Energy Efficiency (2014) 7:271–284 DOI 10.1007/s12053-013-9223-9 ORIGINAL ARTICLE Integrating renewables economic dispatch with demand side management in micro-grids: a genetic algorithm-based approach Ahmer Arif · Fahad Javed · Naveed Arshad Received: 14 February 2012 / Accepted: 6 August 2013 / Published online: 8 September 2013 © Springer Science+Business Media Dordrecht 2013 Abstract Economic dispatch and demand side Keywords Demand side management · management are two of the most important tools Economic dispatch · Smart grids · Micro-grids · for efficient energy management in the grid. It Intelligent energy management is a casual observation that both these processes are intertwined and thus complement each other. Nomenclature Strategies aiming to optimize economic dispatch have implications for demand side management VC Cut-in speed of wind generator techniques and vice versa. In this paper, we VR Rated speed of wind generator present a genetic algorithm-based solution which VF Cutoff speed of wind generator combines economic dispatch and demand side C The set of consumers under the micro-grid management for residential loads in a micro-grid. S The set of currently available energy Our system collects preferences of demand data suppliers to the micro-grid from consumers and costs of energy of various μx Total energy demand by consumer x sources. It then finds the optimal demand schedul- μx,h Total energy demanded by consumer x at ing and energy generation mix for the given time hour h window. Our evaluations show that the given ap- αy,h Total energy available from supplier y at proach can effectively reduce operating costs in a hour h single- and multiple-facility micro-grids for both λy,h Price of energy from supplier y at hour h suppliers and consumers alike. βx,y,h Total energy allocated to consumer x from supplier y at hour h A. Arif · F. Javed · N. Arshad (B) Department of Computer Science, Introduction School of Science and Engineering, Lahore University of Management Sciences, DHA, Sector U, Lahore, Pakistan Economic dispatch and demand side management e-mail: [email protected] (DSM) are two of the most important tools for A. Arif efficient energy management in the grid. Whereas e-mail: [email protected] economic dispatch is the task of finding the right F. Javed and optimal mix of energy for the given demand, e-mail: [email protected] demand side management aims at managing end 272 Energy Efficiency (2014) 7:271–284 user demand to reduce cost. There are various devices and their average generation and con- ways to achieve this goal, the most common way sumption similar to the works of Shivakumar et al. is by shifting movable loads to off-peak time or (2013), Livengood and Larson (2009), Ranade and times of lower cost. The goal is to reduce en- Beal (2010), and Conejo et al. (2010). Study of ergy cost by balancing energy resources against integration of power flows in the micro-grid and demand. integration of different energy sources for a robust It is a casual observation that both these power delivery is the next step of deploying this processes are interlinked. Economic dispatch tries system but is beyond the scope of the current to reduce cost according to available demand and work. DSM tries to manage demand to reduce cost. It The remaining sections of the paper are orga- stands to reason that if both tasks are consid- nized as follows: In “Economic dispatch and DSM ered together, where DSM and economic dispatch opportunities in smart grids,” we discuss the op- planning include their interdependent constraints, portunities for economic dispatch and DSM that a better solution may emerge. have emerged through research in smart grids Specifically, in renewable solution imple- and micro-grids. In “Operational system,” the mented on a small scale by the consumers, this operation system is discussed and in “Problem aspect becomes more evident. In these systems, formulation,” the problem present in this opera- typically, the excess low-cost energy from renew- tional system is modeled and design objectives able sources is either stored in battery systems or are discussed. In “Genetic algorithm,” the ge- sold back to energy retailers up the grid. These ap- netic algorithm formulation and its algorithm is proaches miss the opportunity to take advantage discussed. “Model implementation” specifies im- of the localized aspects of micro-grids and incur plementation details for the various steps for the significant costs associated with often expensive, genetic algorithm, and finally, in “Evaluations and inefficient energy storage systems and transmis- results,” evaluation of the numerical results and a sion losses upstream to the grid. By more immedi- summary of the findings are presented. ately utilizing this excess energy, these losses are minimized and equal or lower energy costs can be achieved with potentially fewer distributed re- Economic dispatch and DSM opportunities newable resources, creating savings in the installa- in smart grids tion, maintenance, and operating costs associated with them. There are two developments in recent times which In this paper, we present a strategy to inte- have provided new opportunities for economic grate DSM and economic dispatch for renewables dispatch and DSM. On one hand, smart grid ini- in a localized setting. In our strategy, user pref- tiatives have provided greater surgical control of erences and generation capacity are encoded as user-side power consumption (Coll-Mayor et al. constraints of the system. With these constraints, 2007; Farhangi 2010; Ipakchi and Albuyeh 2009). an optimization objective function is constructed This is being used as an opportunity by various from the cost of generation for each unit for each researchers to define fine-grained DSM (Conejo hour. Afterwards, a genetic algorithm is used to et al. 2010; Gudi et al. 2011; Livengood and find the optimal solution for the problem. Our Larson 2009; Ranade and Beal 2010) as opposed results show that this integrated methodology to the blanket DSM employed in traditional grids is up to 15 % more cost-effective than state- (Amjady 2007; Gellings and Chamberlin 1993; of-the-art independent DSM and economic dis- Shahidehpour et al. 2002;Weron2006). On the patch models for a neighborhood as compared to other hand, distributed generation and renewable our combined DSM and economic dispatch im- sources have enabled the emergence of micro- plementation for an integrated neighborhood grids within which more efficient and exact eco- approach (Gudi et al. 2011). nomic dispatches are possible (Jiayi et al. 2008; Our study is focused on the abstraction of Lasseter et al. 2002; Tsikalakis and Hatziargyriou houses, renewable energy resources (RES), and 2008; Yalcinoz et al. 2001). Energy Efficiency (2014) 7:271–284 273 Demand-side management programs in tradi- economic dispatch in micro-grids (Lasseter et al. tional grids usually involved manual control of 2002). A genetic algorithm-based approach by end user devices (Gellings and Chamberlin 1993). Yalcinoz and colleagues can also be found in the Automation is proposed by some authors, but the current literature (Yalcinoz et al. 2001). However, control was blanket in the sense that the par- all of these systems cater to economic dispatch ticipating devices were either switched on or off independent of DSM. (Albadi and El-Saadany 2007). There is no con- Recently, a series of researchers have de- cept of intelligent planning for the operation of veloped economic dispatch solution incorporat- these devices. ing the demand response saving. For instance, Fine-grained DSM strategies, on the other Behrangrad et al. presented an economic dispatch hand, can perform complicated automated device model which uses the demand response as spin- scheduling based on consumer preferences and ning reserve (Behrangrad et al. 2011). Parvania electricity prices. Furthermore, with the advent and Firuzabad presented an economic dispatch of more dynamic forms of pricing, they have be- model which used the demand response planners come increasingly flexible in order to adjust for (DRPs) as an integral part of grid and mod- fluctuations in the cost of electricity. These DSM eled the saving realized by DRPs in planning techniques usually manage energy within a sin- optimal economic dispatch (Parvania and Fotuhi- gle household and typically involve capitalizing Firuzabad 2010). However, in both the instances, on variable pricing by energy providers and in- economic dispatch does not actually plan house- tegration of house-based renewables for cheaper hold consumption in the same way as Conejo et al. energy utilization. Several papers have already (2010), Livengood and Larson (2009), or Ranade been published which relate to such techniques. and Beal (2010) do. An example is the work presented by Livengood As demonstrated later in our experiments, and Larson (2009) which uses a stochastic pro- treating DSM and economic dispatch as indepen- gramming solution to optimize user-side demand. dent modules can result in suboptimal power pric- However, the approach focuses on optimizing ing across the grid. Instead, in this paper, we com- for a single household at the granularity of in- bine fine-grained DSM with economic dispatch of dividual consumer devices. On the other end of micro-grids and attempt to find an optimal sched- the spectrum is the probabilistic demand shaping ule for both under a single optimization model. and peak-shaving approach outlined by Ranade This way we are able to leverage the power of and Beal (2010) which concerns itself with reduc- DSM to arrive at better power dispatches, result- ing demand by multiple consumers during peak ing in a lower cost than independent DSM and hours but does not focus on actual scheduling of economic dispatch systems. activities. We specifically consider the micro-grid devel- Economic dispatch in the scope of micro-grids oped within the EU R&D MG project and pre- has also generated a fair amount of interest.