Contrasting Coal Power Trajectories in Southern Africa: Compound Air Pollution, Extreme Heat Exposures and the 2023 Hwange Expansion

Handsome Bongani Nyoni , Rutendo Norman Mandeya, Daniel Tigere, Zororo Kudzai Chinwadzimba , Rutendo Sibanda, Prosper Mthabisi Ncube , Perez Livias Moyo , Webster Gumindoga

Coal-fired power generation is declining at many established sites in South Africa, while new capacity is being commissioned in drought-exposed parts of the Zambezi basin. This uneven transition raises concerns about the coincidence of industrial emissions, extreme heat and socioeconomic vulnerability. We assessed compound air-pollution and heat risk around coal power plants in Zimbabwe, Botswana and South Africa, focusing on atmospheric changes associated with the commissioning of Hwange Units 7 and 8 in Zimbabwe in 2023.
We screened 25 coal-fired power plants using open satellite, meteorological, population and health-care accessibility datasets. Seventeen utility-scale plants had detectable nitrogen dioxide (NO₂) signals and were retained for comparative assessment. Hazard indicators comprised plant-associated excess tropospheric NO₂ and sulphur dioxide (SO₂) columns measured by TROPOMI, modelled fine particulate matter (PM₂.₅), annual days exceeding 35 °C and land-surface heat anomalies. Exposure was characterised by residential populations within 25 km of each plant, while vulnerability was represented by young children, older adults and estimated travel time to health care. These measures were integrated into a compound risk index, with sensitivity analyses assessing rank stability. An interrupted time-series analysis and placebo-in-space comparison examined pollutant changes at Hwange following the commissioning of the additional generating units.
Pollution trajectories and overlapping climate-related hazards differed substantially among plants. At Hwange, post-commissioning NO₂ column excess increased by 9.4 µmol m⁻² (95% confidence interval [CI]: 6.9–11.9), an 80% increase, while SO₂ excess increased by 63.0 µmol m⁻² (95% CI: 37.6–88.4), or 103%. These were the largest increases among the 17 plants, contrasting with declining signals at most South African sites. Communities surrounding Hwange experienced approximately 57 days annually above 35 °C, compared with fewer than two days around plants on the South African Highveld. Hwange also exhibited a night-time land-surface heat excess of 2.3 °C and an estimated travel time to health care of 27 minutes, compared with 2–13 minutes near South African plants. Hwange ranked second among the 17 plants on the compound risk index, ranging from first to fourth in sensitivity analyses. Lethabo ranked highest, with approximately 722,000 residents within 25 km. Across all detectable plants, an estimated 1.86 million people, including 166,000 children under five, lived within 25 km.
The pollution excess at Hwange peaked in 2024, coinciding with the greatest annual frequency of days above 35 °C in the analysed series. This concurrence highlights potential compounding of air-pollution and heat hazards, but does not establish individual exposure, adverse health outcomes or causation by the newly commissioned units alone. Satellite column concentrations reflect atmospheric pollution signals rather than directly measured ground-level exposure; similarly, the compound index prioritises potential concern rather than estimating disease burden. By integrating multiple environmental hazards with population distribution and health-care accessibility, the approach provides a transparent, reproducible screening framework to identify communities where new coal capacity intersects with climatic stress, supporting targeted monitoring, public-health preparedness and equitable energy-transition planning.