Research On Calculation Methods Of Available Transfer Capability In Power Grids

Available transfer capability (ATC) is aquantitative assessment of the power transfer reliabilitybetween interconnected grids and an important index toensure a smooth electric market transaction, which is important for power system operation and planning. In recent years, the problems about ATC such as identification of transmission sections, deterministic calculation of ATC and probabilistic calculation of ATC have received extensive attention. However, with the interconnection of bulk power grids and the access of new energy power generation, the tranditional methods meet with new challenges on two aspects:the balance between the calculation accuracy and speed and how to deal with the uncertainties. First, the weak interconnection of bulk power grids leads to more stability problems. When considering the constraint of transient stability in ATC calculation, the tranditional deterministic methods suffer from difficulties in calculation time to meet the practical demands. Second, the development of power systems leads to more uncertainties. The tranditional probabilistic approaches of ATC encounter a bottleneck in calculation time to deal with them. Besides, the wind farms and the photovoltaic power stations are new uncertainties which require more appropriate probabilistic methods to consider them.In view of this, the fast calculation methods and the probabilistic approaches about ATC problems are chosen as the research topics, and the research work are listed as follows:A novel method for automatic identification of transmission sections using complex network theory is proposed.By systematically analyzing the functions, applications and fundamental conditions of transmission sections, the automatic identification problem is divided into sub-problems. First, a new index named transmission betweenness is presented on the basis of small-world characteristics of power network to identify key transmission links from all transmission lines. Second, the network is partitioned into subareas by the Girvan Newman algorithm, and then the partition sections are obtained from the key transmission links. Third, the best combinations of partition sections are selected as the transmission sections. The numerical results with CEPRI-36system and a provincial system demonstrate that the novel method is very efficient and suitable for online analysis in large-scale power systems. The identified transmission sections can reflect the variation of topological structure and operation state and consider the uncertainties.A fast algorithm for the ATC calculation in bulk interconnected grid is proposed considering the influeces of directions and risks. The calculation framework is based on the repeated power flow(RPF)method with optimizing the constraint checking process as well as controlling the iterations by adaptive step. The static and dynamic constraints are considered in the model comprehensively. Three practical power incremental directions are used to obtain ATC values for different requirements. Furthermore, a series of ATC values are calculated accordance with the risks of transient stability. The results with a provincial system indicate that the proposed method meets the requirements of online calculation in bulk interconnected grid. Different directions and risks provide the operators with more comprehensive information about ATC.An improved algorithm for ATC calculation based on parameterization continuation power flow model is proposed. RPF has better flexibility in considering the influences of several constraints and power incremental directions while CPF has good convergence and steadiness. So CPF is combined into RPF to form an alternate iteration algorithm for ATC calculation so as to improve the tracking process of power flow point. Besides, the parameterization continuation power flow is used in checking the constraints and controlling the power of multiple transmission sections. The results of CEPRI-36system and a bulk interconnected grid system illustrate the speediness and flexibility of the proposed method in poor convergence systems.A new fast calculation method for probabilistic ATC based on Latin hypercube sampling (LHS) and scenario clustering technique is proposed.LHS is used in Monte Carlo simulation (MCS) to select a large number of system states with high sampling efficiency. Cholesky decomposition is combined into the sampling process to deal with thedependencies among input random variables. The sampled scenarios are clustered by vector quantification clusteringalgorithm. Finally, a sensitivitymethod based on optimal power flow is proposed for ATC calculation of the clustered scenarios. The case studies, with the IEEE-RTS, illustrate the advantages of the proposed method that largely reduces the computation burden under thepremise of ensuring its accuracy. The results also verify the obvious enhancement of spatially correlated wind power onthe volatility of ATC, which needs to be considered in the calculation model.A computationally accurate and efficient approach based on stochastic response surface method (SRSM) for probabilistic assessment of ATC is proposed. SRSM is used in ATC model to express the input random variables and output as functions of the standard normal random variables. Then, only a small number of samples are required to calculate the probabilistic distribution of ATC. Polynomial normal transformation combined with orthogonal transformation technique is used to deal with the correlated continuous input random variables with unknown probability distribution functions. To consider the discrete input random variables, a few simulations are added to obtain the discrete distribution part of ATC.The results with the IEE-RTS demonstrate that SRSM achieves high accuracy of evaluating ATC and requires a minimal amount of system simulations compared to MCS.It is suitable for bulk interconnected grid in the presence of wind farms.A Markov chain Monte Carlo method to simulate time series of wind and photovoltaic (PV) power is proposed, which can be used in probabilistic ATC calculation considering chronological characteristics.The method constructs Markov chain for time series of wind and PV power and utilizes Gibbs sampling to obtain the single-or multi-variable time series.Besides, the interrelation among multi-variables is considered. The wind farms and PV power plants located in the control areas of two power companies in Germany are simulated, and through the comparative analysis on statistical feature parameters the validity of the proposed model is verified.