Portal de Periódicos da CAPES

Vehicle‐for‐grid (VfG): a mobile energy storage in smart grid

2018; Institution of Engineering and Technology; Volume: 13; Issue: 8 Linguagem: Inglês

10.1049/iet-gtd.2018.5175

1751-8695

Mehdi Rahmani‐Andebili,

Microgrid Control and Optimization

IET Generation, Transmission & DistributionVolume 13, Issue 8 p. 1358-1368 Research ArticleFree Access Vehicle-for-grid (VfG): a mobile energy storage in smart grid Mehdi Rahmani-Andebili, Corresponding Author Mehdi Rahmani-Andebili mehdir@g.clemson.edu The Holcombe Department of Electrical and Computer Engineering, Clemson University, Clemson, SC, 29634 USASearch for more papers by this author Mehdi Rahmani-Andebili, Corresponding Author Mehdi Rahmani-Andebili mehdir@g.clemson.edu The Holcombe Department of Electrical and Computer Engineering, Clemson University, Clemson, SC, 29634 USASearch for more papers by this author First published: 03 April 2019 https://doi.org/10.1049/iet-gtd.2018.5175Citations: 8AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Vehicle-for-grid (VfG) is introduced as a mobile energy storage system (ESS) in this study and its applications are investigated. Herein, VfG is referred to a specific electric vehicle merely utilised by the system operator to provide vehicle-to-grid (V2G) and grid-to-vehicle (G2V) services. The advantages of VfGs over the ESSs and plug-in electric vehicles (PEVs) include mobility of the VfGs across the distribution system and their complete availability for the system operator, respectively. In this study, VfGs are utilised by the distribution company (DISCO) to minimise the daily operation cost of the distribution system by providing the V2G and G2V services in optimal buses of the feeders. In addition, VfGs are applied by the generation company (GENCO) to minimise the daily operation cost of the generation system by providing the V2G and G2V services at optimal time periods. It is demonstrated that optimal application of VfGs has a considerable potential for cost reduction for both DISCO and GENCO. In fact, the DISCO and GENCO are benefitted because of the minimisation of feeders’ power loss and deferring the expensive generation units, respectively. Additionally, it is proven that cooperation of GENCO and DISCOs in utilisation of the VfGs has more benefit for them. Nomenclature objective function of a GENCO ($) daily value of investment cost for purchasing VfGs ($) daily value of maintenance cost of VfGs ($) daily unit commitment cost ($) lifetime of VfGs (year) price of one VfG ($) v index of VfG V set of VfGs total number of VfGs yearly maintenance cost of one VfG ($) t index of time (hour) T set of hours of a day g index of generation unit G set of generation units total number of generation units fuel cost of generation units ($) emission cost of generation units ($) start-up cost of a de-committed unit ($) shut-down cost of a committed unit ($) fuel cost coefficients of a generation unit ($/MWh2, $/MWh, $) P power of a generation unit (MW) emission coefficients of a generation unit (Ton/MWh2, Ton/MWh, Ton) emission penalty factor ($/Ton) start-up cost of a generation unit ($) s indicator of status of a generation unit shut-down cost of a generation unit ($) D demand of system (MW) efficiency of G2V or V2G service (%) power or demand of a VfG (kW) minimum power constraint of a generation unit (MW) maximum power constraint of a generation unit (MW) spinning reserve of system (%) ramp-up rate constraint of a generation unit (MW/h) ramp-down rate constraint of a generation unit (MW/h) minimum down time of a generation unit (h) minimum up time of a generation unit (h) minimum allowable SOC of a VfG (%) SOC of a VfG (%) maximum allowable power of a VfG (kW) objective function of a DISCO ($) total energy loss cost of feeders ($) f index of feeder F set of all the feeders total number of feeders index of branch set of all the branches total number of branches power loss (MW) electricity price ($/h) resistance of a branch (p.u.) magnitude of current flowing through a branch (p.u.) base MVA defined for the system (MVA) magnitude of apparent power flowing through a branch (MVA) allowable loading constraint of a branch (MVA) acceptable tolerance for the voltage of a bus (%) magnitude of voltage of a bus (p.u.) k index of status of molten metal in a temperature in SA algorithm acceptance indicator of a new status in SA algorithm energy level of molten metal in SA algorithm random value in the range of [0,1) in SA algorithm probability of acceptance of a new status in SA algorithm temperature in SA algorithm z decreasing multiplier of temperature in SA algorithm predefined value for the number of iterations in each temperature in SA algorithm 1 Introduction 1.1 Energy storage systems (ESSs) in smart grid Accommodating all the generation units and ESSs is one of the main goals of a smart grid. ESSs are applied in all parts of a smart grid based on different purposes. Generally, the ESS technologies applied in a smart grid fall into three main categories which are classified based on the form of the energy which is converted in them [1]. These categories include electrochemical, electrical, and mechanical ESSs [2]. The electrochemical ESSs comprise primary cell/battery, secondary cell/battery, reserve cell, and fuel cell ESSs [2]. Moreover, the electrical ESSs include capacitor, supercapacitor, and superconducting ESSs. In addition, the mechanical ESSs include flywheel system, pumped hydro storage system, and compressed air ESS. There are two more ESS technologies, namely chemical and thermal ESSs that their application is not very common in smart grid infrastructure. The chemical ESSs include hydrogen, synthetic natural gas, biofuels, and thermo-chemical ESSs. Also, the thermal ESSs include sensible heat system, latent heat system, and absorption system. The detailed description about the ESSs has been presented in [3]. The main characteristics of an ESS include its rated power and rated capacity (discharge duration). Fig. 1 illustrates the discharge duration of different types of ESSs with respect to their rated powers [4]. As can be seen, each type of ESS has distinct features, thus each ESS is appropriate for the specific part of a smart grid considering their rated power and capacity. ESSs are installed in different parts of a smart grid including generation system, transmission system, substations, distribution system, and end-user-site [5]. Fig. 1Open in figure viewerPowerPoint ESSs in the sense of discharge duration and rated power [4] 1.2 Challenges and opportunities for application of ESSs in smart grid The main challenge for the application of ESSs in a smart grid is concerned with the high capital cost of ESSs, their maintenance costs, and the additional capital costs for their inverters and grid connections [6]. In [4], it is claimed that investment on deployment of ESSs to shave peak demand is more expensive compared to expansion of power system to supply a reliable power. In addition, ESSs have relatively low energy efficiency and short life span. Also, there are considerable power losses in ESSs because of energy conversion/reconversion processes [7]. Moreover, special regulations, standards, and cost-benefit tools about ESSs are needed to be provided to choose the most cost-effective ESSs in different parts of smart grid [7, 8]. The other challenges for prosperity of ESSs in smart grid are related to their large size (ESSs occupy large area) and the hazardous material used in their production [3]. For example, although lithium batteries are non-hazards in most contexts, they have properties that can develop hazardous conditions like arc flash, blast, fire, and vented gas combustibility and toxicity [9]. Industry acceptance is another challenge for prosperity of ESSs in a smart grid. ESSs need to be known as a well-accepted contributor to the realisation of smart grid benefits, since the industry is not well informed yet. The main opportunity for the application of ESSs in smart grid is their integration with renewable energy sources (RESs) as the cost-effective and carbon-free sources of energy. RESs have power intermittency issue which is not favourable for power generation and distribution systems from economic and reliability point of views. In this regard, ESSs play a critical role in smart grid to mitigate the power variability and uncertainty issues of RESs to make them controllable and dispatchable. Large integration of ESSs with RESs can mitigate the energy security and environmental issues concerned with burning fossil fuels [10]. The other opportunities for the application of ESSs in a smart grid are presented as follows [11]: Levelling demand profile by shifting peak demand of system to off-peak and valley periods by optimal charging/discharging patterns. This peak shaving results in postponing the investment on the extension of generation, transmission, and distribution systems. Decreasing power harmonics, and consequently improving power quality and voltage stability. Providing ancillary services including frequency regulation, spinning reserve, and non-spinning reserve. Fast black starting, due to fast response of ESSs even compared to the fastest power plants. Increasing participation potential of customers in different power markets. End-users will be able to have an electricity trade with the local distribution company, in addition to their involvement in demand response programs. 2 Vehicle-for-grid Vehicle-for-grid (VfG) is introduced in this paper as an idea in smart grid infrastructure to be applied as the mobile ESS. In fact, a VfG is a specific electric vehicle utilised by the system operator to provide vehicle-to-grid (V2G) and grid-to-vehicle (G2V) services. In this study, plural form of VfG, that is, vehicles-for-grid is indicated by VfGs, as the defined agreement. Mobility and full availability of VfGs are their advantages over the ESSs and plug-in electric vehicles (PEVs), respectively. Therefore, the VfGs can be utilised by the distribution company (DISCO) to minimise the daily operation cost of the distribution system by providing the V2G and G2V services in the optimal buses of the feeders. Also, they can be utilised by the generation company (GENCO) to minimise the daily operation cost of the generation system by providing those services at the optimal time intervals. 2.1 Current counterpart of VfG The economic, security, and environmental reasons impose the governments around the world to electrify their transportation sector that makes up about 30% of the world's energy consumption and 70% of the global oil demand [12]. In addition, almost 14% of greenhouse gas emissions in the world in 2010 were related to the transportation sector [13]. On the other hand, due to the advancement in the battery manufacturing technology, price of automotive Li-ion battery packs is descending [14]. Moreover, PEVs can be charged by the electricity generated by the RESs such as solar and wind energy, as the free and clean sources of energy [15]. Due to the above-mentioned reasons, the internal combustion-based vehicles are being replaced by the PEVs to mitigate the energy security and environmental issues. Recently, many authors have studied the contributions of optimal management of the PEV fleet, and the impacts of PEVs on the electrical distribution and generation systems [16–30]. The economic and technical features of PEV fleet have been discussed in [16]. In [17–20], parking lot allocation problem has been studied in real power systems such as power system of Germany [17], Portugal [18], Beijing [19], and Seattle [20]. The charging station allocation problems have been investigated in the literature based on different objective functions such as minimum energy and power losses [21–24], maximum reliability of distribution system [25, 26], maximum voltage stability [27, 28], minimum generation system risk [29], optimal power factor of an electrical distribution feeder (DF) [30], and maximum profit of a DISCO [31]. In [32], a charging management framework for the utilisation of photovoltaic system has been presented considering the information exchange between the home and grid energy management systems. Moghaddam et al. [33] have presented a charging strategy that offers the multiple charging options to the PEVs including AC level II charging, DC fast charging, and battery swapping facilities. In [34], the stochastic presence of an aggregator, owning some parking lots, has been studied in the energy market transactions. In this study, the drivers’ behaviour, the hourly capacity of parking lots, the degradation of the PEVs’ batteries, and some economic factors such as inflation and interest rates have been considered. In [35], a dynamic pricing and energy management problem has been solved for the PEVs’ charging service providers considering the RESs and ESSs. In [36], a framework to determine the appropriate level of PEV penetration level for the given electrical distribution system has been presented. In [37], an algorithm for the aggregation of flexible loads with the PEVs has been presented. Liu et al. [38] have presented a locational marginal pricing approach to mitigate the congestion of electrical distribution network with penetration of PEVs. However, PEVs are manufactured for transportation purpose. In other words, they cannot be merely considered as the moveable ESSs, since their availability in the optimal locations and optimal time periods are not guaranteed. 2.2 Potential contribution of VfG for a GENCO The ratio of average load to peak load is defined as the load factor. A high value of load factor in any power system is desirable, since it indicates that the generation system is being used near to its rated capacity over the operation period. In the real power systems, the common values for the load factor are less than unity. In fact, a generation system with a low load factor needs to utilise its fast-starting and expensive diesel generators at peak period [39]. Therefore, increasing the load factor of a generation system by shifting the demand from the peak period to the valley period (and consequently levelling the system demand profile) can decrease the system operation cost. In this paper, VfGs are applied to minimise the operation cost of a GENCO by levelling the system demand profile due to providing the G2V and V2G services at optimal periods of time. This feature of VfGs (their full availability) is their competence over the PEVs. In fact, a GENCO is benefited from the full availability of VfGs. 2.3 Potential contribution of VfG for a DISCO The power systems statistics indicates that significant percentage of energy is wasted in the electrical distribution networks due to their radial structure and high ratio of current to voltage in them [40]. Distributed generation is one of the ideas that can decrease the power loss of an electrical distribution system, since one part of electrical energy is locally generated and delivered to the end users instead of passing through the long electrical feeders. In this study, VfGs are utilised as the distributed loads and generations to minimise the operation cost of a DISCO by providing G2V and V2G services in the optimal buses of the feeder, respectively. In other words, the feeder's load is transferred from some buses to other buses. This characteristic of the VfGs (their movability across the distribution system) is their merit over the stationary ESSs. In fact, a DISCO is benefited from the movability of VfGs. 2.4 Contributions of study Reluctance of the drivers with respect to the request of a DISCO to charge/discharge their vehicles in the suggested buses, as well as, the unavailability of the PEVs for a GENCO (specifically when the GENCO needs to shift the demand from the peak period to other periods) negatively affect the prosperity of application of the PEVs in these kinds of programs. Thus, in this study, VfG, as a new idea in a smart grid, is introduced to a GENCO and a DISCO to minimise their operation costs. Herein, VfGs are the particular mobile ESSs which are utilised by the system operators for the certain tasks including load shifting and distributed generation. Therefore, unlike PEVs and ESSs, VfGs are always available for the system operator and they are not stationary, respectively. In fact, the operation costs of a GENCO and a DISCO are minimised by decreasing the commitment of the most expensive and pollutant generation units and decreasing the energy loss cost of the electrical DFs. In other words, the GENCO needs to utilise the VfGs for G2V and V2G services at optimal time periods of a day and the DISCO needs to apply the VfGs in the optimal locations of the DFs. The remainder of paper is organised as follows. In Section 3, the operation problems of a GENCO and a DISCO are formulated. In Section 4, the optimisation technique for each operation problem is presented and described. Section 5 is concerned with the numerical studies and results analysis, and finally Section 6 concludes the paper. 3 Problem formulation In this section, the operation problems of a GENCO and a DISCO are formulated. 3.1 Formulating operation problem of a GENCO 3.1.1 Objective function of a GENCO The objective function of a GENCO is minimising the total cost of the generation system over the operation period (one day) that includes the scaled value of investment cost to purchase VfGs, the scaled value of maintenance cost of VfGs, and the daily unit commitment cost, as can be seen in the following equation: (1) 3.1.2 Cost terms of a GENCO In the following, the cost terms of objective function of a GENCO is described. Investment cost: The total investment cost for purchasing the VfGs scaled into daily cost is presented in (2). As can be seen, it depends on the number of VfGs () and price of one VfG (). To scale the investment cost, the total investment cost is divided into the number of years related to the lifetime of VfGs () and 365 as the number of days of a year (2) Maintenance cost: The total daily maintenance cost of VfGs is a function of yearly maintenance cost of one VfG (), , and factor for scaling (3) Unit commitment cost: The commitment cost of generation units includes fuel cost of units, emission cost of units, and start-up cost and shut-down cost of units (4)The fuel cost function of a generation unit () is a quadratic polynomial of its power (P) [41]. Herein, , , and are the fuel cost coefficients of a generation unit, as can be seen in (5). Moreover, the greenhouse gas emissions function of a generation unit is a quadratic polynomial of its power (P) [41], as can be seen in (6). Herein, , , and are the emission coefficients of a generation unit and is the emission penalty factor. In addition, the start-up cost of a de-committed unit () and shut-down cost of a committed unit () at every hour of the operation period are presented in (7) and (8), respectively. In other words, starting a generation unit up or shutting a generation unit down imposes cost about and , respectively. Herein, s indicates the status of a generation unit, where ‘1’ and ‘0’ mean ‘on’ and ‘off’, respectively (5) (6) (7) (8) 3.1.3 Problem constraints of a GENCO In the following, the generation system and generation units’ constraints are presented and explained. System power-demand balance constraint: The power-demand balance constraint of the system that must be satisfied at each hour of the operation period (one day) is presented in (9). This constraint indicates the equality of power of the generation units (in ‘on’ status) to the sum of the loads demand and VfGs’ charging demand (or discharging power). Herein, indicates the demand (for positive value) or the power of one VfG (for negative value) corresponding to G2V and V2G services, respectively. In addition, indicates the efficiency of G2V and V2G services provided by a VfG (9) System minimum demand constraint: The generation units that are in ‘on’ status must be able to supply the minimum demand of system, as can be seen in (10). In other words, the commitment of generation units must let the system generate the minimum total demand of system (10) System maximum demand constraint considering spinning reserve: The generation units that are in ‘on’ status must be able to supply the maximum demand of system considering the required spinning reserve of system, as can be seen in the following equation: (11) Generation unit's power constraint: The maximum and minimum power constraints of each generation unit at every hour of the operation period are presented in the following equation: (12) Generation unit's ramp-up rate and ramp-down rate constraints: The ramp-up rate () and ramp-down rate () constraints of a generation unit at every hour of the operation period are presented in (13) and (14), respectively (13) (14) Generation unit's minimum ‘off time’ and minimum ‘on time’ constraints: The duration that a generation unit is continuously ‘off’ and ‘on’ must be more than the allowable minimum ‘down time’ () and minimum ‘up time’ (), respectively. In other words, as it is indicated in (15), if the status of a unit in the previous time step (hour) is ‘off’ () and sum of the statuses of unit over the past period with an interval of is non-zero, the unit must remain in ‘off’ status for the current time step (). In addition, if the status of a unit in the previous time step is ‘on’ () and sum of the statuses of unit over the past period with an interval of is not equal to , the unit must remain in ‘on’ status for the current time step (). In other conditions, the status of the generation unit can be changed. These constraints must be held for each generation unit and at each hour of the operation period (one day) (15) State of charge (SOC) constraint of a VfG: The SOC of every VfG must be more than the minimum allowable limit () at every hour of the operation period to prolong its lifetime. In addition, the SOC level of the VfG cannot be more than its capacity () (16) G2V-V2G balance constraint for a VfG: This constraint is held to use each VfG only for an ESS and not for a generation or load source. In other words, the cumulative values of charged energy (by providing G2V service, where ) must be equal to the cumulative values of discharged energy (by providing V2G service, where ) in the operation period (one day) (17) Input/output powers constraint of a VfG: This constraint is considered to model the charging/discharging rates of each VfG. In other words, the input or output power of a VfG must be within the allowable rages at each hour of the operation period (18) 3.2 Formulating operation problem of a DISCO 3.2.1 Objective function of a DISCO The operation problem of a DISCO deals with the minimisation of daily operation cost of the electrical DFs that comprises the scaled investment cost for purchasing VfGs, the scaled maintenance cost of VfGs, and the energy loss cost of feeders (19) 3.2.2 Cost terms of a DISCO Investment cost: The scaled investment cost for purchasing VfGs and providing G2V and V2G services in the optimal locations of the feeders is presented in the following equation: (20) Maintenance cost: The maintenance cost of VfGs is presented in (21) scaled into daily cost (21) Energy loss cost: The daily energy loss cost of the feeders is presented in (22). In addition, the value of active power loss of a branch at each hour of a day is presented in (23) (22) (23) 3.2.3 Problem constraints of a DISCO Loading limit of branches: The magnitude of apparent power (MVA) flowing through every branch of an electrical feeder must be less than the allowable loading constraint of branch, as can be seen in the following equation: (24) Voltage magnitude limits of buses: The magnitude of voltage of every distribution bus (p.u.) must be within the allowable minimum and maximum limits, as can be seen in the following equation: (25) 4 Optimisation technique In this study, simulated annealing (SA) algorithm is applied to solve the optimisation problems of a GENCO and a DISCO. Other algorithms could be used in these problems; however, the simplicity of SA algorithm along with its powerful search capability in the complex and non-linear environments are its advantages compared to other algorithms [42]. Herein, the value of objective function of a GENCO or a DISCO [total cost of problem over the operation period (one day)] is defined as the value of internal energy of the molten metal () and then the SA algorithm tries to minimise this energy. Herein, the defined matrices in the SA algorithm for the problem of a GENCO and a DISCO are shown in Figs. 2 and 3, respectively. As can be seen in Fig. 2, the dimension of matrix related to the optimisation problem of a GENCO is . Herein, 24 indicates the number of hours of a day and indicates the number of generation units. Fig. 2Open in figure viewerPowerPoint Defined matrix in SA algorithm for the optimisation problem of a GENCO Fig. 3Open in figure viewerPowerPoint Defined matrix in SA algorithm for the optimisation problem of a DISCO In addition, as can be seen in Fig. 3, the dimension of defined matrix for the optimisation problem of a DISCO is . Herein, is the number of buses of a DF. In other words, every bus of a feeder is considered as a candidate for G2V and V2G services. Moreover, 9 bits in a binary system are needed to determine the number of VfGs at a bus, since maximum allowable number of VfGs at a bus is 280 and . In the following, the steps for applying SA algorithm in each problem are presented and described. Step 1: Obtaining the primary data Obtaining the parameters of SA algorithm: These parameters include initial temperature of molten metal (), number of generating new random solutions at every temperature (), and value of coefficient for gradually decreasing temperature of molten metal (z). Obtaining the parameters of system: The value of system parameters and initial data are obtained. Initialising the SA matrix: The related SA matrix is initialised with random binary numbers. Step 2: Generating an acceptable solution Updating the SA matrix: One of the bits in the vicinity of previously modified and accepted matrix is changed. Checking the problem constraints: The problem constraints are checked and if all of them are satisfied, the value of internal energy of molten metal (objective function) is calculated and the algorithm goes on; otherwise, the process is repeated from step 2. Checking the SA algorithm criterion: Based on the SA algorithm criterion presented in (26), the SA matrix resulted in decreased internal energy of molten metal is always accepted; however, the solution with increased value of internal energy is accepted just by an adaptive probability presented in (27). The value of the adaptive probability is decreased as the molten metal is cooled (26) (27) Step 3: Checking the number of iterations for the current temperature If the number of iterations in the current temperature is not equal to the predefined value (), the process is repeated from step 2; otherwise, the temperature of molten metal is decreased based on the following equation: (28) Step 4: Concluding Checking the temperature of molten metal: The temperature of molten metal is measured and if the molten metal is frozen (its temperature is near to zero), the optimisation process is finished; otherwise, the process is repeated from step 2. Introducing the outcomes: The results include the optimal value of SA matrix and minimum value of internal energy of the molten metal (objective function). 5 Numerical study 5.1 Characteristics of the power system under study The geographic single-line configuration of the power system is illustrated in Fig. 4 that includes one GENCO and eight DISCOs. Fig. 4Open in figure viewerPowerPoint Geographic single-line configuration of the power system under study The GENCO includes ten generation units (U1–U10) connected to one generation bus supplying three transmission feeders (TFs 1–3). Each of DISCOs 1–5 and 7–9 has one DF, but DISCO 6 has two DFs. Each DF has different number of distribution buses. The value of resistance and reactance (p.u.) of each branch of every DF are illustrated in Fig. 5. The value of active power (MW) and reactive power (MVAr) demands of each distribution bus of every electrical DF at 9 p.m. (peak demand) are shown in Fig. 6. Furthermore, Fig. 7 illustrates the value of loading limit (MVA) of each branch of DFs. Also, the hourly average demand pattern (p.u.) of each distribution bus of power system is shown in Fig. 8. Furthermore, the technical data of generation units of the GENCO are pres