As South Africans know all too well, the processes involved in drafting and approving an Integrated Resource Plan (IRP) for electricity are emotive, opaque and highly political. It took nearly ten years to update the initial IRP 2010, despite the fact that the document lost relevance almost from the very day of promulgation, as the country’s electricity reality began diverging materially from the demand and technology-cost assumptions contained in the promulgated document.
Various update attempts fell short, particularly as the technoeconomic outcomes began clashing with the narrow interests of a powerful business and political elite. Even the most recent, comparatively coherent and transparent, attempt at finalising an updated IRP (IRP 2018) fell prey to those able to use their political influence to usurp rationality in favour of self-interested lobby groups.
There is a way, however, for South Africa to depoliticise the IRP process and transition to a more transparent, technology-agnostic consultative methodology for determining the best technology mix to procure and, crucially, keep the plan up to date.
First, we need to realise that the technoeconomic groundwork at the core of an IRP is nothing more than a mathematical optimisation problem. The physical world of electricity supply and demand is captured in mathematical formulas that describe the behaviour of the different technologies, usually in hourly time resolution, from today to the end of the planning horizon, typically 30 years into the future.
The model is populated with technical and economical input parameters. The objective function is to ‘minimise cost’, subject to certain boundary conditions. The model can ‘build’ power generators from a predefined set of new-build options whenever they are needed (because of projected increasing demand or because of the projected decommissioning of old power stations). The output is a cost-optimised capacity expansion plan for the electricity sector.
How could that technoeconomic work be depoliticised?
It should all begin by the Department of Energy (DoE) setting firm boundary conditions which must be fulfilled by the mathematical model. For example, the desired level of security of supply and the maximum amount of carbon dioxide emissions, amoung others.
Those boundary conditions should be given to at least three completely independent organisations, with expertise in energy-system planning, being selected by the DoE to develop separate mathematical models for the IRP. Although working independently, everything produced – including input assumptions, source codes (the mathematical formulation of the model) and the model outputs – should be published on a public platform for peer review by the other two organisations, as well as any other individual with the technical skill to interrogate the information.
Full transparency around the technical model is key, otherwise what should be a purely techno-economic optimisation exercise can quickly descend into a negotiation between different technology lobbyists (see also https://www.nature.com/news/energy-scientists-must-show-their-workings-1.21517).
The three different results should then be presented to the public and discussed during stakeholder consultations. As part of the process, the public should be able to request that additional scenarios be tested, even if they deviate from the least-cost mandate. These could range, for example, from ‘30% nuclear energy share by 2050’, to a ‘fully decarbonised electricity system by 2040’, or even ‘building 10 GW of new coal by 2030’. These constraints expressed by the public can be seen as ‘policy deviations’ from an unconstrained least-cost case. Such deviation will naturally lead to an increase in the overall cost of the power system. The cost implications must then be calculated by the three organisations and represented to the public.
A small glimpse of such an approach could be seen during IRP 2018 already, when the cost implications of building two new coal-fired power stations in South Africa (a deviation from least-cost, which does not see any new coal-fired power stations) had been quantified at R23-billion in net present value (see also https://www.engineeringnews.co.za/article/radebe-outlines-additional-cost-of-coal-ipps-to-consumers-2018-10-01). This way the public and policymakers can make informed value-for-money decisions when they decide to purposefully deviate from a pure least-cost pathway.
In parallel to the optimisation of the capacity expansion, which essentially is the outline of an investment plan, an additional independent entity with strong system-operator expertise, could be requested to assess the operational adequacy of the scenarios. In other words, stress-test the operability of the scenarios in an hourly resolution so as to determine whether any additional major cost items, not captured in the models, might be required for stable operation.
A second round of public consultations could follow, where the additional scenarios and their price tags can be debated. In other words, an informed ‘value for money’ discussion could ensue, addressing the cost of the different scenarios, including the effects not optimised by a technoeconomic power-system model, such as the effect on the trade balance, jobs and industrial development.
The outcome should then be synthesised by the DoE into a final IRP for approval by Cabinet.
The approved document should provide visibility of the direction of travel for the coming 30 to 40 years, while the first five to ten years should be hard-wired into an investment plan. This plan would then be implemented through competitive auctions for energy (payment per kilowatt-hour delivered), and for capacity (payment per reliably available kilowatts).
Ideally, the individual new-build technology line items of the IRP should be adjusted yearly to take account of changes in supply and demand, within a predefined narrow band. The transmission system operator should be empowered to make these yearly adjustments, without changing the overall new-build mix though, only the new-build quantities within certain limits, with a rigorous least-cost system review (in line with the process explored above) implemented every two to three years.
Such a process would strike a balance between the need for public consultation and the requirement of ensuring the technical work is produced by independent power-system and energy-planning experts. At the same time, it will address the prevailing trust deficit, as it will involve more than one expert group whose work is fully transparent and open to scrutiny.
The whole process, including the cost of the independent technical experts, could be funded from a separate line item in the multiyear price determination so as to ensure absolute financial independence of the work from the results.
Such an approach would lead to the least-cost, least-risk power system for South Africa, which, owing to its solar- and wind- resource competitive advantages, would regain its position as a go-to place for any electricity- intensive industry. At the same time, South Africa would arguably become the global blueprint for what would be an elegant balance between central planning and market-based, competitive implementation.
Dr Bischof-Niemz is one of the authors of a book titled 'South Africa’s Energy Transition', which offers a roadmap to a decarbonised, low-cost and job-rich energy future. He was the founding head of the CSIR Energy Centre and previously worked for Eskom on the Integrated Resource Plan. He is currently CEO of ENERTRAG South Africa – firstname.lastname@example.org