Improved zonal network models in generation scheduling

Realistic generation scheduling models should take account of the electricity grid constraints. Nodal network models describe the grid in detail but require a high granularity of input data and result in long simulation times. Zonal network models are simplified versions of nodal network models, with lower granularity of input data and shorter simulation times, but less accurate simulation results. Detailed nodal network models are not suited for large scale generation scheduling models due to their complexity. Therefore, it is of importance for generation scheduling models to use an appropriate zonal network model, finding a balance between correctness and convenience. This paper compares a nodal network model with a "simple" zonal network model, a zonal network model with higher granularity and a zonal network model with different Power Transfer Distribution Factors for different types of grid injections. This latter zonal network model gives a sufficiently accurate representation of the grid flows, without increasing the simulation time or the granularity of the input data. All discussed concepts are applied to a case study of the Continental European power system.