Analysis of European res imbalance charges : the impact on PV and wind plants

The increasing share of Renewables Energy Sources (RES) on the energy market has had a great impact on the electricity market evolution. In particular, the high presence of non-programmable RES-E as photovoltaic (PV) and wind plants constitute one of the biggest turns that authorities and system operators have to deal with. The present work focused on the energy unbalance generated by photovoltaic and wind plants. In particular, it was studied the balancing obligation and the gate closure time unit in Germany, Italy and Spain. These three countries were chosen because of the high PV and wind plants penetration on their energy markets. The Italian case was studied more thoroughly, in order to estimate some figures of the nodal and zonal unbalance and the different remuneration mechanism used to penalize the unbalance generators. To do so, a Matlab model was written in order to estimate the photovoltaic and wind plants real and forecasted injections, as well as the forecasted and real load. Given the extent of the data that would be required to represent the whole Italy, it was decided to model just one of the six electric regions composing the country. In this sense, the Sicilian electric region was selected because of its high non-programmable RES penetration and the availability of meteorological data. Additionally, the model was also used to estimate the effect of the gate closure time on the unbalance costs incurred by non-programmable RES-E and the real time balancing cost. To do so, it was required to relate the forecast error to the different gate closure intervals reported on the literature. Then the nodal and zonal unbalances were determined in order to estimate the cash flow of the Transmission System Operator (TSO, in Italy TERNA) and the plants generating the unbalance. Finally, two additional scenarios were considered in order to take into consideration that the forecasting accuracy values might not correspond to the forecasting accuracy of the models used in Sicily. In this regard, both an optimist scenario and a conservative scenario were studied. The first one increases the forecasting model's accuracy, while the second one diminishes it.