Increasing Photovoltaic penetration rate on isolated grid using storage, overbuilding and curtailment: Case study of the Reunion Island

Abstract: In the global context of GHG emissions reduction, insular isolated grids which are the most vulnerable systems in terms of fossil fuel dependency have to conduct their energy transition. Conventional units using fossil fuels are well adapted to those systems because they are dispatchable, reliable and flexible, so they can easily respond to grid requirements. Despite that, their large negative environmental impact and their dependence on geopolitical and economical context encourage islands to develop renewables. The Reunion Island wants to achieve a 100% renewable electricity mix by 2030. This ambitious goal can be achieved using the high solar resource’s potential in the region. However, this renewable potential is limited in the electricity mix because of issues related to uncertainty and variability of grid-connected PV systems. To deal with these issues a solution is to combined PV systems with storage systems to guarantee a firm power production. Considering the high costs for storage and the decreasing costs for PV systems, using overbuilding and curtailment could reduce the needs for storage and thus the cost related to it. The goal of this study is to develop an algorithm that will test and size this solution for the entire power production of the Reunion Island. To do so, satellite based data for solar resource (GHI and DNI) with resolutions of 1km² and 15min over the year 2018 were available from the SUNY model. These data were evaluated by a comparison with 13 ground measurements stations for the GHI and two trackers for the DNI in terms of power and energy. It was shown that estimation errors were quite large (relative RMSE between 27.5% and 78.8%) and that stations with the higher altitudes were the one with highest errors. This solar resource was used as well as temperatures and installed PV capacities to model the entire PV production fleet of the island. Different models were tested using a representative PV installation for the entire island and its parameters were adjusted to fit the real PV power production of 2018. The model with the lowest relative RMSE for power comparison uses only efficiency, tilt and orientation angles with installed PV capacities gathered at each 63kV/15kV transformer’s zone to reach a relative RMSE of 14.3%. Using this model, the algorithm was developed to size PV production fleet, PHES and Li-ion batteries capacities for a given desired production profile. Different production profiles were tested to cover from 10% to 100% of the total electricity demand of the Reunion Island in 2030 with different profile’s shapes. For each of these profiles, best configuration was chosen by minimizing the LCOE resulting in a range from 14.8c/kWh to 22.1c/kWh. Order of magnitude for the different system parts were found to range from 0.5 to 5 GW for PV, 2 to 37 GWh for PHES and 3 to 70 MWh for Li-ion batteries. In some cases, curtailed energy could serve to cover the annual consumption of 44 000 hydrogen cars or the half of 2018 water consumption by using desalinization. Profiles to replace fossil fuels power production were tested and a 100% renewable mix could be achieved with 4 GW of PV, 27 GWh of PHES and 50 MWh of Li-ion batteries for a LCOE of 19.8c/kWh. This profile allows dividing by almost ten the estimated GHG emissions related to the electricity production in 2030. The final conclusion of this study is that variability and uncertainty of solar PV production can be overcome by using overbuilding, curtailment and storage to guarantee a firm power production profile, no matter the desired energy penetration rate and this without increasing the cost of electricity compared to the current mix.