Planning of optimal location and sizes of distribution transformers using integer programming

During planning of the development of medium voltage distribution networks, new distribution transformers medium/low voltage can be developed on more than one location and have more options of connecting to the network. On the other side, planning of development medium-voltage distribution network usually includes optimization of location and rates of substations as well as optimal locations and feeder shaping, while the problem of location optimization and rate of distribution transformers has not been treated in literature so far available. Problem is brought down to associating distribution transformers geographically in to the supply area, associating one supply area to one distribution transformer. The most frequent case is that in one part (observed) of a consumer group appears a need to build one distribution transformer station with a possibility of construction on more than one location. It is enough to apply separate calculations to all possible locations and by comparing to attain the best possible solution. General and rather rare case is when in observed part of a consumer group there is a need to develop more than one distribution transformer with possibility, in addition to be in more locations, that individual distribution transformer also to be installed in different size (for example, 630 kV A and/or 2 x 630 kV A). In that case (general) it is needed to approach with more complex selection way, i.e. with optimization. In this paper a solution has been proposed for optimization of location for a new distribution transformer by associating one consumer area to several distribution transformers that together supply the corresponding area. Optimization is being performed by appliance of network penalty factors and integer programming. Penalty factors are obtained by calculation of medium voltage network state, voltage drops and losses in every medium voltage feeder. The basic idea is to choose from the set of solutions the most qualitative one, from the view point of annual investment costs, voltage drops, load of feeders and substations. At the same time a problem of reinforcement of the existing distribution network is concerned. By observing of one area for more distribution transformers instead of one area for one distribution transformer, the exploitation level of medium voltage network is being raised, since it is being decided to build a one distribution transformer out of the given set, which leads to the least increase of power losses and voltage drops in relation to newly made investment costs. Thus obtained results can be used for more precious definition of locations and powers of distribution transformer, in solving short-term development planning problems after made development plans for distribution system in the way as it has been so far done. Test example shows that developed method can be also used as an additional tool in planning distribution networks development. (C) 2003 Elsevier Science Ltd. All rights reserved.