Elsevier

Energy Conversion and Management

Volume 76, December 2013, Pages 314-322
Energy Conversion and Management

Experimental validation of a real time energy management system for microgrids in islanded mode using a local day-ahead electricity market and MINLP

https://doi.org/10.1016/j.enconman.2013.07.053Get rights and content

Highlights

  • An algorithm is developed to enhance Microgrid performance.

  • Local energy market cost model is proposed to obtain the cheapest price.

  • Several real technical and market scenarios are considered in the study.

  • Simulation and experimental results demonstrate a significant reduction in cost.

Abstract

Energy management systems (EMS) are vital supervisory control tools used to optimally operate and schedule Microgrids (MG). In this paper, an EMS algorithm based on mixed-integer nonlinear programming (MINLP) is presented for MG in islanding mode considering different scenarios. A local energy market (LEM) is also proposed with in this EMS to obtain the cheapest price, maximizing the utilization of distributed energy resources. The proposed energy management is based on LEM and allows scheduling the MG generation with minimum information shared sent by generation units. Load demand management is carried out by demand response concept to improve reliability and efficiency as well as to reduce the total cost of energy (COE). Simulations are performed with real data to test the performance and accuracy of the proposed algorithm. The proposed algorithm is experimentally tested to evaluate processing speed as well as to validate the results obtained from the simulation setup on a real MG Testbed. The results of the EMS–MINLP based on LEM are compared with a conventional EMS based on LEM. Simulation and experimental results show the effectiveness of the proposed algorithm which provides a reduction of 15% in COE, in comparison with conventional EMS.

Introduction

Modern energy management and control systems could help to reduce the cost of energy. However, they are applied either in a complex manner or a too simple way to achieve the desired goal. To maximize energy savings, minimize related costs and obtain a fast payback in MG systems, it is vital and most desirable to optimally operate an aggregated number of micro-sources to incur the lowest possible production cost. In MG, this can be achieved by applying optimization methods and adjusting the generators output to minimize the production costs. The optimization procedure may interact with public network information. For example, energy for storage devices can be bought when prices are low, and sold when required.

Various proposals for EMSs with different optimization methods and different MG structures have already been presented in literature. [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. Some references have developed optimization methods for EMSs aiming to obtain scheduling operations and optimal operating strategy [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11]. The objective function (OF) in these references allows autonomous or grid connected decision-making to determine the hourly optimal dispatch of generators depending on system constraints and market parameters. The economic concepts of EMSs in the MG market and the development of strategies to achieve such benefits are reported in [12], [13], [14], [15], [16], [17], [18].

Another investigated concept in this paper deals with demand response (DR). It is used by electric utilities to manage customer electricity consumption in response to supply conditions. Utilities encourage customers to reduce their consumption at critical times periods or in response to market prices. Currently, generation and transmission system facilities are oversized to cover peak demand plus a margin for forecasting error and unforeseen events. Smoothing such peak demand could lead to cost and size reduction of the plant. Some systems, such as DR, may encourage energy storage to arbitrage within periods of low and high demand (or low and high prices). In the literature, there are several studies investigating DR concept in MGs for different applications, for instance: demand shifting and peak shaving [19], [20], [21], [22], [23], DR exchange in which DR is treated as a public good to be exchanged between buyers and sellers [24], load and generation profiles control [25], [26], incentive based DR regulation considering penalties for customers in case of no load reduction response [27] and the combination of distributed interruptible load shedding and dispatched micro-sources to manage the network by distribution system operators [28]. The study reported in this paper is based on the previous work by the authors, where the design of a modified conventional EMS based on LEM (MCEMSLEM) for a MG system is reported and experimentally tested on IREC’s MG; and no optimization was used in [1]. In this paper, a real-time EMSMINLPLEM is proposed for MG system and a comparison between the MCEMSLEM and the proposed algorithm demonstrates that the proposed algorithm achieves better results.

Section snippets

Problem formulation

The system under analysis encompasses a stand-alone Wind Turbine (WT) – Photovoltaic (PV) – Microturbine (MT) – Energy Storage (ES) system. The mathematical model of the proposed EMSs is introduced in the following subsections. Details of the mathematical implementation of the MCEMSLEM can be found in [1]. For that reason, it is not repeated in this paper.

The proposed algorithms

In this study two algorithms, named MCEMSLEM and EMSMINLPLEM, are considered. These algorithms are presented in the following subsections.

Application to MG Testbed

Simulation and experimental evaluations are performed for a stand-alone WT–PV–MT–ES system shown in Fig. 4. The IREC Testbed is presented in Fig. 5. In short, the Testbed is built from power emulators able to generate or consume any desired power profile. Detailed explanation concerning the structure and configuration setting is found in [1], [34], [35]. As, it is observed from Fig. 4, this system has a central controller to which data will be sent. This data includes the offer of each

Results and discussions

To show the performance of the proposed algorithms, a case study is presented. The studied MG comprises renewable energy sources (WT and PV), and a power and heat combining unit (MT in this study). These are connected to an energy storage system (a battery in this study), as shown in Fig. 4. The outputs of the algorithms are explained in the following four subsections including execution time of the algorithms, the outputs of the EMS unit, the outputs of DR unit and finally the outputs related

Conclusions

This paper has compared two algorithms to implement an EMS based on LEM in MGs. The proposed LEM has been introduced to allow the owner of the generation units to establish their own strategy for participating in MG generation with minimum information shared between micro-sources. Indeed, fulfilling the customer’s requirements with minimum COE is the main theme of this paper. Likewise, DR has been used to avoid the penalty cost due to UP as well as to improve demand side management.

Acknowledgments

This research was supported by EIT and KIC-InnoEnergy under the projects KIC-ASS (30-2011-IP26-KIC ASS) and EVCITY (28-2011-IP24-EVCity) and by the European Regional Development Funds (ERDF, “FEDER Programa Competitivitat de Catalunya 2007–2013”).

The authors thank Jonathan Fournier and Manel Sanmarti for his valuable points of views and language editing.

References (35)

Cited by (213)

View all citing articles on Scopus
View full text