Beitrag in einem Tagungsband
Short-Term Forecast for Reactive Power Compensation - Case Study Photovoltaic in the Medium Voltage Grid
Details zur Publikation
Autor(inn)en: | Junker, B.; Wang, H.; Braun, M. |
Herausgeber: | IEEE |
Verlagsort / Veröffentlichungsort: | Aachen, Germany |
Publikationsjahr: | 2016 |
Seitenbereich: | TBD |
Buchtitel: | Power and Energy Student Summit (PESS) 2016 |
Zusammenfassung, Abstract
The change in energy supply from a centralized to a decentralized power plant structure leads to a higher feed-in of electric power at the medium and low voltage level. With less active and reactive power supply from power plants at the higher voltage levels, reactive power balancing within the distribution grid by controllable distributed generation is a suitable option. In order to accomplish a maximum compensation of the required reactive power, the feed-in of reactive power of photovoltaic (PV) systems can be controlled and optimized. In this paper it is examined to which extent the application of short-term forecasts of PV generation, as predictive component, can contribute to enhance the compensation of the reactive power demand of a medium voltage distribution grid. Using a linear optimization of the forecasted PV feed-in in a power flow simulation, a power factor schedule for the PV systems from 15 minutes up to three hours ahead is created. The results suggest that intelligent control algorithms of the distributed generators can lead to savings of up to 20 percent of the purchased reactive energy of one year from the high voltage grid and that the forecast can indeed be used to predict future reactive power requirements. The results are achieved compared to a reference case without optimization of the PV systems or the tap changer at the network connection point (NCP) to the higher voltage level. In addition the reactive power forecast does not significantly depend on different types and time horizons of the forecast, but the longer forecast enables such a predictive grid control in case of temporal measurement or connection failures.
The change in energy supply from a centralized to a decentralized power plant structure leads to a higher feed-in of electric power at the medium and low voltage level. With less active and reactive power supply from power plants at the higher voltage levels, reactive power balancing within the distribution grid by controllable distributed generation is a suitable option. In order to accomplish a maximum compensation of the required reactive power, the feed-in of reactive power of photovoltaic (PV) systems can be controlled and optimized. In this paper it is examined to which extent the application of short-term forecasts of PV generation, as predictive component, can contribute to enhance the compensation of the reactive power demand of a medium voltage distribution grid. Using a linear optimization of the forecasted PV feed-in in a power flow simulation, a power factor schedule for the PV systems from 15 minutes up to three hours ahead is created. The results suggest that intelligent control algorithms of the distributed generators can lead to savings of up to 20 percent of the purchased reactive energy of one year from the high voltage grid and that the forecast can indeed be used to predict future reactive power requirements. The results are achieved compared to a reference case without optimization of the PV systems or the tap changer at the network connection point (NCP) to the higher voltage level. In addition the reactive power forecast does not significantly depend on different types and time horizons of the forecast, but the longer forecast enables such a predictive grid control in case of temporal measurement or connection failures.
