Planning And Restoration Strategies In Resilient Power Systems

With the development of large-scale power systems and power industry restructuring,modern power systems tend to operate close to their limits.Although well-developed protection and automatic control devices can effectively enhance the robustness of modern power systems,it is almost impossible to completely prevent power outages or blackouts from occurrence.Due to the increasing occurrence of extreme natural disasters and human attacks,the development of resilient power systems has attracted extensive attention,with an objective of quickly and flexibly responding to extreme disturbance events.In order to improve the resilience of a power system,adequate preparation and preventive measures should be taken before a disturbance event actually occurrs,while appropriate power system restoration planning should be made to accelerate the power supply to interrupted customers after an outage or a blackout.Given this background,this dissertation is devoted to addressing potential issues of enhancing the resilience of power systems.The concerned issues are systematically investigated from two perspectives,namely planning methodologies and restoration strategies respectively.Specifically,the main contents of this dissertation can be summarized as follows:1)Optimal placements of smart substations in a distribution system with reliability indices considered.An optimal placement model for smart substations is presented to minimize the upgrade costs of existing conventional substations and the interruption costs of customers.Two reliability indices including the system average interruption duration index(SAIDI)and average energy not supplied(AENS)are considered with tolerable upper limits respectively specified.A fault clearing model is next proposed and the power supply interruption duration of customers following a fault is evaluated.By linealizing the power supply interruption duration and interruption cost function,a mixed integer linear model for the optimal placements of smart substations is attained,and then solved efficiently by commercial solvers.The IEEE RBTS-Bus 4 distribution power system and an actual medium voltage distribution network in Denmark are served for demonstrating the basic characteristics of the presented method.2)A black-start zone partitioning strategy for the parallel restoration after a blackout is proposed.Power system restoration after a blackout or a local outage can be speeded up by employing a parallel restoration strategy.With this in mind,a two-step strategy for black-start zone partitioning is hence proposed.In the first step,a grouping model of generators to be restored is presented for minimizing the restoration time of the generators and hence restarting them as soon as possible.In the second step,a graph partitioning model for a power system is developed to minimize the number of the interconnected lines and to maximize the electrical distance of the interconnected lines among different restoration subsystems.The well-developed commercial tool CPLEX is employed to solve these optimization models.Furthermore,a simplification principle is applied to the network topology with the completeness of line information maintained so as to reduce the computation load.Finally,the New England 10-unit 39-bus system and a part of Zhejiang provincial power system are employed to demonstrate the developed models and method.3)An optimization strategy for network recon.figuration is presented with centralized charging stations of electric vehicles as black-start power sources.Supported by large-scale centralized charging stations of electric vehicles(EVs),the battery swapping mode with battery leasing is a commercially competitive way for the development of EVs.Given this background,a network reconfiguration strategy taking the EV charging stations into account is presented.First,the capacity provided by the batteries in an EV charging station is modeled,and thus the capacity for restarting the non-black-start units can be obtained whenever a power outage occurs.Then,a bi-level optimization based model for the network reconfiguration is proposed.In the upper-level optimization model,the recovery time of a generating unit is determined by maximizing the restored generation capacity,while in the lower-level the restoration path is optimized by minimizing the charging capacitance of the recovered lines.A chance-constrained programming method is employed to address the risks arising from uncertain factors during the system restoration,and then a bi-level optimization model for the network reconfiguration based on chance-constrained programming is presented.An improved particle swarm optimization algorithm is employed to solve the developed model.Finally,the modified New England 10-unit 39-bus power system and a part of the Guangdong power grid in China are employed to demonstrate the basic characteristics of the developed model and method.4)An optimal network reconfiguration strategy is proposed for power system restoration considering generator start-up sequence and load pickup.A mathematical model for skeleton-network restoration is proposed taking the generator start-up sequence and load pickup into consideration,which is decomposed into three mathematical models,including the generator start-up sequence(GSUS)model,transmission line restoration(TLR)model and load pickup(LDP)model.The GSUS model is aimed at maximizing the overall system generation capability with the actual power system topology taken into account.The TLR model is proposed to attain an optimal skeleton-network by identifying the restoration sequence of transmission lines.Certain loads are restored to keep the voltage and frequency within the specified respective thresholds and the load amounts are optimized in the LDP model.The performance of the proposed method is demonstrated by two case studies on the New England 10-unit 39-bus system and an actual power system in Guangdong,China.5)A method for determining the optimal restoration paths for power systems is pretented considering the failure risk of restoring transmission lines.Putting unloaded transmission lines into operation is one of the main steps of network reconfiguration in power system restoration.Whether an unloaded transmission line could be restored successfully or not may have a significant impact on the speed of the whole restoration process.Given this background,an approach to determine optimal restoration paths for a power system is presented with the failure risk of restoring transmission lines taken into account.First,the factors having impacts on restoring transmission lines are analyzed,including the charging capacity and restoration time of the transmission lines.Then,the relative probability of line restoring failure is defined.Based on the severity of line restoring failure,the line restoring risk is defined.Considering that the uncertainty of the restoration time of the lines leads to the uncertainty of the line restoring outcome,two robust optimization models are presented to minimize the total line restoring risks,which are applied to the "series" and "parallel" restoration stages,respectively.The highly efficient commercial solver CPLEX 12.2 is next employed to solve the developed two robust optimization models.Finally,the New England 10-unit 39-bus power system and a part of the modified Guangdong power grid in China are served for demonstrating the basic characteristics of the developed models and methods.The proposed planning theories and restoration strategies provide technical supports for enhancing the resilience of power systems,and enrich the theories and methods for power system planning and restoration.