Impact of Demand Side Management On Unit Commitment Problem Manisha Govardhan Department of Electrical Engineering S.V. National Institute of Technology, Surat, India manishagovardhan@yahoo.com Ranjit Roy Department of Electrical Engineering S.V. National Institute of Technology, Surat, India rr@eed.svnit.ac.in Abstract— Growing energy demand and its remedial solution is a great challenge of power industries. This paper deals with demand side management in unit commitment using two essential approaches of energy balance (i.e. energy shifting) and load reduction during peak hours. Demand side management is a crucial job of reshaping the inconsistent load demand either by shifting demand from peak period to off-peak or low load period or reducing load demand during peak period. This ultimately results in energy saving and cost reduction of a system. A test system of 26 generating units is considered for simulation study using gbest artificial bee colony algorithm. The test results confirm significant savings in cost and energy consumption while peak load demand is reduced. Keywords- Demand response (DR); Demand side management (DSM); Gbest artificial bee colony algorithm (GABC); Unit commitment (UC) . I. INTRODUCTION Demand side management (DSM) is a dynamic solution against emergent energy demand and fuel price, concern for climate change and seasonal uncertainty in load demand. Basically, DSM has been classified as direct load control (DLC), indirect load control and local energy storage [1]. DLC permits utility to regulate isolated customer’s load demand independently; indirect DLC involves customer participation for their load demand deviations according to the change in electricity price and local energy storage allows both customer and utility to preserve energy during off-peak and low load period and utilize the same during peak period. DSM also has a great influence on deregulated electricity market [2]. In general, electricity price is low during off-peak and low load period and high during peak period. DSM encourages end user entities to shift their load demand from peak to off-peak period or reduce energy consumption during peak period which in turn results in energy saving and cost reduction. Furthermore with DSM policies, optimal scheduling of generated power is also again a challenging task. Hence the goal of this paper is to solve UC problem considering DSM. UC can be specifically outlined as a large-scale, nonlinear, mixed integer combinatorial optimization problem which involves two critical sub-problems of power industry namely unit commitment (UC) and economic load dispatch (ELD). UC decides on/off status of generating units and ELD distributes generated power among the committed units while satisfying several constraints. Recently, many demand response program has been reported for DSM. Furthermore, a few demand response based unit commitment (DRUC) models are discussed in literature. A model of emergency demand response (EDRP) and interruptible load contracts (ILC) in UC problem is proposed in [3] to minimize the energy consumption during the critical or peak period of the day. The study has been carried out with a load demand of Iranian grid and developed EDRP cost component added to generation cost. Another UC problem associated with DR program model (UCDR) to study the environment and an economic effect of DR program is suggested in [4]. These DRUC models have considered the cost of DR in addition with the total cost. This paper considers direct load control (DLC) approach in which operator autonomously decides for load flattening and cost reduction. Two basic approaches of energy balance and peak load clipping is considered to enhance the load demand profile and then UC problem has been solved with modified load profile [5]. Gbest Artificial Bee Colony algorithm has been executed for simulation study of 26 generating units test system. The construction of this paper is as follows: structure of UC with DSM policies and associated constraints has been described in Section II. Section III conveys brief introduction of Gbest Artificial Bee Colony algorithm. The results for different test cases are discussed and compared in Section IV. Finally, conclusion is drawn in Section V. II. UC PROBLEM FORMULATION WITH DSM A. Demand Side Management Two different strategies of DSM namely energy balance and peak load saving has been employed and their effect on load demand profile has been studied [5]. This study is carried out with the assumption that DSM has been performed by system operator at remote end. From earlier energy demand and electricity price statistics, operator autonomously imposes varying energy supply according to the change in electricity price to enhance load profile and system cost. a) Energy balance: The significance of energy balance is to shift load demand from a peak period to low and off-peak 2014 International Conference on Control, Instrumentation, Energy & Communication 978-1-4799-2044-0/14/$31.00©2014IEEE 446