are one of the most useful and efficient ways to integrate the renewable energy technologies. As the stability of a microgrid is highly sensitive, an energy storage ...
2013 4th IEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe), October 6-9, Copenhagen
1
Hybrid Energy Storage System with Unique Power Electronic Interface for Microgrids 1
Haritza CAMBLONG1,2 Sylvain BAUDOIN1,2 Stéphane KRECKELBERGH1
Ionel VECHIU1 Aitor ETXEBERRIA1 ESTIA, F-64210, Bidart, France
Abstract— The increasing penetration of Distributed Generation systems based on Renewable Energy Sources is introducing new challenges in the current centralized electric grid. Microgrids are one of the most useful and efficient ways to integrate the renewable energy technologies. As the stability of a microgrid is highly sensitive, an energy storage system is essential and it must satisfy two criteria: to have a high storage capacity and to be able to supply fast power variations. In order to satisfy these two constraints, this paper proposes the association of a Vanadium Redox Battery (VRB) and Supercapacitor (SC) bank in a Hybrid Energy Storage System (HESS). A Three-level Neutral Point Clamped (3LNPC) Inverter is used as a unique interface between the HESS and the microgrid. The paper focuses on the dynamic modelling and validation of the HESS, the power division and the modulation strategies used with the 3LNPC to mitigate the neutral point voltage unbalance effect. Index Terms—Hybrid Energy Storage System, Multilevel Inverter, Microgrids
I.
INTRODUCTION
In the traditional electric grid, the power flow is unidirectional from the centralised big generation plants to a large number of dispersed end-users. Nowadays, the energy market liberalisation has permitted customers not only to consume electricity, but also to generate and sell it. As a consequence, the number of small generation systems that are connected at the distribution line level is continuously increasing. The increasing penetration of this generation systems, called Distributed Generation (DG) is changing the perspective of the grid from a centralised to a decentralised one and creates several challenges that must be carefully addressed in order to keep the proper operation of the electric grid. To take into account all those challenges, a solution capable of guaranteeing a controlled injection of the power generated by DG must be defined. The MG is being analysed as a solution to these issues [1]. From the grid point of view, a MG can be regarded as a controlled entity that can be operated as a single aggregated load or even as a small power source or ancillary service supporting the network. From the customer’s point of view, a MG provides enhanced power quality and reliable energy supply. The most important characteristic of a MG is its ability to operate in grid connected or islanding
2
UPV-EHU, Europa Plaza 1, E-20018, Spain
mode. A MG is inertia-less and the use of storage systems is essential, among other reasons, to be able to maintain power balance when power fluctuations generated by load or generation systems occur. Storage plays a vital role in order to ensure energy balance and also allowing to maintain high power quality. Thus, the use of an Energy Storage System (ESS) is a key element in a MG context, especially if there is a high penetration of renewable sources. An energy storage system helps to smooth the power variations introduced by the Renewable Energy Sources (RES). At the same time, they can store the excess of energy when the generation is higher than the demand and provide the previously stored energy when the generation is lower than the demand. Consequently, an ESS can be controlled in order to maintain the energy balance and also to improve the power quality of the MG [2]. This paper proposes a HESS formed by two storage technologies as the best compromise to satisfy the objectives that were fixed: having a high storage capacity and be able to supply fast power variations. The validation of a dynamic model of the HESS based on the association of a Vanadium Redox Battery/Supercapacitor has been carried using real storage technologies. Once the model of the HESS validated, a Three-level Neutral Point Clamped (3LNPC) Inverter has been used in an original architecture as a unique interface in order to control both storage technologies. Then, the paper focuses on an adapted modulation strategy from the point of view of the power division between a SC bank and a redox battery together with the effect of the neutral point voltage unbalance. II.
HESS MODELLING AND VALIDATION
After a comparative study concerning the storage technologies, it can be concluded that there is no storage technology with a high specific power and at the same time a high specific energy. Consequently, a HESS is needed in order to have a high storage capacity and to be able to supply fast power variations in a MG. Taking into account that the life expectancy of the batteries Lead-acid, NiCd, NaS and Li-ion is several times smaller than the life cycle of the flow batteries, that they have higher self-discharge rates as well as DOD rate limitations and
978-1-4799-2984-9/13/$31.00 ©2013 IEEE
2
other operational constraints, for this work a VRB flow battery has been selected as long-term storage device. On the other hand, among the different high specific power and fast response time storage technologies, the SC has been selected. Compared to other technologies, the SC offers an easier installation, no security issues created by a rotating mass, no maintenance costs and a much smaller self-discharge. Table 1 details the most important characteristics of the selected ESS technologies. Table 1: Characteristics of the analysed ESS technologies VRB SC Technology Characteristics 0.01[7] Energy rating 1.2-60[7] [MWh] 1.5[8] 0.5-20[4] 0.03-3[3] 0-0.3[3] Power rating 0.05-0.25[7] 0.05-100[4] [MW ] 1.5[8] 3[11] 10-30[3] Specific energy 2.5-15[3] 25[13] [Wh/kg] 5[6] 25-35[7] 5-15[10] 30-50[8] 500-5000[3] Specific power 80-150[13] [W/kg] 800-2000[10] 166[7] 10000[6] 23600[7] 70-85[11] Efficiency [%] 85-98[6] 95[9] 80[13] 85[8] Cycle life [cycles] 100k[3] 10000 (75%)[5, 500k (100%)[6] 8] 1M[7] 12000[3] 16000[13] Life time [years] 8-10[10] 5-10[3] 12[6] 8-10[5] 20[3] 15-20[7] Self-discharge 5[10] neg.[13, 11] [%] 5-20[7] very low[7] 20-40[3] (day) Response time very fast n/a (
