OPTIMAL PLACEMENT OF PROTECTIVE DEVICES IN DISTRIBUTION SYSTEM

Abstract

Faults are more frequent in distribution system, as compared to other parts of the power system, leading to supply interruptions to the customers. For maximizing the customer satisfaction and retention, improvement of service reliability is a major concern for electric utilities. The main goal of the power distribution system reliability evaluation is the prediction of the service security of the customers. Service reliability can be improved by placing switches and reclosers at appropriate locations in the distribution system, so that supply from the main substation to the healthy load points can be maintained uninterrupted after isolating the faulted feeder section of the distribution system. Therefore, a strategy for optimal placement of the switches and reclosers needs to be evolved for improving the distribution system reliability. Proper locations of the protective devices (switches and reclosers) must be carefully chosen in order to maximize the benefits of placing these devices in a distribution system. To address this issue, in this thesis, a formulation for optimal placement of switches and reclosers in a distribution system for maximizing distribution system reliability, while minimizing the associated investment and outage costs has been proposed. The proposed formulation has been tested on 13-bus, 58-bus and IEEE 123-bus test systems using evolutionary programming (EP), genetic algorithm (GA), differential evolution (DE) and mixed-integer nonlinear programming (MINLP) methods. The obtained results establish the superiority of the MINLP method over the other optimization methods for the said purpose. However, the initial formulated optimization problem has considered only deterministic values of loads and system data. The input parameters used for reliability evaluation may contain errors as they are derived from historical records. Hence, in order to achieve more realistic reliability indices, system components’ data uncertainties need to be taken into account. To address this issue, Monte-Carlo simulation (MCS) method is used to model the load variation, failure rate ( ) and repair rate ( ) of the system components. The main demerit of MCS is the time consuming iterations making MCS unsuitable for most of the case studies, especially for large systems. Application of point estimate method (PEM) for probabilistic calculations, incorporating uncertainty of parameters, can provide similar results of acceptable accuracy but with less numerical efforts as compared to MCS. The uncertainty associated with the failure rate, outage time (r) and load (L) may be expressed in terms of the expected value (mean) and standard deviation of these quantities with an assumption that they are normally distributed. The PEM, used to calculate the statistical moments of a random quantity which, in turn, is a function of one or several random variables, has three prevalent versions, namely 3-point i estimate method (3PEM), 5-point estimate method (5PEM) and 7-point estimate method (7PEM). It is well established that while for calculating lower order statistics only (mean and variance), 3PEM is sufficient, for calculating higher order statistics (skewness and kurtosis) along with the lower order statistics, 5PEM and 7PEM are more useful. This thesis presents a formulation for an optimal placement of switches and reclosers in a distribution system for maximizing the distribution system reliability considering uncertainties in load data, system failure and repair rates. The uncertainties have been incorporated in the formulation using 3PEM. The proposed formulation has been tested on 13-bus, 58-bus and IEEE 123-bus test systems using DE and MINLP methods. The obtained results establish the effectiveness of the consideration of data uncertainties in maximizing utilities’ profits as also in improving the distribution system reliability by providing the bounds of profit. However, the formulation considered sustained interruptions (caused by permanent faults) only, and hence the scope for inclusion of momentary interruptions (due to temporary faults) has been explored next. In a distribution system, momentary interruptions are more frequent than the sustained interruptions. Till recently, sustained interruptions were the main concern of the utilities and, hence, the protective devices were placed to limit the impact of these. However, these days, loads are more sensitive to momentary interruptions due to proliferation of electronic devices. Due to the increased use of electronic and precision devices, damages due to short-duration voltage disturbances have increased. The utilities employ fuse-save and fuse-blow schemes to decrease the impact of sustained and momentary interruptions, respectively. In the fuse-save scheme, an upstream recloser or circuit breaker operates, before a fuse can trip, to isolate a fault downstream of the fuse. Fuse-save scheme is used with an instantaneous relay or with the fast curve of a recloser associated with a circuit breaker. For temporary faults, service to the customers can be restored immediately by reenergizing the line, resulting in decreased sustained interruptions. The main drawback of fuse-save scheme is that all customers downstream of a recloser or circuit breaker experience momentary interruptions even for permanent faults downstream of the fuse. Because of this, many utilities prefer to use the fuse-blow scheme over the fuse-save scheme. In fuse-blow scheme, the fuse operates for all the downstream faults (temporary and permanent), resulting in sustained interruption for all the customers downstream of the fuse while rest of the system remains uninterrupted. To address the above issue, the effect of temporary faults has also been incorporated next in the optimal placement problem of protective devices in the distribution system. Three different scenarios, for optimal placement of protective devices (switches, reclosers, fuses) in a distribution ii system, considering uncertainties in loads, temporary and permanent failure rates and repair rates have been formulated. The three versions of the formulated problem have been solved for 58-bus and IEEE 123-bus distribution networks using MINLP optimization technique. Apart from the above three scenarios of protective devices’ placement in distribution system, other scenarios pertaining to different combinations of protective devices are also feasible. Each scenario will give a different optimal profit value for a given system, hence, the best scenario needs to be identified. Thus, it becomes necessary to develop a generalized formulation which can simulate any desired scenario and help a utility in deciding the best possible combination and optimal placement of protective devices for profit and reliability maximization. In this thesis, a generalized model has been developed to address the difficulties pertaining to placement of various combinations of protective devices (recloser, switch, fuse-blow fuse and fuse-save fuse) in a distribution network for increasing the profit of the utility through reliability improvement. The uncertainties in temporary failure rates, permanent failure rates, repair rates and load data have been considered in the formulation using 3PEM. The developed objective function is capable of simulating different combinations of the protective devices. The formulated problems have been solved for 58-bus and IEEE 123-bus distribution networks using MINLP optimization technique. After analyzing the test results of the various scenarios for the two test systems, it is concluded that maximum profit to the utility is accrued by using the one involving a combination of all the four protective devices viz. reclosers, switches, fuse-blow fuses and fuse-save fuses. Optimal placement of protective devices in distribution system increases the system reliability by isolating the faulty feeder section of the system and supplying uninterrupted power to healthy feeder sections (upstream of the faulty feeder section). However, the healthy feeder sections downstream of the faulty feeder section remain de-energized until the faulty feeder section is repaired and reenergized. If a distributed generation (DG) is present in the downstream isolated healthy part of the system, it can further improve the system reliability by operating in an islanded mode. For the formation of an island, the DG capacity should be sufficient to avoid load shedding or load prioritization. Thus, integration of DG in distribution networks has added advantages: additional reduction in customer interruption duration and increase in service restoration speed. However, the presence of DG in distribution system increases the complexity of the optimal placement problem of protective devices which has been addressed next. In this thesis, the effect of DG has also been incorporated in the formulation of optimal placement problem of protective devices’ (recloser, switch, fuse-blow fuse and fuse-save fuse) in the distribution iii system. A model has been developed to solve the problem of the protective devices placement in various zones of a distribution system with DG. The uncertainties in temporary failure rates, permanent failure rates, repair rates and load data have been considered in the problem formulation using 3PEM. The formulated problem has been solved for 69-bus and 118-bus distribution systems using MINLP optimization technique. After analyzing the results of the two test systems, it can be concluded that the profit to the utility can be increased if the protective devices are placed optimally in the zones formed due to DGs connected in the system.