Pre-feasibility study of V2G system in the micro-grid of St. Martine Island, Bangladesh.

(1)

Master’s Programme in Renewable Electricity Production- 60 hp

Master Thesis topic-

Pre- feasibility study of V2G system in the micro- grid of St. Martine Island, Bangladesh.

1TE840: Degree Project in Renewable Energy Production-15 hp Report Code:

ELEKTRO-MFE 20001 Md Abu Raihan Chowdhury

Prepared by

Md Abu Raihan Chowdhury

Date of submission: 23-04-2020

(2)

Abstract

The goal of the study was to evaluate the potential of the V2G system as a solution to peak load leveling and integrating more renewable energy in the microgrid of St. Martine Island. Simulink Simscape software was used to model a microgrid with a V2G system for the small community of the Island.

The result of the study shows a V2G system with 100 electric cars could play an important role for peak shaving by supplying up to 0.8 MW of electric power back to the grid during peak hours, where each car contributes 10 kW of electric power. It also demonstrates that the V2G system effectively helps to promote solar power capacity from 1 MW to 2.5 MW, hence increase 23.59% share of solar energy in the total grid energy uses compared with the current microgrid of St. Martine Island.

Key Words: Vehicle to grid, Micro-grid, Electric Vehicle, Simulink Simscape, Feasibility analysis.

(3)

Popular Scientific Summary

The electricity that is generated from non-renewable sources causes environmental pollution and climate changes. Fossil fuel uses leads to the depletion of fossil fuel resources as well as global warming. On the other hand, renewable energy sources can be used to produce electricity with very few or no CO2 emissions. So, now governments are focusing on renewable energy production. But solar, wind, and other types of renewable energy sources have intermittency. They are not continuously available due to natural factors that cannot be controlled. So, renewable energy needs to be utilized when it is available, or its intermittency can be overcome by energy storage.

All Electric vehicle uses a battery pack of large capacity to power the electric motors. These batteries can be used to store the energy that is generated from renewable sources and use them when needed. Besides, the electric grid must always stay in balance. With the development of variable renewable energy production, the management of this balance has become complex. Vehicle to grid is a technology that enables energy to be pushed back to the grid from the battery of an electric car and helps to manage fluctuations on the electricity grid. It helps to balance the grid by charging the battery when renewable energy is available and load demand is low, then sending energy back to the grid when load demand is high.

However, St. Martine Island is a small Island in Bay of Bengal about 9km south of the mainland of Bangladesh. Nearly 6000 people are living there.

Since the island is far away from the mainland, grid connection is almost impossible in terms of cost and geographic location. St. Martine Island has a very high solar power potential, but very low average wind speed. Currently, the electricity demand is fulfilled by stand-alone diesel generators, PV panels, and wind turbines. The current microgrid gets a high load demand during peak hours which is between 6 pm to 11 pm. During this time grid become fully dependent on diesel generators which leads to fossil fuel uses and

environmental pollution.

Here, the project's key objective is to determine the prospects of V2G technology on St. Martine Island to level the peak load during peak hours, given that St. Martine Island is a low windy island with a high average number of yearly peak sun hours. Another goal is to examine the degree to which the share of solar power can be increased by a V2G system in St. Martine Island.

In the project, at first, we have modeled a microgrid using Simulink Simscape software. Simulink Simscape enables modeling of a system by putting direct physical connections between the block diagram. In the microgrid model, there are five main sections, which have been designed by assembling fundamental components in the schematic.

(4)

A V2G system has been modeled which consists of 100 electric cars as a prototype. Each car has a battery of 100 kWh capacity. Considering the condition of St. Martine Island and the objective of the project, we have made some assumptions while modeling the V2G section.

The results of the project showed that the V2G system significantly smoothed out the peak load during peak hours. It also demonstrated that charging electric cars during daytime by solar power and sending energy back to the grid during peak hours enables the V2G system to accommodate more renewable solar energy sources in the microgrid of St. Martine Island.

Finally, the project evident that the V2G system can be integrated into the microgrid of St. Martine Island to level the peak load and to increase the share of solar energy in the total energy uses of the Island.

(5)

Acknowledgments

At first, I would like to express my sincere gratitude to my supervisor Juan de Santiago, Senior Lecturer, Department of Electrical Engineering, Di- vision of Electricity for his continuous guidance, effort and valuable

suggestions throughout the project. I have regarded all of our meetings as very enjoyable and beneficial to the quality of my work. I am grateful to him for all the support and time, this research would not have been possible without your continuous guidance.

Secondly, I would like to thank Valeria Castellucci, Associate Senior Lecturer, Department of Electrical Engineering, Division of Electricity for providing me valuable suggestions to carry out my project successfully.

Thirdly, I would like to thank Hana Barankova, Professor, Department of Electrical Engineering, Division of Electricity, for her review of my final draft and presentation. Her comments have improved the correctness and

readability of this thesis.

Lastly, I would like to express my sincere appreciation to my parents for encouraging and supporting me throughout the study.

(6)

List of Abbreviations

V2G Vehicle to grid

EV Electric Vehicle

PV Photovoltaic

DoD Depth of Discharge

SOC State of Charge

DC Direct Current

AC Alternating Current

LCC Life Cycle Cost

ALCC Annualized Life Cycle Cost OM Operational and maintenance cost LiFePO4 Lithium Iron Phosphate

BEV Battery Electric Vehicle

CO2 Carbon Dioxide

List of Units

Unit Meaning

W Watt

kWh Kilo Watt-hour

Km Kilo Meter

(8)

List of tables

Table 1 Cycle-life of Lithium-ion battery as a function of the depth of discharge.

Table 2 Load estimation of St. Martine Island.

Table 3 Result of comparison of the PV panels capacity in case 1, case 2, case 3 and case 4 of the microgrid of St. Martine Island.

Lists of Figures

Figure 1 Aerial view of St. Martine Island.

Figure 2 Schematic diagram of the V2G microgrid model in Simulink.

Figure 3 Modeling wind power controls based on wind speed.

Figure 4 Car charge and grid regulation controlling Figure 5 Car Profile 1.

Figure 6 Car Profile 2.

Figure 10 Car Profile 6

Figure 11 Load profile of a typical summer day of St. Martine Island.

Figure 12 Load profile of a typical winter day of St. Martine Island.

Figure 13 Yearly simulation diagram.

Figure 14 One-month simulation graph.

Figure 15 Case A: Detail grid scenario of microgrid model on a random summer day.

Figure 16 Case B: Grid scenario with electric vehicles but no V2G facility on a random summer day.

Figure 17 Case 1: Yearly simulation of the current grid with 1MW of PV and without the V2G system.

Figure 18 Case 2: Yearly simulation of modeled microgrid with 1.5 MW PV farm.

Figure 19 Case 3: Yearly grid simulation with 2.5 MW of PV panels.

Figure 20 Case 4: Yearly grid simulation with 3 MW of PV panels.

(9)

9

1. Background

The total area of St. Martine Island is 36 km^2. It is the only coral island in Bangladesh. St. Martin's Island has become a popular tourist spot due to its natural beauty. About 6000 people are living on the Island [1]. Currently, people meet their electricity demand from the microgrid of the island. The existing grid consists of several units of diesel generators, PV panels, and wind turbines. Now, the diesel generator acts as the main source of electricity in the microgrid. St. Martine island has a very low average wind speed but a high average peak sun hour around the year. To make existing micro-grid less fossil fuel dependent, it is important to make sure the proper utilization of available solar power of the island.

Current microgrid gets a high load demand during peak hours which is between 6 pm to 11 pm. During this time grid become fully dependent on diesel generators which leads to fossil fuel uses and environmental pollution.

Figure 1: Arial view of St. Martine Island [2].

As fossil energy resources are depleted and climate change problems became a matter of concern, governments are now focusing on renewable energy production.

Due to intermittency, renewable energy needs to be stored when load demand is low. Vehicle to grid is a potential way to store the energy that is produced from renewable sources and use them when needed. V2G enables energy to be pushed back to the power grid from the battery of an electric car. V2G storage capabilities can enable EVs to store and discharge electricity generated from renewable energy sources such as solar and wind. The electric vehicle can help to balance the grid by charging the battery when demand is low and sending power back to the grid when demand is high. Most modern battery electric vehicles use lithium-ion cells that can achieve round-trip efficiency greater than 90% [3]. A study found that

(10)

10 overall round trip efficiency of V2G is about 70% [4] where most of the losses are in the system rather than in battery.

Most of the popular electric cars use Lithium-ion batteries. Lithium-ion batteries have a finite number of charging cycles, therefore using vehicles as grid storage can impact battery longevity and its capacity. However, Battery cycle life is affected by many different stress factors including temperature, discharge current, charge current, and depth of discharge ranges, etc. [5]. Recent cost reduction of Li-ion batteries raises the expectation that electric vehicles and energy storage will become cost-competitive. It is expected that the price of Li-ion battery will decrease by at least 50 % by 2030 and up to 75 % by 2040, due to learning from mass production driven by electric vehicles [6].

Table 1: Cycle life of Lithium-ion battery as a function of the depth of dis- charge [7].

Depth of Discharge Discharge cycles of Li-ion battery (LiFePO4)

100% DoD 600

80% DoD 900

60% DoD 1500

40% DoD 3000

20% DoD 9000

10% DoD 15000

Table 1 estimates the number of discharge/charge cycles Li-ion can deliver at various DoD levels before the battery capacity drops to 70 percent [7]. The depth of discharge refers to the percentage of the battery capacity that has been

discharged relative to the total capacity of the battery.

Electric car charging strategies can be categorized into two main types; Indirect and direct controlling. Indirect controlling get done by financial or other incentives that influence member charge behavior. And the direct control strategies are get done by a system administrator, either the utility or a third party who controls when and how EVs to charge [8]. In the smart charging system, electric vehicles, charging stations and charging operators share data connections. A study showed that all popular BEV models can completely compensate for the energy cost and generate a positive net profit by observing smart charge scheduling in a V2G system [9]. Finally, this vehicle-to-grid is a promising technology as an approach to help renewable energy to become the base of electrical energy.

(11)

11

1.1 Objective

The main objective of the project is to evaluate the prospects of V2G technology in St. Martine island, considering that St. Martine island is a low windy and a high average number of yearly peak sun hours island. Another objective is to analyze the scope of increasing the share of solar power by a V2G system in the microgrid of St. Martine island.

2. Methodology

In this project, we have modeled a microgrid using Simulink Simscape software to evaluate the V2G system as a potential way of peak load leveling during peak hours as well as to increase the share of renewable energies in the microgrid.

Simscape enables modeling of a system by putting the direct physical connections between block diagrams. Here, the different part of the micro-grid has been modeled by assembling fundamental components in the schematic.

Figure 2: Schematic Diagram of the V2G microgrid model in Simulink.

The designed microgrid consists of five sections. These are, V2G, Diesel generator, Solar PV system, Wind turbine, and consumer loads.

(12)

12

2.1 Residential Load

The size of the microgrid represents a community of 1200 households. Since there is not any available information about the load profiles of St. Martine Island, we have calculated the load profile by the bottom-up method. We have estimated all the appliances that people use and the consumption of all inhabitants sums up the load profile [10]. The result of the load estimation is added to the result section of this report.

2.2 Solar PV system

In the solar PV section, the production of solar power depends on three factors;

Irradiance data [11], the efficiency of the PV panel and the area covered by panels.

The insolation in St. Martine Island varies from 3.84 kWh/m^2 to 5.77 kWh/m^2 on an average of 5 kWh/m^2 in a day [11].

2.3 Wind Turbine

The wind turbine produces electrical power based on provided wind speed profiles [12], where the produced power is proportional to the wind speed; power ∞ (wind speed)^3. When the wind speed exceeds the maximum allowed wind speed value, wind turbines trips and stop producing electrical power until wind speed gets below the nominal value [13].

Figure 3: Modeling wind power controls based on wind speed.

In this project, maximum allowed wind speed is 15 m/s and nominal wind speed value was considered to be 13.5 m/s. St. Martine Island has a low average wind speed, which is less than 5 m/s [12]. During the rainy season, the Island usually goes through several events of the cyclone and very high windy storms which causes the wind turbine to get shut down. A wind turbine rated capacity of 1.5 MW was considered in the model.

(13)

13

2.4 Diesel Generator

In the microgrid model, diesel generator balances the power production with the change of load demand. The frequency deviation can be determined by looking at the rotor speed of its synchronous machine. Any change in the rotor speed from the reference value leads to control the power production. A diesel generator rated capacity of 2 MW was considered in the model.

2.5 V2G

The V2G system is connected to the secondary side of the transformer along with residential loads. It consists of 100 Tesla Model S electric cars with a lithium-ion battery pack of 100 kWh each. Here, the V2G section has two main functions, it controls the charging of EV batteries and uses the available energy in the battery to regulate the grid during peak hours. In this project, we are considering the Co-ordinated centralized charging strategies. A central authority gathers

information about the load demand, electricity generation, and electric cars plugged-in with the grid. Then, depending on the information they control the car charging and regulating the grid [14].

2.5.1 Assumptions

Considering the objectives of the project, some assumptions have been taken while modeling the V2G System of the microgrid. These are:

1. Electric cars are not allowed to charge their battery from 6 pm to 11 Pm.

2. Electric cars can send energy from battery only from 6 pm to 11 pm.

3. Due to battery discharging over 50% DoD significantly reduces its lifetime [7].

4. The electric car will be disconnected from contributing to the grid production when its state of chare goes below 50%.

5. Tesla Model S vehicles come with the Mobile Connector, which allows charging at up to 40 amps and includes adapters for connecting to a variety of electricity sources [15]. Tesla Supercharger is a proprietary DC rapid- charging station that provides up to 145 kW of power, which ensures fast charging facility [16]. In this Project, while modeling the V2G section, we have assumed that the Battery charging rate is double than its discharging rate. While contributing to the grid, each car sends 10 kW power to the grid.

(14)

14 6. Tesla Model S 100D with 100 kWh battery has a range of 540 km [17] and it can drive 100 km by 20 kWh energy [18]. As St. Martine Island is a small Island, it is assuming that the average driving distance covered per day is less than 50 km which consumes around 10 kWh energy from the battery.

Figure 4: Car charge and grid regulation controlling.

The figure shows that the car can participate in grid regulation during peak hours only if the state of charge is higher than 50%. EV batteries are not allowed to charge during peak hours. For both functions, charging batteries and regulating the grid, electric cars need to be plugged in with the grid.

2.5.2 Car Profiles

In this study, we are considering 100 electric cars of the Tesla Model S as a prototype. Different car user has different routines for their work, which leads to several car profiles. Here, we have prepared six different types of car profiles of the cars. The state of charge means the level of charge of a battery relative to its capacity and plug-in state represents the condition when EV is connected to the grid. These car profiles are:

Car Profile -1: 20% of car users use their car to go to the office and after the office, they drive the car to go to do their necessary tasks and return home with some

(15)

15 delay. They have the facility to put their car into the charging station during their office hours.

Figure 5: Car Profile 1.

Car Profile-2: 10% of car users have the same routine with car profile 1, except they return home just after finishing their office.

Car Profile-3: 25% of car users going to work with a longer ride or ride their car several times during their work hours but have the facility to put their car for charging.

Car Profile-4: 5% of people going to work with no possibility to charge their car at work.

(16)

16 Figure 8: Car Profile 4.

Car Profile-5: 20% of cars stay at home for random reasons and contribute to the grid regulation during peak hours.

Car Profile-6: 20% of people working at night shift. They don’t have the facility to charge their car at work.

(17)

17

2.6 Economic analysis

The goal of the economic analysis is to evaluate the scope of increasing the cost- effective share of solar power by the V2G system in the total power production of St. Martine Island.

In the current microgrid, St. Martine Island has a 1 MW of PV farm along with 1.5 MW of wind turbines that produce renewable energies. To increase the share of renewable energy into the total production of a grid, V2G is a potential technology.

It provides the advantage of charging EV batteries by renewable energy and utilize the stored energy to regulate the grid.

In this economic analysis, we will compare PV size of 1 MW, 1.5 MW, 2.5 MW and 3 MW in the case 1, case 2, case 3 and case 4 of the micro-grid of St. Martine Island respectively.

To find out the cost per unit energy for each case, at first, Life cycle cost is calculated.

Life cycle cost= Initial capital cost + total operational and maintenance costs.

The present value of all future operational and maintenance costs is calculated by multiplying the annualized cost of OM and the present worth factor [19]. In this calculation, the annualized OM cost of PV panels is 2% of the initial capital cost.

Present worth factor = ^(1+𝑖)^𝑁⁻¹

𝑖(1+𝑖)^𝑁

Where, i= annual interest rate, N= the service life in the year.

In this analysis, the lifetime of the PV panel is 20 years and the interest rate is considered to be 5%. Thus, the present worth factor results to be 12.46.

Then, the annualized life cycle cost, ALLC = life cycle cost / present worth factor.

Finally, Cost per kWh = ALLC / annual kWh provided by the system.

St. Martine island has a yearly residential load demand of 8471 MWh where 3134 MWh is during the daytime from 7 am to 5 pm. And yearly load demand due to car charging at daytime is 2265 MWh. So, the total yearly load demand from 7 am to 5 pm is 5399 MWh.

(18)

18

3. Results

3.1 Load Estimation

Electricity consumer of the St. Martine Island consists of 1200 households, 2 schools, 2 clinics, 50 shops, 5 resorts and 2 three-star hotel for the travelers, 2 local factories.

Table-2: Load estimation of St. Martine Island [20].

House-1200

Load Power (W) Hours of opera-

tion/day

Energy (Wh/day)

Light 20×8=160 5 800

Ceiling Fan 3×75=225 12 2700

Air-conditioner 2000 (25% of house)

12 24000

Refrigerator 1×150=150 24×30%duty=7.2 1080 Water pump/

Motor

1×750=750 0.5 375

TV 1×60=60 5 300

Washing Ma- chine

1×1200=1200 0.75 900

Total (1200 houses)

3654kW 14586kWh/day

School-2

Light 20×20=400 6 2400

Ceiling Fan 20×75=1500 6 9000

Water pump/

Motor

1×750=750 0.5 375

Computer 10×150=1500 6 9000

Total (2 Schools)

8.3kW 41.55kWh/day

Shop-50

Light 20×5=100 12 1200

Ceiling Fan 2×75=150 12 1800

Air-conditioner 2000 (30% of Shop)

12 24000

(19)

19 Refrigerator 1×150=150 24×30%duty=7.2 1080

Total (50 shops)

50kW 564kWh/day

Clinic-2

Light 20×20=400 24 9600

Ceiling Fan 10×75=150 24 3600

Air-conditioner 2×2000=4000 24 96000

Refrigerator 2×150=300 24×30%duty=7.2 1080

Motor 1×750=750 1 750

Medical Equip- ment

5000 12 60000

Total (2 Clinics)

21.2kW 342kWh/day

Resort-5

Light 3×10×20=600 16 9600

Ceiling Fan 10×75=750 16 12000

Air-conditioner 10×2000=20000 16 320000

Motor 1×2237=2237 3 6711

Others 2000 16 32000

Total (50 shops)

127.93kW 380kWh/day

Three Star hotel-2

Light 25×3×20=1500 16 24000

Ceiling Fan 25×75=1875 16 30000

Air-conditioner 25×2000=50000 16 800000

Motor 2×2237=4474 6 26844

Others 2000 16 32000

Total (2 Hotels) 119.69kW 1825kWh/day

Small Factory-2

Light 20×20=400 24 9600

Ceiling Fan 10×75=750 24 18000

Motor 1×7500=7500 24 180000

Others 1000 24 24000

Total (2 Factories)

19.3kW 231.6kWh/day

In St. Martine, summer lasts for 8 months and winter is 4 months long. Yearly total residential load demand of the Island is 8471 MWh/year.

(20)

20 Figure 11: Load profile of a typical summer day of St. Martine Island.

Figure 12: Load profile of a typical winter day of St. Martine Island.

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

MW

Hours

Load Profile on a Summer Day

House School Shop Clinic Resort Hotel Factory

0 0.2 0.4 0.6 0.8 1 1.2 1.4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

MW

Hours

Load Profile on a winter day

House School Shop Clinic Resort Hotel Factory

(21)

21

3.2 Simulation Results

Simulation for one year has been conducted, which includes load profile, solar irradiance data [11] and wind speed data [12] for one year.

Figure 13: Yearly Simulation graph.

The simulation has been conducted considering the period of hours, which leads to a large number of output data. In the yearly simulation graph, we can see that simulation has introduced some noises which are because of software deficiency.

Here, the simulation started in April as summer starts and it ends in March when winter gets over.

(22)

22 Figure 14: One-month simulation graphs.

From the one-month simulation, we have got a clear grid scenario for June. During monsoon, St. Martine Island usually goes through several events of the cyclone and high windy storm. On the 28th of June, there is a grid situation at 672nd hours of the graph where the wind speed exceeds the maximum allowed wind value 15 m/s that leads to trip the wind turbine.

Figure 15: Case A: Detail grid scenario of microgrid model for St. Martine Island on a random summer day.

Figure 15 shows the grid scenario where most of the cars are charging during the daytime to properly utilize the available solar power. The V2G facility

significantly reduces the amount of peak load during peak hours.

(23)

23 Figure 16: Case B: Grid scenario with electric vehicles but no V2G facility on a

random summer day.

In figure 16, we can see that the electric vehicles charging at any random time which leads to a significant rise in load demand during peak hours.

Figure 17: Case 1: Yearly simulation of the current grid with 1MW of PV and without the V2G system.

Figure 17 shows that 1 MW of Solar PV can’t fulfill total load demand during the daytime. So, more PV can be installed in the current grid of St. Martine Island.

(24)

24 Figure 18: Case 2: Yearly simulation of modeled microgrid with 1.5 MW PV

farm.

Here, total load demand during the daytime has increased due to the addition of the V2G system and the solar PV capacity of 1.5 MW can’t meet the load demand during the day.

Figure 19: Case 3: Yearly grid simulation with 2.5 MW of PV panels.

Here, negative diesel value represents the amount of surplus solar power that is produced by 2.5 MW PV farms.

(25)

25 Figure 20: Case 4: Yearly grid simulation with 3 MW of PV panels.

The negative value of the diesel generator graph in the figure states that 3 MW of Solar PV produces a significant amount of surplus energy after meeting the total load demand during the daytime.

3.3 Result of Economic Analysis

Table- 3: Result of comparison of the PV capacity in case 1, case 2, case 3 and case 4 of the microgrid of St. Martine Island.

(26)

26 Here, 1 MW solar PV farm produces 2055 MWh/year which is 1079 MWh lower than the total residential load demand during the day in the current microgrid of St. Martine Island. In contrast, 1.5 MW of solar PV produces 3082 MWh/year that can fulfill the 98.34% of load demand during the day in the current microgrid.

Therefore, PV capacity increase over the size of 1.5 MW would require storing the surplus energy.

Due to the integration of the V2G system of 100 Tesla Model S car, the load demand is increased by 2265 MWh/year and the total load becomes 5399 MWh/year in the microgrid model during the day.

The PV farm of 1.5 MW can fulfill 57.08% of total load demand during the day in the modeled microgrid. On the other hand, 2.5 MW of PV can meet 95.14% of 5399 MWh/year load demand. And the price of each kWh energy for the cases of the PV capacity of 1 MW, 1.5 MW and 2.5 MW is the same.

Besides, the economic analysis shows if the size of the Solar PV capacity is increased to 3 MW, then 766 MWh of surplus produced energy will be wasted if that is not stored or load demand is not increased. Thus, the price of per kWh energy is increased by 12.69% in the case of 3 MW capacity of PV farm.

Size 1 MW PV

without V2G

1.5 MW PV with or without V2G

2.5 MW PV with V2G

3 MW PV with V2G Residential load

during the day, MWh/year

3134 3134 3134 3134

Battery charging, MWh/year

- 2265 2265 2265

Total load during

the day,

MWh/year

3134 5399 5399 5399

PV production, MWh/year

2055 3082 5137 6165

Amount of solar energy utilized, MWh/year

2055 3082 5137 5399

Capital cost [21] $95000×10

= $950000

$95000×15=

$1425000

$95000×25=

$2375000

$95000×30=

$2850000

Total OM cost $236740 $355110 $591850 $710220

LLC $1186740 $1780110 $2966850 $3560220

ALLC $95243 $142865 $238109 $285731

Cost/kWh $0.0463 $0.0463 0.0463 $0.0529

(27)

27

4. Discussion

Annual electricity consumption of the real grid of St. Martine Island is 8471 MWh/year. Due to the integration of electric vehicles and the V2G system, the overall demand for electricity in the model microgrid is 10736 MWh/year, where 2265 MWh is due to EV battery charging. On the other hand, the V2G system supplies 1493 MWh energy per year to the microgrid which is 13.9% of the total yearly energy demand. The controlled charging system leads most of the EV batteries to be charged by solar power during the daytime. So, the energy that is supplied by the V2G system to the grid results in very low or no CO2 emission.

Figures 15 and 16 demonstrate that the uncontrolled car charging and the grid without V2G facility lead to an increase of 1.5 MW of load demand during peak hours compared with the grid having a V2G system with controlled car charging profiles.

However, the charging of electric cars is controlled by directly or indirectly in the real-life scenario. Indirect controlling controls the behavior of car users by financial or other incentives. And direct control is done by a system administrator who communicates with the charge station and grid to control the charging of EV batteries. In this project, we have considered the direct method of controlling which does not allow an electric car to be charged during peak hours. Although controlled charging methodology shifts the electric load due to EV batteries charging from night to daytime, it opens the door to increase the share of renewable energy in the total electricity uses of St. Martine Island.

Besides, the result of the economic analysis shows that the solar PV capacity of 1.5 MW can produce enough energy to meet the residential load demand during daytime in the current microgrid of St. Martine Island. Further increase of the nominal capacity of Solar PV over 1.5 MW, requires the surplus energy to be stored. The Economic analysis also demonstrates that the Solar PV capacity of 2.5 MW can cost-effectively charge the EV batteries as well as fulfill the residential load during the daytime in the modeled microgrid. However, due to software deficiency, some introduced noise in the simulation leads to a ±5% error.

Moreover, the number of electric vehicles is increasing day by day. More electric vehicles in the system will open the door to integrating more solar power in the system, hence V2G can contribute more portion of renewable energy to the total electricity consumption of St. Martine Island.

Although St. Martine Island has a low potential for wind power, installing more wind turbines will help to reduce the restriction over EV batteries charging regardless of time. V2G system can integrate wind power into the energy system efficiently.

(28)

28

5. Conclusion

The goal of the study was to estimate the advantages of the V2G system in the microgrid of St. Martine Island. The study was performed by designing a microgrid with a V2G network of 100 electric vehicles, considering the controlled charging of the EV battery and the energy can only be sent to the grid during peak hours. A comparison between St. Martine Island's current grid without V2G and the model microgrid with V2G leads to the conclusion that the V2G network can be

introduced in the current grid to level the peak load as well as to properly utilize the available renewable energy of the island. In the microgrid model, charging most of the car during daytime and sending power back to the grid during peak hours levels the peak load from 1.8 MW to 1.1 MW without discharging the batteries below 50% of its full charge capacity.

Moreover, the result of the economic analysis shows if the capacity of PV farm is increased from 1 MW to 2.5 MW, a V2G system of 100 electric cars can effectively be accommodated in the microgrid of St. Martine Island. Thus, EV Batteries can be charged by solar power and can contribute to the grid regulation during peak hours. Finally, several demonstrations of different cases in the project reflect that the V2G system provides more flexibility to the grid, help to smooth out peak load during peak hours and to integrate more renewable Solar energy sources in the microgrid model of St. Martine Island. Therefore, it can be said that the project is successful to evaluate the potential of V2G technology in the microgrid.

6. Future work

With a higher number of electric vehicles in the system, V2G can be used for frequency regulation. For frequency regulation, a finite number of electric cars need to be plugged in with the grid. Further studies required to find out the minimum number of electric vehicles needed to efficiently use the V2G system for frequency regulation in St. Martine Island. Developing the micro-grid model with the features of frequency regulation by the V2G system would be quite interesting future work.

Here, the V2G system provides the advantages of increasing solar power

production, which is used to satisfy some of the load demand during peak hours, hence reducing diesel generator production as well as CO2 emissions. In this project, the amount of CO2 emission reduced by the V2G system was not

considered. Moreover, how the vehicle owner would be financially benefited from the grid authority was not also covered here, since using EV batteries as grid storage significantly impact battery lifetime and capacity. Covering all these issues based on real-life scenarios with a real-life number of electric cars could be an interesting part of future work.

(29)

29

References

[1] A. S. Islam, M. M. Rahman, M. H. Mondal and F. Alam, "Hybrid energy system for St. Martin Island, Bangladesh: An optimized," Evolving Energy-IEF International Energy Congress, 2012.

[2] SaanImages, "Saan Images," [Online]. Available:

https://www.saanimages.com/product/panorama-of-saint-martins-island- bangladesh/.. [Accessed 15 February 2020].

[3] L. O. Valoyen and M. I. Shoesmith, "The effect of PHEV and HEV duty cycles on battery and battery pack performance," Plug-in Highway Electric Vehicle Conference, 2007.

[4] E. Apostolaki-Losifidou, W. Kempton and P. Codani, "Reply to Shirazi and Sachs comments on “Measurement of Power Loss During Electric Vehicle Charging and Discharging”," Energy, vol. 142, pp. 1142-1143, 2018.

[5] J. Wang, P. Liu, J. Hicks-Garner, E. Sherman, S. Soukiazian, M.

Verbrugge, H. Titaria, J. Musser and P. Finamore, "Cycle-life model for graphite-LiFePO4 cells," Journal of Power Sources, vol. 196, pp. 3942- 3948, 2011.

[6] I. Tsiropoulos, D. Tarvydas and N. Lebedeva, "Li-ion batteries for mobility and stationary storage applications- Scenarios for cost and market growth," Publication office of the European Union, vol. EUR 29440 EN, 2018.

[7] Battery University, 2020. [Online]. Available:

https://batteryuniversity.com/learn/article/how_to_prolong_lithium_based _batteries.. [Accessed 13 March 2020].

[8] K. Dayem, C. Mercier, P. May-Ostendorp and X. Consulting, "Electric vehicle charging control strategies.," Business & Technology Surveillance, 2019.

[9] Z. Li, M. Chowdhury, P. Bhavsar and Y. He, "Optimizing the

performance of vehicle-to-grid (V2G) enabled battery electric vehicles through a smart charge scheduling model," International Journal of Automotive Technology, vol. 16, pp. 827-837, 2015.

(30)

30 [10] B. Sicre, J. d. Santiago and O. Rodriguez, "The generation of domestic hot

water load profiles in Swiss residential buildings through statistical

predictions.," Energy and Buildings, vol. 141, pp. 341-348, 15 April 2017.

[11] Solar Electricity Handbook, "Solar Irradiance," [Online]. Available:

http://www.solarelectricityhandbook.com/solar-irradiance.html..

[Accessed 25 January 2020].

[12] Bangladesh Meteorological Department, "Normal Wind Speed," 2014.

[Online]. Available: https://www.bmd.gov.bd/p/Normal-Wind-Speed/..

[Accessed 15 February 2020].

[13] Mathworks, "24-hours simulation of a vehicle to grid system.," [Online].

Available: https://se.mathworks.com/help/physmod/sps/examples/24- hour-simulation-of-a-vehicle-to-grid-v2g-system.html.. [Accessed 01 February 2020].

[14] J. J. A. Saldanha, E. M. d. Santos, A. P. C. d. Mello and D. P. Bernardon,

"Control Strategies for Smart Charging and Discharging of Plug-In Electric Vehicles.," 2016.

[15] Clean Technica, "New Tesla Home Charging Station Offers Faster Charging, Plus You Can Take It With You," [Online]. Available:

https://cleantechnica.com/2019/01/16/new-tesla-home-charging-station- offers-faster-charging-plus-you-can-take-it-with-you/. [Accessed 15 02 2020].

[16] Inside EVs, "Supercharging Tesla Model S 60 kWh Versus 85 kWh - Video + Graphs," [Online]. Available:

https://insideevs.com/news/323718/supercharging-tesla-model-s-60-kwh- versus-85-kwh-video-graphs/.. [Accessed 01 March 2020].

[17] U.S. Department of Energy, "Fueleconomy.gov," [Online]. Available:

https://fueleconomy.gov/feg/Find.do?action=sbs&id=41193. [Accessed 19 2 2020].

[18] Tesla Inc., "European energy level. Tesla Europe," [Online]. Available:

https://www.tesla.com/en_EU/support/european-union-energy-label..

[Accessed 17 February 2020].

[19] N. A. Hamdeh and K. Alnefaie, "Techno-economic comparison of solar power tower system, photovoltaic system, wind turbine, diesel generator in supplying electrical energy to small loads," Journal of Taibah

University for Science, 2018.

[20] Load Profile Generator, [Online]. Available:

https://www.loadprofilegenerator.de/. [Accessed 25 March 2020].

(31)

31 [21] Solar Choice, "Average Commercial Solar panel by city and size,"

[Online]. Available: https://www.solarchoice.net.au/blog/solar-power- system-prices.. [Accessed 05 February 2020].

Pre-feasibility study of V2G system in the micro-grid of St. Martine Island, Bangladesh.