Is Eskom-controlled load-shedding the best way of balancing electrical supply and demand?

February 28th, 2019, Published in Articles: EE Publishers, Articles: Energize

After a life-time of self-imposed exile, providing consultancy services to utilities and government institutions around the world, and occasionally re-visiting South Africa, I detected a noticeable improvement in mood. People are happy and content; business confidence and international co-operation is up. Had South Africa turned the corner? Anything was better than the shambles of Brexit. Then I walked into the reality of Stage 4 load-shedding. What is going on? Here are a few views of what I’ve learned from around the world that might help the debate.

Elwyn Lloyd Jones

An economist should tell you that a fixed price commodity has no stabilising mechanism. If the price is too low, demand will outstrip supply, and load-shedding is inevitable. If it is too high, supply will outstrip demand and a utility crisis is inevitable. In Spain, with a high take-up of lengthy solar PV contracts, but with no interconnects to export surplus electricity, the utility which pays the feed-in-tariff (FIT) soon found the vulture capitalists circling.

In South Africa, where even wind turbines are considered non-despatchable, could the current crisis be the beginning of the end? If one thinks of the early days of electrification, and how fragile the business model was, it was only once they established a network and were able to exploit the economies of scale that the electricity business took off. But decline is also possible, just look to Nigeria to see how its government pays no subsidies but earns a healthy income stream from duties on imported generators.

I listened with considered approval as Cyril Ramaphosa explained that Eskom was going to be split into Generation, Transmission, and Distribution units, and that this was not to be seen as a pre-cursor to the privatisation or selling-off of Eskom. But why not sell off Generation? Very few countries in the modern world can afford state-owned generation facilities. These facilities come and go as they have a finite life; and need continued investment.

Generation is the ideal business for the private sector. There is risk, and there are profits to be made. If properly handled, competition will squeeze out bad design, bad management, and any form of corruption. International build-and-operate companies have deep pockets and access to a wide range of connected skills. Generation (except for storage and some forms of reserve) should all be sold-off in a fire sale regulated by the Competition Commissioner. This should raise enough money to maintain and expand the network.

Going from an integrated utility to a market-based trading system is risky. Most try an intermediate stage of cost-based trading, but this does not work. A taxi driver does not know exactly how much it costs to run his cab (there are too many variable and unknowns), but he does know how much to charge his fare in order to make a profit, and that’s all he needs to know. Only a “big bang” with a single cut-over to market trading will work. There are a few risks in trying to set up an independent power producer (IPP) business in a market dominated by a state-owned monopoly. Subsidies could undercut and force all IPPs out of business, and who’s to say the political mood might not some-day swing in that direction.

The transmission lines on the other hand, are the jewel in Eskom’s crown. Whoever owns the grid knows the constraints, the capacity, and the demand. Whoever owns the grid has to despatch generation and shed load. As a result, who ever owns the grid determines the price of electricity. They can do this according to any one of a multitude of different policies: minimum price, maximum availability, minimum environmental impact, or maximum energy security.

Note that maximising profit is not on this list as there is a limit as to how much profit could be made or for how long. Owning the Transmission business is surely a licence to print money. Transmission (like our roads), is surely the duty of the state. If the transmission grid is privatised, the bureaucracy needed to regulate it will be almost as large as the bureaucracy needed to run it. The UK made the mistake of privatising its national grid, and very quickly, the Saudi State Pension Fund bought a share – and no one can deny that the Saudi’s know a thing or two about energy and money!

Distribution is like Transmission, but just a different part of the network topology. When the UK split off and privatised its regional electricity companies (RECs), and later the retailer or supply companies, this was to deny an income stream for local authorities, as part of the enduring struggle between local government and central government. (It’s easier to cut off someone’s supply for non-payment of electricity bills than it is to evict someone for non-payment of rates or council tax – and some local authorities were believed to be loading the electricity bills as a result.)

South Africa’s tensions between local and central government are not that bad. In the Netherlands where the same type of privatisation occurred, the best electricity supply company (though most expensive) was bought by a consortium of local authorities, who use the council tax to hedge the electricity income stream and are thus able to offer the longest-term fixed price contracts in the market. This is good for pensioners and those on fixed income.

The second reason why the UK split off and privatised the RECs and supply companies was to introduce competition at the point of sale. Unfortunately, all such competition has proved to be bogus as all electricity is supplied to the same quality and delivery standards of voltage, frequency and availability. Any competition has been on a few cents in the price, and in the style of the invoice. Also, customers who have been fiercely loyal to the dominant known brands, don’t like playing the market game, and don’t like having to deal with the plethora of risk-free pseudo-companies which have sprung up to deal with each tiny corner of the supply business. In the UK, significant RECs and supply companies have been bought out by the electricity generating companies which, having accessed the point of sale, are bypassing the competition and market rules established by the transmission business.

A third (and never mentioned) reason for dividing the CEGB into one transmission company and multiple distribution companies or RECs was that in the 1980s when it happened, computer technology was not up to handling the number of customers, metering points, bills and payments needed for the whole country. But with 21st century computer technology, this challenge is easy; the entire network (transmission and distribution) can be run from a single administrative data centre. Think how credit cards, flight bookings, and social media are each run from single data centres on a world-wide basis, without any need for distributed hierarchies. I believe that transmission and distribution should be consolidated and kept under state ownership. South Africa is in the marvellous position of its network still operating under one single Eskom, and not yet having made the mistakes of separation and privatisation.

So, the main objective of Eskom, the grid operator, is to balance supply with demand. Let’s re-cap the various sorts of signalling the industry uses to do this. Since the mid19th century, this was done via frequency, especially applicable to rotating machinery. The interesting thing about frequency is that the whole country operates in synchronisation to exactly the same rhythm. Even in the late 20th century, we had analogue television and electric clocks and timers which were all dependent on an accurately maintained frequency (cycles per day).

But this world is no more; we have digital TV, and digital clocks and timers. West Germany, which is on the same time zone as the Netherlands, typically loses an hour per week, and nobody thinks anything of it. Besides, what is the economic or environmental justification for catching up the lost time?

Then there is also amperage (or power, assuming a constant voltage). This is measured in order to provide some sort of correction such as routing, or sophisticated load support system. But, for the ordinary householder, this is not of much interest. Except, every household has a main circuit breaker rated in Amps. If one halves the rating of this circuit breaker, will you get a significant reduction in costs? If the whole country, including industry, did it, there would surely be more for all, or a reduction in the country’s capacity needs.

Then there is voltage, the most basic of signalling indicators. This uses losses across the network to signal customers to disconnect when the voltage gets too low. Amp-hour billing instead of Watt-hour billing would encourage such behaviour, but currently available under-voltage relays are just not sensitive enough to be useful for such a task. Maybe such a system should be relegated to the so-called nanogrids.

Instead, consider it as a wake-up call. South Africa’s supply voltage standards have gone from ±6% to ±10%. For a resistive load, every percentage point increase in supply voltage results in approximately 2% increase in required capacity and up to 3% increase in the probability of appliance failure. Circular networks (“ring mains”) are more robust than hub-and-spoke networks. Apparently, only the Japanese are looking at these issues seriously and developing and installing intelligent tap-changing LV transformers. I believe that precision supply is the first step to a more efficient and economical future.

This is where we stand at the moment with a command system such as load-shedding and the lesser demand control (geysers which can be switched-off remotely) which although centrally planned, are regionally applied. Whichever way you look at it, these centralised command systems are a PR disaster. They were a PR disaster 40 years ago when they were first trialled, and they are still a PR disaster today. For a lot of people, the cost of electricity is a minor inconvenience, but the damage that a load-shedding command does to the business or quality of life is huge, maybe immeasurable.

If a consumer has cooling capacity (fridges and freezers), heating capacity (geysers and ovens), or lifting capacity (e.g. pumped water) these cost money and are chosen with minimal margin. If Eskom takes some of this capacity for its own use, we should expect fair compensation; and after running down our reserves for a short period, we should expect them to be topped up very quickly, whatever the cost is to Eskom; but, unfortunately, running down reserves is often a pre-cursor to a full-blown cut in supply. This makes customers very angry.

Finally, there is market-based signalling. The price goes up and down (sometimes even negative) as generation is despatched and taken off-line. The decision to disconnect or shed-load is the customer’s, not Eskom’s. I understand Tesla will be coming to South Africa next year, and wherever they enter a new country, they set-up a network of charge points whose operation is based on the variable price of electricity. Individual internet-enabled devices (IOTs) can be programmed to shut down or reduce load; and smart meters can disconnect when the spot price exceeds a threshold.

However, in the overall scheme of things, the financial cost of a short spike in supply price is lost in the annual averages; it’s only when excessive prices are routinely maintained for hours or more that it starts to become noticeable. Financial signals are too small and too slow to be used for short-term balancing of supply and demand; and one or more of the previous technical signals still have to be used.  In the UK and Europe where time-of-use (TOU) metering has been used for decades, they have made the mistake of having such small differentials between high and low prices that it’s not even worth thinking about.

However, internally, the UKs national grid uses a parameter called “value of lost load” (VLL), which is basically a cap on the wholesale price of electricity. That is, the maximum that the grid operator is allowed to pay any generator under any circumstance. This is effectively the point at which state approved load-shedding is allowed.

I’ve found this an incredibly useful number that can be used for all sorts of PPA tender selection, reserve planning, investment, and operational calculations. It was originally set at about 40 times the average price of electricity, and to my knowledge, it’s only once ever got anywhere near that value, and then only for an hour or so.

So, let’s summarise costs as I see them. Generators have capital costs, stand-by costs, and running or production costs, each of which is different for each generator, and can vary from time-to-time. The capital cost is part of the risk and in my opinion should be amortised and included with the stand-by and production costs, so the network operator only sees the stand-by and production costs of the generators selected. The network operator (transmission and distribution) has various operational and maintenance costs sometimes charged as “connection fees” irrespective of which way the connection is made (producer or consumer); but these in my opinion should all be bundled into the uplift (or mark-up between buying and selling price).

The end-customer is, in my opinion, only interested in the fixed charges, energy charges, availability, and future price stability. Fixed charges should rather be thought of as a “right to switch on” or take load; should be used to pay for the stand-by costs; and should be refunded in the case of involuntary disconnection. Energy charges should be used to pay for the production costs; and the price of availability should be determined by the maximum instantaneous price one is prepared to pay for electricity on an individual or community basis. Future price stability can in my opinion only be achieved by de-monetising your energy supply by technological means such as off-grid and other renewable solutions; for in South Africa, whatever the wiles of the market and other political machinations, the wind will still blow, and the sun will still shine.

The opinions expressed above are private opinions and do not represent the views of the IEEE.

Contact Elwyn Lloyd Jones,


Related Articles

  • Now Media acquires EngineerIT and Energize from EE Publishers
  • Making smart grids smart, makes smart cities smarter
  • Evolving 4IR technologies and digital substations
  • ICT infrastructure to support SA’s utilities of the future
  • New report assesses SA companies and banks’ response to climate risks