After several delays, the draft Integrated Resource Plan (IRP) 2019, updated by the Department of Energy (DoE) to incorporate the comments arising from the public participation process which closed on October 26, 2018, was finally submitted to Nedlac on March 6.
This now enables Nedlac to commence its work of reviewing the updated draft IRP 2019 and making input to the cabinet before the IRP is subject to policy intervention by government, and thereafter finalisation and publishing of the final IRP 2019 in the Government Gazette.
In this way government’s so-called “social partners”, namely business, labour and communities, are given a last opportunity to make input on important, planned, economic and social interventions by government.
In light of the work still to be done by Nedlac, and “policy adjustment” by the cabinet, it appears increasingly unlikely that the final IRP 2019 will be completed and gazetted before the general elections on May 8.
Some analysis, notes and observations on the updated draft IRP2019 that was presented to Nedlac are as follows:
There does not appear to be a table clearly stating and summarising the assumptions made in the updated draft IRP 2019, with links or references to pages in the IRP, or to external documents, supporting or explaining the assumptions.
Nor is there a clear table showing the assumptions in the updated draft IRP 2019 that have changed from the earlier draft IRP 2018 that was issued for public comment, to highlight the changes and reasons for the changes, with links or references to pages in the IRP, or to external documents, supporting or explaining the changes.
The updated draft IRP 2019 shows clear signs of political policy interventions and appeasement of various stakeholder interests by DoE officials. The IRP drafts should be apolitical, techno-economic studies and policy interventions should not find expression in the updated draft IRP prepared by officials. This should be left for the review and recommendations by Nedlac, and the policy adjustment by politicians.
The immediate and most important issue of identifying the critical supply and demand-side interventions required to get South Africa through the next five years remains unspoken in the updated draft IRP 2019.
As in the draft IRP 2018 issued for public comment, the updated draft IRP 2019 does not show the unconstrained least-cost scenario as the base case.
In addition, the base-case scenario, the unconstrained least-cost scenario and the other scenarios modelled in the updated draft IRP 2019 are not costed.
So neither Nedlac, the Cabinet, nor anyone else for that matter, is able to evaluate the cost/benefit of each of the scenarios modelled, over that of the base-case or the unconstrained least-cost scenarios.
In the updated draft IRP 2019, the demand forecast still appears to be significantly too high, if the comments by the Energy Intensive User Group (EIUG), whose members use about 40% of all the electricity generated in South Africa, are anything to go by.
On the other hand, the energy availability factor (EAF) projected for the Eskom fleet of coal-fired power plants appears over-optimistic. However, to some extent the overly high demand forecast may be offset by the over-optimistic Eskom EAF, so that any variances may not impact new capacity requirements that much.
An additional 750 MW of new coal-fired power generation is included in the updated draft IRP 2019 scheduled for 2027, over and above the 750 MW of new coal power from the Thabametsi and Khanyisa IPPs scheduled for 2023.
While the updated IRP 2019 acknowledges that banks and financial institutions are no longer able to fund fluidised-bed, sub-critical, coal-fired power plants (like Thabametsi and Khanyisa), it does not indicate the technology that should or will be used for the 750 + 750 = 1 500 MW of new coal power included in the updated draft IRP 2019.
Nor does the updated draft IRP 2019 clarify the cost premium of using ultra super-critical (USC) steam generation technology for new coal power, above the cost of the fluidised-bed technology used for the first 750 MW of new coal power from the Thabametsi and Khanyisa IPPs in the updated draft IRP 2019.
In the updated draft IRP 2019, the life of the Koeberg nuclear power plant has been extended to 60 years from 2024. This was to be expected, as steam generator upgrades at Koeberg are in progress, and further upgrades and life-extensions are planned. No other new nuclear power plants are planned in the period to 2030.
Power from the Inga hydro-power project in the DRC is still included in the updated draft IRP 2019 in 2030, although it is unlikely to happen. Presumably, the power resulting from a formal international treaty between South Africa and the DRC has to be shown somewhere.
Battery energy storage is now shown explicitly in the updated draft IRP 2019, although the timing in only two large chunks of 513 MW in 2022, and 1575 MW in 2029, seems impractical and unlikely. It would be better to spread these over a number of years.
In addition to its other system benefits (such as frequency and voltage control, demand management, peak shifting, arbitrage, capital deferment, micro grids, etc.), battery energy storage will become more and more cost effective as a quick source of flexible generation capacity.
Gas-to-power, on the other hand, is reduced in the updated draft IRP 2019 to 6 380 MW by 2030, perhaps reflecting the slower pace of gas-to-power decisions and construction, and the fact that gas-to-power and battery energy storage are both flexible power generation proxies.
The artificial annual constraints on renewable energy in the updated draft IRP 2019 remain, namely 1600 MW per year for wind power, and 1000 MW per year for solar PV.
Although the DoE says the artificial annual renewable energy constraints do not impact the roll-out of wind and solar PV, this is not true. In fact, the constraints force 1 500 MW of new coal power into the mix by 2030. However, due to funding constraints, whether the new coal generation capacity will actually happen, is another story.
New wind and solar PV beyond the REIPPP Bid Window 3.5 and Bid Window 4 has been brought forward by three years. This is positive, and reflects the short-term new capacity needs of South Africa resulting from the declining EAF of Eskom’s aging coal-fired fleet, and its problems at Medupi and Kusile.
The Updated Draft IRP 2019 indicates high diesel-driven open-cycle gas turbine (OCGT) load factors in next few years, way beyond their intended design limits. This shows that emergency generation capacity will be used for operational purposes, at very high cost. This is not wise, and other sources of new flexible generation capacity, like battery energy storage and gas engines, should be considered.
In the updated draft IRP 2019, embedded generation is now called “distributed generation”. An open-ended and unquantified allocation of distributed generation is now shown in the years to 2022 “to the extent [of meeting] the short-term capacity and energy gap”.
After 2022, the annual allocation of distributed, customer side, “behind-the-meter”, own generation has been increased from 200 MW per year to 500 MW per year in the updated draft IRP 2019, with no total given for distributed generation for the years to 2030.
This speaks volumes, and indicates that there is a growing realisation that, for Energy Transition 4.0, embedded, distributed generation and the customer form a most critical part of the solution to meeting the electricity supply needs of the future.
My conclusion is that this updated draft IRP 2019 has several significant flaws, even coming as it does after several iterations since the first draft IRP update in 2013. The CSIR could and would have done a significantly better, non-political, techno-economic IRP study in three months, to enable Nedlac and the cabinet to do their reports and policy adjustments properly.
South Africa has the necessary scientific, engineering and planning competencies to do a proper IRP update every two years, and we should not have to accept years of uncertainty, or be forced to accept a flawed IRP in the interests of expediency, and at the expense of security of supply.
Chris Yelland is investigative editor at EE Publishers.