Design & Implementation of MPPT Algorithm for Battery Charging with Photovoltaic Panel Using FPGA Joydip Jana1, Konika Das Bhattacharya2, Hiranmay Saha3 Indian Institute of Engineering Science & Technology (IIEST), Shibpur Botanic Garden, Howrah, West Bengal, India 1
[email protected],
[email protected],
[email protected] Abstract— Solar photovoltaic system as renewable energy source is being significant due to several advantages such as the absence of fuel cost, no carbon footprints, and little maintenance. There are two primary constraints to the use of solar photovoltaic systems: low energy conversion efficiency and high installation cost. As the number of photovoltaic systems increase, so does the demand for high-efficient battery chargers for energy storage. Most systems in the market today use batteries without proper charge controllers. Direct battery charging with solar panels will result in premature battery failure or capacity loss, due to improper charge procedure. And due to the power wastage because of inefficient use of solar energy needs to install more panels for same power requirement. This is the reason of requirement of proper charge controller. In this work, an attempt has been made to develop a charge controller working on maximum power point tracking for the solar panel. This controller controls the output power of the solar panel and the charging current for the battery. It works on an embedded system (FPGA) and aims to use the solar panel efficiently for battery charging. Simulation and experimental results demonstrate the effectiveness and validity of the developed controller. Keywords—Photovoltaic module; MPPT; Buck converter; Battery; FPGA I.
converter used in this charge controller circuit is a buck converter. The MPPT algorithm in this charge controller is implemented by changing the dc-dc converter duty cycle. The developed controller uses the perturbation and observation (P&O) MPPT algorithm with the objective to extract maximum energy from solar panel for battery charging. The MPPT algorithm is incorporated in one of the stages of battery charging. II.
BATTERY CHARGE CONTROLLER DESIGN
The goals of the developed charge controller are: 1) to convert the solar power into electricity as much as possible under the varying weather condition (solar irradiance and temperature); 2) to charge the battery as fast as possible obeying the proper charge procedure to increase the battery life-cycle. As shown in figure1, the total battery charging system contains the following parts: 1) A photovoltaic panel, which converts the solar power into DC electricity; 2) A MPPT based charge controller, which controls the output power of the PV panel and the charging current for the battery; 3) A lead-acid battery
INTRODUCTION
The performance of batteries in solar photovoltaic (PV) applications is not as good as data given by battery manufacturers which is typically based on tests conducted at more favorable conditions [1]. Premature failure or capacity loss of batteries is therefore a big challenge to very high share of the running cost in such applications. The main aim of a battery charge controller in PV systems is to maintain the highest possible state-of-charge while preventing battery overcharge and avoid battery over-discharge. Solar panel is a non-linear type of dc power source. The power generation of solar panel depends on the solar irradiance and the operating temperature. To use the solar power efficiently and to reduce the installation cost, it is important that PV panel operates at its maximum power point (MPP) for a particular weather condition. This paper describes a lead acid battery charge controller which will maintain a proper charge procedure and also track MPP of the panel. The dc-dc 978-1-4799-6042-2/14/$31.00 ©2014 IEEE
Fig1:- Block diagram of the total battery charging system
Before starting to design the charge controller it is necessary to describe each and every parts of the battery charging system. It starts by describing the solar panel characteristics, describing various MPPT algorithms, describing the DC-DC buck converter needed to implement the MPPT algorithm and describing lead-acid battery charging characteristics. A. Solar Panel Characteristics The power generation of solar panel depends on the solar irradiance and the operating temperature.
Though fuzzy and neural network based algorithms are developing in the present days, the efficiency remains high in P&O method. [3]. In this work, P&O MPPT algorithm is implemented, due to its simplicity [4]. In the developed controller, this algorithm is operational from a FPGA based digital platform.
Fig2. Current-Voltage curve and Power-Voltage curve of a solar panel for different values of irradiance at a temperature of 25 ºC.
Figure 2 illustrates the current-voltage curve and powervoltage curve of a panel for different solar irradiance. It consists of two regions: one is the current source region, and the other is the voltage source region. The right side of the curve is voltage source region and the left side of the curve is current source region. The MPP of the panel is located at the knee of this curve. According to the maximum power transfer theory, the power delivered to the load is maximum when the load line intersects this point. From figure 2, it is found that power generation increases because of current on increasing of solar irradiance. B. MPPT Algorithms for PV array The most discussed MPPT algorithm is the perturbation and observation (P&O) algorithm, but this algorithm does not work in rapidly changing weather condition. In this algorithm the panel voltage is perturbed by changing the converter duty cycle and it is observed that whether the power is increased or not. From figure 2, it is easy to understand that decreasing voltage on the right side of the MPP increases power and increasing voltage on the left side of the MPP increases power. This is the main idea of P&O algorithm. But there is an inherit problem in this algorithm that the operating voltage oscillates around the MPP voltage. The simplest form of the algorithm uses a fixed step to increase or decrease voltage. The size of the step determines the size of the deviation while oscillating about the MPP and the speed of tracking. The P&O algorithm is shown in figure 3.
C. DC-DC Buck Converter A DC-DC buck converter is used to control the output power of the PV panel and to control the charging current for the battery. In this topology the input voltage (PV voltage) is higher than the output voltage (battery voltage). Figure 4 shows the schematic of the DC/DC converter implemented in hardware.
Fig4:- Buck converter
The buck converter is designed for 12 volt output as the battery voltage is 12 volt here. The input voltage and output load current of the converter are varying, but it is still true that ܸ௨௧ = Dܸ where ܸ௨௧ is the output voltage, ܸ is the input voltage of the converter and D is the duty cycle of the MOSFET switch, if the inductor value L ܮ௧ which is the condition that converter will never operate in discontinuous mode. With the active switch (MOSFET) on, D = ͳʹൗܸ , and ௗ
ܸ – ܸ௨௧ = Lௗ௧, ܸ - 12 = Lሺଵଶ
௱
ൗ ሻ்
(1)
Under one particular set of conditions, the inductor L = ܮ௧ will produce the familiar triangle wave with a minimum at zero and a maximum at 2݅ۃ௨௧ ۄ, where ݅ۃ௨௧ ۄis the average load current. Under this conditions, ܸ - 12 = ܮ௧
ଶۃೠ ۄ ቀଵଶൗ ቁ்
,
݅ۃ௨௧ = ۄ
ೠ ଵଶ
(2)
Where ܲ௨௧ is the output power. The critical inductance can be found in equation 3, ܮ௧ = (12T -
ଵସସ்
)
ଵଶ
ଶೠ
(3)
In this case, the worst-case combination is for the minimum output power and maximum input voltage. An inductor equal to 100ܮ௧ will produce ±1% current ripple under worst case conditions. The output capacitor value can be found in equation 4, ܥ௨௧ =
Fig3:- The P&O MPPT algorithm
் మ ሺଵିሻ ଼
೩ೇ ೇ
(4)
D. Lead-Acid Battery Charging The battery information is very important in designing battery charge controller [5]. Thus in this paper, the correct charge
curves are presented in figure 5 and different stages of charging are described. This work also explains that if the correct charge curves are respected in the charge periods, then the operation period of the lead acid battery will be longer. The charge time of a lead-acid battery is typically ten hours or more for higher capacity batteries.
IV.
SIMULATION AND EXPERIMENTAL RESULT
The MATLAB SIMULINK model of the buck converter is shown in figure 7.
Fig7:- MATLAB/SIMULINK model of buck converter
Fig5. Current and voltage curves of battery charge.
In bulk charging stage, the battery is charged with a constant current until its terminal voltage reaches topping voltage as shown in stage 1 in figure 5. The time taken for this stage should be set about five hours by adjusting the current for a C/10 charging rate. From this point, the rate of rise of the voltage slows down, with the consequent decrease in current. The time taken to reach from topping voltage to full voltage takes another five hours as indicated in stage 2. This action of topping the charge is essential for the well being of the battery. If omitted, the battery would eventually lose the ability to accept a full charge. The battery is supposed to be fully charged when the current drops to 3% of the rated current. Stage 3 is where the incoming ‘trickle’ charge merely compensates the charge lost during self-discharge. III.
PROPOSED ALGORITHM
The proposed algorithm is shown in figure 6. In this algorithm, only 1st and 2nd stages are implemented. The 3rd stage is omitted. When battery voltage ܸ is less than the topping voltage ܸ , then it is charged with constant current. In this stage MPPT algorithm is implemented. In 2nd stage of charging, if the current drops to 3% of the rated current, ܫ௦௧ௗ௬ , then the PWM is stopped and it waits for the drop of voltage below floating voltage for restart the operation.
Fig6:- The proposed algorithm
MPPT based charge controller is designed in MATLAB as shown in Figure 8, where the first block is the model of solar panel, the second block is for voltage and current sensing and for charge control operation and the third block is the model for buck converter.
Fig8.MATLAB/SIMULINK model of the proposed system
In order to validate the proposed controller, several simulations and experiments are carried out. Figure 9 shows the I-V characteristics of a solar module used for charging the battery and figure 10 shows the P-V characteristics of the same solar module got from the simulation. For experiment purpose, the same specification of solar module as shown in table 1 is generated in the software interface of the hardware photovoltaic simulator. Table1:- Specifications of the solar module used for experiment Parameter Value Open Circuit Voltage 20V Optimum Operating Voltage 17V Short Circuit Current 2.5A Optimum Operating Current 2.2A Power at STC 37W -0.34% Temperature Coefficient for ܸ /Ԩ -0.43% Temperature coefficient for Power/Ԩ
controller tries to maximize the watts output from the photovoltaic simulator by controlling the duty cycle to keep the photovoltaic simulator operating always at its Maximum Power Point.
Fig9:- IV characteristics of solar module simulated in MATLAB/SIMULINK Fig12:- Buck converter circuit with MOSFET driver
C. FPGA-The converter is controlled by a cyclone FPGA device. It calculates the solar watts generated by reading the voltage and current of the photovoltaic simulator through the Analog to Digital converter port. It also calculate the load side voltage and current in the same way and send corresponding control signal to the converter to increase, decrease or turn off the converter accordingly.
Fig10:- PV characteristics of solar module simulated in MATLAB/SIMULINK
A prototype of the charge controller has been built. Experiments have been conducted on the prototype. The following components are used in the prototype.
D. Voltage and Current Sensor-The current and voltage signals from the PV array are monitored by using the current and voltage sensing circuit. The sensing circuit that acts as transducer-converts the current reading to the voltage signals so that the controller can understand and use the information to process and hence performs the PWMs. A simple voltage divider network to sense the voltage and a shunt to sense the current is used as shown in figure 13.
A. Photovoltaic simulator- In this experiment, a solar module is created according to the specifications stated before in table 1 and a varying solar irradiance profile is generated in the software interface of the photovoltaic simulator as shown in the figure 11.
Fig13:- Current and voltage sensing circuit
E. Battery-A 12V/20AH lead-acid battery is used in this experiment set up.
Fig11:- Software interface of the photovoltaic simulator
B. High Frequency Buck Converter- A buck converter operating at high frequency (20 KHz) is designed as shown in figure 12 which is controlled by the PWM signal generated from the FPGA platform and also a MOSFET driver circuit is developed to drive the power MOSFET used in the converter as a switch. Duty cycle is changed by FPGA programming according to the temperature and solar radiation level. The
The voltage and current of solar panel and battery is read during the charging of the battery and in table 2 some of those recorded values are shown during the bulk charging stage of the battery. It is showing in table 2 that, the developed controller is operating at very near of the maximum power point at different solar irradiance and it is drawing almost 95% of the power it would draw at MPP.
Table2:- Readings taken during the bulk charging stage PV current PV voltage Power Power at Solar at operating radiation at operating at operating point of the point of the MPP point of (W/݉ଶ ) controller controller (Watt) the (Amp) (Volt) controller (Watt) 100 16.6 0.19 3.48 3.31 200 0.38 16.5 6.83 6.42 300 0.61 16.9 11.08 10.43 400 16.7 0.80 14.31 13.48 500 1.0 17.2 19.28 18.30 600 1.30 17.1 23.42 22.25 700 1.35 16.9 24.41 22.95 800 17.2 1.71 31.46 29.57 900 1.8 17.3 33.90 32.54 1000 2.0 17.3 37.36 35.49 In figure 14, solar radiation intensity vs. output power is shown (data are taken from table 2). It is seen in the software interface of the hardware photovoltaic simulator that using the prototype of the proposed controller almost 95% efficiency is achieved.
installation and maintenance cost of any solar battery charge controller by adding MPPT and proper charging algorithm. REFERENCES [1]
[2]
[3]
[4]
[5]
J. P. Dunlop et al., ''Batteries and Charge Control in Stand-Alone Photovoltaic Systems'', Sandia National Laboratories, Cocoa, FL. Jan. 15, 1997. K.H. Hussein, I. Muta, T. Hoshino and M. Osakada, “Maximum photovoltaic power tracking:an algorithm for rapidly changing atmospheric onditions,”IEEEploc.-Gener. Transmission and Distribution, Vol. 142, No. 1 , Jan. 1955. Kumaresh.V, Mridul Malhotra, Ramakrishna N and SaravanaPrabu.R, “Literature Review on Solar MPPT Systems”, Advance in Electronic and Electric Engineering, ISSN 2231-1297, Volume 4, Number 3 (2014), pp. 285-296 Knop H. (1999). Analysis, Simulation, and Evaluation Of Maximum Power Point Tracking (MPPT) Methods for a Solar Powered Vehicle, Mater of Science Thesis in Electrical and Computer Engineering, Portland State University, 1999 S. Harrington, J. Dunlop, Battery charge controller characteristics in photovoltaic systems, Aerospace and Electronic Systems Magazine, IEEE,Volume:7,Issue:8,p:15–21,1992 Joydip Jana was born in Howrah, West Bengal on 2nd May, 1987. He received his B.Tech and M.tech degree in Electronics & Communication Engineering in 2009 and 2012, respectively. Currently he is pursuing PhD in Indian Institute of Engineering Science & Technology (formerly Bengal Engineering & Science University), Shibpur in Solar Photovoltaic Systems. Prof. Konika Das Bhattacharya is a professor in Electrical Engineering Department in Indian Institute of Engineering Science & Technology, Shibpur. Her areas of interests are energy saving devices and smart grid, recent trend in power systems and protections etc.
Fig14:- solar radiation vs. output power curve
V.
CONCLUSION
In this paper it is shown that, solar power can be efficiently used using MPPT and as a consequence the initial panel installation cost can be reduced and also respecting correct charge procedure the battery life can be increased and in this way battery maintenance cost can be reduced. MPPT algorithms are simple, but there should be a clear idea to the designer about the dc-dc converter needed to implement these algorithms and also for design of proper charge controller the battery curve should be known. It is possible to optimize the
Prof. Hiranmay Saha received his M. Tech degree in Radiophysics and Electronics from University of Calcutta in 1967 and Ph. D degree in Solar Cells and Systems from Jadavpur University in 1977. He was former Chairman of Solar Energy Division (Eastern Region), the Ministry of New and Renewable Energy, Government of India and Advisor of WBREDA (Dept. of Power and N.E.S., Govt. of West Bengal). Currently he is the Professor-in-charge of 'the Centre of Excellence for Green Energy and Sensor Systems', Indian Institute of Engineering Science & Technology, WB. He is Fellow, IEE (UK), IETE and Member IEEE, NCS (WB), Dept. of Science & Technology (WB), WBREDA. Dr. Saha is the chairman of the Board of Directors of Agni Power & Electronics Pvt. Ltd.