Stochastic Optimization for Security-Constrained Day-Ahead Operational Planning under PV Production Uncertainties: Reduction Analysis of Operating Economic Costs and Carbon Emissions

International audienceIntermittent renewable energy sources (RES) generate variable power that cannot be fully predicted by advanced forecasting tools. Fortunately, re-schedulable conventional generators can contribute to the power balancing. However, the rescheduling usually leads to abrupt operations of fuel-based fast generators, which will induce an increase of fuel costs and CO2 emissions. This paper presents a general operational planning framework for controllable generators, one day ahead, under uncertain renewable energy generation. The planning objective consists on minimizing operating costs and/or equivalent carbon dioxide (CO2) emissions. The uncertainty is handled by a two-stage method: past error predictions are analyzed by a probabilistic approach and then possible future error predictions are considered through scenarios. Based on distributions of forecasting errors of the net demand, a loss of load probability-(LOLP) based risk assessment method is proposed to determine an appropriate amount of operating reserve (OR) for each time step of the next day. The effect of photovoltaic (PV) power generation uncertainty on operating decisions is examined by incorporating expected possible uncertainties into a two-stage unit commitment optimization. In a first stage, a deterministic optimization within a mixed-integer linear programming (MILP) method generates the unit commitment of controllable generators with the day-ahead PV and load demand prediction and the prescribed OR requirement. In the second stage, possible future forecasting uncertainties are considered. Hence, a stochastic operational planning is considered and optimized in order to commit enough flexible generators for reserve provision to handle unexpected deviations from predictions. Most of all, through the presented security-constrained two-stage optimization, OR is optimally scheduled (by using PV generators and conventional generators) to reach the minimum costs and CO2 emissions by considering PV panel characteristics and stochastic nature of PV production. The proposed methodology is implemented for a local energy community. Results regarding the available operating reserve, operating costs and CO2 emissions are established and compared. About 15% of economic operating costs and environmental costs are saved, compared to a deterministic generation planning while ensuring the targeted security level. INDEX TERMS Decision making, generator scheduling, probabilistic modelling, renewable energy, reserve allocation, stochastic optimization, uncertainty, unit commitment, urban energy system NOMENCLATURE A. ACRONYMS BPNN Back-propagation neural networks CCUC Chance-constrained unit commitment CHP Combined Heat and Power CI Confidence intervals DUC Deterministic unit commitment LOLP Loss of load probabilit