Research On Load Supply Capability For Distribution Networks With Distributed Generations

Rapid development of economy contributes to continuous growth of electricity consumption. However, the construction investment of distribution network lags behind load growth, which causes many networks in the bottleneck of power supply. As a result, system security is threatened. Load supply capability (LSC) is an important index of distribution network reliability, which can not only reflect the available load supply capability of the network, but also guide the planning and construction work. The access of distributed generation (DG) has brought new node types to traditional power systems, so there are new requirements put forward on the power flow algorithm. Furthermore, the power output of some DGs has a certain randomness and correlation influenced by natural factors. Undoubtedly, these changes have brought a huge challenges to the researches of LSC.This paper takes the common DGs as different node types according to their interface types and control strategies. Then different mathematical models of the nodes are established, which can be processed in power flow calculation. Node incidence matrix and path matrix are used to describe the network topology. On this basis, direct solving matrix is derived by the law of KCL and KVL. And a forward/backward sweep of the whole distribution network can be completed through an operation of the matrix. The proposed algorithm owns rather strong processing ability for different node types of DGs and looped networks.Specific to the randomness and correlation of the power output of photovoltaic generation and wind power generation, Latin hypercube sampling-based Monte Carlo simulation is used to get the light intensity and wind speed from Beta distribution and Weibull distribution respectively, from which the stochastic model of the power output can be obtained. And the order of the samples needs to be adjusted to reach the desired related level through permutation based on Spearman rank correlation coefficient. In the current operation mode, step-varied repeated power flow is developed to calculate the load growth margin. And LSC of the distribution network with DGs is evaluated by some indices. Judging from the comparison of the results, the LSC decreases in general and the fluctuation degree of the LSC increases after taking the correlation into account while there is a negative correlation between the power output of the DGs, and vice versa.