Power grid simulation model for long term operation planning

• Comprehensive mixed-integer power grid simulation model is developed.
• Proposed solution strategy aims to lower computational times of large power grids.
• Efficient formulation of power unit outage scheduling algorithm is presented.
• Proposed approaches are demonstrated on an annual simulation of Czech power grid.
• Influence of hydrothermal decomposition on the solution quality is analysed.

In this paper a comprehensive mixed-integer linear programming (MILP) model of transmission network and power plants is presented. The model is capable of forced and regular outage simulation with respect to the spinning reserve requirements. The model formulation includes thermal, pumped-storage and conventional hydro plants as well as renewable resources. Fast and efficient outage scheduling algorithm suitable for prospective studies is presented. We also show how the overall price minimisation can lead to unrealistic pumped-storage plants dispatch and how can this behaviour be corrected using hydrothermal decomposition. To overcome this issue a dedicated MILP water planning model is proposed and its results are compared to the real day production profiles of pumped-storage and conventional hydro plants. The simulation times and results of models with and without hydrothermal decomposition are compared on an annual simulation of Czech transmission network (61 power plants, 110 units and 31 nodes connected by 53 lines). Simulations produced by the model presented in this paper exhibit realistic transmission system behaviour, which was confirmed by the Czech TSO. The annual simulation times stay below reasonable values allowing multiple scenario prospective studies.