sustainability Article
Design of a Bidirectional Energy Storage System for a Vanadium Redox Flow Battery in a Microgrid with SOC Estimation Qingwu Gong and Jiazhi Lei * School of Electrical Engineering, Wuhan University, Wuhan 430072, China;
[email protected] * Correspondence:
[email protected]; Tel.: +86-27-6877-2281 Academic Editor: Tomonobu Senjyu Received: 1 December 2016; Accepted: 14 March 2017; Published: 17 March 2017
Abstract: This paper used a Vanadium Redox flow Battery (VRB) as the storage battery and designed a two-stage topology of a VRB energy storage system in which a phase-shifted full bridge dc-dc converter and three-phase inverter were used, considering the low terminal voltage of the VRB. Following this, a model of the VRB was simplified, according to the operational characteristics of the VRB in this designed topology of a VRB energy storage system (ESS). By using the simplified equivalent model of the VRB, the control parameters of the ESS were designed. For effectively estimating the state of charge (SOC) of the VRB, a traditional method for providing the SOC estimation was simplified, and a simple and effective SOC estimation method was proposed in this paper. Finally, to illustrate the proper design of the VRB ESS and the proposed SOC estimation method, a corresponding simulation was designed by Simulink. The test results have demonstrated that this proposed SOC estimation method is feasible and effective for indicating the SOC of a VRB and the proper design of this VRB ESS is very reasonable for VRB applications. Keywords: vanadium redox flow battery; energy storage system; SOC estimation; simplified equivalent model; parameter design
1. Introduction An energy storage system (ESS) which can be used for peaking regulation in power plants or power stations, and which can be used as an energy storage device in new energy generation systems, has the potential to be very widely applied in power systems [1,2]. As an effective technique for enhancing integrating intermittent renewable energy into a power grid, battery energy storage has become one of the focal points of development, due to its unique performance [3]. A Vanadium Redox flow Battery (VRB), as a new storage battery, can be used as the energy storage unit in an ESS. In an ESS, the topology should consider the terminal voltage of the VRB. In addition, to design reasonable control parameters for an ESS, the model of a VRB should be simplified, which has not been included in previous research. The parameter of the state of charge (SOC), as one of the most important parameters of a VRB, represents the residual capacity of a VRB and is the key parameter to regulate or control the operating states of an ESS. Hence, the effective estimation of the SOC is very significant. Recently, many researchers have concentrated on the design of a VRB energy storage system [4–10]. In [4,5], a control strategy of an ESS, based on a bidirectional DC-DC converter for a novel stand-alone photovoltaic power system, was proposed. But, this ESS was connected to a DC bus, without considering the power flow between the energy storage battery and the AC grid, resulting in great limitations for the energy storage system. The ESS in [6,7] is composed of a three-phase grid-connected inverter and a power frequency transformer, in which reasonable control of the battery is very difficult Sustainability 2017, 9, 441; doi:10.3390/su9030441
www.mdpi.com/journal/sustainability
Sustainability 2017, 9, 441
2 of 15
to realize. In [8–10], a large scale ESS based on various energy storage units, in which a buck-boost converter is used, was proposed. However, the ESS topologies involved in these papers are not suitable for a VRB ESS, considering the low terminal voltage of a VRB. The phase-shifted full bridge dc-dc converter in [11–13] uses a low current and voltage for the power switch and is very suitable for a VRB ESS. In the designed topology of a VRB energy storage system, there are various VRB models which have been developed for dealing with the control parameters of an ESS. However, most of these are based on how to establish an objective function of a VRB for describing the performance characterization of a VRB. In [14–18], an equivalent model of a VRB was comprehensively introduced, to obtain the operational characteristics of a VRB. In [19,20], a comprehensive equivalent circuit model of a VRB was presented for system level analysis, in which the concentration over potential is modeled as a function of flow rate, in an order to determine determine an appropriate variable flow rate. But these proposed models of VRBs are not suitable for the design of the control parameters in a VRB energy storage system, considering the complexity of these models. Moreover, the SOC estimation has become an important research topic for VRBs in ESSs. In [21–24], different effective management methods based on the parameter of the SOC of the battery were introduced, but the SOC estimation method of a VRB is not included. In the past decade, several methods have been proposed for SOC estimation in many battery applications [25–28]. A very common method for estimating the SOC of a redox flow cell is the installation of an open-circuit cell at the entry or exit ports of the electrolytes in the cell stack. By using the Nernst equation, the SOC can be readily obtained from the measured open circuit voltage (OCV) [25]. However, as the OCV gives an overall value of the two half-cells which are assumed to be balanced, this procedure does not accurately indicate the SOC of a VRB. In [26], for the estimation of Zero Emission Battery Research Activities (ZEBRA) battery, a hybrid neural model is proposed considering the state of health and the discharge efficiency parameters. In [27], two state-of-charge monitoring methods and their use in the VRB are investigated. In [28], the SOC estimation of VRBs was realized by an extended Kalman filter (EKF) method. However, these SOC estimation methods make stack configuration more complex as one open flow cell needs to be spared for each stack and additional voltage sensors need to be installed. Therefore, all of these methods for SOC estimations are too complex and are not suitable for application. Considering these points, this paper firstly designs a two-stage topology of a VRB energy storage system considering the low terminal voltage of the VRB, in which a phase-shifted full bridge dc-dc converter and three-phase inverter are used and the bidirectional power flowing between the energy storage battery and AC grid is achieved. Following this, the model of a VRB is simplified according to the operational characteristics of the VRB in this proposed topology of a VRB ESS and the control parameters of the ESS are designed. For effectively managing the control center of the microgrid (MG), a traditional method for SOC estimation is simplified, and a simple and effective SOC estimation method is established. Finally, the feasibility and effectiveness of the proposed SOC estimation method are verified by the designed VRB ESS, and the test results demonstrate that the proposed SOC estimation method is reasonable for indicating the SOC of a VRB and that the proper design of the VRB ESS is very reasonable for VRB applications. 2. Topology of a VRB ESS in a Microgrid With the increasing depletion and fluctuation of fossil fuel prices, residential and industrial pollution, etc, energy storage system can be used as a sorted renewable energy system for clean energy [29]. Figure 1 shows the designed topology of a VRB energy storage system in a microgrid. In Figure 1, the structure of a microgrid (MG) is shown. Photovoltaic generation units and wind generation units (WPG) are respectively connected to the AC grid by power electronic devices. Due to the randomness of the PV and WPG, it is essential to configure the VRB ESS, to maintain the power balance in the MG. The control center of the MG gathers the electrical quantities of all of the converters and operates the power flow of the MG by controlling these converters. It should be noted that the
Sustainability 2017, 9, 441
3 of 15
SOC of the VRB is one of the most important parameters for the control center of the MG to control these converters. Therefore, the SOC of the VRB is effectively estimated in this paper. 3 of 15 Sustainability 2017, 9, 441 The most important part of a battery energy storage system is the power conversion system most important part of a battery energy storage system the of power conversion system system, (PCS), which The is actually a voltage source inverter. Through theisuse a power conversion (PCS), which is actually a voltage source inverter. Through the use of a power conversion system, bidirectional power flow is achieved between the energy storage battery and the AC grid. Discharging bidirectional power flow is achieved between the energy storage battery and the AC grid. and charging management of the VRB, andofload power oftracking the ACofgrid, can also realized by Discharging and charging management the VRB, andtracking load power the AC grid, canbe also be arealized controlling PCS. by controlling a PCS.
1. Topology of the VRB energy storage system in the microgrid. FigureFigure 1. Topology of the VRB energy storage system in the microgrid.
In this paper, a 5 kW VRB is used as the storage battery. Considering the voltage level of the battery, a two-stage of the PCS, which is composed a bidirectional DC-DC converter In this paper, a 5 kW topology VRB is used as the storage battery.of Considering the voltage level of the and a three-phase inverter, is used as shown in Figure 1. This designed topology of the PCS can battery, a two-stage topology of the PCS, which is composed of a bidirectional DC-DC converter and realize the flexible control of the VRB, as it can be controlled by the DC-DC converter and the a three-phase inverter, is used as shown in Figure 1. This designed topology of the PCS can realize the inverter. In addition, its size is relatively small as no power transformer is needed. flexible controlInofthis thedesigned VRB, as VRB it canESS, be controlled by the and the inverter. In addition, a phase-shifted fullDC-DC bridge converter dc-dc converter is used as the bidirectionalsmall DC-DC andtransformer its design is introduced in Section 4, associated with the control its size is relatively asconverter no power is needed. strategy of the VRB VRB ESS. In this designed ESS, a phase-shifted full bridge dc-dc converter is used as the bidirectional DC-DC converter and its design is introduced in Section 4, associated with the control strategy of the 3. Vanadium Redox Flow Battery VRB ESS. 3.1. Structure and Operating Principles of the VRB
3. Vanadium A Redox Flow Battery VRB, as an electrochemical cell, is divided into two compartments. One is the positive and negative tanks containing the electrolyte, and the other is the pump and piping for circulating the
3.1. Structure and Operating Principles of the electrolyte from the tanks to the cell. TheVRB vanadium ion VO2+ is in the positive electrode and V3+ is in the negative electrode. They transform themselves to VO2+ and V2+, respectively, by charging the
A VRB, as an electrochemical cell, is divided into two compartments. One is the positive and VRB. At the same time, H+ moves from the positive electrode to the negative, through the negative tanks containing the electrolyte, and the other thefollows pumpa and piping forFigure circulating the ion-exchange membrane. The discharging process of the is VRB similar process. 2 structure andthe operating principles of the VRB [14].2+ is in the positive electrode and V3+ is in electrolyteshows fromthe the tanks to cell. The vanadium ion VO + by the Nernst 2+ In Figure 2,They the potential of thethemselves individual cells given equation and depends on the negative electrode. transform toisVO 2 and V , respectively, by charging the VRB. the vanadium+species concentrations and the proton concentrations [15]: At the same time, H moves from the positive electrode to the negative, through the ion-exchange c 2 H ac similar membrane. The discharging process of the process. Figure 2 shows the structure RT VRB cVOfollows V [V ] ln E E (1) c F and operating principles of the VRB [14]. cVO V In Figure 2, the potential of the individual cells is given by the Nernst equation and depends on where E is the formal potential, which is set as 1.255 V; T is the stack temperature; R is the the vanadium species concentrations−1and −1the proton concentrations [15]: −1 2
2
2
3
universal gas constant 8.314 J·K ·mol ; and F is Faraday’s constant 96,485.3399 mol . c is the ( value of the SOC:! concentration of ions and it depends on the )
E = EΘ +
RT ln F
cVO2 + · c2 H + cVO2+
cV 2 + cV 3 +
[V ]
(1)
where EΘ is the formal potential, which is set as 1.255 V; T is the stack temperature; R is the universal gas constant 8.314 J·K−1 ·mol−1 ; and F is Faraday’s constant 96,485.3399 mol−1 . c is the concentration of ions and it depends on the value of the SOC:
Sustainability 2017, 9, 441 Sustainability 2017, 9, 441 Sustainability 2017, 9, 441
4 of 15 4 of 15 4 of 15
c c SOC c V 22 c VO22 SOC ,c c SOC (2) cVO2+ 1 SOC ccVV23+ cVO SOC , c HH c HH (( )) SOC (2) = = , c = c + SOC (2) V VO 2 + + SOC c c 1 H (o ) 2 VO cVV33+ cVO 1 − SOC H 2+ is the initial concentration of H+. The sum of the equilibrium potential E of the where c where c H ( ) is the initial concentration of H+. The sum of the equilibrium potential E of the where c H +H (o()) is the initial concentration of H+ . The sum of the equilibrium potential E of the individual individual cells presents the stack potential. individual cells the stack potential. cells presents thepresents stack potential.
-V 2、V 3 V 2、V 3
VO 2、VO2 VO 2、VO2
Figure 2. Structure and operating principles of the VRB. Figure 2. Structure andand operating principles of the VRB. Figure 2. Structure operating principles of the VRB.
When a net current is flowing through the stack, the equilibrium conditions are not met and the When a net current is flowing through the stack, the equilibrium conditions are not met and the a net current is flowing through the stack,between the equilibrium conditions are not met stackWhen voltage Ustack is then given by the difference the equilibrium potential Ueq and the stack voltage Ustack is then given by the difference between the equilibrium potential Ueq and the stack voltage U is then given by the difference between the equilibrium potential U internal losses Uloss [2,15]. eq and the internal losses Ustack loss [2,15]. internal losses Uloss [2,15]. Ustack Ueq Uloss nE Req,chargr/discharge I (3) Ustack Ueq Uloss nE Req,chargr/discharge I (3) Ustack = Ueq − Uloss = nE − Req,chargr/discharge I (3) where n is the number of cells in the stack, Req,chargr/discharge is the equivalent charge resistance of the is the equivalent charge resistance of the where n is the number of cells in the stack, Req,chargr/discharge VRB under charging orofdischarging, and I is the charging orisdischarging current. In Equation of (3), E where n is the numberor cells in the stack, the equivalent charge VRB under charging discharging, and I Riseq,chargr/discharge the charging or discharging current. Inresistance Equation (3),the E can be calculated by Equation (1), in which the concentration c is calculated by the SOC in Equation VRB under charging or discharging, and I is the charging or discharging current. In Equation (3), E can can be calculated by Equation (1), in which the concentration c is calculated by the SOC in Equation (2). be (2).calculated by Equation (1), in which the concentration c is calculated by the SOC in Equation (2). 3.2. Simplified Method for SOC Estimation 3.2. Simplified Method for SOC Estimation The SOC of the VRB is an important parameter reflecting the residual capacity of the VRB and it The SOC of the VRB is an important parameter reflecting the residual capacity of the VRB and it is the thethe operating states of aofVRB ESS.ESS. The The traditional method for SOC the key keyparameter parameterfor forregulating regulating operating states a VRB traditional method for is the key parameter for regulating the operating states of a VRB ESS. The traditional method for estimation is based on theon equivalent circuit circuit of the VRB, which shownisinshown Figurein3 Figure [16,17].3In[16,17]. FigureIn 3, SOC estimation is based the equivalent of the VRB,iswhich SOC estimation is based on the equivalent circuit of the VRB, which is shown in Figure 3 [16,17]. In the stack by a voltage-controlled source, which is which influenced by the value of Figure 3, voltage the stackUvoltage Ustack is imitated by a voltage-controlled source, is influenced by the stack is imitated Figure 3, the stack voltage Ustack is imitated by a voltage-controlled source, which is influenced by the the SOC andSOC the cell the VRB, as shown Equation (3). value of the andvoltage the cellofvoltage of the VRB, asinshown in Equation (3). value of the SOC and the cell voltage of the VRB, as shown in Equation (3).
Rfixed
U battery Rfixed U battery
Celectrode C electrode I pump I pump
U stack U stack Rreaction R reaction
Rresistive R resistive
Figure 3. 3. Equivalent Circuit of of the VRB. VRB. Figure Equivalent Circuit Circuit Figure 3. Equivalent of the the VRB.
Generally, the system state of charge can be defined as Equation (4): Generally, the system state of charge can be defined as Equation (4):
Sustainability 2017, 9, 441
5 of 15
Generally, the system state of charge can be defined as Equation (4): Sustainability 2017, 9, 441
SOC =
5 of 15
Ecurrent Ecapacity
(4)
Ecurrent (4) E In line with Equation (4) and the equivalentcapacity circuit of the VRB, Equation (5) can be acquired: SOC
In line with Equation (4) and the equivalent circuit P · T of the U VRB,· Equation I · T (5) can be acquired:
∆SOC =
∆E
Ecapacity E
SOC
Ecapacity
=
stack
step
=
stack
stack
step
PstackEcapacity Tstep U stack I stackPrating Tstep · Trating Ecapacity Prating Trating
(5)
(5)
Accordingly, the value of the SOC can be calculated by Equation (6): Accordingly, the value of the SOC can be calculated by Equation (6): Z U · Istack U stack Istack SOC = SOC + dt stack o SOC SOCo dt P · Trating Prating rating Trating
(6)
(6)
In [14], the the SOC estimation on Equation Equation(6)(6)is isproposed proposed In [14], SOC estimationmethod method based based on andand this this VRBVRB SOC SOC estimation method is named thethe traditional method this paper. It should be noted estimation method is named traditionalSOC SOC estimation estimation method in in this paper. It should be noted that traditional this traditional SOC estimationmethod methodis is very very complex considering the required that this SOC estimation complextotoimplement, implement, considering the required integration operation in Equation (6). Besides, the stack current stack is difficult to accurately integration operation in Equation (6). Besides, the stack current Istack Iis difficult to accurately measure. measure. In line with (3), the stack voltage Ustack can be calculated and the corresponding parameters are line with (3), the voltage Ustack can be calculated corresponding parameters are can given inInTable 1. Then, thestack terminal voltage Ubattery of the 5and kWthe VRB under a constant current given in Table 1. Then, the terminal voltage Ubattery of the 5 kW VRB under a constant current can be be caught by the simulation of the VRB equivalent circuit, and the parameters of the VRB equivalent caught by the simulation of the VRB equivalent circuit, and the parameters of the VRB equivalent circuit are given in [15,16]. Figure 4 shows the curves of Ubattery and Ustack of the 5 kW VRB in pace circuit are given in [15,16]. Figure 4 shows the curves of Ubattery and Ustack of the 5 kW VRB in pace with with the SOC under a constant charging A. the SOC under a constant chargingcurrent current of of 80 80 A.
Figure 4. Voltage curves of the VRB under a charging current of 80 A. Figure 4. Voltage curves of the VRB under a charging current of 80 A.
As can be seen from Figure 4, the curves of Ustack and Ubattery are approximately parallel, and the
As can be seen from Figurealong 4, thewith curves Ustackof and approximately parallel, and the stack voltage Ustack increases, the of increase theUcharging current Ibattery. Hence, the value battery are the stack voltage Ustack under charging canincrease be approximated as Equation (7): Ibattery . Hence, the value stackof voltage Ustack increases, along with the of the charging current of the stack voltage Ustack underUcharging as Equation (7): Ucan be approximated I R stack,charge
battery,charge
battery
charge
(7)
=the Ubattery,charge Ibattery · Similarly, Rcharge the stack voltage Ustack where Rcharge is the chargeUresistance VRB under− charging. stack,chargeof under discharging can be approximated as (8):
(7)
where Rcharge is the charge resistance of the VRB under charging. Similarly, the stack voltage Ustack Ustack,discharge Ubattery,discharge I battery Rdischarge (8) under discharging can be approximated as (8): where Rdischarge corresponds to the discharge resistance of the VRB under discharging. According to Equations (7) and (8) and Figure 5, Rcharge Rdischarge are set the same value, 0.0713 Ω, for the 5 kW (8) Ustack,discharge = Uand + Iatbattery · Rdischarge battery,discharge VRB in this paper. In Figure 3, ignoring to thethe impact of Rfixed and Ipump,ofthe stack of the VRB can be where Rdischarge corresponds discharge resistance thestack VRBcurrent under Idischarging. According to approximated as I battery . As a result, the traditional VRB SOC estimation method based on (6) can be Equations (7) and (8) and Figure 5, Rcharge and Rdischarge are set at the same value, 0.0713 Ω, for the
5 kW VRB in this paper.
Sustainability 2017, 9, 441
6 of 15
In Figure 3, ignoring the impact of Rfixed and Ipump , the stack current Istack of the VRB can be approximated as Ibattery Sustainability 2017, 9, 441 . As a result, the traditional VRB SOC estimation method based on (6) 6 ofcan 15 be simplified and the stack voltage Ustack can be calculated by Equations (7) and (8). Moreover, 6the stack Sustainability 2017, 9, 441 of 15 simplified the stack voltage Ustack can current Istack isand approximated by Ibattery . be calculated by Equations (7) and (8). Moreover, the stack current Istackand is approximated by IUbattery . simplified the stack voltage stack can be calculated by Equations (7) and (8). Moreover, the stack
3.3. current ProposedIstack SOC Estimation Method is approximated by Ibattery. 3.3. Proposed SOC Estimation Method
In this section, a simple and effective SOC estimation method is proposed. According to
3.3. Proposed Estimation Method In this SOC section, a simple and effective SOC estimation method is proposed. According to Equation (3), the curves of the stack voltage Ustack of the 5 kW VRB in pace with the SOC under Equation (3), the curves of the stackeffective voltage U stack of the 5 kW VRB in pace with the SOC under a In charging this section, adischarging simple and estimationTherefore, method is proposed. According tostack a constant and current canSOC beacquired. acquired. surface charts of the constant charging and discharging current can be Therefore, thethe surface charts of the stack Equation (3), the curves of the stack voltage U stack of the 5 kW VRB in pace with the SOC under a voltage Ustack , terminal current I Ibattery,, and and SOCunder undercharging charging and discharging, can be fitted, shown voltage Ucharging stack, terminal currentbattery and discharging, can be fitted, constant and discharging currentSOC can be acquired. Therefore, the surface charts of theshown stack in Figures 5 and 6, respectively. in Figures 5 and 6, respectively. voltage Ustack , terminal current Ibattery, and SOC under charging and discharging, can be fitted, shown in Figures 5 and 6, respectively.
Figure Surfacecharts charts of of the the stack stack voltage Figure 5.5.Surface voltageunder undercharging. charging. Figure 5. Surface charts of the stack voltage under charging. It can be seen from Figures 5 and 6 that the stack voltage Ustack increases with the increase of the It can becurrent seen from Figures and 6ofthat the stack voltage Ustack current increases with the increase of the terminal Ibattery the5value When the terminal Iwith battery is determined, the It can be seen fromand Figures 5 and 6 the thatSOC. the stack voltage Ustack increases the increase of the terminal Ibattery andcharging the valueand of the SOC. When therapidly terminal current Ibased is determined, valuecurrent of the SOC under can the be the valuethe of the battery on terminal current Ibattery and the value of discharging the SOC. When terminalcalculated, current Ibattery is determined, value ofstack the SOC under and discharging can be rapidly calculated,Inbased on the value of the voltage U stackcharging and the surface charts of Figures 5 and 6, respectively. addition, the stack value of the SOC under charging and discharging can be rapidly calculated, based on the value of voltage Uvoltage stack under charging and discharging can be estimated (7) and (8), respectively. Namely, the stack voltage Ustack and surface charts of Figures and6,by 6,respectively. respectively. addition, the stack the stack Ustack andthe the surface charts of Figures 55and In In addition, the stack the value of the SOC can only be estimated by the value of the terminal current I battery and terminal voltage Ustack under charging canbe beestimated estimated and respectively. Namely, voltage Ustack under chargingand anddischarging discharging can byby (7)(7) and (8),(8), respectively. Namely, battery , which can be accurately measured. This SOC estimation method aforementioned is voltage U the value of the SOC can only be estimated by the value of the terminal current I and terminal the value of the SOC can only be estimated by the value of the terminal current Ibattery and terminal battery the proposed SOC estimation method in this paper. This proposed SOC estimation method ignores voltage Ubattery , which ThisSOC SOCestimation estimation method aforementioned , whichcan canbe beaccurately accurately measured. measured. This method aforementioned is is voltage Ubattery the proposed balance ofSOC two half-cells and isinvery simple to implement. Further simulations in Section V the estimation method this paper. This proposedSOC SOCestimation estimation method ignores the proposed SOCthe estimation method in this paper. This proposed method ignores the will illustrate the effectiveness of this proposed SOC estimation method. the balance of the two half-cells and is very simple to implement. Further simulations in Section V
balance of the two half-cells and is very simple to implement. Further simulations in Section 5 will will illustrate the effectiveness this proposed estimation method. illustrate the effectiveness of thisof proposed SOC SOC estimation method.
Figure 6. Surface charts of the stack voltage under discharging.
Figure 6. 6.Surface voltageunder underdischarging. discharging. Figure Surfacecharts charts of of the the stack stack voltage The simulation parameters of the 5 kW VRB used in Figures 4–6 are shown in Table 1.
The simulation parameters of the 5 kW VRB used in Figures 4–6 are shown in Table 1.
Sustainability 2017, 9, 441
7 of 15
Sustainability 2017, 9, 441
7 of 15
The simulation parameters of the 5 kW VRB used in Figures 4–6 are shown in Table 1. Table 1. Parameters of the 5 kW VRB. Table 1. Parameters of the 5 kW VRB.
Name Name of cells Ncell number number ofRcells N/Ω eq,charge cell Req,charge /Ω Req,discharge/Ω Req,discharge /Ω electrolyte flowrate Q/(L·s−1) electrolyte flowrate Q/(L·s−1 ) −1) initial concentration of H+·/(mol·L initial concentration of H+ /(mol L− 1 ) −1) electrolyte vanadium concentration/(mol·L electrolyte vanadium concentration/(mol ·L− 1 ) tank tank size Vsize /L Vtk/L tk −1) initial concentration of vanadium/(mol ·L−1 ) initial concentration of vanadium/(mol·L
Value Value 39 39 0.047 0.047 0.049 0.049 2 2 66 22 260 260 11
3.4. Equivalent of VRB VRB In Figure 1, the VRB should be modeled for the control parameters design of an energy storage system. The model model of ofthe theVRB VRBbased basedon onFigure Figure3 3is is very complex and model is not suitable for very complex and thisthis model is not suitable for the the modeling ofenergy an energy storage system. Namely, the model of VRB the VRB should be simplified. modeling of an storage system. Namely, the model of the should be simplified. In accordance Figure 5, the equivalent resistance Req ofReq theofVRB eq = accordance with withEquation Equation(7)(7)and and Figure 5, the equivalent resistance the (R VRB Ubattery ) under a charging can be acquired, as shown as in shown Figure in 7. Figure 7. (R Ubattery /Ibattery ) under a state charging state can be acquired, eq =/I battery
Figure 7. The The equivalent resistance Req ofthe the VRB under a charging state. eq of
As can be be seen seen from from Figure Figure 7,7,RReqeq varies varies substantially substantially with with the theSOC SOCand andcurrent currentI Ibattery under As can battery under charging, which cannot be modeled as a fixed component. In this paper, the VRB energy storage storage charging, which cannot be modeled as a fixed component. In this paper, the VRB energy system designed under under aa constant constant charging charging current current and and constant constant charging charging power. power. The system was was designed The VRB VRB should be considered in the worst case scenario, namely, when R eq is selected as the minimum value. should be considered in the worst case scenario, namely, when Req is selected as the minimum value. As an an example, example, under under aa constant 1 As constant charging charging current current of of80 80A, A,RReqeqisisconsidered consideredasasthe theminimum minimumvalue value 1ΩΩfor forthe themodeling modelingofofananenergy energystorage storagesystem. system. On the other hand, the VRB is asas a variable voltage source Ueq U under discharging in a On the other hand, the VRB ismodeled modeled a variable voltage source eq under discharging VRB energy storage system. The variable amplitude Ueq can U beeqascertained by Figureby 6 and Equation in a VRB energy storage system. The variable amplitude can be ascertained Figure 6 and (8). As an example, under a constant discharging current of 80 A, U eq varies from 41.285 V to Equation (8). As an example, under a constant discharging current of 80 A, Ueq varies from 41.285 V 56.251 V. V. In In this state, the VRB to 56.251 this state, the VRBisismodeled modeledasasa avoltage voltagesource sourcewith withaaconstant constantvoltage voltageof of41.285 41.285V, V, considering the worst case scenario (the SOC is considered as the minimum value, 0.025). considering the worst case scenario (the SOC is considered as the minimum value, 0.025). According to this thissimplified simplifiedequivalent equivalent model VRB under charging discharging, a According to model of of thethe VRB under charging and and discharging, a VRB VRB ESS was modeled and the control parameters of the ESS were reasonably designed in Section 4. ESS was modeled and the control parameters of the ESS were reasonably designed in Section 4.
Sustainability Sustainability 2017, 2017, 9, 9, 441 441 Sustainability 2017, 9, 441
88 of of 15 15 8 of 15
4. Design of an Energy Storage System 4. Design of an Energy Storage System 4.1. Structure of an Energy Storage System 4.1. 4.1. Structure Structure of of an an Energy Energy Storage Storage System System In a VRB ESS, a bi-directional DC-DC converter is required to raise the terminal voltage of the In aa VRB ESS, bi-directional DC-DC converter is required to the voltage of the ESS, aaflexible bi-directional converter required to raise raise the terminal terminal the VRB In and VRB to achieve controlDC-DC of the battery. In is this paper, the phase-shifted fullvoltage bridge of dc-dc VRB and to achieve flexible control of the battery. In this paper, the phase-shifted full bridge dc-dc VRB and to control of the battery. In this paper, the phase-shifted full bridge converter is achieve used asflexible the bi-directional DC-DC converter, to overcome the drawback of thedc-dc low converter isisused asas thethe bi-directional DC-DC converter, to overcome the drawback of the low converter used bi-directional DC-DC converter, to overcome the drawback of terminal the low terminal voltage of the VRB, as shown in Figure 8. voltage the VRB, as shown in shown Figure in 8. Figure 8. terminalofvoltage of the VRB, as
L L
ibattery ibattery S1 S1
U battery U battery
ilk ilk
A A
C C
T4 T4
T1 T1
S4 S4
Llk n T n2 Llk n1* T n* 2 1 *
C C
*
D D
B B S3 S3
Cax Cax
T3 T3
S2 S2
U ax U ax
T2 T2
Figure 8. Bi-directional DC-DC converter in the ESS. Figure Figure 8. 8. Bi-directional Bi-directional DC-DC DC-DC converter converter in in the the ESS. ESS.
In Figure 8, T is a high frequency transformer which matches the voltage Ubattery and Uax, Llk is Figure 8, TTisisaahigh high transformer whichmatches matchesthe thevoltage voltageUbattery Ubatteryand andUUaxax,, LLlklk is In Figure 8, transformer which the leakage inductance, andfrequency Lfrequency is the output filter inductance. inductance, and L L is is the the output filter inductance. inductance. the leakage and filter When inductance, the VRB is charged, theoutput phase-shifted full-bridge converter works in a buck state, by When the VRB is charged, the phase-shifted full-bridge converter works inthe astate, buck by When the VRB is charged, the phase-shifted full-bridge converter in a buck bystate, driving driving the switches T1~T4. On the other hand, when the VRB isworks discharged, phase-shifted driving the switches T 1 ~T 4 . On the other hand, when the VRB is discharged, the phase-shifted the switchesconverter T1 ~T4 . On the other hand,state, when VRB is discharged, full-bridge full-bridge works in a boost bythe driving the switches S1the ~S4.phase-shifted Hence, the bidirectional full-bridge converter works in a boost state, by driving the switches S 1~S 4. Hence, the bidirectional converter works in a boost state, by driving the switches S ~S . Hence, the bidirectional power flowing 1 4 power flowing between the VRB and the AC grid is realized. power flowing between the AC grid is realized. between the VRB and thethe ACVRB gridand is realized. 4.2. Control Strategy of Energy Storage System 4.2. 4.2. Control Control Strategy Strategy of of Energy Energy Storage Storage System System The VRB ESS is mainly comprised of an on-grid mode and off-grid mode. In the on-grid mode, The VRB ESS is mainly comprised of an on-grid mode and off-grid mode. In the on-grid mode, the VRB is discharged and supplies its chemical energy to the AC grid. In the off-grid mode, the VRB the VRB is discharged and supplies its chemical energy to AC grid. In the off-grid mode, the VRB supplies to the thecan AC be grid. is used for discharged storing the excess energy ofchemical the AC energy grid, which generated by renewable energy is used for excess energy of of thethe ACAC grid, which can can be generated by renewable energy such used forstoring storingthe the energy grid, which be generated by renewable energy such as photovoltaic orexcess wind power. as photovoltaic or wind power. such Figure as photovoltaic or wind power. 9 shows the control diagram of the bidirectional DC-DC converter, in which the VRB is Figure 9 shows the control diagram of the bidirectional DC-DC converter, in which the VRB is charged and discharged under a constant current or power. Figure 10 shows the control strategy of charged and discharged under a constant current or power. Figure 10 shows the control strategy of the three-phase grid-connected inverter, in which the control strategy of the outer current loop and the three-phase grid-connected inverter, in which the control strategy of the outer current loop and inner voltage loop were likely used. The phase of the grid connected current is controlled, remaining inner voltage loop were likely used. The phase of the grid connected current is controlled, remaining consistent with the phase of the grid voltage [30,31]. consistent with the phase phase of of the the grid grid voltage voltage [30,31]. [30,31].
I ref I ref Pref Pref
i t i t
i
PI PI
pbattery battery pbattery ibattery
Figure 9. Control diagram of the bidirectional DC-DC converter. Figure 9. Control diagram of the bidirectional DC-DC converter. Figure 9. Control diagram of the bidirectional DC-DC converter.
Sustainability 2017, 9, 441
9 of 15
Sustainability Sustainability2017, 2017,9,9,441 441
99of of15 15
Q1 Q1
U ax
A
Q3
Q5
Q3
Q5
Lf
ia
Lf
A
iaib
B
U ax
ib ic
C
B
Q4
Q6
Q2
Q4
Q6
Q2
ic
Cf
C U B* U B*
Cf
id
iq
UiB
d
iq UB
Figure 10. Control strategy of the three phase grid-connected inverter. Figure 10. Control strategy of the three phase grid-connected inverter. Figure 10. Control strategy of the three phase grid-connected inverter.
4.3. Parameters Design of Controller 4.3. Parameters Design of Controller 4.3. Parameters of Controller In line withDesign the model of the VRB and the control strategy of the ESS, the modeling of the VRB In line with the model of the VRB and the control strategy of the ESS, the modeling of the VRB ESS can be analyzed. In line with the model of the VRB and the control strategy of the ESS, the modeling of the VRB ESS can analyzed. full-bridge DC-DC converter can be evolved from the Buck-Boost circuit, and Thebe phase-shifted ESS can be analyzed. The phase-shifted full-bridge DC-DC be evolved from thethe Buck-Boost circuit, and its small-signal equivalent circuit has been converter shown in can [32,33]. In a buck state, VRB is modeled a The phase-shifted full-bridge DC-DC converter can be evolved from the Buck-Boost circuit,as and its small-signal equivalent circuit has been shown in [32,33]. In a buck state, the VRB is modeled as resistance Req inequivalent the worst circuit case. The function of theInoutput Ibattery inispace with as theaa its small-signal hastransfer been shown in [32,33]. a buckcurrent state, the VRB modeled resistance in worst case. The transfer function duty ratio dRReq in thethe case of the continuous current mode of is: the output current Ibattery in pace with the resistance eq in the worst case. The transfer function of the output current Ibattery in pace with the duty ratio d in the case of the continuous current mode is: duty ratio d in the case of the continuous current Rmode is: 1 nU ax [ sC ( c 1) ] R R R1eq c eq nU · [ sC ( + ) R 1 + R1eq ] ax c Req Gid ( s) (9) nU [ sC ( 1) ax Gid (s) = (9) R R Rc R L] R LC (( Rcc +1)1s)2s2 [+ CR[CR ( dc( R 1) CR eqCRc]s+ (L d]s +1)( Rd + 1) eq + LC 1 ) + d c R R R R Gid ( s) eq eq eq eq (9) RReq RReq Req RReq L LC ( c 1) s 2 [CRd ( c 1) CRc ]s ( d 1) 2 Req Req R4n Rnisisthe eq 2Llk where R CCand RRdd== 4n the turns turns ratio ratio of of the high Rcc is the the resistance resistance of ofcapacitor capacitor andresistance resistance lkffs.s .neq frequency transformer T.T.The parameters of the ESS ESS are presented in Table 2.Table Under 80 A charging The parameters of resistance the are 2. an Under A where Rc istransformer the resistance of capacitor C and Rd =presented 4n2Llkfs. n in is the turns ratio ofan the80 high current, Rcurrent, as the minimum value 1value Ω and theand corresponding parameters of PI of in charging R eq is considered as the minimum 1 Ω the corresponding parameters eq is considered frequency transformer T. The parameters of the ESS are presented in Table 2. Under an 80 A Figure 10 can10becan designed, as shown in Figure 11. 11. PI in Figure designed, as shown in Figure charging current, Rbe eq is considered as the minimum value 1 Ω and the corresponding parameters of
PI in Figure 10 can be designed, as shown in Figure 11.
Figure 11. Mode of the DC-DC converter in a buck state. Figure 11. Mode of the DC-DC converter in a buck state. Figure 11. Mode of the DC-DC converter in a buck state. In a boost state, the VRB is modeled as a voltage source with a constant voltage, and the In a boost state, the VRB is modeled as a voltage source with a constant voltage, and the parameters parameters of the canVRB be isdesigned analogously. Besides,with the design a three-phase boost state,PIthe modeled voltageofsource constantofvoltage, and can the of theIn PI acan be designed analogously. Besides,as thea design a three-phaseagrid-connected inverter parameters of the PI can be designed analogously. Besides, the design of a three-phase be carried out by the small-signal modeling method [34], which is omitted as a space limitation.
Sustainability 2017, 9, 441
10 of 15
Table 2. Parameters of the designed bidirectional ESS for the VRB. Name
Value
inductance Lf /µH capacitor Cf /µF turns ratio n of Transformer T leakage inductance Llk /µH Switching frequency fs /kHz filter inductance L/µH capacitor C/µF Resistance of capacitance Rc /Ω
200 100 1:8.35 25 5 80 20 0.2
5. Simulation 5.1. Simulation Model of the VRB According to the equivalent circuit in Figure 3 and the parameters of the 5 kW VRB ESS, the corresponding simulation is set up by Simulink, to illustrate the proper design of the VRB ESS and the proposed SOC estimation method. In simulation, the calculated VRB parameters are based on estimating losses of 21% in the worse case operating scenario, for a minimum voltage of 42 V and a current of 112 A [14]. Internal losses of the VRB account for 15%, in which the reaction loss Preaction is 9% and the resistive loss Presistive is 6%. The parasitic loss of the VRB accounts for 6%, in which the fixed loss Pfixed is 2% and the pump loss Ppump is 4%. The parameters of the 5 kW VRB are calculated as follows: Pstack =
5000 = 6329.11(W ) 1 − 0.21
422 422 = = 13.936(Ω) Pfixed 6329.11 × 2% I Pparasitic = Pfixed + k stack , k = 42.5 SOC
Rfixed =
(10)
(11) (12)
Rreaction =
Preaction 569.62 = ≈ 0.045(Ω) 2 1122 Istack
(13)
Rresistive =
379.75 Presistive = ≈ 0.030(Ω) 2 1122 Istack
(14)
5.2. Simulation Results Based on the designed control parameters in the VRB ESS, the simulation results are shown in Figures 12–17, where Figures 12–14 show the simulation results under a charging and discharging current of 80 A. In Figure 12, the quantitative deviation of the current Ibattery and Istack is less than 5 A. Therefore, the current Ibattery can be used to approximate Istack in practice. Besides, the curves of Ustack and Ubattery under charging or discharging are approximately parallel. In Figure 13, the efficiency of the VRB varies from 0.8 to 0.85, and the efficiency of the VRB reaches its maximum value when the SOC is 0.975. In addition, the efficiency of the VRB under a constant charging current of 80 A is higher than the efficiency of the VRB under a constant discharging current of 80 A. This is because more energy is consumed when the VRB operates in a discharging state.
Sustainability Sustainability 2017, 2017, 9, 9, 441 441 Sustainability 2017, 9, 441
11 11 of of 15 15 11 of 15
discharging discharging current current of of 80 80 A. A. This This is is because because more more energy energy is is consumed consumed when when the the VRB VRB operates operates in in aa discharging current of 80 A. This is because more energy is consumed when the VRB operates in a 11 of 15 discharging state.
discharging state. Sustainability 2017, 9, 441 discharging state.
Figure Figure 12. 12. Current Current and and voltage voltage of of the the VRB VRB under under aa charging charging and and discharging discharging current current of of 80 80 A. A. Figure Figure 12. 12. Current Current and and voltage voltage of of the the VRB VRB under under aa charging chargingand anddischarging dischargingcurrent currentof of80 80A. A.
Figure 13. Power and and efficiency of the VRB under charging and Figure discharging current of 80 A. Figure 13. 13. Power Power and efficiency efficiency of of the the VRB VRB under under aaa charging charging and and discharging discharging current current of of 80 80A. A. Figure 13. Power and efficiency of the VRB under a charging and discharging current of 80 A.
In was obtained Equation In Figure Figure 14, 14, the the actual actual value value of of the the SOC SOC was was obtained obtained by by Equation Equation (4). (4). As As can can be be seen seen from from In Figure 14, the actual value of the SOC was obtained by Equation (4). As can be seen SOC from Figure 14, the proposed SOC estimation method very effective, the estimation value Figure the estimation value ofof thethe SOC is Figure 14, 14, the theproposed proposedSOC SOCestimation estimationmethod methodisis isvery veryeffective, effective,asas as the estimation value of the SOC Figure 14, the proposed SOC estimation method is very effective, as the estimation value of the SOC is very close to the actual value of the SOC. Besides, the charging time is longer than the discharging very close to to thethe actual value ofof the is very close actual value theSOC. SOC.Besides, Besides,the thecharging chargingtime timeisislonger longerthan thanthe the discharging discharging is very close to the actual value of the and SOC. Besides, the charging time is4longer than the discharging time, time is about time is about h. time, as the charging hh and the discharging This is is because because the the stack stack time, as as the the charging charging time time is is about about 555 h and the the discharging discharging time time is is about about 44 h. h. This time, as thedischarging charging time is about 5h and the discharging time is about 4 h. This is because the stack power in state is more expended. power power in in aaa discharging discharging state state is is more moreexpended. expended. power in a discharging state is more expended.
SOC of the Figure Figure 14. 14. SOC SOC of of the the VRB VRB under under aa charging charging and and discharging discharging current current of of 80 80 A. A. Figure 14. SOC of the VRB under a charging and discharging current of 80 A.
Sustainability 2017, 9, 441 Sustainability2017, 2017,9,9,441 441 Sustainability
12 of 15 12of of15 15 12
Similarly,Figures Figures15–17 15–17show showthe thesimulation simulationresults results under a charging charging and discharging power Similarly, Figures 15–17 show the simulation results under and discharging power Similarly, under a acharging and discharging power of of 5 kW. of 5 kW. 5 kW.
Figure15. 15.Current andvoltage voltageof ofthe theVRB VRBunder underaaacharging chargingand anddischarging dischargingpower powerof of555kW. kW. Figure 15. Current and and voltage of the VRB under charging and discharging power of kW. Figure
Figure16. 16.Power ofthe theVRB VRBunder underaaacharging chargingand anddischarging dischargingpower powerof of555kW. kW. Figure 16. Power of of the VRB under charging and discharging power of kW. Figure
Figure17. 17.SOC SOCof ofthe theVRB VRBunder underaaacharging chargingand anddischarging dischargingpower powerof of555kW. kW. Figure Figure 17. SOC of the VRB under charging and discharging power of kW.
InFigure Figure15, 15,the thequantitative quantitativedeviation deviationof ofthe thecurrent currentIIbattery batteryand andIIstack stackis isless lessthan than55A, A,which whichagain again In In Figure 15, the quantitative deviation of the current Ibattery and Istack is less than 5 A, which verifies that the stack current I stack of the VRB can be approximated as I battery . From Figure 16, the verifies that the stack current Istack of the VRB can be approximated as Ibattery. From Figure 16, the again verifies that the stack current Istack of the VRB can be approximated as Ibattery . From Figure 16, efficiencyof ofthe theVRB VRBunder underaaconstant constantcharging chargingpower powerof of55kW kWisisabout about0.85 0.85and andthe theefficiency efficiencyof ofthe the efficiency the efficiency of the VRB under a constant charging power of 5 kW is about 0.85 and the efficiency of VRB under a constant discharging power of 5 kW is about 0.8. The efficiency of the VRB reaches its VRB under a constant discharging power of 5 kW is about 0.8. The efficiency of the VRB reaches its the VRB under a constant discharging power of 5 kW is about 0.8. The efficiency of the VRB reaches its maximumvalue valuewhen whenthe theSOC SOCisis0.975. 0.975.In Inaddition, addition,the theefficiency efficiencyof ofthe theVRB VRBincreases increasesin inpace pacewith with maximum the increase of the SOC, as shown in Figures 13 and 16. the increase of the SOC, as shown in Figures 13 and 16.
Sustainability 2017, 9, 441
13 of 15
maximum value when the SOC is 0.975. In addition, the efficiency of the VRB increases in pace with the increase of the SOC, as shown in Figures 13 and 16. In Figure 17, the charging time of the VRB under 5 kW power is about 5 h, and the discharging time of the VRB under 5 kW power is about 3.25 h. Compared with Figure 14, the error between the estimation value of the SOC and the actual value of the SOC is much bigger, which is mainly due to the variability of current Ibattery under a constant charging or discharging power of 5 kW. However, this error is still very small, which again verifies the rationality and efficiency of the proposed SOC estimation method for a VRB ESS. Moreover, the curves of the charging and discharging current Ibattery in Figure 12 and the curves of the charging and discharging power Pbattery in Figure 16 demonstrate a fast response, strong adaptation, and high steady precision, which reasonably verify the proper control parameters design of the VRB ESS. 6. Conclusions DGs connected to a distribution system could critically cause some security risks for distribution systems. Considering the low terminal voltage of a VRB, this paper firstly designed a two-stage topology of a VRB energy storage system, in which a phase-shifted full bridge dc-dc converter was used and the bidirectional power flowing between the energy storage battery and the AC grid was achieved. Following this, according to the operational characteristics of the VRB in this proposed topology of a VRB ESS, the model of the VRB was simplified and the control parameters of the ESS were designed. For effectively estimating the SOC of the VRB, a traditional method for SOC estimation was simplified, and a simple and effective SOC estimation method was proposed in this paper. Finally, the feasibility and effectiveness of the proposed SOC estimation method were verified by the designed VRB ESS, and the test results demonstrated that this proposed SOC estimation method is reasonable for indicating the SOC of a VRB, and that the design of the VRB ESS is suitable for VRB applications. Acknowledgments: This work was supported by The National Key Technology R&D Program of China (2013BAA02B01) and the Qinghai Province Key Laboratory of Photovoltaic Grid Connected Power Generation Technology (2014-Z-Y34A). Author Contributions: Qingwu Gong proposed the concrete ideas of the proposed optimization method. Jiazhi Lei performed the simulations and wrote the manuscript. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5. 6. 7.
Qiu, X.; Crow, M.L.; Elmore, A.C. A balance-of-plant vanadium redox battery system model. IEEE Trans. Sustain. Energy 2015, 6, 1–8. [CrossRef] Díaz, A.; Ramos-Real, F.; Marrero, G.; Perez, Y. Impact of electric vehicles as distributed energy storage in isolated systems: The case of tenerife. Sustainability 2015, 7, 15152–15178. [CrossRef] Williams, B.R.; Hennessy, T. Energy oasis [vanadium redox battery system in power distribution application]. Power Eng. 2005, 19, 28–31. [CrossRef] Li, L.; Huang, Z.; Li, H.; Lu, H. A high-efficiency voltage equalization scheme for supercapacitor energy storage system in renewable generation applications. Sustainability 2016, 8, 548. [CrossRef] Azib, T.; Bethoux, O.; Remy, G.; Marchand, C. An innovative control strategy of a single converter for hybrid fuel cell/supercapacitor power source. IEEE Trans. Ind. Electron. 2011, 57, 4024–4031. [CrossRef] Wu, J.C.; Wang, Y.H. Three-phase to single-phase power-conversion system. IEEE Trans. Power Electron. 2011, 26, 453–461. [CrossRef] Xia, T.; He, L.; An, N.; Li, M.; Li, X. Electromechanical transient modeling research of energy storage system based on power system security and stability analysis. In Proceedings of the IEEE International Conference on Power System Technology, Chengdu, China, 20–22 October 2014; IEEE Press: New York, NY, USA, 2014; pp. 221–226.
Sustainability 2017, 9, 441
8.
9. 10.
11.
12. 13. 14.
15.
16.
17.
18.
19.
20. 21.
22. 23. 24.
25. 26. 27.
14 of 15
Ni, L.; Patterson, D.J.; Hudgins, J.L. A high power, current sensorless, bi-directional, 16-phase interleaved, DC-DC converter for hybrid vehicle application. In Proceedings of the IEEE Energy Conversion Congress and Exposition, Raleigh, NC, USA, 12–16 September 2010; IEEE Press: New York, NY, USA, 2010; pp. 1141–1151. Jin, K.; Ruan, X.; Yang, M.; Xu, M. Power management for fuel-cell power system cold start. IEEE Trans. Power Electron. 2009, 24, 2391–2395. [CrossRef] Li, X.; Zhang, W.; Li, H.; Xie, R. Design and control of bi-directional DC/DC converter for 30 kW fuel cell power system. In Proceedings of the IEEE 8th International Conference on Power Electronics and ECCE Asia (ICPE & ECCE), Jeju, Korea, 30 May–3 June 2011; IEEE Press: New York, NY, USA, 2011; pp. 1024–1030. Sabate, J.A.; Vlatkovic, V.; Ridley, R.B.; Lee, F. Design considerations for high-voltage high-power full-bridge zero-voltage-switched PWM converter. In Proceedings of the IEEE Applied Power Electronics Conference and Exposition (APEC), Los Angeles, CA, USA, 11–16 March 1990; IEEE Press: New York, NY, USA, 1990; pp. 275–284. Jeon, S.J.; Cho, G.H. A zero-voltage and zero-current switching full bridge dc-dc converter with transformer isolation. IEEE Trans. Power Electron. 2001, 16, 573–580. [CrossRef] Schutten, M.J.; Torrey, D. Improved small-signal analysis for the phase-shifted pwm power converter. IEEE Trans. Power Electron. 2003, 18, 659–669. [CrossRef] Barote, L.; Marinescu, C. A new control method for VRB SOC estimation in stand-alone wind energy systems. In Proceedings of the IEEE International Conference on Clean Electrical Power, Capri, Italy, 9–11 June 2009; IEEE Press: New York, NY, USA, 2009; pp. 253–257. Blanc, C.; Rufer, A. Multiphysics and energetic modeling of a vanadium redox flow battery. In Proceedings of the IEEE International Conference on Sustainable Energy Technologies in Singapore, 24–27 November 2008; IEEE Press: New York, NY, USA, 2008; pp. 696–701. Li, M.H.; Funaki, T.; Hikihara, T. A Study of Output Terminal Voltage Modeling for Redox Flow Battery Based on Charge and Discharge Experiments. In Proceedings of the Power Conversion Conference, Nagoya, Japan, 2–5 April 2007; IEEE Press: New York, NY, USA, 2007; pp. 221–225. Barote, L.; Weissbach, R.; Teodorescu, R.; Marinescu, C.; Cirstea, M. Stand-alone wind system with Vanadium Redox Battery energy storage. In Proceedings of the IEEE International Conference on Optimization of Electrical and Electronic Equipment, Brasov, Romania, 22–24 May 2008; IEEE Press: New York, NY, USA, 2008; pp. 407–412. Barote, L.; Marinescu, C.; Georgescu, M. VRB modeling for storage in stand-alone wind energy systems. In Proceedings of the IEEEBucharest PowerTechin, Bucharest, Romania, 28 June–2 July 2009; IEEE Press: New York, NY, USA, 2009; pp. 1–6. Zhang, Y.; Zhao, J.; Wang, P.; Skyllas-Kazacos, M.; Xiong, B.; Badrinarayanan, R. A comprehensive equivalent circuit model of all-vanadium redox flow battery for power system analysis. J. Power Sources 2015, 290, 14–24. [CrossRef] Tang, A.; Bao, J.; Skyllas-Kazacos, M. Studies on pressure losses and flow rate optimization in vanadium redox flow battery. J. Power Sources 2014, 248, 154–162. [CrossRef] Lawder, M.T.; Suthar, B.; Northrop, P.W.C.; De, S.; Hoff, C.M.; Leitermann, O. Battery energy storage system (bess) and battery management system (bms) for grid-scale applications. Proc. IEEE 2014, 102, 1014–1030. [CrossRef] Santhanagopalan, S.; White, R.E. State of charge estimation using an unscented filter for high power lithium ion cells. Int. J. Energy Res. 2010, 34, 152–163. [CrossRef] Guggenberger, J.; Elmore, C.; Tichenor, J.; Crow, M. Performance prediction of a vanadium redox battery for use in portable, scalable microgrids. IEEE Trans. Smart Grid 2012, 3, 2109–2116. [CrossRef] Nguyen, T.A.; Xin, Q.; Guggenberger, J.D.; Crow, M.L.; Elmore, A.C. Performance characterization for photovoltaic-vanadium redox battery microgrid systems. IEEE Trans. Sustain. Energy 2014, 5, 1379–1388. [CrossRef] Corcuera, S.; Skyllaskazacos, M. State-of-charge monitoring and electrolyte rebalancing methods for the vanadium redox flow battery. Eur. Chem. Bull. 2012, 1, 511–519. Gharavian, D.; Pardis, R.; Sheikhan, M. ZEBRA battery SOC estimation using pso-optimized hybrid neural model considering aging effect. IEICE Electron. Express 2012, 9, 1115–1121. [CrossRef] Skyllas-Kazacos, M.; Kazacos, M. State of charge monitoring methods for vanadium redox flow battery control. J. Power Sources 2011, 196, 8822–8827. [CrossRef]
Sustainability 2017, 9, 441
28.
29.
30.
31.
32. 33.
34.
15 of 15
Xiong, B.; Zhao, J.; Wei, Z.; Skyllas-Kazacos, M. Extended kalman filter method for state of charge estimation ofvanadium redox flow battery using thermal-dependent electrical model. J. Power Sources 2014, 262, 50–61. [CrossRef] Ghazanfari, A.; Hamzeh, M.; Mokhtari, H.; Karimi, H.R. Active power management of multihybrid fuel cell/supercapacitor power conversion system in amedium voltage microgrid. IEEE Trans. Smart Grid 2012, 3, 1903–1910. [CrossRef] Shen, H.; Zhang, Y.; Shi, Y.L.; Sun, L. Research on control strategy of three-phase grid-connected inverter under distorted and unbalanced voltage conditions. In Proceedings of the IEEE Transportation Electrification Asia-Pacific, Beijing, China, 31 August–3 September 2014; IEEE Press: New York, NY, USA, 2014; pp. 1–6. Wu, Y.; Li, N.; Yao, Z.; Xue, Y.; Wu, D. Control strategy based on double coordinates for three-phase grid-connected inverters. In Proceedings of the International Power Electronics and Motion Control, Harbin, China, 2–5 June 2012; IEEE Press: New York, NY, USA, 2012; pp. 2272–2276. Vlatkovi´c, V.; Sabaté, J.; Ridley, R.B.; Lee, F.C.; Cho, B.H. Small-signal analysis of the phase-shifted pwm converter. IEEE Trans. Power Electron. 1992, 7, 128–135. [CrossRef] Hu, X.; Nan, G. The research of modeling and simulation for phase-shifted full-bridge ZVS DC/DC converter. In Proceedings of the IEEE International Conference on Intelligent Information Technology Application, Nanchang, China, 21–22 November 2009; IEEE Press: New York, NY, USA, 2009; pp. 549–552. Liu, X.D.; Wu, J.; Liu, S.C.; Fang, W.; Liu, Y.F. A current control strategy of three-phase grid-connected inverter with LCL filter based on one-cycle control. In Proceedings of the 17th International Conference on Electrical Machines and Systems (ICEMS), Hangzhou, China, 22–25 October 2014; IEEE Press: New York, NY, USA, 2014; pp. 939–943. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
