High-resolution multi-model benchmarking and validation of SATIMGE power-sector outcomes for South Africa's Second NDC using PyPSA, PLEXOS and FlexTool3

Abstract

This report documents the high-resolution, multi-model validation of the power-sector outcomes produced by SATIMGE for South Africa's Second Nationally Determined Contribution (NDC). Three scenario-matched 2035 power-system configurations were evaluated at full 8,760-hour resolution in PyPSA (UCT-ESRG) and triangulated against complementary analyses in PLEXOS (Stellenbosch University) and FlexTool3 (IRENA). The central finding is that the SATIMGE-derived systems are considered to comfortably meet typical system adequacy benchmarks and are operationally feasible with near-zero unserved energy, however power-sector CO2 emissions are 2.6% to 6.4% higher than the SATIMGE outcomes and up to 2.8% of VRE curtailment once full hourly chronology and the modelled operational constraints are represented.The remaining deviations in CO2 outcomes are modest, and driven primarily by the observed renewable curtailment and the lower achievable utilisation of the most efficient coal units when operated under full annual hourly dispatch - of which neither effect is fully captured by the reduced-resolution representation in SATIMGE. The PLEXOS and FlexTool3 models independently corroborate the findings with reported system adequacy metrics well above national benchmarks and similar VRE curtailment outcomes to PyPSA.For these NDC-aligned pathways, emissions feasibility under hourly chronology and the tested operational constraints became more binding than adequacy. The results translate into specific soft-linking refinement recommendations for further development and iteration such as: adjusted flexible dispatchable plant minimum utilisation factors, dispatch-informed coal utilisation or availability bounds on certain groups of plants, or curtailment-constrained VRE capacity factor adjustments, among others.