Investigation of Demand-Side Management Techniques for a Sample Residential Load Profile Improvement in Pakistan /

Abstract

This research work is based on the Demand Side Management (DSM) of the residential load of Pakistan through peak shaving using a Battery Energy Storage System (BESS). The secondary data set used in this study is the PRECON data set that is composed of per-minute electric power usage of 42 houses of different demographics in Lahore, Pakistan. The in-depth analysis of the PRECON data set is made based on demographics and also through K-Means clustering of metadata. A distribution network synthesis is performed in MATLAB to design unstructured communities consisting of a user-specified number of households and feeders. DSM is done at household and community levels in System Advisor Model (SAM). At the household and unstructured community level, DSM is also done through an automated algorithm that determines peaks and valleys from the load curve of a given house for battery scheduling. At the structured community level, a third-party distributor, other than the consumer and utility, is supposed to buy electricity from the utility at off-peak hours and provide it during peak hours at a cost lower than the official peak-hour cost. The original data is classified into three different scales, representing three structured communities, through K-means clustering and feeder synthesis is done through uniform distribution in a Monte-Carlo simulation. After feeder synthesis, peak shaving is done using BESS for each feeder by performing a techno-economic feasibility analysis in SAM. The results of this research work show that it is economically and technically beneficial to install the BESS at the community level than at the household level. For individual households, the DSM using an automated algorithm resulted in 76% of houses with 2.5 to 29% better load factors. At the structured community level, this work is equally valid to profit the distributor, reduce the electricity billing costs and improve the load factor from 1 to 7%.