Modelling of grid connected geographically dispersed PV systems for power system studies
The growth of the photovoltaic market indicates that in the near future PV electricity generation may rise to a significant power source. As the proportion of electric power generated from PV systems becomes significant, the effect of these sources on transmission and distribution networks must be considered. This research work has investigated suitable representations of the PV resource and the output power of dispersed PV systems to study the effects of large-scale deployment of PV systems on the grid operation. The representation of solar radiation is very important since this dictates the output power of PV systems. In this work, the simple and reliable Markov Transition Matrix (MTM) method was selected to generate synthetic horizontal solar radiation data. A single MTM was developed to generate half-hour horizontal solar radiation data for different locations in the UK. Large-scale inclusion of PV systems in the UK electricity supply is expected to take the form of a large number of small, geographically dispersed building integrated PV systems. The study also developed a detailed PV cluster model to represent these dispersed PV systems. The variation of PV output power may impact the demand and generation balance on the network requiring additional reserve generation to ensure the system security. In this work, the variation of PV output power and the impact on the reserve requirement was analysed for different penetration levels. This is also the first study to analyse the correlation of solar radiation for different locations in the UK in regard to the impact on reserve requirements. Using data from three locations and according to the National Grid Company (NGC) requirements, it was found that PV capacities of 3750 MW could be added to the present network without additional reserve requirements. The additional reserve required is not on the basis of "MW of reserve per MW of PV capacity". Rather it is based on the aggregation of load demand and of PV output from all regions. The reduction in the reserve requirement by forecasting the weather profile of the day was also illustrated. In this case, a PV capacity of 22,500 MW, which can generate a little over 5% of the UK electricity demand, can be added with minimal increase in system cost. Therefore, the variation of PV output power is unlikely to be a threat to the system security.