The objective of this paper is to quantify the magnitude of energy transition risks to employment for coal miners, and to make a first step in identifying potential solutions. To achieve this, we explore the current characteristics (age and education) of the coal mining workforce according to existing data, in particular the PALMS dataset of Datafirst. We use plausible scenarios from a linked South African bottom-up (SATIM) and top-down (eSAGE) energy economic model to estimate the impact of timing of a reduction in coal mining output on employment and job losses in coal mining.