This is the case in the work of CSIR for South Africa, where work described by the authors as “thought experiment” generated headlines and belief that the lowest-cost electricity for South Africa involves some 86% annual contribution from renewables – onshore wind and solar photovoltaic (PV) .
But a series of simplifications means the suggested costing is not even wrong, its irrelevant. It’s based on a system that will not work in a future that should not happen.
A process of systematic under-estimation begins in the assumed demand being 261 TWh per year, approximately +15% on current levels.
Energy transitions for large nations occur, at best, over decades, so a plausible range of demand scenarios must be considered for South Africa in the 2050-2060 period.
Currently, South Africans per capita electricity consumption is low, and the pattern of consumption is drastically uneven. Only 40% of the working age population of South Africa is currently employed, and the population of South Africa is expected to grow by 24% from 2015-2050.
Under the CSIR assumption, South Africa would use less electricity per capita than it does today, sending poverty, unemployment and inequality even higher. Merely the per capita electricity consumption of an efficient economy like Denmark would require closer to 400 TWh per annum by 2050, or +60% from today.
Might it be higher than that? Certainly. A prosperous South African economy would likely have more focus on energy intensive activity like mining, mineral beneficiation and manufacturing. Moves to a more sustainable economy through the use of electric vehicles and desalination technology would nudge electricity consumption higher still.
Researchers need to consider many factors to create a plausible range. The CSIR assumption is only plausible as a scenario of ongoing and worsening poverty.
That assumption drives the proposed unit cost of electricity artificially low. At the lowest point of wind and solar supply in the CSIR modelling, the residual demand that must be met by the highest priced peaking gas is 34GW.
Under a realistic demand scenario, that figure would easily be 60GW. Greater installed gas capacity would be needed, to supply a larger proportion of the demand.
The correlated nature of the wind and solar supply, shown in the modelling, means additional 'cheap' wind and solar would achieve virtually nothing in periods of low supply, and would add to the surplus in periods of high supply; adding much cost for little return.
The modelled system treats nearly all South Africa as a supply catchment for wind and solar, to derive the benefit of geographic 'smoothing' of the variable supply. That demands an expanded transmission network.
Literature based on the European grid suggests transmission needs to be 5-6 times greater to derive 98% of the benefits of spreading out the renewable generation. That capital would often be idle; for example, every night no solar PV electricity will be transmitted.
It would nearly always be operate well-below capacity, for example when wind in any individual location is either idle or generating at below the rated installed capacity.
This is yet more inefficient capital expenditure South Africa can ill-afford. It is unmapped, unquantified and uncosted in the CSIR work, yet those fixed costs would need to be recouped on the price of the sold electricity.
The trade-off from trying to constrain the transmission network would be installing more megawatts of generation and adding electricity storage to achieve the reliability. Anyway you turn, there is more cost.
Finally, the CSIR modelling shows much and often all of the demand at any time provided by wind and solar PV.
Without additional remedies, which at such a scale have not been proposed or modelled anywhere in the world, that’s technically impossible. Wind and solar PV technologies provide asynchronous sources of generation.
This means they provide none of the inertia that offers the essential instantaneous frequency control that keeps an electricity system within its operational limits.
If a system falls outside of those limits, the whole system trips and you have a total blackout..
There are remedies to provide frequency control without turbine-based electricity generation. For example, batteries with the necessary electronics can provide this service. Synchronous condensers, devices that operates somewhere between a motor and a generator, can also contribute. We know this can work.
We don’t yet know at what scale and we certainly don’t have confidence to throw out synchronous generation altogether. Any such remedy, being the use of batteries, modified wind operations or synchronous condensers, add further cost that is not reflected in the simple prices applied in the CSIR report.
So the true cost of high penetration renewables based on 'low cost' on shore wind and PV begins to reveal itself. In such as system, supply, reliability and stability all have to be purchased separately, and managed through a greatly expanded, inefficiently operated transmission network. To date, it is beyond the bounds of operational knowledge.
As potentially publishable research there is much of interest in what CSIR delivered. Their work gives a great illustration of the potentially available supply from onshore wind and solar PV, as well as illustrating the challenges of variable and highly correlated supply spread over such a large country.
However it is essential to understand that it is, as originally advertised, a thought experiment. It’s not plan, it’s not a blueprint, it’s not 'proof' of any kind as the ripple of headlines and reporting suggests.
The potential future for renewables in South Africa is strong. On-shore wind, solar PV and solar hot water systems will all have a larger role to play. Overstating the case is not the way forward.
What South Africa needs is an overall system that is reliable, scalable to the essential needs of growing the economy, as clean as possible to boost competitiveness, and optimised for cost when all the components of a reliable system are considered. That means thinking well-beyond just wind turbines and solar panels.
Featured image: 123rf
Ben Heard is Founder and Executive Director of environmental non-profit Bright New World. He is also a doctoral energy researcher at the University of Adelaide, South Australia. His most recent paper, Burden of Proof: A comprehensive assessment of the feasibility of 100 % renewable electricity systems, will shortly be published in the peer reviewed journal Renewable and Sustainable Energy Reviews.