Model And Algorithm Of Power Flow Tracing And Its Application In Transmission Congestion Management

The reformation of electric power industry has gradually changed the operation and management of electric power system from the traditional vertical monopoly mode to transmission open access mode by splitting an original integrated system into three parts: generation, transmission, and distribution. While trying to introduce competitions and to optimize resource utilization, this reformation has brought a number of challenges to power system operation and planning, including the transmission costs distribution, network loss allocation, and congestion management. In the environment of deregulation and electricity market, it is extremely important to quickly and accurately determine the contributions of individual generators and loads to line flows, active power transfers between individual generators and loads, and distributions of power losses. Research on how to accurately determine the system usage of different market participants and fairly allocate the transmission costs and network losses among the market participants is of significant importance both in theory and in practice to mitigate transmission congestions and to assure a secure operation of the power systems.Sponsored in part by the Chongqing Natural Science Foundation projects of"Research on power flow analysis theory and its application in electricity markets"and"Research on model and algorithm of transmission congestion management in electricity markets", this thesis has completed a detailed study on the power flow tracing (PFT) technique and its application to transmission congestion management.In order to overcome the shortcomings of existing PFT methods in published literatures so far, such as time-consuming search and tracing processes and requirement for a specific tracing principle, an effective and fast analytical tracing model and its solution algorithm for PFT were proposed. The construction approach of the down-stream distribution matrix (DSDM) and the properties of DSDM, such as a diagonally dominant and upper triangular matrix were discussed. The quantitative relationship of DSDM associated with the bus power injection, generation and load power vectors was derived. The concept of distribution coefficient matrix (DCM) of the generation bus to the load bus was presented and the analytical model of power distribution between generations and loads was built. The correctness of PFT algorithm based on the DSDM was shown using the matrix theory. This method was applied to the IEEE-RTS, NewEngland-39, and IEEE-118 power systems to demonstrate the power distribution among generators, loads and transmission lines. The case studies have shown promising application results of the proposed approach.Usually, a power system can be equivalent to a lossless network for PFT studies as there are always power losses in power transmission lines. As we know, the current in power systems is lossless and its real and imaginary parts are orthogonal. Based on this property, an effective analytical model and algorithm for tracing current using extended incidence matrix (EIM) were introduced to improve the accuracy of power flow tracing and loss allocation. The concept and construction approach of EIM were presented. The properties of the EIM, such as the property of the sum of the row and column elements, were discussed. A network can be decomposed into two separate networks: the real and the imaginary current networks based on an AC power flow solution. The EIM corresponding to each current network can be also derived. Based on the generation injections and load current vectors resulted from the AC power flow solution, an analytical current tracing model between generators and loads can be established. More accurate results of PFT and loss allocation can be obtained from this current tracing model. The correctness of the current allocation and extraction can be proved using the basic electric network theorem. In other words, after current tracing, the currents injecting to a node are subjecting KCL and the voltages across branches satisfy the KVL. The major advantage of the proposed method is that the matrix theory is directly used without proportional sharing assumption on the flow distribution. The tests on a four bus power system and the IEEE-RTS power system indicated that the developed technique can be applied to any power system with or without loop flows.The power or current tracing model described above is to analyze the power allocation and extraction under specific operating conditions. This model can not meet the requirement of the real-time operation of a power systems since the loads in a power system vary with time. To keep power balance in the system, the generation outputs are also changing with time. It is necessary to re-calculate the power flow tracing in a timely manner. However, this is very inefficient or impractical in the real-time operation environment. In order to avoid the repeated calculations of PFT, the rule of power distributions and extractions varying with the change of generation outputs was analyzed according to PFT or current tracing model. The curve of power flow tracing (CPFT) model was built based on the polynomial theory and was solved by using the least square method. The Taylor series theory was used to estimate the approximate solution of the dynamic PFT model. The PFT sensitivity was presented based on the CPFT model, which can be used to express the change of line flow varying with generation outputs. The congestion on a line can be eliminated efficiently by determining the generating units and the amounts of adjustment according to the PFT sensitivity and the overload of line flows. In turn, the congestion management can be realized by eliminating the line congestion one by one. The tests on the IEEE-RTS power system and other practical systems have shown correctness and effectiveness of the proposed method.A non-linear regression model (NRM), which reflects the relationship between the results of PFT and generation outputs, was proposed according to a non-linear characteristic of power systems. The NRM was transformed to linear regression model by using the idea of transposition. The Monte Carlo method was used to simulate the examples based on PFT results. The NRM was solved by the least square method using the simulated examples. The relations between the disturbance of congested lines and that of generation outputs were obtained based on the Taylor series theory and the NRM, which are formulated to the equality constraints of the congestion management model (CMM). The CMM was formulated to minimize the total costs and was solved by using the Lagrange multiplier method. Tests on the IEEE-RTS power system and other practical systems show that the proposed method is effective.