Cross-correlation of output fluctuation and system-balancing cost in photovoltaic integration

Abstract

The authors analysed the cross-correlation of photovoltaic (PV) output fluctuation for the actual PV output time series data in both the Tokyo area and the whole of Japan using the principal component analysis with the random matrix theory. Based on the obtained cross-correlation coefficients, the forecast error for PV output was estimated with/without considering the cross-correlations. Then the operation schedule of thermal plants is calculated to integrate PV output using the proposed unit commitment model with the estimated forecast error. The system-balancing cost of PV system was also estimated with or without demand response. Finally, validity of the concept of ‘local production for local consumption of renewable energy’ and alternative policy implications were discussed.

Keywords

• photovoltaic power systems
• power generation economics
• power generation scheduling
• load forecasting
• principal component analysis
• time series
• estimation theory
• matrix algebra
• photovoltaic output fluctuation cross-correlation
• system-balancing cost estimation
• photovoltaic integration
• actual PV output time series data
• Tokyo area
• Japan
• principal component analysis
• random matrix theory
• cross-correlation coefficients
• forecast error estimation
• thermal plant operation scheduling
• unit commitment model
• policy implications
• local production concept
• local renewable energy consumption