Design and Simulate an Off-Grid PV System with a Battery Bank for EV Charging

This paper presents the control and design of a stand-alone photovoltaic (PV) system with a battery bank for an electric vehicle (EV) battery charging. It also describes the necessary requirements and power electronic converters design for the EV battery charging station. The design of a bidirectional converter to allow for bidirectional power flow control to regulate the charging and discharging of the battery bank is presented. The PV system connected to the battery bank system is used to enhance the power output of renewable energy sources, regulate electrical power to effectively charge batteries, draw maximum power from solar panels, and provide a high-quality DC output for electric vehicle charging. The PV is the primary power source of the system, and a battery bank is used as a backup storage system. A comparison between the performance both of lithium-ion batteries and lead-acid batteries based on using it as a backup storage system for the off-grid PV system is presented. The proposed system has been designed and simulated in the MATLAB-Simulink environment. The effective performance of the proposed station for EV battery charging is achieved through the simulation results. A proposed charging station cost estimation is performed to present a powerful index of the proposed charging station's effectiveness.


Introduction
With the continual increase of environmental worries, countries of the world have made many standards for carbon emissions. Electric vehicles (EVs), which fundamentally depend on electricity, have the chance to replace traditional gasoline and diesel vehicles. EVs provide an efficiency increase in energy savings through effective fuel economy, particularly when the electricity is produced from renewable resources, like solar and wind. Several suggestions have been made when dealing with EVs in smart grids, which are not just an electrical load, but as a power resource [1][2][3]. A methodology for modeling EV as an extra load on the distribution network was proposed in [4][5][6][7][8][9][10]. Many researchers introduced the grid-connected PV system for EVs charging [11][12][13][14]. Generally, the grid-connected system does not need a backup battery. The grid-connected system has two configurations. The first: the PV system participates with the utility grid in supplying the load with energy. This means that the PV system power is often less than the power required for the load. The second: the PV system is designed to supply the load with the required power and the utility grid in this configuration is a backup source during solar power fluctuations. With the continuous growth of demand for using EVs, it became necessary to increase charging stations. EVs need a large number of charging stations. Thus, EVs will be a significant burden on the distribution network system in the future. The off-grid PV system or standalone PV system term means the system has no connection to the main grid. Off-grid systems are appropriate for the electrification of small societies and it is feasible for remote areas. Off-grid systems are suitable for EV charging stations in faraway roads. Many papers presented the off-grid system design [15][16][17]. For facing renewable power fluctuations, the control of the charging process of EV has been discussed in [18][19][20]. In [21] introduces EV fast-charging technologies. Banks of backup batteries are used most mainly for the off-grid system for energy storage [22][23]. A comparison between Lead-Acid and Lithium-Ion Batteries for energy storage in off-grid energy systems is presented in [24]. In [25] shows a study of the backup battery charging control techniques for PV systems. This controller maintains the good advantages and solves the drawbacks and restrictions of the conventional controllers. Lithium-ion batteries [26] are one of the best energy storage systems because of the many advantages of high storage capacity, and therefore the solar charging station is a fully renewable energy system. In this case, the PV-Battery Energy Storage System needs a good controller to regulate the charge and discharge of the battery bank.

Contribution of This Paper
The present paper describes a very practical approach in designing a model of a charging station of electric vehicles shown in Fig. 1 consists of an off-grid system with a battery bank.
In a simple way, the principal contributions of this paper may be described as follows: 1. The present work offers an environmentally-friendly charging station model for charging electric vehicles that suit remote areas far from the utility grid. 2. The design of the proposed charging station in step by step is explained. 3. DC-DC converters design used for the proposed charging station is presented to ensure the best performance. 4. The effectiveness test of the logic controller for the bidirectional converter [27][28] is done to regulate the charge and discharge of the battery pack. 5. The performance test of the lithium-ion batteries and the lead-acid batteries as an energy storage system for the proposed charging station. 6. A practical approach in cost estimation of the proposed charging station is carried out.

The Organization of This Work
The rest of the paper is documented in the following order. Section 2 presents the estimation of the required charging power of an electric vehicle. The back-up batteries' storage capacity and the sizing of the PV system are calculated in sections 3 and 4 respectively. Section 5 displays the design of the boost converter, the buck converter, and the bidirectional converter. A summary of the parameters that are used and that are calculated is listed in section 6. The charge and discharge controller modes for the battery pack are described in section 7. Section 8 discusses the simulation results. The performance of the lithium-ion batteries and the lead-acid batteries with the proposed charging station is shown in section 9. Finally, the proposed charging station cost is estimated in section 10.  To achieve the least charging time of EV battery, enough large power chargers are essential, which follows physics fundamental laws: ℎ = / ℎ , where: ℎcharger power in W, -battery energy in watt-hour, ℎ -charging time in h. Thus, for instance, to charge 100% of the 100 kWh battery (Tesla S 100D cars) in 20 min, the required charger power is 300 KW.

Specifying the Battery Storage Capacity of the Off-Grid System
Batteries are the main component of off-grid systems. The storage battery supplies the load with the required power during the night and cloudy weather periods or with the participation of the PV system during limited sunlight intervals. The periods in which the system doesn't depend on energy generated by PV modules are known as 'days of autonomy'. Days of autonomy ( ) usually are 4 days.
Other parameters are taken into account when calculating storage capacity: = the electric vehicle battery power, i.e. 100 kWh battery (Tesla S 100D cars) which is chosen as a load to design the proposed charging station.
= Upper limit allowable depth of the battery bank discharge i.e. 0.8.

Sizing of PV Array
The major steps which are necessary for determining an off-grid PV system sizing as the following: 1. Deciding the nominal operating voltage of the PV system ( − ).
The nominal Voltage i.e., − = 600 V is chosen for the design.
2. Specify the total load current.
The total load current is the sum of both the charge current required for the electric vehicle battery and the charging current required for the battery bank. The charger power is 300 KW and − is 600 V consequently; the charge current ( ) is 500 A. The battery bank storage capacity is 1500 Ah and the charging current of the storage system ( ) is almost 20 % of the battery bank capacity, i.e. 300 A. The total charge current is 800 A.

System losses are added.
The consumed energy by the system components is system losses. Consequently, the required total current, which was specified in step 2, is increased with 20 for system losses compensating. The total current ( ) after adding the system losses is approximately 1000 A.

Determining the number of PV modules.
In the best arrangement of the panels, the required total PV array load in ampere (as shown in step 3) is got with the least number of panels. The solar array is formed of modules that are connected either in parallel or in series.
The demanded number of panels in parallel ( ) is computed by dividing the whole current demanded from the solar array (shown in step 3) by the current generated by the panel at maximum power ( ) shown in Table 2. Consequently, is almost 180 modules.
(The number of modules in series) is calculated by dividing 600 V ( − ) by the nominal panel voltage ( ) shown in Table 2. Thus, is approximately 11 modules. The total number of modules is = * = 1980.

Design the Proposed Charging Station Converters
The essential technology linked with PV systems is the power converter. A perfect PV power converter should drag the maximum power from the PV module and provide it to the load. For the grid-connected systems, this should be achieved with the least harmonic content in the current and at a PF (Power Factor) value is higher than 0.9. In the off-grid systems, the output voltage should also be adjusted to the required value. Three DC-DC converter topologies are used for the proposed charging station shown in Fig 4. The boost converter is used to convert the variable voltage from the modules into a stable voltage while the buck converter is used to reduce the DC Bus voltage to the electric vehicle battery voltage. The bidirectional converter is used to regulate both the charging and discharging of the battery bank. The design of three DC-DC converters is described below.

DC-DC Boost Converter
The boost converter is shown in  The output voltage is related to the input voltage according to the following equation [30]: Where D is the duty cycle.
To calculate the duty cycle of the boost converter, Eq (2) is used. Through the simulation, solar radiation varies from 0 to 1000 W/m 2 and temperature varies from 25 to 75 o C. The boost converter duty cycle is determined for a cell temperature of 25 o C and irradiance of 1000 W/m 2 . As noticed from Figs. 3 and 4, = 600 V. is equal to 600 V which gives, 600 = 1.
During the irradiance 1000W/m 2 and at the temperature 100 o C, the total modules' voltage decreases to 448 V, as noticed in Fig. 3. Therefore, the duty cycle should be decreased to obtain the required DC-link voltage. The computed value of the duty cycle using Eq. 2 is, The increase and decrease of the power and voltage of the array are checked by applying the Perturb and Observe Algorithm [31] for MPPT using the direct duty cycle control method.

Design of Boost Inductor and PV Array Capacitor
The inductance for continuous current in the boost converter is calculated as follows [30]: is the switching frequency which is chosen to be 40 kHz. V IN is the converter input voltage = 448 V. ∆ _ (The ripple current) is calculated as follows [30] (taken as 13% of the total current): The ripple of the output voltage as follows: The PV array is a current source so a capacitor ( ) is connected in parallel to its output, determined using Eq. (4) [30]. (4)

DC-DC Buck Converter
The buck converter shown in   The inductance for continuous current in the buck converter is calculated as follows [32]: ∆ − (The ripple current) is calculated as follows (taken as 13% of the charge current): The output voltage ripple (Δ / ) must not increase by 0.5%. The buck capacitor is calculated using Eq. (7) [32].

The Bidirectional Converter
To smooth power storage generated from off-grid systems during excess generated, the storage battery is provided as a backup service during electricity loss.
This paper suggests a bidirectional buck-boost converter, as shown in Fig. 7 [33] to control the storage and supply of power between the PV system and the battery bank as a result of the following advantages:  It offers the most efficient solution with the smallest external components.  It performs step-up or step-down of voltage using minimum components.  It offers a lower operating duty cycle.  It offers high efficiency across wide input and output voltage ranges.  It is less expensive compared to most of the converters.
Two operation modes can be designed for the bidirectional converter based on the direction of the power flow: boost and buck modes. These modes are presented in detail in [33].  Table (3) presents a summary of the parameters that were used and that were calculated. The charging current of the storage system 300 A

The Parameters Summary
The total current after adding the system losses 1000 A

The number of modules in parallel 180
The number of modules in series 11 The total number of modules 1980 The switching frequency 40 kHz The inductance of the boost converter 16.3

−
The DC-link capacitor 1375 The PV array capacitor 10 The EV battery voltage 350 V The inductance of the buck converter 96.15 The capacitor of the buck converter 67.76 The inductance of the bidirectional converter 128.2 The bidirectional converter capacitor on the side of the battery pack 1375

SOC %
The initial charge state of the battery bank 80 % The initial charge state of the EV battery 40 %

The Charge and Discharge Controller Modes
The charging and discharging controller are used to perform power transfer from and to the battery bank. The stand-alone PV system with the battery bank using the bidirectional converter shown in Fig. 7 has four different control modes based on the magnitude of power provided by the PV array which relies on the temperature and irradiance.
Mode 1: The first control mode is operated when the generated PV power is less than the power demanded by the EV battery, i.e., < ℎ and the charge state of the storage battery (SOC) is more than 40%. In this mode, the bidirectional converter turns into the buck mode in which the controller gives a control signal to make switch 1 ( 1 ) to be in a disconnected state and at the same time gives a control signal to make switch 2 ( 2 ) to be in a connecting state. Storage batteries provide supplementary power through this mode.

Mode 2:
The second control mode is operated when the generated PV power is less than the power demanded by the EV battery, i.e., < ℎ and the charge state of the storage battery (SOC) is less than 40%. The controller in this mode provides a control signal to make both of switch 1 ( 1 ) and switch 2 ( 2 ) to be in a disconnected state.
Mode 3: The third control mode is operated when the generated PV power is more than the power demanded by the EV battery, i.e., > ℎ and the charge state of the storage battery (SOC) is less than 90% and more than 40%. During this mode, the PV array not only provides the power to the EV battery under maximum power point enabled control, but also the surplus power produced by the PV array is used to charge the battery bank. The controller in this mode gives a control signal to make switch 1 ( 1 ) to be in a connecting state and at the same time gives a control signal to make switch 2 ( 2 ) to be in a disconnected state, i.e., the bidirectional converter turns into the boost mode.

Mode 4:
The fourth control mode is operated when the generated PV power is more than the power demanded by the EV battery, i.e., > ℎ and the charge state of the storage battery (SOC) is more than 90%. In this mode, the PV array supplies power to the EV battery under maximum power point enabled control and it is also maintained that the charge state of the storage battery remains more than 90% through constant voltage charging so that the battery does not discharge itself. The controller in this mode provides a control signal to make both of switch 1 ( 1 ) and switch 2 ( 2 ) to be in a disconnected state. The flowchart in Fig. 8 illustrates the four operating modes.

Results and Comments
The proposed charging station is shown in Fig. 4 is simulated using MATLAB, Simulink. The irradiance and temperature time-varying are input signals to the PV modules as described in Fig. 10 as follow: The irradiance starts with zero value until the time 1.5 sec, then gradually increases until equals 1000 W/m2 at the time 2 sec. The temperature starts with 25 o C until the time 2.2 sec, then gradually increases until equals 75 o C at the time 2.5 sec. Fig. 11 describes the discharging and charging modes as follows: With no irradiance and at temperature 25 o C until a time 1.5 sec, as shown in Fig 10, the battery bank power ( ) provides the charging energy for the EV battery in which the battery bank is in the discharging mode. After a time 1.5 sec, the irradiance starts increasing and the PV array power starts increasing. By the time 2s, the battery bank power ( ) decreases in which the battery bank charging mode starts. As noticed in Fig. 11, the power demanded by the EV battery( ℎ ) increases by increasing the PV array power.  Fig. 13.
It is observed in Fig. 12 that the SOC of the battery bank decreases with no solar radiation until the time 1.7 sec, and with the gradual increase of solar radiation, the battery bank begins to gradually enter the charging mode. It is also noted that the transition from the discharge mode to the charging mode is relatively slow. In Figure 13, it is observed that the SOC of the EV battery increases in all modes but in the charging mode, it increases faster.

The Performance of the Lithium-ion Batteries and the Lead-Acid Batteries with the Proposed Charging Station
In this section, the lead-acid batteries will be used instead of the lithium-ion batteries with the same rated capacity as well as the voltage rated to make a comparison between the performance of each type with the proposed charging station.

The SOC of the Lithium-ion and the Lead-acid Batteries during the Charging and Discharging Mode
The SOC of the lithium-ion and the lead-acid batteries during the charging and discharging mode are presented in Fig. 14. It is noted in Fig. 14 that lithium-ion batteries are deeper discharging than the lead-acid batteries while in charging mode the lead-acid batteries are faster charging than the lithium-ion batteries. The charging and discharging mode of the battery bank directly affects the charging time of the electric vehicle battery and also affects the power of the PV modules destined for the load, and this is what will be shown in the following sections.

The SOC of EV Battery with the Lithium-ion and the Lead-acid Batteries
From Fig. 15, it is clear that the EV battery is charging faster with the lithium-ion batteries. The charging time is important for charging the EV battery, so the choose the lithium-ion batteries as a back-up storage system for the stand-alone PV system is the best.

PV Performance with the Lithium-ion and the Lead-acid Batteries
From Fig. 16, it is evident that the PV modules provide the electric vehicle battery with greater charge power, in the presence of lead-acid batteries, in contrast, lithium-ion batteries help reduce the energy generated by the stand-alone PV system. For further clarification, Fig. 17 is a magnification of a section of Fig. 16 to demonstrate the difference between the power of the solar panels with both lithium-ion batteries and lead-acid batteries. 284 Design and Simulate an Off-Grid PV System with a Battery Bank for EV Charging

The Proposed Charging Station Cost Estimation
The cost estimation at this stage can present a powerful index of the proposed charging station's effectiveness and also determines if the decision of installing such a charging station practical or not. To complete the costing process, the design of the rest of the main charging station components must be performed, this is presented in the following section.

Determination the System Protection Rating
The PV modules in parallel need DC fuses to protect each module apart. DC fuse rating is calculated as follows: Which SF is a safety factor. The nearest standard value for 7.4 is10 . The system needs 180 DC fuses. The modules in parallel are divided into 18 groups, each group contains 10 modules with CB for each group and its rating is calculated as follows: = * 10 = 5.95 * 10 = 59.5 The closest standard value for 59.5 is 63 . The system needs the number 18 units of the circuit breaker rating 63A.
The main circuit breaker rating for total PV modules' protection is calculated as follows: The nearest standard value for 1071 is 1250 . The circuit breaker rating for boost converter protection is calculated as follows: = * = 1000 * 1.25 = 1250 Thus the system needs the number 2 units of the circuit breaker rating 1250 .
The buck converter current and the bidirectional converter current are the same (500 ), so the CB rating for buck converter protection and bidirectional converter protection is the same and is calculated as follows: = * = 500 * 1.25 = 625 Thus the system needs the number 2 units of the circuit breaker rating 625 A.

The Number of Chassis and the Cables Sizes
The chassis carries 4 panels, so the total number of the chassis is calculated as follows:

The Cables Sizes between the PV Modules and the Boost Converter
The wires between the PV modules in parallel and the boost converter must withstand the maximum current ( ) produced by these modules; the wire cross-sectional area is given by, Let the length of the wire ( − ) = 15m. Where ρ = resistivity of copper wire which was taken as 1.724 10 −8 Ωm. The maximum DC wiring voltage drop ( ) for the off-grid system is 4% of the system voltage ( − ) [34]. It means that, The length of the wire cross-sectional area 10 2 = 2* − * = 5400 .

The Cables Sizes between the Boost Converter and the Buck Converter
The wires between the boost converter and the buck converter must withstand the boost converter current; the system needs 6 boost converters, each one has 100kW and 200A rated. Let the length of the wire ( − ) = 5m. The wire cross-sectional area is given by Eq (12)  The length of the wire cross-sectional area 10 2 between the boost converter and the buck converter = 2* − * 6 = 60 . Thus, the total length of the wire cross-sectional area 10 2 = 5400 + 60 = 5460 .

The Cables Sizes between the Boost Converter and the Battery Bank
Let the length of the wire ( − ) = 5m, the wires between the boost converter and the battery bank must withstand the bidirectional converter current ( ); The wire cross-sectional area is given by Eq (13)

The Cables Sizes between the Buck Converter and the EV's Battery
The wires between the buck converter and the EV's battery must withstand the buck converter current; the system needs 3 buck converters, each one has 100kW and 100A rated. Let the length of the wire ( − ) = 3m. The wire cross-sectional area is given by Eq (14)

Determination of the Payback Period
Let the proposed charging station is located in a remote area where electricity tariff is US$ 0. 23

Conclusions
A charging station using the stand-alone PV system with an energy storage system to charge an EV battery has been successfully designed and simulated in this paper. The MPPT controller with an Incremental Conductance algorithm has been applied. A charge and discharge stages using the bidirectional converter of the battery energy storage was also successfully modeled and simulated. A logic controller for the bidirectional converter has also been designed and simulated. The results have proven the logic controller's effectiveness with the bidirectional controller. A comparison between the lithium-ion and the lead-acid batteries as a power storage system for the proposed charging station has been achieved. Using lithium-ion batteries is better than the lead-acid for the proposed charging station, in which the charging time of the EV battery is less. A proposed charging station cost estimation has been performed. Lastly, the model, in general, has been successfully simulated in Matlab, Simulink and satisfactory simulation results have also been achieved. Although the use of a bidirectional converter can adjust the charging and discharging of the battery pack and reduce the amount of battery usage; however, when the load current changes instantaneously, the battery needs to quickly provide energy to the load. Will the existence of the converter cause resistance to the speed of energy supply? This point needs to be clearly studied in future work. In this work the results were presented only in simulation; however, it is recommended in future work that experimental results should be added to compare between the simulation and the practice.