An increasing amount of low carbon technologies(LCT)such as solar photovoltaic,wind turbines and electric vehicles are being connected at medium and low voltage levels to electric power networks.To support high-level ...An increasing amount of low carbon technologies(LCT)such as solar photovoltaic,wind turbines and electric vehicles are being connected at medium and low voltage levels to electric power networks.To support high-level decision-making processes,the impacts of the LCTs on large numbers of different types(e.g.,rural,suburban,urban)of distribution networks need to be fully understood and quantified.However,detailed modeling of large numbers of real-world networks is challenging for two reasons.First,access to real-world network data is limited,and second,cleaning the data requires a significant amount of time,even before modeling of the networks.This paper offers a novel systematic methodology aimed at identifying and quantifying the key electrical properties of medium-voltage level distribution networks.The methodology allows for characterizing different types(e.g.,suburban,urban)of distribution networks and obtaining'depth'dependent electrical properties of the models of the networks.Two key sets of(electrical)data were used for the study.The first set was installed capacities of distribution substations;and the second set was the conductor cross sections of the distribution lines.In the graph models of real-world networks,'nodes'represent the distribution sub-stations,switchgears,busbars and consumers locations of the network.'Links/edges'stand for the connections between the nodes through distribution lines.The results of the investigation of the real-world networks showed that,the substation capacities and the conductor cross sections could characterize the electrical properties of suburban and urban type distribution networks.The resulted probability density functions(PDF)of the electrical properties of suburban and urban type distribution networks have the potential to be directly used in generating realistic distribution network models.展开更多
基金This work was supported in part by the EPSRC Supergen Energy Networks Hub(EP/S00078X/1)UKRI EnergyRev Plus project(EP/S031898/1)EPSRC-NFSC MC2 project(EP/T021969/1).
文摘An increasing amount of low carbon technologies(LCT)such as solar photovoltaic,wind turbines and electric vehicles are being connected at medium and low voltage levels to electric power networks.To support high-level decision-making processes,the impacts of the LCTs on large numbers of different types(e.g.,rural,suburban,urban)of distribution networks need to be fully understood and quantified.However,detailed modeling of large numbers of real-world networks is challenging for two reasons.First,access to real-world network data is limited,and second,cleaning the data requires a significant amount of time,even before modeling of the networks.This paper offers a novel systematic methodology aimed at identifying and quantifying the key electrical properties of medium-voltage level distribution networks.The methodology allows for characterizing different types(e.g.,suburban,urban)of distribution networks and obtaining'depth'dependent electrical properties of the models of the networks.Two key sets of(electrical)data were used for the study.The first set was installed capacities of distribution substations;and the second set was the conductor cross sections of the distribution lines.In the graph models of real-world networks,'nodes'represent the distribution sub-stations,switchgears,busbars and consumers locations of the network.'Links/edges'stand for the connections between the nodes through distribution lines.The results of the investigation of the real-world networks showed that,the substation capacities and the conductor cross sections could characterize the electrical properties of suburban and urban type distribution networks.The resulted probability density functions(PDF)of the electrical properties of suburban and urban type distribution networks have the potential to be directly used in generating realistic distribution network models.