Virtual energy storage modeling based on electricity customers’ behavior to maximize wind profit

Highlights

• A new virtual storage is designed based the customers behavior.

• Utility function is used to estimate electricity customers’ behavior and price-elasticity of electricity demand.

• The proposed virtual storage is utilized to compensate wind generation uncertainty.

Abstract

Nowadays utilizing renewable energy resources (RES) has become one of the main features of the modern power systems. Despite the many benefits of these resources, the output power uncertainty limits RES competitive ability with the other conventional power producers. Using the energy storage system (ESS) is an effective solution to resolve the output power uncertainty problem. However, ESS remains to be an expensive technology although there are declinations in the cost in recent years. To this end, this paper utilizes demand response resources as a virtual energy storage (VES) in which incentive and discount payment are applied to convince the customers to reduce or increase their consumptions, respectively. The consumption decreasing and increasing provide functions similar to discharging and charging an ESS. Customer behavior plays an important role in designing an effective VES. So, this paper employs the concept of the utility function considering different risk aversion coefficients to model the different customers’ behavior. In the numerical Section, the proposed VES is implemented considering different customer types with different risk aversion coefficients to improve the wind generation profit in the day-ahead. From the results, as the risk aversion coefficient increases, a high incentive/discount is needed to convince the customer to participate in the VES.

Introduction

Nowadays, due to the finite nature of fossil fuels and environmental concerns, many energy providers are motivated to use renewable energy resources (RESs) [1], [2]. The main drawback of the most RESs is their random nature resulting in uncertainty of the output power [3]. Similar to the other power producers, RES should submit bidding offer to the day-ahead market one day before. The power producers are responsible to provide the submitted power. If the submitted power is not generated in settlement time, the power producer will be penalized economically [4]. Hence, Due to the power generation uncertainty, providing the committed power is the main challenge of RESs [5].

Numerous studies have been conducted to resolve market problems of RESs in which renewable resources are operated in joint with the other power energy resources. Among the different resources, energy storage systems (ESSs) are more suitable to increase RES compatibility with the electricity market. The ESSs have the advantage to be supportive both of overproduction and underproduction problems of the RESs. Fig. 1 illustrates different types of ESS which are classified into conventional and virtual forms. The conventional form of ESS is classified as electrochemical, mechanical, electrical and thermal energy storage [6]. Although there are declinations in the cost in recent years, the installation cost of conventional ESS is still considerable. In addition to the conventional ESS, demand response (DR) resources can be utilized as a virtual energy storage (VES). Similar to ESSs, DR can provide charging/discharging functions by intelligently managing the electricity customers’ consumption [7]. The DR resources have been considered as a VES in [8] to maximize the micro-grid profit in the real-time day-ahead electricity markets. The results have been illustrated that not only the VES does not suffer physical constraints, but also enjoys the best performance in saving the operation cost. Optimization of energy storage/demand response taking into account consumer behavior has been considered in [9], [10]. The references have considered industrial, commercial and residential customers’ behavior with the objective of guaranteeing thermal comfort.

The concept of VES has been utilized in some studies from different power system operation aspects. In [11], the authors have studied the general benefits of VES. The benefits are low carbon emission, fast response, low operational and investment cost. Ref. [12] has aggregated and controlled the air-conditioning loads to manage the network loading through the concept of the VES. To achieve the maximum efficiency and profit, ESS and VES can be operated jointly. In Ref. [11], the VES and ESS have employed the domestic refrigerators and flywheel energy storage systems to store and release energy in response to regulation signals. Also, Ref. [13] has proposed a hybrid energy management strategy for residential consumers using virtual and actual storage systems to achieve certain benefits to the consumers. Ref. [14] has analyzed the VES using distributed electric loads with thermal storage. The loads are treated as thermal cells and various analysis has been carried out. Because of the fast response of the VES, it has been utilized to mitigate unscheduled interchange caused by wind generation in Refs. [15] and [16]. Authors in [17] have utilized a VES and an ESS to support the distribution network voltage. The voltage support allows more distributed generation installation in the network.

In terms of the energy storage viewpoint, the electricity loads are divided in different groups. First group of loads, e.g. lighting loads, do not have capability to store the electricity in any energy forms. The second group of loads, e.g. refrigerators, can store the electricity energy in the other energy forms, especially thermal energy. In this case, although the stored energy cannot return to the grid, the consumption time would be flexible [18]. The third group of the loads, e.g. electrical vehicle, include battery therefore, they are capable to store energy and use the charged energy whenever it is necessary. Through the smart grid infrastructure, the stored energy also can be return to the grid [19].

In addition to the mentioned load groups, some electricity demands are sensitive to the price changes. For example, to reduce the electricity bill, consumption (e.g. washing machine) can be transferred from the peak to the off-peak periods. This group of loads cannot store the energy, but can decrease or increase the consumption with respect to the electricity price changes. The consumption increment and decrement provide functions similar to charging and discharging an ESS. In this case, welfare level of the consumers must be considered. This paper designs a novel VES model in which price signal is used to convince the customers to modify their consumption and provide charging/discharging functions.

Although various VES models have been surveyed, but most of them consider the load, especially thermal demands, management aspects of the consumption. Designing proper load management method to utilize the customers’ appliances is technically difficult. Plus, like the physical energy storage, these types of VES suffer the physical constraints such as maximum charging/discharging rate. Indeed, the previous studies have aggregated and operated virtually the physical loads as a VES limited to the physical constraints. Hence, this paper proposes a new VES model based on the incentive DR programs in which, the price signal is used to convince the customers to reduce (discharge) or increase (charge) their demand. In this case, the objectives of VES can be obtained only by propagating the price signal among the customers without involving with the technical issues. Success in implementing the DR program depends on the ability of the customer to modify his/her demand which is measured using demand price-elasticity. Elasticity evaluates customers’ behavior and demand changes in response to the price signal. Customers will participate voluntarily in the DR program if they are offered the appropriate incentives. The offered incentives can be optimized if the customers’ behavior can be estimated properly. This paper utilizes the concept of the utility function to estimate the customers’ behavior. This function represents how a rational customer would make consumption decisions. At first, a new load model is designed to measure the price signals effect on the electricity consumption and estimation the price elasticity of electricity demand. Then, affective price signals are generated to assure the VES will be operated properly which is used to maximize the wind generation profit. The main contributions of this paper are summarized as follows:

• A novel virtual energy storage is designed based on the incentive/discount signal.

• The proposed virtual energy storage does not suffer from the physical constraints like the maximum charging/discharging rate and has an immediate answer.

• The incentive/discount signal is determined considering the customers’ behavior.

• The customers’ behavior is modeled considering the concept of the utility function.

The rest of this paper is organized as follows. Section 2 describes the proposed VES and its charging/discharging model. Section 3 describes jointly operation of the VES with wind generation. Section 4 presents numerical results of the case study. Finally, Section 5 presents the conclusions.