Rural Distribution Reactive Power Planning Research Considering The Total Life Cycle Cost

Rural power grid is the important part of the electric system and is also the basic condition of the rural economy, agricultural production and rural social development. With the rapid development of social economy, the demand and operation of rural distribution power system are undergoing tremendous changes. With the rapid increase of the social demand for electric power, distributed renewable energy generation connection in distribution network, the rural distribution network structure and operation mode are becoming more and more complicated. Comparing with developed countries, the distribution network in China is backward and the investment is inadequate. The phenomenon of that the voltage quality is low caused by the unreasonable reactive power device configuration and the low rate of operation, and the problems of a high network power loss are existing for a long time. The strategy of sustainable development was presented in our country in the new period, and the principles of saving energy, reducing power loss and lean management are needed in the important field of energy electric power construction of the people’s livelihood. The objective of the reactive power optimization is to ensure the system reactive power balance, voltage controlled and the power loss reduction, and it is also one of the important means to solve the economic and security operation of the power system. It is important for the distribution network safety, reliability and economic operation that the new ideas, new technology and new management method are required in the distribution network reactive power planning and planning scheme evaluation. In this paper, the theory of life cycle management is applied in the rural distribution network reactive power planning. The main research results as follows:1. The reactive power planning mathematical model is proposed with the objective function of maximum planning scheme life cycle net benefit in present value, and the model shows the benefit of power loss reduction vividly. An adaptive method to determine the reactive power compensation nodes was presented to ensure maximum benefit of the investment in the distribution network reactive power planning. Regarding to the defects of traditional sharing niche genetic algorithm applied to the reactive power planning of distribution network, the improved sharing niche genetic algorithm is put forward. The improved sharing niche genetic algorithm is applied to reactive power planning of distribution network which enhanced the ability of global optimization and the stability of results. The new model and the algorithm are feasible and practical by a realistic reactive power planning project.2. The planning cycle of the medium and long term planning of distribution network planning is long, so there are many uncertainties. In order to meet the load growth needs with less investment, the actual situation of distribution network planning is usually divided into several stages. The multi-stage distribution network reactive power planning model based on the lift cycle net benefit in present value was built. The initial investment, operation and maintenance cost, disposal cost of planning scheme were considered in the new model that based on the life cycle net present value of the planning scheme. The model not only showed the benefit of power loss reduction vividly, but also considered the distribution network reactive power demand in every stage, so the project investment was optimized. The particle swarm optimization based on classification and correlate to solve the mathematical model. The new algorithm’s global searching capability was improved. Finally, the new model is feasible and practical by a realistic reactive power planning project.3. It is one of the basic requirements of the smart distribution grid that improving the adaptability to distributed generation and having high economy and efficiency for distribution equipment. The distributed generation has a significant impact on the network loss and node voltage. The probabilistic distribution and the power output ability of the grid-connected distributed generation system are analyzed, and that some types of the grid-connected distributed generation have a strong reactive power capability is demonstrated. The reactive power planning model based on the life cycle net present value and the reactive power output ability model is established. The reactive power demand of the distribution network can be met with less reactive compensation. The improved algorithm based on the chaos multi-swarms particle swarm optimization algorithm is presented. The new algorithm’s global searching capability is improved. The new model and the algorithm are feasible and practical by a realistic reactive power planning project.4. In the traditional distribution network reactive power planning, the reactive power planning compensation is not enough or wasteful usually, due to load model is not accurate, and the expected loss reduction targets and earnings cannot to be gotten. The reactive power planning model based on the lift cycle net benefit in present value was built. The active power and reactive power time sequence characteristics of different types of the typical day power loads in different seasons(agricultural load seasonality is obvious) were considered in the model. Considering the entire sequence of the different types of the typical day loads in different seasons with reactive power planning model can accurately deal with the load uncertainty, and the planning result is not only closer to the reality, but also in favor of a true reflection of the economic and technical indicators. The time sequence complementary characteristic of the power load was considered in the new model, so the investment of reactive power compensation equipment is saved vividly. The model solving is more complex because the timing sequence characteristics are considered. The two-stage distribution network reactive planning solving method is presented, and the search space and difficulty of solving the problem is greatly reduced. Finally, the distribution reactive power planning model is solved by chaotic particle swarm algorithm, and the planning results prove the correctness and effectiveness of the model and method.