lmplementation of Efficient B2G and V2G in Practical Cases

MD Shahrukh Adnan Khan＊ | Kazi Mahtab Kadir | Md.lbrahim lbne Alam | Md.Khairul Alam | Jianhui Wong | Aseef lqbal

1. lntroduction

A next generation evolvement is knocking at the world door and the people today are eagerly welcoming it.The smart city living is not a dream anymore rather it is rapidly rushing to its success story.The next generation smart city planning is one of the most eagerly awaited technological advances that young generation wants currently.The world is running towards smart enhancements.Smart building, smart vehicle, smart technology, intelligent city, and green city all are part of the next generation smart city planning[1]-[4].Building to grid (B2G) and vehicle to grid(V2G) are the most advance level technologies of smart city planning.Both of them are newly innovative concepts, which are yet to be implemented practically in different cities.Even though, experimental implementations are taking place in different parts of the world, there is still an ample space to conduct elementary research and define both the B2G and V2G clearly and their influence in smart city living.The current generation technology includes G2B (grid to building) and G2V (grid to vehicle).For a bidirectional energy flow, B2G and V2G are being considered for the next generation technology.For a unity bidirectional energy flow, all the four elements B2G, V2G, G2V, and G2B are clustered together which is called “crowd energy”.In [5] the authors defined crowd energy to be the endeavor of profit/non-profit corporations or individuals combining assets through the help of information and communications technology (ICT) to achieve a societal,economic, and political change for the move from centralized non-renewable energy production to decentralized renewable energy production.This next generation technology has not yet been practically implemented commercially although new research is being conducted in this area[2]-[6].This paper tries to fill up the research gap in this respect and tries to give a clear concept on these two advanced parts of smart city planning, strategical work flow analysis of B2G-V2G, and case examples with current and future scope including limitations.

2. Building to Grid (B2G)

2.1. lntroduction

Smart grid (SG) is the next generation of electric grid and this smarter grid needs B2G as its principal part to operate properly and serve the future of power network.The existing unidirectional electricity grid is facing numerous challenges, which cannot be dealt with the current grid topology.The future grid needs to be bidirectional where the grid is a digital network with lots of sensors and will be a self-monitoring and self-healing entity[1].Therefore, a need for integrating B2G emerged and this integration is expected to keep the power system stable by allowing the buildings to contribute in the changes of electricity supply and/or demand.Therefore, B2G can be defined as the coherent part of the future grid system having reliable interactive environment, where benefits and motivations of the customers lead them to offering services to the power distribution system.

The consistency of the B2G integration is hoped to ensure a secure, dependable, and robust energy distribution system which is able to support high penetrations of a demand and green environment[2]-[4].The B2G concept is not only about providing energy to the grid; but more so about giving maximum achievable support to the grid from the buildings.There are mainly two segments of B2G, in one segment the research is oriented on the demand or load related optimization, whereas the other segment is concentrating on using buildings to supply energy to the grid[3].With ongoing research on smart grid and B2G, some new terms have materialized like intelligent building,prosumers, demand management system (DMS), and demand response (DR)[4]-[9].The collaboration and efficient communications of all these with building and grid are hoped to make the whole B2G concept successful.In Fig.1,B2G communications network has been depicted where it can be seen that, single or multiple buildings need to maintain the communications with the grid through a building energy management system (BEMS).The BEMS provides information mainly to the market operators who are responsible for the balancing of bidirectional energy transmission between the building and grid.The grid with its different demands (e.g.peak shaving, voltage and frequency regulation, etc.) asks the market operators to provide support and the market operators try to comply.Also in the diagram the renewable energy sources (RESs) and loads are shown; these energy sources and intelligent loads are optimized by different optimization methods.One such optimization method is shown in Fig.2 where the methodology is similar to the found in [6] to [10].

2.2. Case Study

The idea of B2G emerged at the start of this millennium and became highly popular in the research and power sector along with the smart grid.After that, a boon can be observed in the last decade in the field of B2G research.There are a lot of journals and papers in the field of B2G; in the following paragraphs, some of them are summarized chronologically.

Fig.1.Generalized B2G network.

Fig.2.Sample flow chart of optimization technique.

Buildings in the city consume a large amount of energy from the grid (around 33%); therefore, BEMSs are important for the power system.The authors in [11] discussed about the smart grid BEMS (SG-BEMS) in the Jeju Island that was developed by Korean Telecomm (KT), whose architecture is shown as Fig.3.BEMS can provide significant support to grid by reducing energy consumption and by load management to perform peak shaving and voltage and frequency regulation.As ancillary services, BEMS helps to reduce the emission of carbon and provide support to RESs to be used widely.BEMS is also very important for the customers who want to join the energy market[10]-[12].

Fig.3.SG-BEMS architecture[11].

To help the buildings connect with the grid properly building agent scheme was proposed in [12].The proposed building agent is supposed to work as a load management gateway by proper utilization of load models and communications.The main idea behind the building agent was that SG could not function properly if the control of loads is handed over to the grid even on a small scale.Therefore, the agent hides the complexity of the building from the grid and provides only the information needed by the SG to perform flawlessly.

Commercial and private buildings can accommodate photovoltaic (PV) cells and with optimal power management mechanism, these PVs can provide a range of support to the grid.Intensive PV penetration of the grid to offer peak shaving services at minimal cost was the main objective in [13].They proposed an optimal predictive power-scheduling algorithm and after extensive real time simulation, it was found that 13% electricity bill could be minimized.Also the algorithm needed a high level of forecasting accuracy to provide expected results.In[14], the objective was to increase the monetary value of the PV solar production by providing ancillary services(i.e.active power supply, demand side management (DSM)).

Integration of buildings to the grid cannot be done with the current status of the buildings; these buildings need to be automated through building automation system (BAS), which is an essential element of building management system (BMS)[6].With BAS, the users have the ability to control energy and media consumption that is important to optimize the energy management system (EMS).In [6], they delineated the importance and applications of BAS in smart grid and showed that DR can be attained easily with the help of BAS.Also in that paper they hoped to perform further research on the implementation of BAS in AutBudNet laboratories and other AGH-University of Science and Technology (AGH-UST) buildings and afterwards they came up with further findings in [15].The importance of BAS was also emphasized by [7], the authors indicated that typical BASs do not consider the user activities and behavior, which are responsible for the wastage of almost one third of the consumed energy.Therefore, intelligent buildings must have the technology to recognize user activities and behavior and with these taken into account, minimizing energy consumption should be tried.

A study on B2G was done by Lawrence Berkeley National Laboratory[4], where the feasibility of implementing B2G in India was checked.Motivating the local power markets towards the implementation of smart grid was the main objective of the study.The study also promoted DMS to be incrementally used in existing and yet to be built commercial buildings.DR is one of the key features of DMS, the authors in [8] extensively examined all models of DR up to that time and concluded that the demand and resource both are highly diversified; therefore a single model for DR is impractical.In another work with DR[9], the authors studied a DR model which was based upon bidirectional energy trading facility and hybrid energy system.They argued that conventional DR with its unidirectional topology could not support the users efficiently.The users can store energy for future use or can sell energy to the power system with the presented model.Another control approach, model predictive control (MPC)was proposed in [16]; the approach was to regulate frequency by exploiting the flexibility of heating ventilation and air conditioning (HVAC) on the demand side.The MPC method used load forecasts and ramping rates of different providers to perform the frequency regulation service.

Bidirectional optimization model to achieve higher load factors with reduced power consumption was proposed in [17].The case study on an office building in Michigan Technological University (MTU) showed that the presented model works properly and can reduce electricity cost by 25% with a better load factor.The same authors in their subsequent work proposed a novel B2G indexing technique based on the building energy cost and load factors of the nodes[18].The objective of the paper was to facilitate the users by minimizing the energy cost and the grid by ensuring the maximum load penetration with a greater load factor.The model parameters were collected from MTU and the performance was evaluated for 33-node test feeder supporting B2G based commercial buildings; a promising performance was observed.Nonetheless, the model performance was observed to greatly depend on the price of energy, load flexibility, customers’ choice, grid controlling ability of the equipment, and weather forecasting accuracy.

Thermal electric storage (TES) devices have the capability of shifting the demand in time and form energy arbitrage[19].Hence, TES merged with B2G can provide DSM to perform better.The study in [19] was done for residential buildings in Ireland and showed that if TES works on 01:00 to 04:00 in the night, then it not only shifts the demand but also creates energy arbitrage.Moreover, the B2G model with its rolling optimization approach allows preservation of some energy in the TES, which can be used in the next day.It was also observed that by using the TES, the total demand could be reduced substantially.Further studies on the all island power system(AIPS) of Ireland were conducted by the same authors in [20] to achieve further optimization.

With the increment of unpredictability in demand, the grid cannot rely on traditional DR methods, which are mostly manual and rule based.DR-advisor is a data driven approach proposed in [21], which is quite reliable(93%), cost effective, and fast to cope with the demand fluctuations.With real time scenario and tariff structure DRadvisor was simulated in University of Pennsylvania; it was observed that it could save around 38% of the summer energy bill.The main challenge was to evaluate data rapidly, take quick decisions about the refrainment of energy usage, and provide a monetary benefit.

In recent years, Brazil is harnessing a large amount of energy through its installed PV cells which are currently supporting the national grid by providing a maximum of 51% coverage of the overall demand[22].To increase the efficiency of the PVs and find the feasibility of investment, the authors in [22] performed a simulation for net plus energy buildings with energy plus software.After performing the simulation for four metropolitan areas, they suggested that rooftop systems are more efficient and viable.Also they found that the energy compensation system of Brazilian PV could be quite beneficial for the prosumers.Installing PV in buildings may seem quite lucrative after the example of Brazil, however in [23] the authors discussed some facts, which might limit the willingness to invest in grid connected PV cells mounted in buildings.They carried out a case study in Beijing and found that many people residing there were not interested in investing in PV, as the static payback period (SPBP)is low.The case study showed that even with political incentives and available investment area for building integrated grid connected PV (BIGCP) installations (AIABI), support from locality may not be favorable.They also pointed out that single family residing in Beijing should not be encouraged to carry out BIGCP installations.

In some recent studies of B2G, DR with ancillary services and predictive power flow control has been discussed in [10] and [24].They suggested that for a faster DR, there is no better way but to improve communications latency.Also the consumers need to be aware of their preferences, as a single change in the control inputs of HVAC can create drastic effects in performance.In the MPC framework presented in [10], the method can solve the duck-curve problem by reducing the maximum load ramp rate.The authors used a Monte-Carlo based simulation to simulate the control of power flow between grid, solar panels, and energy storage systems.The authors in [25] discussed techno economic methods to get a clear idea of investment feasibility.They argued that traditional engineering economic methods should be reformed to comprehend market based PV investment.Fig.4 shows the techno economic structure from prosumer’s perspective.

Fig.4.Techno economic configuration from prosumer’s perspective[25].

3. Vehicle to Grid (V2G)

3.1. lntroduction

V2G is one of the avenues through which smart grid is implemented.V2G involves the exchange of power and information between electric vehicles (EVs) and power grid.This allows the EVs to provide several DR services to the grid while allowing EV users to enjoy different monetary and other form of incentives[26].

Control and maintenance of EVs through aggregators/utility bodies can be connected in three configurations.These are vehicle-to-house (V2H), vehicle-to-vehicle (V2V), and vehicle-to-grid (V2G).V2H involves energy exchange between EV and local home micro-grid, V2V involves energy exchange between grid-enabled group of EVs, and V2G employs energy interchange between grid and EV clusters through the control of aggregators.This paper mainly focuses on the V2G concept.

Power flow between EV and grid can happen in two different methods, namely unidirectional V2G (or G2V) and bidirectional V2G (true V2G).

Unidirectional V2G is simply the flow of power from grid to EV for charging the EV battery.It does not require any special apparatus except a charging outlet and it does not add to EV battery degradation due to cycling.DR services can be included if a simple and cheap controller is added to handle the charging rate.EV owners need to be incentivized to make them take part in these services thus ensuring G2V during off-peak hours and restricted during peak hours.However, for other important DR services, true V2G is necessary.

Bidirectional V2G, as the name implies, relates to the bidirectional power flow between grid and EV.The EVs require bidirectional chargers which have a bidirectional ac/dc converter (allowing power factor correction) and a bidirectional dc/dc converter (which manages battery charge/discharge current)[27].

3.2. V2G Advantages

V2G advantages include active power and reactive power management, valley filling, harmonics filtering, peak shaving, reduction of utility expenses, enhancement of load parameters, reduced of carbon footprint, tracing of RESs[28], frequency management, power failure recovery, etc.[26].

Unidirectional V2G (G2V) enables auxiliary services for the grid by changing EV charging rate according to power generation companies.This is handled by entities called aggregators, which combine and manage the charging process for a large group of EVs.These auxiliary services can be power grid management, which allows grid frequency balancing between the production and load, and allocation of spinning reserve that provides additional quick response generation capacity to meet sudden losses in generation.Thus the EVs become distributed energy assets[28].Reactive power support can also be achieved through voltage and frequency management.It also enables power factor correction, which limits power line losses and overloading of power devices[26].Frequency management can be done by turning large generators on/off but these can be expensive so fast charging/discharging EVs can provide a cheaper alternative[28].Grid connected EVs (GEVs) can provide reactive power support because of capacitors in their chargers[26].

Traditional power generation produces many emissions.RESs can limit these emissions but their generation is sporadic and dependent on environmental factors.But V2G can help mitigate the sporadic nature of RES[26].Studies have also shown that V2G could reduce greenhouse gases more than if only plug-in EVs (PEVs) were implemented[27].

3.3. V2G Challenges

V2G challenges include the following.EVs taking part in V2G (bidirectional V2G) will have more charging/discharging cycles compared to EVs not taking part or only employing G2V.So these EVs will have faster than normal battery degradation.The degradation will be worse with deeper battery depth of discharge and frequency of battery cycles.V2G control mechanisms can be formulated to reduce the impact of these processes but a balance should be sought between the financial gain and longevity reduction[26].

V2G implementation requires high investment costs regarding the smart grid and bidirectional-charging infrastructure.Also repeated battery charging/discharging cycles will increase conversion losses.Increase in PEV demand will require extra generation capacity[27].

EV owners might get anxious whether the amount of charge left will be sufficient for the trip back to home, if they take part in V2G activities.This can be mitigated using properly planned and appropriately distributed charging stations[26].

Also the smart grid, which will be the backbone for V2G, will require sufficient monitoring to detect anomalies,capability to resist hacking of communications and power networks necessary for V2G, improvement in power quality, enhanced reliability and efficiency, etc.[28].

3.4. Case Study

In [29] the authors evaluated the capability of adjustable PEV control for bidirectional charging using intermittent wind power to enhance the grid energy distribution and stabilization of power generation and demand without compromising PEV user demands.For this purpose, the following three energy distribution methods have been proposed.The first one is called ‘valley searching dispatching method’, where charging or discharging cannot be suspended and where it looks for valley of wind generation in a 24 h period and subsequently raises the valley by bidirectional charging of PEVs.The second one is ‘interruptible dispatching method’ where charging/discharging rates are same as before but can be suspended if necessary.Here the cut level, i.e.power generation/utilization magnitude that needs to be selected to get time periods for balancing of load/generation, should be less than load profile.In each iteration cut level is increased till iteration limit is reached or all PEVs reach the minimum battery state of charge (BSOC).The third method is‘variable-rate dispatching method’, which is similar to the second method but the rate can be changed according to charging/discharging generation and utilization rather than the number of PEVs that can take part in the process.The numerical simulation was carried out based on MATLAB using deterministic plus stochastic model and assessed by checking coordination between power production & demand and user contentment.The results showed that the coordination of power production/utilization got better for all three methods but the latter two had better results regarding the decrease of wind energy wastage at night,minimizing daytime flow of energy from grid and overall increase of user contentment.Furthermore, the proposed methods can be implemented with nominal technical obstacles.Although the models were proposed for dispersed wind power and PEVs, it could be adapted for other generation/load scenarios.For future work, further research is necessary to evaluate whether the harmful effect on battery longevity due to that the charging/discharging outweighs the above mentioned benefits[27],[28].

The authors of [30] explained peak-shaving and valley-filling in the context of a V2G control scheme.Peak shaving and valley filling involve the distribution of loads between peak and off-peak hours using bidirectional charge/discharge of V2G enabled EVs.In order for the target power curve and V2G plan curve to closely follow each other, an objective function was proposed with the aforementioned peak-shaving and valley filling systems.A simulation was devised with parameters including the numbers of EVs, EV batteries, user settings, etc.and data from representative cities.The results of the simulation pointed to the fact that if the EV number is increased or if the mean target is decreased, the effectiveness of peak shaving and valley filling is increased.Quantitatively if the standard deviation and root mean square of the difference of the two curves are below 10 then the curves follow each other well.

In [31] the authors discussed several technologies related to gridable electric vehicles (GEVs) i.e.V2H, V2V,and V2G.V2H comprises of a singular GEV and residence where it makes the daily residence load curve uniform through active power exchange and utilizes a charger capacitor for reactive power support.V2V entails several GEVs and smart homes where an aggregator is used for synchronized control of V2V and power grid.V2H is a sub-set of V2V.V2G involves a greater number of GEVs compared to the previous cases and employs smart homes, parking lots, rapid charging stations, etc.for energy exchange.GEV aggregators handle the distribution of reactive power, active power, and other grid optimization activities.For modelling purposes, household devices were assumed to be daily load profiles, GEVs were modelled as mathematical equations and V2H, V2V, and V2G systems were modelled according to their aims and limitations which include peak load reduction, reactive power compensation, etc.The authors put forward a mathematical model to improve the quality of GEV power distribution systems using the target function and limitations.The simulation showed that V2G causes load shift by increasing load demand.It was seen that V2G could reduce peak load and valley in the load curve could be increased.Furthermore, night charging of GEVs coincides with the excess power production.Also GEVs can reduce line losses, voltage variation, etc.

The self-governing V2G control scheme suggested by the authors in [32] is a spread out spinning reserve system useful for sporadic RESs based on droop control.In the control scheme, balance control is used for handling BSOC; i.e.smart charging for vehicle user’s planned charging.The purpose of these schemes was to enable a lower carbon footprint energy system by aggregating sporadically available RESs.The simulation involved two interconnected grids, two V2G groups, and a simple lithium ion battery model.One of the V2G groups had two types, medium (Mitsubishi i-MiEV/EV1) and large (Nissan Leaf/EV2) sized batteries and the other group had only small (Toyota Prius/PHV).The results showed that frequency deviations caused by the RES variation were counteracted by the V2G in a reasonable amount of time.Smart charging of EV1 did not change thermal power generation and EV1 did not provide any spinning reserve for the grid, but EV2 and PHV supported good quality frequency control.The simulation quantitatively established that V2G control had a quicker response than governor-less control of thermal power generation.Also the capacity of the PHV battery was found to be adequate for the spinning reserve.Limitations of the study included further necessary research regarding the efficiency of V2G control, effect on battery longevity, secure connection to grid, etc.

The authors of [33] enquired into the likelihood of diminishing load inconsistency in residence micro-grid by controlling charging schemes of PHEVs.A mathematical model was formulated with two PHEVs where Case 1 had a typical initial condition and Case 2 had optimized parameters.In both cases, the two PHEVs had bidirectional power flow between micro-grid and them.In Case 1, initial BSOC was assumed whereas Case 2 had optimal battery charge.For Case 2, optimization resulted in the reduction of load power curve’s mean and standard deviations.Furthermore, the routing of power towards the micro-grid when required resulted in the further reduction of these parameters.It was found that energy losses arose due to that the chargers and subsidies could be provided because they can encourage proper usage and controlled charging of PHEVs,which can benefit the grid.For simplification, some parameters were assumed to be known in advance e.g.the load curve of house appliances.The authors hope to use dynamic programming and other details for better accuracy in the future.

In [34] the authors were focused on V2G control involving primary frequency control (PFC) of energy grids.A decentralized V2G control (DVC) scheme was put forth where a charging/discharging apparatus manages bidirectional power between grid and EV to limit frequency variation, sustaining of BSOC, and fulfillment of charging demand.For this purpose, BSOC holder (BSH) was formulated, which was based on starting state of charge(SOC), to maintain EV BSOC using adaptive droop to sustain remaining battery energy levels.The control scheme is adaptive enough to maintain different staring SOC levels with frequency control.For charging demand, another V2G scheme was used called charging with frequency regulation (CFR) which consists of frequency droop control(to enhance frequency quality) and scheduled charging power (to reach charging demand).CFR is based on EV plugged-in time and typical SOC.It was shown in the simulation that the proposed DVC can limit frequency deviation and meet the charging demand in a two region connected power grid.Furthermore, the DVC is shown to be better than autonomous distributed V2G control discussed in Section 1.

The authors of [35] discussed a framework for lithium ion battery degradation dependent on different EV load outlines using the simulation of varying conditions.The battery degradation model was based on calendar aging and cycle aging.Calendar aging is the loss of usable lithium ions due to unwanted chemical reactions between electrode and electrolyte, which is influenced by electrochemical voltage and temperature.Cycle aging involves battery charging/discharging cycle depth and the total number of said cycles where degradation increases for both cases.The simulation parameters had different driving cycles (i.e.type of battery usage based on synthetic loads and real life car driving patterns), charging strategies and grid services (i.e.charging/discharging patterns based on the typical duration of travel and grid demand), daily duty cycle (i.e.a combination of the above mentioned parameters), and battery environment temperature.The result of the simulation showed that battery longevity could be extended by time-managed charging and demand-controlled charging i.e.delaying charging until remaining BSOC is not sufficient for the next scheduled trip.The simulated grid services for peak shaving decreased longevity and increased costs due to higher cycling.

The authors in [36] presented a control scheme for V2G with EV aggregation to allow supplementary frequency regulation (SFR).The proposed model consists of an EV aggregator, EVs, and their charging locations.The EV charging locations evaluate frequency regulation capacity (FRC) and issue management assignments to the EVs.In the case of frequency control, EVs perform the management assignments as mobile storage equipment.The FRC evaluation is based on available V2G power and expected V2G power is implemented, taking into account both frequency management and charging demand.The simulation is done in simulink environment and Monte Carlo sampling is performed to calculate BSOC level.The results for a model, based on real life two region connected grids in China, indicate that variations of area control error and grid frequency can be efficiently inhibited.Also required BSOC levels can be obtained by stabilizing regulation-up and regulation down functions.

Table 1: Current world scenario of V2G and B2G with limitations and further scope

4. B2G and V2G: Current World Scenario

Fig.5.Generalized V2G network.

Fig.6.World map scenario of B2G and V2G (Project places are indicated as per our findings, more projects may be active currently).

B2G and V2G are the inseparable part of smart grid, which is the key to our better living in the future.Therefore, to ensure a better and greener place to live for our next generation, more research should be done and for that the current world scenario needs to be known.After the thorough literature review done by the authors,Table 1 is presented here with some of the pilot projects being operated at present throughout the world[37]-[40].Also the limitations and scope of research are given in a concise and clear way.The authors hope that researchers will find the table helpful for their future endeavors.

5. Conclusion: Summary, Limitations, and Future Scope

The conclusion of this research comes up with few important findings and proposals.First, knowledge sharing and distribution are the next key-step to make the B2G and V2G like advanced technology an essential part of smart city planning.Renewable energy such as wind turbines, solar panels, and hybrid storage devices like supercapacitor can play a major part in next generation energy sector.Most of the researchers have not yet been properly motivated and informed about this new field of research—crowd energy and smart city living.It is believed that upon getting the concept of these technologies being implemented into next generation smart living, many researchers and scientists will take active part in this.

The primary limitation faced in this field of study is to find adequate information to perform research.As, both B2G and V2G are the next generation technologies for smart living, not enough verified cases and information could be found to conduct further study currently.Therefore, conducting the research is challenging.Apart from it,few projects are currently being implemented in Asian countries.However, information again could not be gathered,as not all of them were in operational.Governments as well as research institutes, and other related organizations should come forward to sponsor more projects in this area.In Bangladesh, the present government has taken this technology into key-consideration and already the Ministry of Power Energy and Mineral Resources has asked research proposals from individuals as well as academic institutes and other organizations for funding the research grant in smart vehicle charging, energy management, and power and renewable sector.Furthermore, The Ministry of Power Energy and Mineral Resources in Bangladesh formed a separate council namely EPRC (Bangladesh Energy and Power Research Council) to focus in energy and power sector research[37]-[40].Proper initiatives from countries like Bangladesh and others, hopefully within the next decade, world will be emerging to the next generation smart city where B2G and V2G based crowd energy will be implemented.

Acknowledgment

The authors would like to acknowledge the Institute of Energy, Environment, Research and Development(IEERD), Department of Electrical and Electronic Engineering, and overall University of Asia Pacific, Bangladesh to make the platform for this research.