## Detailed programme

##### 2019 TR Medal winning paper

###### 20 October 2020

###### 17:00 - 18:30 (SAST)

## Tradeoffs between levelling the reserve margin and minimising production cost in generator maintenance scheduling for regulated power systems

# BG Lindner, R Brits, JH van Vuuren, J Bekker

# Department of Industrial Engineering, Stellenbosch University

One of the key focus areas for a power utility is the planned preventative maintenance of the power generating units in its power system. The well-known generator maintenance scheduling (GMS) problem involves finding a schedule for the planned maintenance outages of generating units in a power system. A novel bi-objective model is proposed for the GMS problem in which demand reliability is maximised, by minimising the sum of squared reserves (SSR), and electricity production cost (predominantly fuel cost) is minimised. A novel production planning module is proposed to estimate the production cost associated with an energy generation plan, using a linear programming (LP) model to solve the economic dispatch (ED) problem, which precedes application of a simple unit commitment (UC) algorithm. A dominance-based multi-objective simulated annealing approach is then adopted to determine trade-off solutions to the model. Parallel computing is also utilised to increase the efficiency of approximating the Pareto front. The modelling approach is demonstrated in the context of a case study involving the 32-unit IEEE Reliability Test System. The results are compared to the best known singleobjective solution in the literature, which only minimises the SSR, and the conflicting relationship between the two model objectives is investigated. It is found that more non-dominated trade-off solutions result if the load demand increases (i.e. the gap between installed capacity and load demand decreases). Therefore, if the installed capacity is sufficiently high, the reliability objective of minimising the SSR produces sufficiently small production cost solutions. Fuel cost savings of 0.41% are achieved in respect of a most “reliable” solution in the literature, but considerable cost savings are possible (up to 7.11%) if the maintenance duration and crew constraints are relaxed.