structure) of supplying these sectors.

The next step would be to get the renewable energy procurement programme going again and consider whether the last incomplete round, which went as far as completed bidding and allocation, could continue.

Also, at the national level, a 2013 study by the University of Cape Town found that there is 430MW of co-generation capacity available immediately and within a relatively brief time period, and another 4,000MW could be brought into operation relatively quickly if existing regulatory and pricing hurdles were removed. To put this in perspective, 4,500MW is about what Medupi should (but isn’t able to) produce.

These national-level capacity additions should address the pain of scheduled load shedding.

Below that, in our cities and towns, we should be focusing on-demand response to address unscheduled load shedding. Demand response is a change in the power consumption of an electric utility customer to better match the demand for power with the supply.

This could work in two ways.

First, using standby generators owned by businesses and individuals. In 2009, Eskom produced a study looking at just this and referenced well-established programmes in the UK and France where privately owned standby generators are extensively used. The question then for Eskom, was whether harnessing (and paying for) private standby generators was preferable to having Eskom commissioning new open cycle gas turbines.

In an earlier 2007 study by the UCT’s Energy Research Centre, A Case for Incorporating Standby Generators into the South African Electricity System, the question arose of incorporating standby generators into our electricity system. At the time, it was thought that at least 3,000MW (three load shedding stages) worth of private standby generators existed. It is difficult to get any accurate figures but one can be confident, 11 years on and several severe bouts of load shedding later, that the current capacity is now triple that amount.

Of course, very little of this capacity can be connected or be safely synchronised to the electricity grid, but at the moment standby generators are only used if they are in a zone where load shedding takes place. An option would be to pay the owners of standby generators to run them and thereby reduce the amount of electricity they are drawing from the grid.

The second part of a demand response programme is looking at curtailing demand. Municipal distributors, especially in our larger metros, have to change the way they run their distribution networks. This is not just about addressing load shedding. As more businesses and households install rooftop PVs that are also connected to the grid, the question of how much a particular circuit serving part of a suburb can handle becomes important. If electricity flows within the system, right down to secondary substations, types of electrical substations and functions are monitored and measured and the data made available, then it is clearer where opportunities for optimisation can be made.

Once consumption within zones is better understood, should load shedding still be unavoidable, one could have a better-designed load shedding schedule which might include much smaller non-contiguous zones spread across a city – perhaps even according to whether the secondary substation is serving mostly a residential, mixed-use or commercial area.

Much more preferable though would be to get sufficient electricity consumers to reduce their load in response to a load shedding call. Most of us can and would be able to shift and instantly reduce our consumption of electricity considerably and would gladly do so if we could retain some power – the amount needed to keep our businesses going and certainly to help keep the traffic lights working. If enough of us, encouraged by a sufficiently attractive tariff, could choose a single circuit in our businesses with typically low loads, but have that power guaranteed, there may be sufficient interest.

Knitting all this together is not a trivial matter. Where it works in other countries, private sector aggregators are used. Licensed aggregators, armed with a mandate to pay for the use of an electricity consumer’s standby generator and an attractive tariff to have a participating consumer reduce consumption, would accrue to themselves the ability to reduce electricity demand at any point. Of course, aggregators would need to be able to guarantee the demand reduction, but the technical details of how to do this are already well developed in more sophisticated markets.

If it is of any comfort, all the aforementioned initiatives should already have been explored and implemented. The reluctance to have done so might have something to do with the idea that somehow, given the right kind of management, Eskom will be restored to the entity people thought it once was in the 1990s.

It should be obvious to everyone that Eskom’s monopoly has been deeply damaging to South Africa’s prospects. Whatever the future of Eskom is going to be, it will be very different from its past. Let’s accept that and move on by making our electricity system the flexible and robust system it ought to have been. DM