South Africa once again faces a familiar dilemma. If IRP modelling does not allow for nuclear, politicians can’t permit either the gas or renewable energy processes to proceed, as their proliferation exposes the poor prospects of nuclear from a cost perspective, while it erodes the need for it. The result: South Africa is prevented from removing the electricity constraints we face. By DIRK DE VOS.
It can’t have escaped anyone’s notice that the future of our electricity supply system is once again the focus of strenuous debate, with Eskom now criticising the role of Independent Power Producers. Part of the problem is that the planning function, under the Department of Energy, has failed to produce an updated integrated development plan (IRP) and the previous one, in 2013, was never adopted. The one before that is obviously completely out of date.
It is not that work on the new IRP isn’t being done, it is just that there is no scenario that allows for any reasonably costed nuclear build plan to proceed and the modelling tools used is simply not open to political “interpretation”. Yet, one cannot simply hardcode a nuclear powered future into the model and retain the model’s integrity. It is not as though this can go on for much longer. There is 30GW of coal-fired power to replace in the next 25-30 years. The jam is this: if IRP modelling does not allow for nuclear, then our politicians can’t permit either the gas or renewable energy processes to proceed as their proliferation exposes the poor prospects of nuclear from a cost perspective while it erodes the need for it. The result: South Africa is prevented from removing the electricity constraints we face.
The future of our electricity supply is among the most important decisions the country has to make and in a democracy, perhaps all of us need to know what our options are and how the system works. This is hard to do: One of the striking things about the broader energy sector is how preconceived ideas drive people into different camps. On one level there are the strong divides between adherents of different political and economic systems, those who believe in strong centralised systems that require state ownership and those that believe in private ownership and open, competitive markets wherever possible. On another level, there are the adherents of different technologies: there are those convinced that the future is all about renewable energy, then there are those who are convinced that fossil fuels will be at the centre of our energy systems as they have been for over a century – after all, what about securing the baseload? Then, of course, there are the nuclear power types, convinced that their technology is a viable solution to our future electricity requirements.
There is a scattering of the full political-economic spectrum within the different generation technologies and it can all become rather tangled. Part of the problem is that each camp is absolutely convinced of their own merits and know little about the nature of the merits of the other.
One also must take into account the electricity system we have. Any electricity generated has to be delivered into the South African grid and that grid is owned and operated by Eskom. Eskom inevitably has its own agenda and interests that do not necessarily align with those of South Africa as a whole. This lack of alignment is also known as the Principal-Agent problem.
For a sensible debate, one needs a basic technical understanding of how any large electricity system works:
- Electrical power in a simple direct current circuit is measured in Watts and derived by multiplying the Volts with the current or Amps (W=V x A). For the same power, you can increase Volts by decreasing Amps. In practice, there are losses with a step up or step down in Voltage;
- For the most part, electricity generated has to be consumed at the same time. Generation and load have to be balanced. Storage, mostly in the form of pumping water up a mountain when demand is low and releasing the water through generators when demand is high, allows for some mismatch between generation and load at a cost of some losses;
- The whole system uses alternating current (AC) which means that the electrical power calculation above has an additional multiple, known as the power factor (any number from 0 to 1). As such AC power is measured using a different measure kVA or Kilovolt Amps. For maximum efficiency, the power factor has to be as close to 1 as possible. To do this, the voltage and current need to be as close as possible in phase (i.e. keeping it above 0.85);
- There are three main parts of the whole system: Generation, Transmission and Distribution. Generators connecting to the Transmission System need to do so at very high voltages as this is more efficient for transporting electricity over long distances – still there are unavoidable losses over long distances via resistance in the wires.
- Different customers connect at distribution grid level by which time the voltage is progressively stepped down until the 220V level used at the residential level – each step down results in losses. There is another important point – in times of high demand, when more electricity is flowing through the wires, there are greater losses as a share of total electricity than at periods of low demand.
Now, looking at different generation technologies, one can make the following basic points:
- Renewables (wind and solar PV): Over a period of, say, a year, their output is highly predictable but over much shorter time scales (minutes, days, weeks) they are less predictable and so they are called variable sources of power. The marginal cost of electricity from these facilities is almost zero – no additional costs are incurred when they generate electricity. It is simply a function of whether the sun is out or the wind blowing. The cost of electricity derived from them is baked into the cost of building them.
- So-called thermal technologies: These include nuclear, coal and concentrated solar (CSP) with storage. These technologies work by heating water to steam and using the steam to turn a turbine which can then generate electricity. For investment planning purposes on a per kWh basis: about 75% of the costs of electricity produced by a nuclear power plant is derived from the cost to build it and less than 10% for the fuel. For coal-fired power stations, capital costs represent roughly 40-60% of the kWh’s generated and the fuel between 25%-40% depending on one’s view of carbon taxes. The downside of all these thermal technologies is that they are inflexible. If you set them up so that their output averages daytime power requirements, they struggle to ramp up to peak demand times and produce excess in the late evenings and weekends, not all of which can be stored.
- Gas-fired power: Like thermal technologies, they also work via a spinning turbine but do so directly, not via water/steam. There are two types of gas-fired turbines. An Open Cycle Gas Turbine (OCGT) and a Combined Cycle Gas Turbine (CCGT). The CCGT is more efficient because it traps the waste heat to extract more power from the gas supply but is less flexible, operating more like a thermal power station. The OCGT are highly flexible and are designed to be so. They are cheap to build but depending on the price of gas, 60% + of the cost of electricity is in the cost of the fuel itself.
- Storage: Energy storage is not a source of energy as such. At present, over 99% of all electricity storage today is in the form of hydroelectricity via pumped storage. In low demand periods, water is pumped up a hill and in peak periods the water is released to drive turbines to produce electricity. Pumped storage is also reasonably efficient at as much as 80% efficiency. Pumped storage is limited to places where there is available water and a suitable slope.
However, the whole field is undergoing rapid change. CSP with storage (via heating a salt solution) is one technology but there are multiple developments elsewhere including Lithium Ion batteries that have come to dominate electric vehicles. Other projects are in the area of compressed air, flow batteries, synthetic gas/hydrogen, giant flywheels that can spin at enormous speeds, super-capacitors and others. Other than pumped storage, current storage technologies are still very expensive but that is rapidly changing. Between 1991 and 2005, the cost of storage with Lithium ion batteries fell by a factor of 11 and fell by another 40% between 2010 and 2013. Based on learning rates, the consensus is that batteries need to fall by another factor of 10 before they, along with renewables, could displace all fossil fuels on price. On current trends, that could happen within 20 years, well within the investment horizon of conventional fossil fuel power plants and certainly within five years of the likely first commercial operating date of a go-ahead of any nuclear power plant investment made today. Of course, this is speculation, but it can’t simply be ignored in future planning.
Electricity supplied into an AC system via turbines (spinning at speeds between 2,000 and 7,000 rpm) differ in one important respect from renewable power delivered from solar PV and wind. In addition to the power fed into the grid, the spinning turbine effect helps keep the whole system in phase, maintain high power factors so increased system efficiency.
Unique features of South Africa’s transmission and distribution grid
In comparison to most other countries that generate as much electricity as South Africa, a few things are striking. First, the whole system is run by one entity, Eskom, a state-owned vertically integrated monopoly that is responsible for 95% of total generation, runs the transmission grid and has a big part to play in the distribution grids. On a technical level, there are considerable inefficiencies. Power, from mostly remote and very large coal-fired power stations, is supplied into a single national grid that must be one of the largest single system in terms of geographical distances anywhere. The only non-Eskom part of the system are the municipal distribution grids but all of these are tightly integrated with Eskom and Eskom is their only supplier.
There are historical reasons for all this. Eskom was originally set up to provide huge amounts of electricity to the mining and heavy industry (metallurgy) sectors from the 1930s on (the so-called mining-industrial complex). Back then and until the late 1950s, our cities were responsible for generating their own electricity but this electricity was only a small fraction supplied to Eskom’s main customers. Due to massive scale and surplus capacity, it was easy for Eskom to extend its grid to our cities which were happy to get cheap electricity and close down their own generating capacity. The efficiencies of scale outweighed the technical inefficiencies being baked into the system.
These inefficiencies have only got bigger as South Africa has become more urbanised and the shift away from being primarily a mining and heavy industry based economy. In short, Eskom’s physical infrastructure was put in place for a country and economy that does not exist any more.
It is in the residential and commercial sectors where one can best observe the embedded inefficiencies. Unlike continuous processes in mines and industry, the loads in the residential sector can be highly variable. In South Africa, the residential sector represents less than 20% of all electricity consumed but at peak times it represents around 35% of all demand. Rooftop PV and energy efficiency will make the variations over a day and between seasons only more pronounced.
The distance of loads in cities from the generators results in losses in transmission but also reductions in the power factor as current gradually shifts out of phase with voltage. This is particularly a problem during peak periods when there is more “contention” on all electrical wires and circuits. Eskom has to build and maintain much higher capacity at the generation side and supply more MWh’s of generation to supply a MWh of load at peak times and then has to carry this excess capacity in off-peak times. The result is an over-investment in power generation capacity because of remote generators, particularly to meet peak demand. It is very wasteful.
Although there is premium of electricity supplied to customers (the municipalities) at peak times, the simple requirement of meeting the peak creates costs for the system as a whole. Put another way, the cost of meeting the peak is not carried only by the peak energy and demand charges.
The potential of gas
On a technical level, the one source that we don’t have is gas. It is worth looking at what gas can do and how gas-fired electricity should be procured. Having it as just another source of electricity next to others would be a mistake. Instead, we should fully exploit the potential of gas to ramp up or down instantaneously in response to changes in load or generation caused by variable sources and use the spinning turbine effect to maximum advantage by siting gas-fired power plants near the loads. The R/MWh is a secondary factor if it is deployed in the right way. Simply put, the MWh provided by gas-fired power stations located close to our major cities can do more than the MWhs provided by other sources.
If gas plays its “filler” role, Eskom will be able to scale back on its investment strategy based as it is on remote generators. At the same time, Eskom could become more efficient by supplying only a fixed baseload without having to manage the daily peaks, the variability of renewable energy and to have to maintain power factors across the vast grid as well.
Some final thoughts
As the future is uncertain, energy planners have to ask themselves: What is the route of least regret decision for the country? This takes the debate away from stylised and often incorrect arguments of cost per kWh. Also, looking at planning in this way means that one always wants to retain optionality or the right to change direction if something better comes along. Also, there is the question of whether one wants any direct private sector investment in the electricity system or not.
Addressing the issues in this way often has the benefit of answering the technology questions as well. Big mega-projects like Medupi and Kusile, but even more so in the case of nuclear, lock out private investment and because they are invariably delivered well over budget and over time, only governments have the funding and are able to shift the risks of having these mega-projects inevitably go wrong onto the current and future taxpayers.
It therefore comes down to whether one focuses on a handful of state-funded/guaranteed projects with little price certainty and a technology lock-in for two generations hence, or hundreds of smaller projects that incrementally build capacity in which the funding and risks are those carried by the private sector. DM
Photo: Darling Wind Farm (Warren Rohner via Flickr)
Daily Maverick © All rights reserved