Mengqi Yao

Public Safety Power Shutoffs (PSPS), also known as proactive de-energizations, proactively de-energize a portion of power systems to mitigate the risk of catastrophic wildfires caused by electric infrastructure. Since the first practice in Southern California in 2012, PSPS have been widely discussed in government and industry, but seldom in the academic literature. This paper surveys the PSPS program in California, including its history, policies, and practices. In practice, PSPS present strong interactions between electric utilities and customers, yielding PSPS a trade-off problem to balance the risk of power-line-ignited wild-fires (i.e., wildfire risk) against the harms of power shutoff (i.e., PSPS risk). In this regard, this paper summarizes the industry-standard and research methods for PSPS studies, including models, data sources, and test systems. It is suggested to integrate engineering solutions with social-economic science, such as energy innovation, energy equity, and PSPS uncertainties, in future PSPS studies.

Growing penetrations of single-phase distributed generation such as rooftop solar photovoltaic (PV) systems can increase voltage unbalance in distribution grids. However, PV systems are also capable of providing reactive power compensation to reduce unbalance. In this paper, we compare two methods to mitigate voltage unbalance with solar PV inverters: a centralized optimization-based method utilizing a three-phase optimal power flow formulation and a distributed approach based on Steinmetz design. While the Steinmetz-based method is computationally simple and does not require extensive communication or full network data, it generally leads to less unbalance improvement and more voltage constraint violations than the optimization-based method. In order to improve the performance of the Steinmetz-based method without adding the full complexity of the optimization-based method, we propose an integrated method that incorporates design parameters computed from the set-points generated by the optimization-based method into the Steinmetz-based method. We test and compare all methods on a large three-phase distribution feeder with time-varying load and PV data. The simulation results indicate trade-offs between the methods in terms of computation time, voltage unbalance reduction, and constraint violations. We find that the integrated method can provide a good balance between performance and information/communication requirements.

Maintaining balanced voltages across distribution networks is becoming more challenging with increasing deployment of single-phase distributed generation and larger single-phase loads. The paper develops a reactive power compensation strategy that uses distributed solar photovoltaic (PV) inverters to mitigate such voltage unbalance. The proposed strategy takes advantage of Steinmetz design and is implemented via both decentralized and distributed control. The latter coordinates PV inverters through a communication network. We demonstrate the performance of the controllers on the IEEE 13-node feeder and a much larger taxonomy feeder (617 nodes and 1196 triplex nodes), and consider different connections of loads and PV systems. Simulation results demonstrate the trade-offs between the controllers. It is observed that the distributed controller achieves greater voltage unbalance reduction than the decentralized controller, but requires communication infrastructure.

The increasing penetration of renewables has driven power systems to operate closer to their stability boundaries, increasing the risk of instability. We propose a multiperiod optimal power-flow approach that uses demand responsive loads to improve the steady-state voltage stability, which is measured by the smallest singular value (SSV) of the power-flow Jacobian matrix. In contrast to past work that employs load shedding to improve the stability, our approach improves the SSV by decreasing and increasing individual loads while keeping the total loading constant to avoid the fluctuation of the system frequency. Additionally, an energy payback period maintains the total energy consumption of each load at its nominal value. The objective function balances SSV improvements against generation costs in the energy payback period. We develop an iterative linear programming algorithm using singular value sensitivities to obtain an ac-feasible solution. We demonstrate its performance on two IEEE test systems. The results show that demand response actions can improve the static voltage stability, in some cases, more cost effectively than generation actions. We also compare our algorithm's performance to that of an iterative nonlinear programming algorithm from the literature. We find that our approach is approximately six times faster when applied to the IEEE 9-bus system, allowing us to demonstrate its performance on the IEEE 118-bus system.