Delivered by: The Carbon Trust, Energy Systems Research Group (ESRG) of the University of Cape Town, Presidential Climate Commission (PCC)
Beneficiary: PCC
Country: South Africa
Challenge
South Africa has 40,000 MW of coal-fired power capacity, accounting for 70% of annual electricity generation. Most of this is from 13 coal-fired power plants (CFPPs) owned and operated by Eskom, a vertically integrated monopoly wholly owned by the state.
The country holds climate change mitigation obligations in the near term (Nationally Determined Contributions out to 2030). It has set a longer-term target to achieve Net Zero by 2050 (SA-LEDS 2050), which requires urgent attention on the future role of coal in the national energy supply. The carbon budget is critical in determining the pace of coal closure.
In addition, South Africa’s Air Quality (AQ) standards are considered low and not fully enforced, resulting in severe health impacts from continued and long-term exposure to air pollutants.[1] Pollution from South Africa’s CFPPs is also geographically concentrated in areas with higher poverty and unemployment rates, exacerbating the impact of pollution on vulnerable communities.[2]
As such, this technical assistance project explored the potential for an accelerated phase-out of the CFPP fleet in South Africa in response to the implications of AQ compliance decisions and the country’s climate commitments. The project further considers the overarching imperative of ensuring that the transition of South Africa’s energy sector is just and fair to all, especially across the coal value chain.
[1] South Africa – Clean Air Fund; Health impacts of Eskom’s non-compliance with minimum emissions standards
[2] The Implications of Climate Change on Health among Vulnerable Populations in South Africa: A Systematic Review – PMC
1. Deploy
2. Design
3. Scope
Support Provided
The Coal Asset Transition Accelerator (CATA) and the Africa Climate Foundation (ACF) co-funded the following activities:
- Prioritisation framework assessment
The Carbon Trust conducted a high-level prioritisation framework assessment of the CFPP fleet, whose preliminary recommendations indicated the order for retiring each CFPP in the fleet. This assessment considered South Africa’s national context, including energy security, cost, environmental and Just Transition factors. - Techno-economic modelling
ESRG undertook a techno-economic modelling exercise using the South Africa TIMES (SATIM) model, a linear least-cost optimisation model of South Africa’s entire energy system.
Scenarios
- Both technical pieces of work compared the reference scenario to two scenarios. The scenarios were chosen in consideration of the impacts of continued failure to comply with Minimum Emissions Standard (MES), the options available for compliance, and the effects of the potentiality and timing of investments required.
- 2030 Closure scenario – Most CFPPs would need to shut down (compliance by decommissioning) by 2030
- 2030 Retrofit scenario – CFPPs would need to comply with the MES through significant retrofitting of plants with technologies to reduce air pollutants by 2030 and then continue to operate as planned.
- Reference Scenario – follows Eskom’s Emission Reduction Plan (ERP) 2022 where CFPPs run till their technical life.
* Currently, none of Eskom’s plants are compliant with the Minimum Emissions Standard (MES). It is assumed that the two newer plants, Medupi & Kusile, will be retrofitted to comply with these standards by 2030. The closure scenario assumes that the remaining non-compliant plants will be decommissioned by 2030.
** The retrofit scenario assumes that plants will be retrofitted with AQ abatement technologies by 2030 to comply with the MES and continue to operate as planned.
Impact
The results demonstrate that South Africa should aim to accelerate coal closure and aggressively ramp up renewable energy to at least 6GW a year to ensure energy security, and not retrofit plants
This is because decommissioning CFPPs, coupled with accelerating the renewable energy rollout (i.e., under the 2030 Closure scenario), has lower direct costs and risks than retrofitting the plants with air quality abatement technologies to comply with the MES (i.e., the 2030 Retrofit scenario),or continuing to operate non-compliant plants (Reference Scenario).
Retrofitting costs are high and could enhance this risk and exacerbate the economic and developmental impacts of load shedding if retrofits are not undertaken promptly. Furthermore, the difference in investment in RE between the Reference Scenario and the Closure Scenario is small, increasing the rate of variable renewable energy rollout from 5.5 GW to 6 GW per year.
These findings have added to the PCC’s evidence base for its ongoing work to ensure a Just Transition to a low carbon and climate-resilient economy and society in South Africa and support shaping its continued recommendations to Eskom.
Methodology detail and Key takeaways
Reference Scenario
The Reference Scenario maps the installed capacity by power plant through to 2050, with no cap of GHG emissions, in accordance with the decommissioning timeline of Eskom’s Emission Reduction Plan (ERP) 2022.
Assumptions
- Plants are decommissioned at their end of life. The reference CFPP retirement schedule assumes over 15 GW of coal-fired capacity remains in 2040 and over 5 GW in 2050
- Currently there are no plans to retrofit plants with AQ abatement technology except for two of the newer plants.
Key Takeaways
- The Reference Scenario shows 26% reduction (10.4 GW) in coal capacity by 2030, 41% (16.2 GW) by 2035 and 87% (34.4 GW) by 2050, relative to 2023 levels.
- MES compliance cannot be achieved as limited abatement retrofits are planned in this scenario, with no projects to deal with SOx emissions.
- Matla, Duvha, Lethabo, Matimba, Kendal, and Majuba CFPPs would be operating without compliance (SOx) beyond 2030 under this scenario.
2030 Retrofit Scenario
The 2030 Retrofit scenario analyses the decommissioning schedule, as per Eskom’s ERP 2022, with retrofits of the non-MES compliant plants with technologies to reduce air pollution and comply with the MES.
Assumptions
- Arnot, Camden, Duvha, Grootvlei, Hendrina, Matla, Tutuka, and Kriel CFPPs are assumed to be decommissioned in line with current plans (IRP2019; draft IRP2023; and Eskom plans), so no retrofits were modelled for these plants.
- This scenario includes technologies to reduce Particulate Matter (PM) and NOx (nitrogen oxides), but not technologies to reduce SO2 emissions due to high costs.
Key takeaways
- Decommissioning CFPPs does not accelerate notably compared to the Reference scenario
- Retrofitting non-MES compliant plants comes at a high cost. It would cost ZAR 30-50 billion per plant to retrofit AQ technologies to comply with the MES requirements.
- We note that while the retrofit is being implemented, a plant may need to be taken offline for at least 6 months (some estimate it could take 12 months or more). This decrease in installed capacity could prolong system operation challenges, a risk mitigated by building more infrastructure for renewable energy.
- In addition, the cost of continued operation and maintenance of these plants is significant, and continued SO2 emissions mean that the health costs are not decreased by this intervention.
2030 Closure Scenario
The 2030 Retrofit scenario analyses the decommissioning schedule, as per Eskom’s ERP 2022, with retrofits of the non-MES compliant plants with technologies to reduce air pollution and comply with the MES.
Assumptions
- Except for Medupi and Kusile, retrofit costs are not included in this scenario.
Key takeaways
Overall fleet operation and maintenance costs and adverse health impacts and costs would also be avoided. However, shutting down a plant without replacement capacity cannot be supported and significant new capacity (RE) must be brought online. Additionally, the costs of urgently implementing a Just Energy Transition must be weighed against the cost of doing nothing or implementing the 2030 Retrofit Scenario.
Sensitivity analysis – 9Gt Carbon Budget
The project also modelled a scenario for meeting South Africa’s NDC (i.e. implementing a 9Gt carbon cap, with no additional retrofit/closure action for MES) as sensitivity analysis on the Reference scenario.
The results showed that the carbon cap drives the pace and scale of the closure of coal-fired power plants post-2030. Wind and solar penetration in particular need to substantially increase regardless of the remaining coal capacity operating.
This scenario would smooth new capacity additions and could support reducing pollutants and GHG emissions. While investment needs are higher than Eskom’s ERP, most of the new investment in generation capacity does not depend on Eskom or deepen its indebtedness.