DISTRIBUTION SYSTEM LOAD MANAGEMENT

Abstract

Distribution Automation System (DAS) provides means for maximizing utilization of distribution system so that the available generating capacity is optimally utilized causing minimum inconvenience to the consumers. It includes load factor improvement, minimization of losses and maximizing system reliability. To incorporate the above, DAS includes automation of various functions such as load management by switching and monitoring individual loads at each of the customer's end and load transfer,!.e. switching load from one feeder to another. Load management is the ability of power system to monitor and control a variety of loads in a network without creating any inconvenience to the customers in the residential, commercial and industrial sectors. This is done by shifting the time and amount of use of electric power and energy of the distribution system. For maximizing the energy delivery capability of power systems, the individual customer loads are managed. During severe emergencies which result in insufficient resources to meet load, distributed intelligence in load control^i.e. an automatic load shedding program throughout the affected area can prevent load area collapse. By rescheduling the dispersed generators and controlling the storage devices (commonly known as dispersed generators and storage devices or DSGDs), the need for new generating capacity can definitely be delayed for few years. In a country like India where generation capacity is already low if the load management is implemented as one of the several distribution automation function, the cost/benefit position will be better than for a simple management system only. Direct load control is considered as a possible tool to help such load management. As the part of the work a scheme of automatic, intelligent load shedding by directly controlling consumer loads is presented. The Demand Limiters are installed at each medium and large power consumer end which are centrally controlled by the Digital Computer (DC) installed at substation. Demand Limiter takes decision for load shedding/restoration, considering the priorities set by the consumer asynchronously. The Demand Limiter unit is developed around 8085 microprocessor system and the scheme is tested in the laboratory for three categories of loads. The change of priorities by consumer asynchronously is realized by providing the priority selector switches, one for each category of load. Past history of load shedding is also taken into account so that a low priority load will not be cut for a very long duration of time. The scheme presented specifically suits Indian conditions. This scheme will be more effective if the demand limiters are installed in the areas having significant number of medium and large power consumers. Besides direct load control, there are other methods to reduce the system load when generation and outside purchases of a power utility can not meet the demand. These are Reduction in voltage, Disconnection of interruptable devices, Voluntary cut-back of consumption by large industrial customer (typically in the case of foreseen shortage of generation), Disconnection of loads by underfrequency relays and Change in configuration by sectioning the system. Voltage reduction is used as an emergency measure for reducing the peak demand. However, effect of lowering voltage differs depending upon the composition of loads on a feeder. In many feeders constraint of voltage profile to be within the lower 5% of an acceptable band leads to energy reduction. With all the feeders having different composition of loads, the voltage profile in some of them can not be maintained as specified without significant investment of capital cost. Therefore, the effect of change in voltage on a feeder is to be studied precisely by performing system analysis before taking a decision for voltage reduction. The distribution networks of the present days are very complex, hence to carry out a complete analysis of such a system is nonviable and also it is not necessary to perform the detailed analysis because the shape of the profile is less important than knowing the extreme, as there are customers connected to all points of the feeder and no customer vpltage should be too high or too low. Instead of performing the analysis of the complete system the distribution networks can be reduced and analysis can be performed for this reduced equivalent network.

In real time control of distribution system in case of voltage regulation, automatic switching capacitors are to be switched on, after obtaining the voltage profile in primary distribution system. Requirement for fast decision in real time control operation of distribution system necessitates the need for a method to analyze the system which gives reasonably accurate solution in shortest possible time. A new improved reduction method is developed to get reduced equivalent of a distribution network (Hybrid model-I). The hybrid model-I represents the distribution feeder circuit by a point load at the end of an equivalent length with an ideal transformer; the primary of the ideal transformer is connected at the end of the equivalent length and total load on secondary side. The transformer ratio of the ideal transformer and equivalent length are so adjusted that voltage at the end point and power losses in the equivalent are same as in original feeder with dispersed load. Few examples of practical distribution network having both series configuration and parallel laterals are considered to show better relative accuracy of this developed hybrid model-I. The results are compared with the existing hybrid model. The parameters of the hybrid model-I developed are dependent upon system voltage. To improve upon this drawback a new reduction technique for distribution network analysis is developed and called hybrid model-II. This hybrid model represents the distribution feeder circuit by a point load along with a fictitious section connected at equivalent length. The fictitious section may represent load or generation unit. A computer simulation program (NETRD) is developed in FORTRAN to reduce a radial distribution network circuit (Which branch out into tree type structure) using both the developed hybrid models ,viz, hybrid model-I and hybrid model-II. Few examples are considered to show much better accuracy of both the developed hybrid models in comparison to the available hybrid model for network reduction published in literature. While performing system studies of a complex distribution network many times it is required to reduce the network keeping certain

important buses intact. Whenever the voltage reduction procedure is to be adopted as a means of reducing demand, for examining the effect of voltage, the network is reduced for fast solutions. The selective reduction procedure for retaining the sensitive or desired buses is developed introducing the new idea of bus absorption procedure. The developed reduction technique can also take care of Normally Open Points which are present in the distribution network to recover loads, Overcome voltage problem and to Prevent customer disconnection in emergency situation. The developed feeder reduction technique can also be used to reduce few feeders which are not to be considered in detail, while performing load flow analysis to get the faster solution for direct load control operation to implement load control strategies. The principle objectives of load management are to reduce the average cost of electricity, generally improve load factor and to reduce the need for generation capacity by shifting electricity use from peak to off peak periods. A new method is developed to obtain optimal operating strategy for a distribution system having various load control options like dispersed generators, dispersed storage devices and load curtailment option at selected buses. All the constraints and their derivatives w.r.t. controlled variables, are formulated for the above optimization problem; Simplex method, neglecting nonlinearities in the system, is used to start with. Next, the above problem is solved by using Charlambous least p algorithm which is a part of a software package FL0PT5. FL0PT5 is a very powerful package developed by Bandler and Sinha to solve Minimax optimization problem. An important feature of FL0PT5 is that it gives solution even when one starts from an unfeasible solution point and the constraints violated at a particular instant are only considered in the next iteration instead of all the constraints in every iteration. A complete software package (OPLCP) is developed which can be used to determine all possible optimum operating strategies with multiple iv objectives. The weight factors corresponding to different load control alternatives' in the overall objective function can be set to zero to eliminate desired objective terms. Also by choosing appropriate weight factors the desired load shapes can be developed. Two distribution systems namely 4 bus available in literature and 26 bus system that from Kanpur, a city of India are considered for study. Different load control alternatives and their effect on operating cost and load factor is evaluated.Significant improvement is obtained in these quantities. Drastic reduction in CPU time for obtaining the optimum operating strategy is achieved when nonlinear optimization problem is formulated for the system and solved through the developed program OPLCP using FL0PT5 software. For a large distribution network the requirement of memory and computation time necessitates to evolve a decomposition method in which the above problem can be solved in parts to obtain overall solution of the system. A decomposition method is developed to divide the large network containing DSGDs into small blocks which are solved individually and then the results are integrated to obtain the complete solution. The memory requirement and computation time is reduced drastically by using developed decomposition method. The usefulness of this decomposition method is illustrated using the same 26 bus distribution system containing DSGDs.