Study On Optimal Allocation Of Intermittent Generation Based On Probabilistic Load Flow

Coupled with economic development and scientific advancement, human have placed greater stress on environmental protection while trying to meet the increasing demand of energy. New energy resources such as WG, PV, based on renewable clean resources, have thus received wide attention and application for their capacity to conserve energy and reduce gas emission. However, with obvious intermittence and randomicity of power output in consequence of natural climate changes, more uncertain factors are brought into system after integration decreasing stability and reliability. As a result, this paper are dedicated to probabilistic power flow and optimal allocation of WGs considering the correlation of wind velocity in various wind farms; optimal allocation of WGs and PVs based on output complementary; and comprehensive optimal allocation of intermittent DGs considering reactive compensation. The details are as follow:1. As for the correlation of wind velocity in various wind farms, this paper has put forward a simulation calculation method considering this correlation. Assuming the wind velocity in wind farms are all in line with normal distribution, probabilistic power flow taking wind velocity correlation into account can be calculated through random sampling technology, linear transformation and Monte Carlo simulation method. And on the basis of steady state equivalent circuit of induction generator, a coefficient-variable quadratic polynomial model is proposed to describe its reactive consumption in power flow calculation. The influence of different wind velocity correlations is analyzed on probabilistic density of integration voltage and probabilistic distribution of transmission power on related branches, and also the allocation of reactive compensation capacity at integration point. Results have demonstrated relatively considerable changes occurred in integration voltage and power probabilistic distribution on branches closer to wind farms under the circumstance of different correlations.2. With introduction of semi-invariant random variables, an analytical algorithm of probabilistic power flow considering correlation is proposed in this paper. In other words, several independent random variables are introduced in the first place. Then, according to those correlation coefficients of wind velocity in each wind farm, the wind velocity in the rest wind farms are fitted as the sum of that of a reference wind farm and the introduced random variables. After discretizing wind velocity, system power flow and its probability of different wind velocity combinations are worked out. Finally, with Von Mises function, the system probabilistic power flow under of all wind velocity combinations can be figured out. Assuming the wind velocity in wind farms are all in line with normal distribution and making comparison to the results of the simulation method considering wind velocity correlation, the validity and accuracy of this analytical algorithm are testified. Further this analytical model in the solution when wind velocity in each wind farm is under Weibull distribution, with different types of wind-fueled generator, calculation of system probabilistic power flow under low, median, strong positive and strong negative correlations are conducted. And the results have shown remarkable changes in system probabilistic power flow in compliance with parameters of different WGs.3. Based on the probabilistic power flow analytic algorithm considering wind velocity correlation proposed above, optimal allocation of wind farms is proposed. Under three different optimization objective functions-allocation for system maximum wind power installation capacity, allocation for optimal economic operation of wind farms and allocation for optimal reactive compensation of wind farms, research on different types of WGs are conducted setting voltage limitation violation probability at integration point as constraints. Besides, the optimal allocation scheme and compensation scheme of three objective functions under different correlations between all the wind farms. The optimization has demonstrated relatively remarkable changes in optimal scheme due to the difference of WG and wind velocity correlations.4. Complementarity analysis of the power output of intermittent DG on time domain and regional domain has given birth to its optimal allocation model. Given the probabilistic distribution of power output of DG and PV in each period of a typical day and the distribution level of system load in different periods, a typical day is divided into many parts. Then, an objective function aimed at comprehensive optimization of electricity sale, investment, line loss, gas emission and system voltage in an entire period is established with opportunity-constrained planning method applied in each part. The results have shown the great impact that difference and complementarity of the probabilistic distribution of system load and intermittent DGs’ power output in the whole time-and-space area have on the optimal allocation scheme of DG and system operation.5. Intermittent DG and reactive compensating capacitor can both help to improve system voltage and reduce network loss, nevertheless, separate independent allocations of them have hardly turned out to be satisfying. In order to solve this problem, the comprehensive optimization of DG and reactive compensating capacitor has been conducted through opportunity-constrained planning method, which has taken full advantage of compensating capacitor to improve system probabilistic power flow, raise expectation of node voltage and increase its probability of remaining in the normal scope. Results have favored this method for reducing overall investment cost of intermittent DG and compensating capacitor and achieving a comprehensive optimization of economy, environment and voltage quality in operation.