Coal-fired power stations have stockpiles that act as buffers for any variations in the coal usage rates, coal delivery rates, and coal supply interruptions. The level of the stockpile is influenced by how much coal is delivered and how much coal is burnt. This is in turn affected by a range of factors, including coal quality, planned and unplanned outages, load forecast uncertainty and the merit order scheduling interactions between stations.

With the ever increasing load growth in the country, a shrinking reserve-margin and insufficient installed capacity, the load factor at the power stations has to increase which, by inference, places pressure on the coal supplier contractual limitations. Ancillary suppliers are therefore required, which in turn come with their risks such as availability, reliability and costs, increasing the risks on the stockpile.

The decisions required in managing the stockpiles need to come from an informed position, not only from a historical perspective, but also from a forward looking one, supplemented with ‘what-if’ scenarios. These aspects are particularly important for Eskom’s Primary Energy Division (PED), responsible for the planning, procurement and delivery of coal to the power stations.

A stochastic simulation model was required to further inform stockpile level management decisions, specifically so that the stochastic interaction of stations, due to planned and unplanned outages, could be included in purchasing additional, or moving coal around. The specific contribution sought was the ability to cater for situations where planned coal usages are altered on the fly, in the future, due to the fact that when one station has an outage, another station must ‘take-up’ the production. As a result, additional coal is used at the second station, where normal linear planning modules (e.g. linear optimisation in terms of reducing transport costs, contract purchase cost) are not able to cater for these stochastic interactions.