Energy Storage System Sizing Based on a Reliability ... - MDPI

sustainability Article

Energy Storage System Sizing Based on a Reliability Assessment of Power Systems Integrated with Wind Power Nian Shi * and Yi Luo State Key Laboratory of Advanced Electromagnetic Engineering and Technology, Huazhong University of Science and Technology, Wuhan 430074, China; [email protected] * Correspondence: [email protected] Academic Editor: Tomonobu Senjyu Received: 18 January 2017; Accepted: 3 March 2017; Published: 7 March 2017

Abstract: The available capacity is a major factor that influences the reliability contribution of energy storage in power systems integrated with wind power. This paper presents the capacity value of the energy storage metrics to quantitatively estimate the contribution of energy storage to the generation adequacy. A method in accordance with EFC approach has been introduced to model the capacity value of energy storage. The adequacy-oriented model of the energy storage available capacity is proposed for the energy storage system, regarding the roles of the key parameters for the CVES analysis. The case study results indicate that the capacity value of energy storage quantitatively weigh the contribution of the energy storage to system reliability. The sensitivity analysis of the impact factors for the CVES is conducted. Keywords: generation adequacy; capacity value; energy storage system; wind power

1. Introduction The utilization and development of distributed generators to satisfy electrical load demands has received considerable attention in the last decade. Generation adequacy is utilized to weigh the ability of the available supply capacity to satisfy the load demand. Owing to the fast increase in installed distributed resources, such as wind turbines and energy storage systems, researchers have become more concerned about power system reliability and its economics [1] after the distributed resources are integrated to the system. Energy storage is a promising technology for improving the generation adequacy of a power system integrated with intermittent energy, and also brings challenges to the security and reliability of the microgrid [2]. Regarding the time scale, the role of an energy storage system (ESS) is recognized either for reliability improvement [3], or for smoothing the fluctuations of the intermittent energy output [4,5]. The energy storage system can respond by suppressing the changes caused by wind and photovoltaic power. Therefore, allocating the capacity of energy storage system is a fascinating challenge to achieve better performance [6–8]. Capacity value is the metric typically used to quantify the contribution of generation on system reliability and for long-run resource adequacy planning [9]. Capacity value is often estimated using reliability-based methods [9]. In these studies, the chronological method and the probability method are used. In addition, the two methods have preferred solutions for different people within the power grid. System operators mainly use the chronological method, while system planners use the probability method [10]. For the probabilistic method, there are four typical capacity value definitions [11], i.e., the equivalent firm capacity (EFC) [12,13], equivalent conventional power plant [10,14], effective load

Sustainability 2017, 9, 395; doi:10.3390/su9030395

www.mdpi.com/journal/sustainability

Sustainability 2017, 9, 395

2 of 20

carrying capability (ELCC) method [15–18] and guaranteed capacity. The EFC and ELCC methods are the two most commonly used methods [19]. With respect to the assessment of the contribution of generation devices to the reliability of a power system, current work is mainly centered on intermittent energy such as wind power [3]. Besides, with the solar power as a not negligible part of the power supply, the capacity value of the photovoltaics [20] and the concentrated solar power [14] are also investigated. The studies are concerned with modeling the capacity value of wind power and photovoltaics themselves [15,21]. However, energy storage, for its part, also makes important contributions to the generation adequacy in a power grid. Many factors influence the contribution of the energy system to the grid. First, the characteristics of energy storage systems have added complexity to this problem. In addition, with the development of interactive smart grids, peak loads on power grids can be shifted to off-peak hours using distributed optimization and control solutions [22,23]. Demand side management ultimately influences the energy storage allocation demand [24–26]. Therefore, the contribution of energy storage is often over-estimated, if the contribution has been simply measured by the installed capacity of the energy storage. Thus far, there has been no comprehensive framework to quantitatively assess the energy storage available with a capacity metric. Furthermore, no developed method has been proposed to systematically evaluate the available capacity for energy storage and consider the influencing factors and parameters. On these premises, intent on how to develop an assessment model to measure the contribution of energy storage to the generation adequacy of a power grid, the authors have worked out an assessment method to quantitatively assess the available capacity of energy storage, providing a basic method for planners to allocate the energy storage capacity. This paper presents a novel and comprehensive framework to evaluate the available energy storage capacity, which effectively establishes a bridge between the reliability contribution and installed capacity for energy storage. The key contributions of this systematic framework include the following: (1)

(2)

(3)

Introduction of the definition of the capacity value of energy storage (CVES) metrics, in terms of the available capacity; the renewable energy-oriented available capacity metrics are extended to access the energy storage system generator adequacy contribution. Moreover, new CVES metrics are defined to quantify the energy storage that can displace the generation capacity. The adequacy-oriented models of the energy storage available capacity. In addition to the CVES metrics, an assessment model is proposed for the energy storage system, regarding the roles of the key parameters for the CVES analysis. An algorithm based on a cooperative equivalent firm capacity to evaluate the CVES metrics in detail by applying this framework to the CVES. Two case studies provide general insights into the CVES concept with different influencing factors.

The rest of this paper is organized as follows: Section 2 presents the definition, the model and the solution of the CVES. Then, the load model, the energy storage model and the wind power model related to the CVES calculation are illustrated in Section 3. The proposed method is applied in the test system IEEE-RTS in Section 4, and the influencing factors of the CVES are analyzed. Finally, the sensitivity analysis is provided. 2. Capacity Value of Energy Storage The energy storage system stores the surplus energy to satisfy the load demand when power shortages occur, and it effectively improves the generation adequacy. This section describes the definition of CVES. The value of storage from the capacity value is utilized to weigh the capacity contribution of energy storage devices to the generation adequacy. To better improve the system reliability, the energy storage devices are installed near the wind farm in this paper, as shown in Figure 1.

Sustainability 2017, 9, 395 Sustainability 2017, 9, 395

3 of 20 3 of 20

Figure Figure 1. Schematic Schematic diagram diagram of of the the wind-integrated system with energy storage. storage.

2.1. Definition Definition of of the the Capacity Capacity Value Value of of Energy Energy Storage Storage 2.1. The CVES CVESrepresents representsthe theequivalent equivalentcapacity capacity energy storage is utilized to satisfy the The of of energy storage thatthat is utilized to satisfy the load load demand in a power The energy ofstorage energydevices storageisdevices is from different demand in a power system.system. The energy offeringoffering mode ofmode energy different that from that of wind turbines. Energy storage improves the generation adequacy by storing of wind turbines. Energy storage improves the generation adequacy by storing conventional energy conventional energy to satisfy the added load demand. In addition, also describes the sources to satisfy the sources added load demand. In addition, CVES also describes CVES the capacity of the load capacity satisfied of the load demand peak load storage shiftingsystem of the energy system with the demand by peak loadsatisfied shifting by of the energy with thestorage same reliability index. sameThe reliability index. reliability index of a power system is correlated to the added installed capacity (conventional The reliability index ofand a energy power storage system devices) is correlated the added installed capacity power plants, wind turbines and thetoadded load under the condition (conventional power plants, wind turbines and energy storage devices) and the added load under of the invariable load curve and generator capacity. The index measures the effect on the supply the condition of the invariable load curve and generator capacity. The index measures the effect adequacy of a power plant with different properties and load demands. Thus, the contributions on of the supply adequacy a power with different and demands. Thus, the different types of powerofplants to theplant supply adequacy canproperties be assessed by load the index. contributions types ofBaleriaux–Booth power plants toformula the supply adequacy can be assessed byload the Accordingoftodifferent the well-known [27], the corresponding equivalent index. duration curve (ELDC) is as follows, considering the influence of the installed energy storage on the According to the well-known Baleriaux–Booth formula [27], the corresponding equivalent load generation adequacy: 0 duration curve (ELDC) is as follows, of the installed energy storage on the f ES ( x )considering = pkES f 0 ( x )the + qinfluence (1) kES f ( x − CES ) generation adequacy: where f 0 (x) is the original load duration curve and CES is the capacity of the installed energy storage. ES )  pkES f 0 ( xthat )  qthe f 0 ( x  CES ) of the energy storage is 0, and (1) In addition, qkES is supposed to befthe( xprobability kES output power pkES is given by pkES = 1 − qkES . where f0(x) is the original load duration curve and CES is the capacity of the installed energy storage. The output power of an energy storage system is related to its stored energy and the load level of In addition, qkES is supposed to be the probability that the output power of the energy storage is 0, the power system. When the wind power is sufficient to satisfy the load demand or only the minimum and pkES is given by pkES = 1 − qkES. energy is stored in the energy storage devices, the output power of energy storage is 0. The output power of an energy storage system is related to its stored energy and the load level Then, the expression of LOLP is given by: of the power system. When the wind power is sufficient to satisfy the load demand or only the minimum energy is storedLOLP in the energy storage devices, theESoutput (Cw +power CES ) of energy storage is 0. (2) ES = P ( Ek > Cw + CES ) = f Then, the expression of LOLP is given by: where Ek is the equivalent load and Cw is the output power ofESthe wind turbines. LOLPES  P ( E k  C w  C ES )  f ( C w  C ES ) (2) To obtain the equivalent capacity of the energy storage system, it is supposed that a unit is installed of storingload energy power system. Therefore, considering is the equivalent andin Cwthe is the output power of the wind turbines. the influence of the where Ek instead added on the corresponding ELDC is: Tounit obtain thegeneration equivalentadequacy, capacity the of the energy storage system, it is supposed that a unit is

installed instead of storing energy in the power system. Therefore, considering the influence of the f C ( x ) =the pkCcorresponding f 0 ( x ) + qkC f 0 (ELDC x − Cxis: ) (3) added unit on the generation adequacy, 0 0 where f 0 (x) is the original load duration curve of the added unit. In addition, f C ( x)  pkC f and ( x) Cqx kCis fthe ( x capacity  Cx ) (3) pkC is the forced outage rate of the unit, and qkC is defined as qkC = 1 − pkC . is the original of load duration curve where f0(x)the Thus, expression LOLP is given by: and Cx is the capacity of the added unit. In addition, pkC is the forced outage rate of the unit, and qkC is defined as qkC = 1 − pkC. Thus, the expression of LOLP LOLPCis=given P( Ek by: > Cw + Cx ) = f C (Cw + Cx ) (4)

C

LOLP  P( Ek ofCC  Cx ) satisfies  f (Cwthe  Cfollowing: ) C a value w x that x From Equation (4), we can obtain From Equation (4), we can obtain a value of Cx that satisfies the following: LOLPES = LOLPC LOLPES  LOLPC

(4) (5)

(5)


4 of 20

The value of Cx that achieves Equation (5) represents the capacity value of the added energy storage system without changing the shape of the load duration curve. Hence, the capacity value of the energy storage (CVES) presented in this paper can be defined as: f ES (Cw + CES ) = f C (Cw + CVES)

(6)

P( PL > PWind + PES_output ) = P( PL > PWind + CVES)

(7)

It can also be given by:

In Equation (7), PWind is the total output power of the wind turbines installed in this power system. The variable PES_output is the output power of the energy storage system. The calculation of CVES is based on certain reliability criteria. Therefore, the factors that affect the reliability of the power grid also affect the storage value through the capacity value. According to Equation (7), the characteristics of the wind turbine, load demand and energy storage system will affect the CVES. In this paper, the CVES represents the supply adequacy added in the system with the same reliability index after the energy storage devices are installed. The CVES is a value to quantitatively measure the influence of the added energy storage on the generation adequacy. 2.2. EFC Method for Calculating the Capacity Value of Energy Storage In a traditional reliability evaluation, a forced outage of the generating unit is one of the main causes of a power supply shortage. After the integration of a wind turbine, the intermittent characteristics of wind power are an additional important cause of a power shortage. There are many approaches of different equivalent ways to calculate the value of storage using the capacity value. From the perspective of the power supply, the CVES represents the conventional power plant capacity that can be replaced by an energy storage system. From the load perspective, the CVES is the load capacity that is supplied by the added energy storage system. The equivalent firm capacity (EFC) method is one of the common methods for capacity value assessment. In this paper, the EFC is utilized to calculate the value of storage from the capacity value. After the energy storage devices are installed, the equivalent firm capacity of GEFC is defined as the capacity of a unit, when maintaining the same HLOLE without changing the load duration curve. Furthermore, the added firm capacity is the evaluation measure for the system reliability, and the total load rejection hours are the reliability index. In the previous studies, the LOLP [28,29] and LOLE [30–33] were mainly utilized as the power system reliability indexes to assess the generator capacity adequacy [19,34]. However, the power supply mode of energy storage is different from conventional generators. The output power of the energy storage system is affected by the status of the energy storage devices during the previous time step. Thus, the time interval of the wind speed and the reliability index are the same. In this paper, an hourly loss of load expectation (HLOLE) is presented as the reliability evaluation index for the CVES assessment. The time interval of the HLOLE is 1 h. The HLOLE represents the number of hours during which the generators cannot meet the load demand in a given period n. It is calculated by: n

HLOLE =

∑ Pi (Gi < Li )

(8)

i =1

where n is the total number of hours of the given period, Pi (Gi < Li ) is the load loss probability at hour i, and the value of Pi is given in Table 1.


5 of 20

Table 1. Load loss probability for different system statuses. System Status

Load Loss Probability

The load demand is lower than the power supply Other conditions

0 1

Before the energy storage devices are installed, the original system integrated with intermittent energy has a certain generation adequacy. After the energy storage is added, the power supply system can satisfy more load demands. The generation adequacy of the power system is improved, and the added generation adequacy is supplied by the installed energy storage. To quantitatively represent the capacity value of energy storage, the EFC method is utilized in this paper. The capacity value of energy storage indicates the equivalent capacity that replaces the added energy storage while maintaining the same value of HLOLE. After energy storage is added, to calculate the value of the EFC of GES , the HLOLE0 of the system can be given by: n

HLOLE0 =

∑ Pi (Gi + PES_output,i < Li )

(9)

i =1

After the firm capacity, GEFC , replaces the added energy storage, the HLOLEf of the system is calculated by: n

HLOLE f =

∑ Pi (Gi + GEFC < Li )

(10)

i =1

The capacity of the unit, GEFC , is adjusted until: HLOLE f = HLOLE0

(11)

The GEFC that satisfies Equation (11) is the capacity value of the added energy storage. Therefore, the value of storage from the capacity value can be calculated by: n

n

i =1

i =1

∑ Pi (Gi + PES_output,i < Li ) = ∑ Pi (Gi + CVES < Li )

(12)

where Gi is the output power of the generators at hour i, PES_output,i is the output power of the added energy storage at hour i, Li is the load demand at hour i, and n is the number of hours in the time periods considered in this equation. Thus, the CVES in Equation (12) is the capacity value of energy storage. 2.3. The Algorithm Flowchart for the Capacity Value Based on the EFC Method This algorithm is used for calculating the value of storage from the capacity value on the basis of the EFC method and encompasses calculation process of the reliability index. Furthermore, for this analysis, the calculation of reliability index is needed, before and after the energy storage devices are installed. The calculation flowchart of the value of storage from the capacity value based on the EFC method is shown in Figure 2. The algorithm developments of CVES are detailed as follows: (1) (2) (3)

Input the data of the original power system, such as the load level, the wind speed, the wind turbine model and related parameters, and the energy storage model and related parameters. Calculate the load loss probability of the wind-integrated system at hour i, after the energy storage is added. Calculate the HLOLE0 of the power system integrated with wind power within the time period using Equation (9). When the load demand is larger than the available capacity at hour i, the


6 of 20

HLOLE0 is accumulated. In addition, when the load demand is smaller than the available capacity, the value of HLOLE0 remains the same. (4) Save the value of HLOLE0 of the wind-integrated power system with energy storage devices. (5) Replace the capacity of the energy storage system with the firm capacity. (6) Adjust the value of the firm capacity, CEFC . (7) Calculate the load loss probability of the wind-integrated system at hour i, after the firm capacity Sustainability 2017, 9, 395 6 of 20 is added. (8) Calculate the value of the HLOLEf of the power system integrated with wind power within the (8) Calculate theusing valueEquation of the HLOLE time period (10). f of the power system integrated with wind power within the period using Equation (10). (9) time Check whether Equation (11) is met. (9) Check whether Equation (11) is met. (10) If Equation (11) is not met, return to step (6) to adjust the value of the firm capacity; (10) If Equation (11) is not met, return to step (6) to adjust the value of the firm capacity; if Equation if Equation (11) is met, then the value of the HLOLE0 is equal to HLOLEf . In addition, the (11) is met, then the value of the HLOLE0 is equal to HLOLEf. In addition, the value of the firm value of the firm capacity is equal to the capacity value of energy storage. capacity is equal to the capacity value of energy storage.

Figure 2. 2. The flowchart. Figure The CVES CVES calculation calculation flowchart.

2.4. Communication System 2.4. Communication System For this this study, study,the theproposed proposedalgorithm algorithmcan canbebeused usedbybysystem system planners allocate capacity For planners to to allocate thethe capacity of of energy storage to make the full use of renewable energy. The calculation process is finished energy storage to make the full use of renewable energy. The calculation process is finished offline. offline. Theofmodel of practical for the calculation, eachcan scheme can beHowever, verified. The model practical system issystem set up is forset theup calculation, and each and scheme be verified. However, for engineering application, the power of wind turbines, the load demand, the SOC and for engineering application, the power of wind turbines, the load demand, the SOC and available available energy ofstorage the energy system change every next hour, thus an excellent energy of the energy systemstorage change every next hour, thus an excellent communication system communication system is an important part in the evaluation tool. To ensure the real ittime of this is an important part in the evaluation tool. To ensure the real time of this communication, is possible communication, it is possible for big areas having multiple ESS, and the area has to be divided into for big areas having multiple ESS, and the area has to be divided into small units. smallTo units. control the operation and management of area units, a local energy management system is To control the operation and data management of[2]. areaThe units, a local energy management system is needed for each area unit to which will be sent information received by the central control needed for each area unit to which data will be sent [2]. The information received by the central unit includes the fuel power, value of predicted renewable power, load management and the value control includes fuel power, value oftime. predicted renewable load management and of storedunit energy in thethe battery in the previous The main goal ofpower, the local energy management the value of stored energy in the battery in the previous time. The main goal of the local energy system is to ensure that the customer demand is met at all times, allocate the dispatchable load demand management is to ensure that the renewable customer resources. demand is met at all times, allocate the and maximize system the utilization of the available dispatchable load demand and maximize the utilization of available renewable[35]. resources. A central controller is needed to be responsible for thethe overall coordination It collects the A central controller is needed to be responsible for the overall coordination [35]. It collects the general properties of each area units [35,36]. The controller can make the required decisions for energy general properties eachunits areatounits [35,36]. The controller can make the required decisions for exchange within theofarea decrease production or increase consumption in situations when energy exchange within the area units to decrease production or increase consumption in situations when there is a surplus of electricity in the system based on offer prices [35]. The communication system is addressed in detail in [2,13]. 3. System Models


7 of 20

there is a surplus of electricity in the system based on offer prices [35]. The communication system is addressed in detail in [2,13]. 3. System Models The models for the load, wind power, energy storage and communication system are established in this section to analyze the CVES in the power system integrated with intermittent energy. As the reserve power supply of the intermittent power, energy storage effectively improves the generation adequacy by peak load shifting. Meanwhile, the factors that influence the output power of energy storage include the output characteristics of the intermittent generator, the characteristics of the load, and the control strategy for energy storage. In addition, the available output power of energy storage at hour (h + 1) is correlated to the energy stored in the energy storage system at hour h; thus, time-sequential models are utilized below. 3.1. Modeling Load The load models are approximate results of the actual facilities. To assess the CVES, an hourly load model is used to describe the actual load demand. The IEEE-RTS load model [30] and constant load model [30] are also utilized to analyze the effect of the load demands on the CVES. Data for the weekly peak loads as a percent of the annual peak load, a daily peak load cycle as a percent of the weekly peak and weekday and weekend hourly load models for each of the three seasons are provided in the IEEE-RTS load model. We can determine the load value at every hour according to the three parts. Otherwise, to minimize the production costs and the market clearing prices and to improve the utilization of renewable energy resources, optimal operation programming of electrical systems has been proposed. To reach this goal, energy management systems have been studied in many research studies. Moreover, a demand response expands customer participation in power systems and results in a paradigm shift from conventional to interactive activities in power systems due to the progress of smart grid technology [23]. Meanwhile, the power consuming behavior of users has been directed by the adjustment of the load management electricity structure. During power shortages, the users are prompted to reduce their electricity load. When the available capacity is abundant, users are encouraged to maximize the use of electricity so that the load curve is smoother for achieving the purpose of peak-load shifting. 3.2. Modeling the Wind Power The wind power output is related to the wind speed and the wind turbine parameters. The ARMA model [32,33,36] and Weibull model [29,30] have been widely used to model the wind speed. The parameters of these wind speed models are estimated using historical data. A wide range of wind speed data was obtained from data measured during a long period in the same area, in [35]; the historical data cover a long span of time and are used for the wind speed characteristic analysis. The wind speed output probability model is estimated by the cluster of the wind speed history data. The yearly historical chronological wind speed data are directly utilized to describe the wind power model in this paper. According to the wind speed, the output can be divided into four ranges, including the input speed, output speed and rated speed [30]. The output power of a single wind turbine can be calculated as:

PWind

  0,    P ( C + C v + C v2 ), r 2 3 1 =  Pr ,    0,

0 ≤ v < Cin Cin ≤ v < Crate Crate ≤ v < Cout Cout ≤ v

(13)


8 of 20

In this study, 10-year-old historical wind speed data from Prince Albert in Canada is adopted. Based on the wind speed data, Figure 3 describes the probability distribution of the wind speed, while Sustainability 2017, 395 88 of Figure 4 shows the Sustainability 2017, 9, 9, 395wind power probability of a single wind turbine for each year. of 20 20

Figure Figure 3. 3. Cumulative Cumulative distribution distribution curve curve of of the the wind wind speed. speed.

Figure 4. single wind turbine output Figure 4. The The output power power probability. probability. Figure 4. The single single wind wind turbine turbine output power probability.

3.3. the Energy Storage 3.3. Modeling Modeling the the Energy Energy Storage Storage 3.3. Modeling Intermittent such as wind and photovoltaic energy, is considerably influenced by Intermittent energy, energy, such such as as wind wind and and photovoltaic photovoltaic energy, energy, is is considerably considerably influenced influenced by by Intermittent energy, environmental conditions with intermittent factors that decide when the intermittent energy is environmental conditions with intermittent factors that decide when the intermittent energy is environmental conditions with intermittent factors that decide when the intermittent energy is unstable unstable energy for output. The energy storage is capable of effectively providing reserve energy unstable output. energy storageofiseffectively capable of effectively providing energy forenergy output.for The energyThe storage is capable providing reserve energyreserve during energy power during power supply shortages due to the intermittency of the power. during power supply shortages due to the intermittency of the power. supply shortages due to the intermittency of the power. The model of the energy storage system includes the control approach, energy stored approach The model model of of the the energy energy storage storage system system includes includes the the control control approach, approach, energy energy stored stored approach approach The and maximum available power. The contribution of the stored energy to the power supply and maximum maximumavailable available power. contribution the stored to the power supply and power. TheThe contribution of theofstored energy energy to the power supply adequacy adequacy is dependent on the output power of the stored energy, the restricted and maximum adequacy is dependent on the output power of the stored energy, the restricted and maximum is dependent on the output power of the stored energy, the restricted and maximum permission permission charging power, the power of storage and its strategy. Their permission charging power,power the surplus surplus power of energy energy andstrategy. its operation operation Their charging power, the surplus of energy storage and itsstorage operation Their strategy. relationship is relationship is described as: relationship described as:is described as: PES_output = , ,PPStatus_discha, ,EEBat, P (14) Available P P )) ) (14) PES f (ffP((Status_cha PStatus , PStatus , EBat , ,PPAvailable (14) ES __ output output Status __ cha cha Status __ discha discha Bat Available Therefore, the output power of the stored energy is constrained by the output power of the wind Therefore, the power of stored energy is constrained by output power of Therefore, the output output power of the the the stored energy is maximum constrained by the the output In power of the the generator, the power differences between loads and the energy allowed. addition, it wind generator, the power differences between the loads and the maximum energy allowed. In wind generator, the power differences between the loads and the maximum energy allowed. is related to the control strategy [32] and the status and the output function of the energy storage. It In is addition, it is related to strategy [32] and the and the output addition, is energy relatedlimitations to the the control control strategy andlimits the status status theenergy. output function function of of the the limited byitthe and the output[32] energy of the and stored energy storage. It is limited by the energy limitations and the output energy limits of the stored energy storage. It is limited by the energy limitations and the output energy limits of the stored The output power of the energy storage system is also affected by its failure rate. The energy energy. energy. battery module is composed of a series of parallel batteries and a battery management storage The output power of energy storage system is affected by failure rate. The The[37]. output power of the the energy system is also also affected by its itssystem, failure the rate.components The energy energy system With regard to the gridstorage connected battery energy storage storage storage battery battery module module is is composed composed of of aa series series of of parallel parallel batteries batteries and and aa battery battery management management system system [37]. [37]. With With regard regard to to the the grid grid connected connected battery battery energy energy storage storage system, system, the the components components are are classified as a series system for reliability analysis. The power conversion system classified as a series system for reliability analysis. The power conversion system is is an an important important component component of of the the grid-connected grid-connected energy energy storage storage system. system. Therefore, Therefore, the the failure failure rate rate of of the the grid-connected battery energy storage system is related to the failure rate of the energy storage and grid-connected battery energy storage system is related to the failure rate of the energy storage and


9 of 20

are classified as a series system for reliability analysis. The power conversion system is an important component of the grid-connected energy storage system. Therefore, the failure rate of the grid-connected battery energy storage system is related to the failure rate of the energy storage and the power conversion system. In this paper, it is assumed that the failure rate is λbat , and the repair rate is µbat . 3.3.1. Charging and Discharging Intent on the reliability assessment of the wind-integrated power system with energy storage, a simple control approach is utilized in this model without considering the energy price, operational reserve opportunities, etc. [38]. To clearly discuss the CVES, the energy storage system is set near the wind turbines, to carry out charging with the full use of wind power. Once the wind power is no longer sufficient to satisfy the load demand, the stored energy in the ESS will be utilized to afford the demand. The operation strategies for energy storage are stated as follows. When the wind power is able to satisfy the load demand (PWind − PL ≥ 0) at hour i, the ESS stores the remainder of the electric energy from the power grid, and when the wind power cannot satisfy the load demand (PWind − PL ≤ 0) at hour i, the energy storage supplies power back to the grid. Thus, the amount of charging power of the stored energy at hour h is: ( PStatus_cha =

PEScha_Max , PWind − PL ,

PWind − PL ≥ PEScha_Max PEScha_Max ≥ PWind − PL ≥ 0

(15)

The amount of discharging power of the stored energy at hour h is: ( PStatus_discha =

PWind − PL , 0 ≥ PWind − PL ≥ − PESdischa_Max − PESdischa_Max , − PESdischa_Max ≥ PWind − PL

(16)

The variables PEScha_Max and PESdischa_Max are the constraints on the charging and discharging power of the stored energy. The maximum charging and discharging power should not exceed these values. 3.3.2. Energy Stored in the Energy Storage System The energy stored in the energy storage system is a piecewise function related to the power status of the ESS. The energy of the ESS at a certain time is related to the energy of the previous time step. The output energy of the ESS is calculated as:  −1   Eoutput (h) + TPES_output , Emin ≤ Eoutput (h) + T PES_output ≤ Emax − 1 Eoutput (h + 1) = Emin , Eoutput (h) + T PES_output ≤ Emin   E , E −1 max max ≤ Eoutput ( h ) + T PES_output

(17)

where Emax is the highest energy stored in the energy storage system. The stored energy should not exceed Emax . In addition, Emin is the constraint for the lowest energy stored in the ESS, that is, the stored energy should not be lower than Emin . The variable T represents the sampling period. Compared to other time periods, the response time of the battery energy storage is very small; thus, it is negligible in this model [39]. Furthermore, the rated capacity of the energy storage system is relative to the reference temperature. The maximum discharge capacity, the maximum charge/discharge current and the working life of the energy storage system will change with the ambient temperature [40]. Figure 5 shows the relationship between the temperature and the charge/discharge factors [37]. The charge/discharge factors in Figure 5 are the maximum charge and discharge current, respectively. The maximum charge and discharge current changes with the changing of the ambient temperature.


10 of 20

According to Figure 5a, if the temperature is in the range of 0–30 ◦ C, the charging current maintains the rated value. When the temperature increases to the range of 30–40 ◦ C, the charging current of the battery rapidly increases to 0. Furthermore, the battery cannot be charged when the temperature is higher than 40 9,◦ C. Sustainability 2017, 395In addition, according to Figure 5b, the discharge current is the rated value10when of 20 the temperature is in the range of 0–40 ◦ C. In addition, if the temperature is in the range of −20–0 ◦ C in the40–60 range◦ C, of the −20–0 °C and current 40–60 °C, the battery discharge current of the battery changes temperature and discharge of the changes with temperature linearly.with In particular, it is ◦ Crange linearly. In particular, it is not the battery the temperature of not advisable to discharge theadvisable battery if to thedischarge temperature is in theif range of less thanis−in20the or more ◦ less than −20 °C or more than 60 °C. than 60 C.

(a)

(b)

Figure Figure 5. 5. (a) (a) Effect Effect of of temperature temperature on on the the maximum maximum charge charge current; current; and and (b) (b) effect effect of of temperature temperature on on the maximum discharge current [40]. the maximum discharge current [40].

The charge/discharge capacity is also related to the temperature [40]. The reference The charge/discharge capacity is also related to the temperature [40]. The reference temperature is temperature is 25 °C. Therefore, the percentage factor of the charge capacity is 100% under the 25 ◦ C. Therefore, the percentage factor of the charge capacity is 100% under the reference temperature. reference temperature. When the ambient temperature is 5 °C lower than the reference temperature, When the ambient temperature is 5 ◦ C lower than the reference temperature, the capacity of the battery the capacity of the battery decreases nearly by 5% [40]. When the temperature decreases to 0 °C, the decreases nearly by 5% [40]. When the temperature decreases to 0 ◦ C, the capacity of the battery capacity of the battery decreases to 75%. decreases to 75%. 3.3.3. Available Power of Energy Storage 3.3.3. Available Power of Energy Storage The charging and discharging power of the energy storage system is affected by the different The charging and discharging power of the energy storage system is affected by the different values of the wind power and load. The available power from the energy storage system can be values of the wind power and load. The available power from the energy storage system can be described by: described by: 1 (  ( Eoutput ( h)  Emin ), T 1 ( Eoutput ( h)  Emin )  PESdischa _ max T − 1 − ), T 1 ( Eoutput (h) − Emin ) ≤ PESdischa_max P ( h  1) T  ( Eoutput (h) − Emin 1 PAvailable (hAvailable + 1) = P , T ( Eoutput(h( h) )− E Emin )) ≥PP − 1  ESdischa _ max PESdischa_max  ESdischa _ max, T ( Eoutput min ESdischa_max

(18) (18)

The The available available power power of of the the energy energy storage storage at at hour hour hh ++11isisrelated relatedto tothe theoutput output energy energy at at hour hour hh and the sampling period T. The variable P ESdischa_max is the constraint of the maximum discharging and the sampling period T. The variable PESdischa_max is the constraint of the maximum discharging power and the the available available power power should should not not exceed exceed this this constraint. constraint. power of of the the stored stored energy, energy, and 3.3.4. 3.3.4. Charging Charging Power Power Period Period of of Energy Energy Storage Storage The main process processofofcharging charging includes stages, constant current charging and constant The main includes twotwo stages, constant current charging and constant voltage voltage charging. Figure 6 gives a typical charging curve of batteries. As Figure 6 shows, the charging. Figure 6 gives a typical charging curve of batteries. As Figure 6 shows, in the initial in stages initial stagesprocess, of charging process,voltage the charging voltage until value. to the The set charging power value. The of charging the charging rises until to therises set power process is charging process is changed to constant power charging until to make the SOC of batteries to a changed to constant power charging until to make the SOC of batteries to a certain value. Then it is certain Then voltage it is changed constant voltage charging. State ofcapacity charge of (SOC) is the changedvalue. to constant charging.toState of charge (SOC) is the remaining a battery. In remaining of a we battery. the process charging, we need to take notice overcharging. of the SOC of the processcapacity of charging, need In to take notice ofofthe SOC of energy storage to avoid energy storage to avoid overcharging. In this paper, the sampling period is set to be 1 h. Therefore, as Figure 6 shows, the charging power can be approximately regarded as the rated power of ESS.

Sustainability Sustainability 2017, 2017, 9, 9, 395 395

11 11 of of 20 20

Figure 6. Charging curve of energy storage device. Figure 6. Charging curve of energy storage device.

In this paper, the sampling period is set to be 1 h. Therefore, as Figure 6 shows, the charging 4. Simulation Results power can be approximately regarded as the rated power of ESS. 4.1. System Analysis 4. Simulation Results The capacity assessment of the energy storage system contribution to the power supply is examined applying the IEEE-RTS [41] and RBTS [42]. The studied system comprises wind 4.1. System by Analysis turbines, a power unit and energy storage system (a battery in this study). The wind turbines are The capacity assessment of the energy storage system contribution to the power supply is located where the wind speed is the same as that in Prince Albert in Canada from year 2001 to 2012. examined by applying the IEEE-RTS [41] and RBTS [42]. The studied system comprises wind turbines, The wind speed data can be found on a Canadian website [43]. The rated wind power is 1000 kW. a power unit and energy storage system (a battery in this study). The wind turbines are located where The energy storage system is installed near a wind farm. The peak yearly peak power is 1800 kW, the wind speed is the same as that in Prince Albert in Canada from year 2001 to 2012. The wind speed and the capacity of fuel generator is 1000 kW. The SOC initial value is set at minimum value. data can be found on a Canadian website [43]. The rated wind power is 1000 kW. The energy storage The proposed method for estimation CVES value is implemented in MATLAB 7.0 platform system is installed near a wind farm. The peak yearly peak power is 1800 kW, and the capacity of fuel and executed with G2030 CPU, 3.0 GHz desktop computer. The calculations with different generator is 1000 kW. The SOC initial value is set at minimum value. discharge power employ the same maximum iteration settings (is set to 15,000 iterations). The The proposed method for estimation CVES value is implemented in MATLAB 7.0 platform and algorithm is explained in the following including execution time of the algorithm and the CVES executed with G2030 CPU, 3.0 GHz desktop computer. The calculations with different discharge power related to different Pdischa_max. The CCVE value and the computation time of the proposed algorithm employ the same maximum iteration settings (is set to 15,000 iterations). The algorithm is explained in with different discharge power are maintained in Table 2. The execution time increases with the the following including execution time of the algorithm and the CVES related to different Pdischa_max . increase of discharge power of energy storage system. However, the proposed algorithm can be The CCVE value and the computation time of the proposed algorithm with different discharge power finished in a few seconds. Therefore, the proposed algorithms should be able to fulfill this are maintained in Table 2. The execution time increases with the increase of discharge power of energy requirement in a short period. storage system. However, the proposed algorithm can be finished in a few seconds. Therefore, the As shown in Table 2, the CVES value of energy storage is closely related to the discharge proposed algorithms should be able to fulfill this requirement in a short period. power of energy storage system. The CVES values are approximately equal to the discharge power of EES, when the capacity is lower than 300 kW. This means that the ESS is fully used to feeding the Table 2. The CVES of energy storage with different discharge power. load demands at this working status. When the capacity keeps on increasing, the CVES values are less than the installed capacity.PThis is means that the adding Execution capacity of ESS(s)plays a minor role in Pchmax CVES Time dchmax improving the reliability of system. As a result, by this way, less expenses will be paid for feeding 250 200 199 13.109 them by consumers. 300 250 245 13.890 350 300 293 14.500 400 2. The CVES 350 of energy storage 332 with different discharge 15.454 power. Table 450 400 380 15.672 Execution Pchmax Pdchmax 500 450 424 CVES 16.719 Time (s) 250 13.109 550 500 200 460 199 17.156 600 550 250 489 245 17.438 300 13.890 350 300 293 14.500 400 350 332 15.454 As shown in Table 2, the CVES value of energy storage is closely related to the discharge power of 450 400 380 15.672 energy storage system. 500The CVES values 450 are approximately 424 equal to the discharge 16.719 power of EES, when the capacity is lower 550 than 300 kW. This500 means that the 460 ESS is fully used to17.156 feeding the load demands at 600 550 489 17.438


12 of 20

Sustainability 2017, 9, 395 hourly The aggregate

12 of 20 average load demand and available output power in the power system is presented in Figure 7a, and the charging/discharging status is presented in Figure 7b during 24 h timeworking interval.status. As observed, the wind turbine is faced with athe shutdown during 0:00– this When the capacity keeps on increasing, CVES values are 19:30–23:00 less than theand installed 6:00 periods and the ESS is also faced with a shutdown during 11:00–14:00 and 15:00–19:00 periods capacity. This is means that the adding capacity of ESS plays a minor role in improving the reliability as system. shown As in Figure Asthis seen in less Figure 7b, allwill of be thepaid available powerthem by ESS has been used in of a result,7.by way, expenses for feeding by consumers. algorithm during 11:00–12:00, 14:00–15:00 and 19:00–23:00 periods. During the 0:00–8:00, 11:00–14:00 The aggregate hourly average load demand and available output power in the power system and 15:00–19:00 of the system operation, ESS is always operated in charging mode. In the is presented in Figure 7a, and the charging/discharging status is presented in Figure 7bmethod, during ESS is operated about 67% period in charging mode and 33% in discharging mode during 24 and h of 24 h time interval. As observed, the wind turbine is faced with a shutdown during 19:30–23:00 system operation. During the 11:00–14:00 and 15:00–18:00 periods, the wind power is abundant to 0:00–6:00 periods and the ESS is also faced with a shutdown during 11:00–14:00 and 15:00–19:00 feeding the load demand and to charging the ESS, and the wind power curtailment occurs during periods as shown in Figure 7. As seen in Figure 7b, all of the available power by ESS has been used in the periods. algorithm during 11:00–12:00, 14:00–15:00 and 19:00–23:00 periods. During the 0:00–8:00, 11:00–14:00 The SOC and thesystem windoperation, power during 24-hoperated system operation shown in method, Figure 7c,d, and 15:00–19:00 of the ESS isthe always in chargingare mode. In the ESS respectively. SOC of energy storage is shown in Figure 7d for the algorithm. As a result, the value is operated about 67% period in charging mode and 33% in discharging mode during 24 h of system of SOC in this algorithm reaches about 60% in theperiods, daily operation. if it can use the restthe of operation. During the 11:00–14:00 and 15:00–18:00 the wind Therefore, power is abundant to feeding the generated power for charging the ESS, the total generation cost will be minimized. load demand and to charging the ESS, and the wind power curtailment occurs during the periods.

(a)

(b)

(c)

(d)

Figure 7. 7. (a) (a) The The power power flow flow of of test test system; system; (b) (b) output output power power of of energy energy storage; storage; (c) (c) output output power power of of Figure wind turbines; and (d) SOC of ESS. wind turbines; and (d) SOC of ESS.

4.2. Sensitive Analysis The SOC and the wind power during the 24-h system operation are shown in Figure 7c,d, The factors affect the reliability will also affect7d thefor CVES. These factors discussed in this respectively. SOCthat of energy storage is shown in Figure the algorithm. As aare result, the value of section. They include the load yearly peak load, maximum permissible energy stored for storage SOC in this algorithm reaches about 60% in the daily operation. Therefore, if it can use the rest of the power, thepower installed wind the wind speed series, the temperature and generated for capacity chargingof thethe ESS, the turbines, total generation cost will time be minimized. the failure rate and repair rate of the energy storage. The parameters of the test system are shown in 4.2. Sensitive Analysis analysis of these factors is investigated in the following case studies. Table 3. A sensitivity The factors that affect the reliability will also affect the CVES. These factors are discussed in this section. They include the load yearly peak load, maximum permissible energy stored for storage power, the installed capacity of the wind turbines, the wind speed time series, the temperature and


13 of 20

Sustainability 9, 395repair rate of the energy storage. The parameters of the test system are shown 13 of in 20 the failure 2017, rate and Table 3. A sensitivity analysis of these factors is investigated in the following case studies.

Table 3. The main parameters of the test system. Table 3. The main parameters of the test system.

Wind Turbines Parameters Rated power (PTurbines r) 18,000 kW Wind Parameters Cut-in wind speed (Cin) 3.75 m/s Rated power (Pr ) 18,000 kW rate) m/s Rated wind speed (C Cut-in wind speed (Cin ) 3.7512m/s m/s Cut-out wind (C speed Rated wind speed 1223 m/s rate ) (Cout) Cut-out wind speed 230.1203 m/s C1 (Cout ) C1 C2 0.1203 −0.08 C2 −0.08 0.0128 C3 C3 0.0128 Energy Storage System Parameters Energy Storage System Parameters Maximum charging power (PEScha_Max) 800 kW Maximum charging power (PEScha_Max ) 800 kW Maximum discharge power (PESdischa_Max) 250 kW Maximum discharge power (PESdischa_Max ) 250 kW Sampling time period (T) 1h Sampling time period (T) 1h 500 kWh Lower energy Lower limitlimit energy storedstored (Emin ) (Emin) 500 kWh Upper limitlimit energy storedstored (Emax ) (Emax) 10,000 kWh 10,000 kWh Upper energy

4.2.1. Impact of the Load 4.2.1. Impact of the Load In this study, the impact of the load yearly peak value on the CVES is analyzed. Figure 8 In this study, the impact of the load yearly peak value on the CVES is analyzed. Figure 8 demonstrates the trends of the CVES with different load yearly peak powers. The values of the demonstrates the trends of the CVES with different load yearly peak powers. The values of the storage storage from the capacity value when the yearly peak power increases from 1800 kW to 3200 kW from the capacity value when the yearly peak power increases from 1800 kW to 3200 kW with a 200 kW with a 200 kW step are shown in Figure 8. When the yearly peak load is low, the wind power is step are shown in Figure 8. When the yearly peak load is low, the wind power is adequate to satisfy the adequate to satisfy the load demand to maintain a certain reliability level. Thus, the energy storage load demand to maintain a certain reliability level. Thus, the energy storage system contributes little system contributes little to the generation adequacy, and the value of storage from the capacity value to the generation adequacy, and the value of storage from the capacity value is low. Therefore, from is low. Therefore, from Figure 8, when the yearly peak load is lower than 1900 kW, the value of Figure 8, when the yearly peak load is lower than 1900 kW, the value of storage from the capacity value storage from the capacity value is 2 kW. Then, the storage values from the capacity value is 2 kW. Then, the storage values from the capacity value significantly increase when the load yearly significantly increase when the load yearly peak power increases to 2400 kW. This is because the peak power increases to 2400 kW. This is because the wind power cannot satisfy the load demand that wind power cannot satisfy the load demand that keeps increasing, and the energy storage devices keeps increasing, and the energy storage devices afford the rest of the load demand. Therefore, the afford the rest of the load demand. Therefore, the storage values from the capacity value have a high storage values from the capacity value have a high growth rate. If the load yearly peak power keeps growth rate. If the load yearly peak power keeps on increasing to 3200 kW, the energy storage on increasing to 3200 kW, the energy storage system cannot supply the load demand any longer, and system cannot supply the load demand any longer, and the growth rate of the CVES will slow down. the growth rate of the CVES will slow down. Then, the energy storage from the capacity values will Then, the energy storage from the capacity values will finally reach a steady level. finally reach a steady level.

Figure 8. Variations Variations of the CVES against the load yearly peak power. Figure power.

With different maximum discharge powers, PESdischa_Max, the CVES values have the same trends. When the load yearly peak power is low, the CVES values remain constant with varying PESdischa_Max.


14 of 20


14 of 20

With different maximum discharge powers, PESdischa_Max , the CVES values have the same trends. When the load peak power is low, to theincrease, CVES values remain constant varying PESdischa_Max In addition, if yearly the peak load continues the CVES values withwith different PESdischa_Max will. In addition, the peak load continues to increase, the CVES values with different PESdischa_Max will reach steadyifvalues. reach steady values. 4.2.2. Impact of the Wind Power 4.2.2. Impact of the Wind Power (1) Impact of the installed wind power capacity of the wind turbines (1) Impact of the of installed wind power capacity of changes the windinturbines The trends CVES are represented by the the renewable energy penetration rate with The different discharge output powers of the energy storage system, as energy shown penetration in Figure 9. rate with trends of CVES are represented by the changes in the renewable The discharge curves with different discharge output storage powerssystem, of the energy storage system different output powers of the energy as shown in Figure 9. have the same shape. When the rated output power of the wind turbines is low, the discharge power of the The curves with different discharge output powers of the energy storage system have theenergy same storageWhen system has littleoutput influence on of CVES values. Therefore, thethe contribution the added energy shape. the rated power the wind turbines is low, discharge of power of the energy storage system with different the generation is the same. an storage has littledischarge influence powers on CVEStovalues. Therefore,adequacy the contribution of theThen, addedwith energy increase in the rated output power of the wind turbines, the CVES values will increase to a stable storage with different discharge powers to the generation adequacy is the same. Then, with an increase value, and the energy storage with a higher discharge power has awill larger CVES. in the rated output power of the wind turbines, the CVES values increase to a stable value, and Figure 9 shows that when the rated output power of the installed wind turbines increases from the energy storage with a higher discharge power has a larger CVES. 1800 Figure kW to 92700 kW, the CVES values will rapidly increase. The energy stored in the energy storage shows that when the rated output power of the installed wind turbines increases from system has been fully utilized, and no more stored energy is in excess of the load demand, which 1800 kW to 2700 kW, the CVES values will rapidly increase. The energy stored in the energy storage will be has satisfied by the increased wind rated power of will the system been fully utilized, and no morepower. stored Therefore, energy is inwhen excessthe of the loadoutput demand, which installed wind turbines keeps increasing 3600 kW,when the CVES reaches a maximum value. Then, be satisfied by the increased wind power. to Therefore, the rated output power of the installed with the sustainable growth of the rated output power of the wind turbines, the CVES values rapidly wind turbines keeps increasing to 3600 kW, the CVES reaches a maximum value. Then, with the decreases. growth of the rated output power of the wind turbines, the CVES values rapidly decreases. sustainable

Variationsofof CVES values against the different rated output of the wind Figure 9.9.Variations thethe CVES values against the different rated output powers powers of the wind turbines. turbines.

(2) ofof the wind speed time series (2) Impact Impact the wind speed time series Figure 10 demonstrates demonstrates the theCVES CVESatatthethe same location in Canada for different Figure 10 same location in Canada for different years.years. The The comparison results are shown in Figure 10. In this study, the wind speed time series at Prince comparison results are shown in Figure 10. In this study, the wind speed time series at Prince Alert, Alert, Canada in and 2006 2012 and 2012 are applied. Therefore, influence windspeed speedtime time series series is Canada in 2006 are applied. Therefore, the the influence of of thethewind is studied. From Figure 10, the changing trend of the capacity value of energy storage against the E of max studied. From Figure 10, the changing trend of the capacity value of energy storage against the Emax energy storage is the With the increasing of the Eof of energy systemthe reliability maxthe energy the storage, system of energy storage is same. the same. With the increasing Emax ofstorage, increases. When the When Emax ofthe energy storage is storage determined, the capacity of energy at energy is determined, thevalue capacity value storage of energy reliability increases. Emax of different years is similar. Because the same location had similar climate conditions for different years, storage at different years is similar. Because the same location had similar climate conditions for the CVESyears, valuesthe have the values same trends withsame respect to different Emax to values for those two years. for those different CVES have the trends with respect different Emax values

two years. However, as shown in Figure 10, the capacity value of energy storage in 2012 is higher than the capacity value of energy storage in 2012. Because the fluctuation of wind turbine output in 2012 is higher than the fluctuation in 2006.

Sustainability 2017, 9, 395 Sustainability 2017, 9, 395

15 of 20 15 of 20


15 of 20

Figure 10. 10. Variations Variationsofofthe theCVES CVES values against different values different Figure values against different EmaxEmax values withwith different windwind speedspeed time time series. series.

instorage Figure 10, the capacity value of energy storage in 2012 is higher than the 4.2.3.However, Impact ofas theshown energy capacity value of energy storage in 2012. Because the fluctuation of wind turbine output in 2012 is Figure 10. of theenergy CVES values max values with different wind speed time (1) Impact of Variations the maximum storedagainst in thedifferent energy Estorage system higher than the fluctuation in 2006. series. In this test, the maximum energy stored in energy storage system, Emax, is set from 0 kW to 4000 4.2.3. Impact11ofshows the energy storage kW. the trends of the CVES values with different Emax values. In Figure 11, with the 4.2.3.Figure Impact of the energy storage increase of Emax, there are three stages for the trends of the CVES. (1) ofof thethe maximum energy stored in in thethe energy storage system (1) Impact Impact maximum energy stored energy storage system In test, the themaximum maximumenergy energystored storedininenergy energy storage system, , is from set from kW to is set 0 kW0 to 4000 In this test, storage system, EmaxE,max 4000 kW. Figure 11 shows the trends the CVES with different Emax In values. Figure 11, FigureIn11, with the kW. Figure 11 shows the trends of the of CVES valuesvalues with different Emax values. with the increase of Emax three for the of the CVES. are, there three are stages forstages the trends oftrends the CVES. increase of Emax, there

Figure 11. Variations of CVES against different Emax values of energy storage.

In the first stage, because of the intermittent characteristics of wind power, the wind turbines cannot commendably satisfy the load demand without energy storage. Therefore, after the energy storage devicesFigure are added, the system reliability is effectively improved, the CVES rapidly 11. of against EEmax of storage. Figure 11. Variations Variations ofCVES CVES againstdifferent different max values values of energy energyand storage. increases during this stage. Then, as the load demand supplied by the energy storage system is restricted thestage, output of the of wind and conventional generators, the growth rate of CVES In the theby first stage, because of the thepower intermittent characteristics wind power, power, the wind wind turbines In first because intermittent characteristics of wind the turbines decreases. During this satisfy stage, the of the energy storage system to improve theafter power cannot commendably commendably the ability load demand demand without energy storage. Therefore, the supply energy cannot load without energy storage. energy max , the CVES increases slowly in the second adequacy is limited. Therefore, with the increase in E storagedevices devicesare are added, system reliability is effectively improved, CVESincreases rapidly storage added, thethe system reliability is effectively improved, and the and CVESthe rapidly stage until it reaches stable value. third stage, restricted thesystem operation mode of increases this astage. as In thethe load demand supplied bybythe energy system is during thisduring stage. Then, as theThen, load demand supplied by the energy storage isstorage restricted by the the energy storage devices, maintains constant value. generators, the growth rate of CVES restricted by the outputthe of CVES the wind power aand conventional decreases. During this stage, the ability of the energy storage system to improve the power supply adequacy is limited. Therefore, with the increase in Emax, the CVES increases slowly in the second stage until it reaches a stable value. In the third stage, restricted by the operation mode of the energy storage devices, the CVES maintains a constant value.


16 of 20

output of the wind power and conventional generators, the growth rate of CVES decreases. During this stage, the ability of the energy storage system to improve the power supply adequacy is limited. Therefore, with the increase in Emax , the CVES increases slowly in the second stage until it reaches a stable value. In the third stage, restricted by the operation mode of the energy storage devices, the Sustainability 2017, 9, 395 16 of 20 CVES maintains a constant value. With different maximum discharge powers, P , the CVES values have the same trends, With different maximum discharge powers, PESdischa_Max ESdischa_Max, the CVES values have the same trends, and finally, they reach their stable values. In addition, the energy storage system with a higher and finally, they reach their stable values. In addition, the energy storage system with a higher discharge output power will have a higher CVES with the same E . max. discharge output power will have a higher CVES with the same Emax (2) Impact of the temperature (2) Impact of the temperature The systems under 25 ◦ C and 0 ◦ C are discussed in this section. Figure 12 shows the CVES values The systems under 25 °C and 0 °C are discussed in this section. Figure 12 shows the CVES for the same location in Canada at different temperatures. The values of energy storage from the values for the same location in Canada at different temperatures. The values of energy storage from capacity values have the same shape with respect to different temperatures with different load yearly the capacity values have the same shape with respect to different temperatures with different load peak powers. When the load yearly peak power is in the range of 2000 kW to 2200 kW, the CVES yearly peak powers. When the load yearly peak power is in the range of 2000 kW to 2200 kW, the values at different temperatures exhibit no significant changes. This is because energy storage is not CVES values at different temperatures exhibit no significant changes. This is because energy storage used to the fullest when the load power is low. Therefore, the change in the capacity of a battery has is not used to the fullest when the load power is low. Therefore, the change in the capacity of a little influence on the values of energy storage from the capacity values. In addition, when the load battery has little influence on the values of energy storage from the capacity values. In addition, yearly peak power increases, the energy storage is put into full application. When the capacity of the when the load yearly peak ◦power increases, the energy storage is put into full application. When the energy storage system at 0 C from the capacity value is reduced to 75% of the rated value, the energy capacity of the energy storage system at 0 °C from the capacity value is reduced to 75% of the rated storage capacity value below 0 ◦ C is lower than that at 25 ◦ C. value, the energy storage capacity value below 0 °C is lower than that at 25 °C. The steady value region yet shows some CVES oscillation with respect load yearly peak power, The steady value region yet shows some CVES oscillation with respect load yearly peak power, that because the capacity value of energy storage system is an integer programming. The relationship that because the capacity value of energy storage system is an integer programming. The between energy storage capacity and peak load is not simple monotone increasing. relationship between energy storage capacity and peak load is not simple monotone increasing.

Figure 12. of the valuesvalues against against load yearly peak powers withpowers differentwith temperatures. Figure 12.Variations Variations of CVES the CVES load yearly peak different temperatures.

(3) Impact of the failure rate of the energy storage system (3) Impact of the failure rate of the energy storage system A comparison is provided for the CVES values between considering the failure rate and repair A comparison is provided for the CVES values between considering the failure rate and repair rate and not considering them, as shown in Figure 13. The failure rate of the grid-connected energy rate and not considering them, as shown in Figure 13. The failure rate of the grid-connected energy storage system is λ = 0.307, and the repair rate is µbat = 20. With an increase in the load yearly peak bat = 0.307, and the repair rate is μbat = 20. With an increase in the load yearly peak storage system is λbat power, the values of energy storage from the capacity value have the same shape. When the load power, the values of energy storage from the capacity value have the same shape. When the load yearly peak power is in the range of 1800 kW to 2000 kW, the CVES values considering the failure yearly peak power is in the range of 1800 kW to 2000 kW, the CVES values considering the failure rate and repair rate of the energy storage system are the same as those without such considerations. rate and repair rate of the energy storage system are the same as those without such considerations. Because the peak load is low, the energy storage system is not used to the fullest. In addition, the Because the peak load is low, the energy storage system is not used to the fullest. In addition, the grid-connected energy storage system has a lower failure rate and repair rate. Therefore, the fault occurrences have minimal effects on the CVES values. When the load yearly peak power continues to increase, the energy storage utilization increases, and occurrence of the faults affect the CVES values. Considering the failure rate and repair rate, the energy storage system has a lower value of energy storage from the capacity value. This is because the potential reliability risk increases with


17 of 20

grid-connected energy storage system has a lower failure rate and repair rate. Therefore, the fault occurrences have minimal effects on the CVES values. When the load yearly peak power continues to increase, the energy storage utilization increases, and occurrence of the faults affect the CVES values. Considering the failure rate and repair rate, the energy storage system has a lower value of energy storage from the capacity value. This is because the potential reliability risk increases with the increasing of failure rate of energy storage system. As the peak load shifting of energy storage, Sustainability 2017, 9, 395 17 of 20 system reliability will be effectively improved by the energy storage capacity value increasing, and the increasing of theofinstalled capacity value of energy storage storage will compensate for the potential the increasing the installed capacity value of energy will compensate for the reliability potential risk of distribution network. reliability risk of distribution network.

Figure 13. 13. Effect Effect of of different different failure failure rates Figure rates in in the the energy energy storage storage system system on on the the CVES CVES values. values.

5. Conclusions In this paper, the definition of the CVES is presented to quantitatively weigh the contribution of the energy storage to the generation adequacy. For this purpose, a method in accordance with EFC approach has been introduced to calculate the energy energy storage storage values values from from the the capacity capacity value. value. The developed model can also be used as a useful tool for the CVES value estimation to save the cost allocated Moreover, it can be used by planners to allocate the energy allocated to toenergy energystorage. storage. Moreover, it can be used by planners to allocate the storage energy capacity. storage The approach validatedisexperimentally on a test system. results have demonstrated capacity. The is approach validated experimentally on a The test obtained system. The obtained results have the effectiveness the proposedofalgorithm in the algorithm CVES value results demonstrated theofeffectiveness the proposed in estimating. the CVES The valuesimulation estimating. The have shownresults that CVES presents available capacity energy storage, and estimate the simulation haveeffectively shown that CVEStheeffectively presentsofthe available capacity of energy capacity contribution precisely. Furthermore, theprecisely. impact factors on the CVES are investigated storage, and estimate of theESS capacity contribution of ESS Furthermore, the impact factors on in caseare study, such as the yearly peaksuch power, the load maximum energy stored in thethe CVES investigated in load the case study, as the yearlypermissible peak power, the maximum the system, the renewable temperature, thepenetration, failure rate and repair rate of the permissible energy storedenergy in the penetration, system, the the renewable energy the temperature, grid-connected and the wind energy speed time series. The impact factors the failure rate andenergy repairstorage rate ofsystem the grid-connected storage system and the windinfluence speed time supply-demand relationship of the actual storage output. The sensitivity analysis of thestorage impact series. The impact factors influence the energy supply-demand relationship of the actual energy factors thesensitivity CVES is conducted. Therefore, the contributions and aims is of conducted. the presented paper canthe be output.for The analysis of the impact factors for the CVES Therefore, summarized as follows. contributions and aims of the presented paper can be summarized as follows.

theCVES CVESto to quantitatively weigh the contribution EESgeneration to the generation (1) Presenting (1) Presenting the quantitatively weigh the contribution of EES of to the adequacy. adequacy. (2) The CVES provides a feasible solution to quantitatively weigh the contributions of energy storage. CVES provides a feasible solution quantitatively weigh the of contributions of and energy (2) The In addition, it is related to the system loadtodemand, the output power wind turbines the storage. In addition, it is related to the system load demand, the output power of wind turbines characteristics of energy storage devices. and the characteristics of energy storage devices. (3) The peak load and the rated power of the wind turbines are comprehensively considered for (3) The peak load and the rated power of the wind turbines are comprehensively considered for the the energy storage capacity optimization allocation. If the supply power is adequate, energy energy storage capacity optimization allocation. If the supply power is adequate, energy storage system is not used to the fullest. Therefore, the CVES value is low. In addition, when the storage system is not used to the fullest. Therefore, the CVES value is low. In addition, when the energy storage is sufficiently utilized, the CVES value reaches a maximum. There is no point in continuing to increase the energy storage capacity. Therefore, using the CVES, we can obtain a suitable energy storage capacity allocation scheme. (4) When the energy storage devices are integrated to the grid, the system reliability is effectively improved, and the CVES values rapidly increases during this stage. Then, with the increase in


(4)

(5)

18 of 20

energy storage is sufficiently utilized, the CVES value reaches a maximum. There is no point in continuing to increase the energy storage capacity. Therefore, using the CVES, we can obtain a suitable energy storage capacity allocation scheme. When the energy storage devices are integrated to the grid, the system reliability is effectively improved, and the CVES values rapidly increases during this stage. Then, with the increase in Emax , the CVES increases until reaching a stable value. In addition, the Emax increases as well, but the contribution of the energy storage to the grid reliability did not increase. Therefore, the CVES values remain constant. The failure rate and repair rate of the energy storage and temperature affects the CVES. With the same parameters, if the energy storage system is not used fully, the influence of the temperature and failure rate of energy storage on the CVES is not apparent. However, if the energy storage system is more fully used, the CVES decreases if the ambient temperature and the failure rate and repair rate are considered.

Acknowledgments: This work was supported by the Key Project of Chinese National Programs for Research and Development (No. 2016YFB0900100). Author Contributions: All authors contributed to this work. Nian Shi performed the numerical modeling and theoretical modeling and executed the numerical work. Yi Luo performed the theoretical and experimental modeling. Conflicts of Interest: The authors declare no conflicts of interest.

Nomenclature The following abbreviations are used in this manuscript: Capacity Value Estimate Parameters and Variables: Pwind_total Total output power of the wind turbines (kW) PES Output power of the energy storage system (kW) HLOLE Hourly loss of load expectation Gi Available capacity at hour i (kW) Li Actual load level at hour i (kW) PL Load power (kW) n Total number of hours in a year Wind Power Model Parameters: v Wind speed (m/s) Pr Rated output power of wind turbines (kW) Cin Parameter related to the cut-in wind speed of wind turbines Crate Parameter related to the rated wind speed of wind turbines Cout Parameter related to the cut-out wind speed of wind turbines C1 , C2 , C3 Parameters related to the wind turbines PWind Output power of a wind turbine (kW) Energy Storage Model Parameters: SOC State of charge PES_output Power of the energy storage system (kW) Pstatus_cha Charging power of the energy storage system (kW) Pstatus_discha Discharging power of the energy storage system (kW) Ebat Energy stored in the energy storage system (kWh) PAvailable (h) Available power of the energy storage system at hour h (kW) PEScha_Max Maximum charging power of the energy storage system (kW) PESdischa_Max Maximum discharge power of the energy storage system (kW) Eoutput (h) Output energy of the energy storage system at hour h (kWh) T Sampling time period (h) Emin Lower limit energy stored in the energy storage system (kWh) Emax Upper limit energy stored in the energy storage system (kWh)


19 of 20

References 1.

2.

3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

23.

Marzband, M.; Ardeshiri, R.R.; Moafi, M.; Uppal, H. Distributed generation for economic benefit maximization through coalition formation-based game theory concept. Int. Trans. Electr. Energy Syst. 2017, 1–16. [CrossRef] Marzband, M.; Parhizi, N.; Savaghebi, M.; Guerrero, J.M. Distributed smart decision-making for a multimicrogrid system based on a hierarchical interactive architecture. IEEE Trans. Energy Convers. 2016, 31, 637–648. [CrossRef] Hu, P.; Karki, R.; Billinton, R. Reliability evaluation of generating systems containing wind power and energy storage. IET Gener. Transm. Distrib. 2009, 3, 783–791. [CrossRef] Wang, J.C.; Wang, X.R. A control strategy for smoothing active power fluctuation of wind farm with flywheel energy storage system based on improved wind power prediction algorithm. Energy Power Eng. 2013, 5, 387–392. [CrossRef] Sebastián, R. Smooth transition fromwind only to wind dieselmode in an autonomous wind diesel system with a battery-based energy storage system. Renew. Energy 2008, 33, 454–466. [CrossRef] Bahmani-Firouzi, B.; Azizipanah-Abarghooee, R. Optimal sizing of battery energy storage for micro-grid operation management using a new improved batalgorithm. Renew. Energy 2014, 56, 42–54. Francesco, M.; Guangya, Y.; Chresten, T.; Jacob, O.; Esben, L. A decentralized storage strategy for residential feeders with photovoltaics. IEEE Trans. Smart Grid 2014, 5, 974–981. Danny, P.; Marko, A.; Predrag, D.; Goran, S. Whole-systems assessment of the value of energy storage in low-carbon electricity systems. IEEE Trans. Smart Grid 2014, 5, 1098–1109. Sioshansi, R.; Madaeni, S.H.; Denholm, P. A dynamic programming approach to estimate the capacity value of energy storage. IEEE Trans. Power Syst. 2014, 29, 395–403. [CrossRef] Castro, R.; Ferreira, L. A comparison between chronological and probabilistic methods to estimate wind power capacity credit. IEEE Trans. Power Syst. 2001, 16, 904–909. [CrossRef] Amelin, M. Comparison of capacity credit calculation methods for conventional power plants and wind power. IEEE Trans. Power Syst. 2009, 24, 685–691. [CrossRef] Haslett, J.; Diesendorf, M. The capacity credit of wind power: A theoretical analysis. Sol. Energy 1981, 26, 391–401. [CrossRef] Zachary, S.; Dent, C.J.; Zachary, S. Probability theory of capacity value of additional generation. Proc. Inst. Mech. Eng. Part O J. Risk Reliab. 2012, 226, 33–43. [CrossRef] Usaola, J. Capacity credit of concentrating solar power. IET Renew. Power Gener. 2013, 7, 680–688. [CrossRef] Zhu, S.; Zhang, Y.; Chowdhury, A.A. Capacity credit of wind generation based on minimum resource adequacy procurement. IEEE Trans. Ind. Appl. 2012, 48, 730–735. [CrossRef] Garver, L.L. Effective load carrying capability of generating units. IEEE Trans. Power Appar. Syst. 1966, PAS-85, 910–919. [CrossRef] Gao, Z.X.; Mao, A.J.; Chen, D.Z.; Song, Y.T. A wind farm capacity credibility calculation method based on parabola. Appl. Mech. Mater. 2014, 472, 953–957. [CrossRef] Handschy, M.A.; Rose, S.; Apt, J. Is it always windy somewhere? Occurrence of low-wind-power events over large areas. Renew. Energy 2017, 101, 1124–1130. [CrossRef] Da Silva, A.M.L.L.; Sales, W.S.; Da Fonseca Manso, L.A.; Billinton, R. Long-term probabilistic evaluation of operating reserve requirements with renewable sources. IEEE Trans. Power Syst. 2010, 25, 106–116. [CrossRef] Perez, R.; Taylor, M.; Hoff, T.; Ross, J.P. Reaching consensus in the definition of photovoltaics capacity credit in the USA: A practical application of satellite-derived solar resource data. J. Abbr. 2008, 1, 28–33. [CrossRef] Zhang, N.; Kang, C.; Kirschen, D.S.; Xia, Q. Rigorous model for evaluating wind power capacity credit. IET Renew. Power Gener. 2013, 7, 504–513. [CrossRef] Marzband, M.; Javadi, M.; Domínguez-García, J.L.; Moghaddam, M.M. Non-cooperative game theory based energy management systems for energy district in the retail market considering DER uncertainties. IET Gener. Transm. Distrib. 2016, 10, 2999–3009. [CrossRef] Marzband, M.; Azarinejadian, F.; Savaghebi, M.; Guerrero, J.M. An optimal energy management system for islanded microgrids based on multiperiod artificial bee colony combined with markov chain. IEEE Syst. J. 2015. [CrossRef]


24.

25.

26.

27. 28. 29. 30. 31. 32. 33. 34.

35. 36. 37. 38.

39. 40. 41. 42. 43.

20 of 20

Marzband, M.; Yousefnejad, E.; Sumper, A.; Domínguez-García, J.L. Real time experimental implementation of optimum energy management system in standalone Microgrid by using multi-layer ant colony optimization. Int. J. Electr. Power Energy Syst. 2016, 75, 265–274. [CrossRef] Marzband, M.; Majid, G.; Sumper, A.; Domínguez-García, J.L. Experimental validation of a real-time energy management system using multi-period gravitational search algorithm for microgrids in islanded mode. Appl. Energy 2014, 128, 164–174. [CrossRef] Marzband, M.; Sumper, A.; Ruiz-Álvarez, A.; Domínguez-García, J.L.; Tomoiagă, B. Experimental evaluation of a real time energy management system for stand-alone microgrids in day-ahead markets. Appl. Energy 2013, 106, 365–376. [CrossRef] Booth, R.R. Power system simulation model based on probability analysis. IEEE Trans. Power Appar. Syst. 1972, 91, 62–69. [CrossRef] Bagen; Billinton, R. Incorporating well-being considerations in generating systems using energy storage. IEEE Trans. Energy Convers. 2005, 20, 225–230. [CrossRef] Wang, M.Q.; Gooi, H.B. Spinning reserve estimation in microgrids. IEEE Trans. Power Syst. 2011, 26, 1164–1174. [CrossRef] Bhuiyan, F.A.; Yazdani, A. Reliability assessment of a wind-power system with integrated energy storage. IET Renew. Power Gener. 2010, 4, 211–220. [CrossRef] Bahramirad, S.; Reder, W.; Khodaei, A. Reliability-constrained optimal sizing of energy storage system in a microgrid. IEEE Trans. Smart Grid 2012, 3, 2056–2062. [CrossRef] Wangdee, W.; Billinton, R. Considering load-carrying capability and wind speed correlation of WECS in generation adequacy assessment. IEEE Trans. Energy Convers. 2006, 21, 734–741. [CrossRef] Caralis, G.; Zervos, A. Value of wind energy on the reliability of autonomous power systems. J. Abbr. 2010, 4, 186–197. [CrossRef] Marzband, M.; Sumper, A.; Domínguez-García, J.L.; Gumara-Ferret, R. Experimental validation of a real time energy management system for microgrids in islanded mode using a local day-ahead electricity market and MINLP. Energy Convers. Manag. 2013, 76, 314–322. [CrossRef] Marzband, M.; Ghazimirsaeid, S.S.; Uppal, H.; Fernando, T. A real-time evaluation of energy management systems for smart hybrid home Microgrids. Electr. Power Syst. Res. 2017, 143, 624–633. [CrossRef] Billinton, R.; Karki, B. Well-being analysis of wind integrated power systems. IEEE Trans. Power Syst. 2011, 26, 2101–2108. [CrossRef] Cheng, K.W.E.; Divakar, B.P.; Wu, H.; Ding, K. Battery-management system (BMS) and SOC development for electrical vehicles. IEEE Trans. Veh. Technol. 2011, 60, 76–88. [CrossRef] Vargas, L.S.; Bustos-Turu, G.; Larraín, F. Wind power curtailment and energy storage in transmission congestion management considering power plants ramp rates. IEEE Trans. Power Syst. 2015, 30, 2498–2506. [CrossRef] Wen, Y.; Guo, C.; Kirschen, D.S.; Dong, S. Enhanced security-constrained OPF with distributed battery energy storage. IEEE Trans. Power Syst. 2015, 30, 98–108. [CrossRef] Zhou, L.D. Discussion on sensitivity of value-regulated lead-acid battery on temperature. Commun. Power Technol. 1998, 4, 16–20. Subcommittee, P.M. IEEE reliability test system. IEEE Trans. Power Appar. Syst. 1979, PAS-98, 2047–2054. [CrossRef] Tripak, O.; Kongsiriwong, S. A reliability test system for educational purposes-basic data. IEEE Trans. Power Syst. 1989, 4, 1238–1244. Historical Climate Data. 2012. Available online: http://climate.weather.gc.ca (accessed on 28 February 2017). © 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/).

Energy Storage System Sizing Based on a Reliability ... - MDPI

Energy Storage System Sizing Based on a Reliability ... - MDPI

Suggest Documents

Reliability-Based Sizing of Backup Storage

A Stochastic Sizing Approach for Sharing-based Energy Storage ...

A Review on Flexible and Transparent Energy Storage System - MDPI

Sizing Hybrid Energy Storage Systems for Distributed Power ... - MDPI

Optimized Sizing and Scheduling of Hybrid Energy Storage ... - MDPI

Hybrid energy system sizing incorporating battery storage - IEEE Xplore

Optimal Siting and Sizing of Energy Storage System for Power ...

Sizing of Energy Storage System for Microgrids - IEEE Xplore

Optimal sizing of battery energy storage system in ... - Semantic Scholar

Distributed Renewable Generation and Storage System Sizing ... - MDPI

Renewable energy based microgrid system sizing and energy ... - raiith

A Flywheel Energy Storage System Based on a Doubly Fed ... - MDPI

Design of a Bidirectional Energy Storage System for a ... - MDPI

Design of a Bidirectional Energy Storage System for a ... - MDPI

A Review of Flywheel Energy Storage System Technologies - MDPI

Application of a LiFePO4 Battery Energy Storage System to ... - MDPI

Optimal Scheduling of a Battery Energy Storage System with ... - MDPI

Energy Management for an Onboard Storage System Based on Multi ...

Energy Storage System Based on Cascaded Multilevel Inverter with ...

Battery Energy Storage System Load Shifting Control Based on Real

Hybrid Energy Storage System based on Compressed Air ... - CiteSeerX

Development of a hybrid energy storage sizing ... - Springer Link

Sizing/Optimization of a Small Satellite Energy Storage and Attitude ...

Sizing of integrated renewable energy system based ...