Microgrid-level reliability assessment of mid-term electricity provision under intermittency of renewable distributed generation: A probabilistic conditional value at risk modeling

In critical and sensitive load sites like medical and educational microgrids (MGs), the reliability index holds paramount importance, yet it has not received necessary attention. This paper introduces a scheduling strategy focused on maximizing the MG's annual reliability index. Here, a power-based failure rate (PFR) concept to enhance the reliability assessment of MGs, establishing a relationship between component failure rates and production, is proposed. Moreover, MGs face various uncertainties, including renewable energy resources’ generation and consumption patterns. These uncertainties are modeled using a scenario-based stochastic approach with well-known probability density functions. Furthermore, the potential risks of the worst-case scenario present a significant obstacle to the reliability-focused improvement process; to safeguard the suggested framework from undesirable conditions, a conditional value at risk (CVaR)-based framework is developed. This framework assists the MG operator (MGO) in managing the analyzed system during worst-case scenarios. Ultimately, the proposed model encountered a non-convex challenge due to the exponential nature of reliability and PFR curves, which transformed into a mixed integer linear programming model utilizing piecewise linearization and the MacCormack relaxation techniques. Simulation findings indicate that under full-risk conditions, the annual reliability index of MGs slightly decreases due to the MGO's conservative policies.