Risk Analysis of Optimal Design of a VPP in Risk-Seeking/Risk-Averse Modes Using IGDT and Considering Wind, Solar, and Load Uncertainties

Abstract

This paper presented a risk analysis using information gap decision theory (IGDT) in risk-seeking and risk-averse approaches to invest in the construction of virtual power plants (VPPs), considering the uncertainties of wind speed, solar radiation, and load demand. The construction of VPP has financial and operational risks, such as investment costs, maintenance, power exchange with upstream grid, unsupplied energy, and combined heat and power (CHP) system fuel for investors in terms of uncertainties related to load demand and various sources of distributed energy generation (distributed energy resources [DERs]). Investing risk analysis on an experimental system, including photovoltaic (PV) and wind, thermal, and combined power plants (CHP), battery energy storage system (BESS), and electric and thermal loads, using a multiobjective function to minimize the total cost of VPP construction in risk-seeking modes (πo) and risk aversion (πc), was performed. The risk analysis involved 8% reducing/increasing of the total VPP construction investment cost in 2% steps for the risk-seeking owner (ρ)/risk-averse owner (σ). The results of this article showed that despite high uncertainty in the studied VPP, it is still possible to make a decision for risk-seeking and risk-averse investors to build VPP. In the risk-averse mode, due to the increase in cost, there is more flexibility in choosing and using equipment, as well as in determining its amount. On the contrary, in risk-seeking mode, VPP designer should determine the smallest radius of uncertainty (αwind, αpv, and αload) in ideal conditions, because to formulate a suitable investment proposal to use equipment and their quantity in a way that leads to cost savings is vital. The key numerical results include the following: (1) The base investment cost is 18,731.2 monetary units; (2) a risk-averse scenario achieves a critical cost of 20,229.7 monetary units at a risk tolerance of σ = 0.08; (3) for a risk-seeking scenario, the critical cost is 17,232.7 monetary units with a risk parameter ρ = 0.08; (4) in the risk-averse case, optimal robustness ensures critical cost πc = 17,232.7 monetary units at risk parameter σ = 0.06 assurance at an uncertainty radius where αwind = 2.353%, αpv = 1.176%, and αload = 14.118%; (5) in the risk-seeking case, optimal opportunity ensures critical cost πo = 17607.3 monetary units at risk parameter ρ = 0.06 assurance at an uncertainty radius where αwind = 2.353%, αpv = 17.647%, and αload = 18.824%; and (6) the harmony search algorithm (HSA) algorithm demonstrated an 18% faster convergence speed compared to genetic algorithm (GA). This approach makes sure VPP design stays strong and full of chances even when things get tough. It is a useful way to handle investment risk and get the best results when there is a lot of uncertainty. Finally, the actual numerical results applying the proved test case prove the introduced approach to be effective.