P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

22

Photovoltaic Energy Conversion System for Water Pumping Application P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy ABSTRACT: Solar power is the most abundant renewable energy source. By using PV cells we can convert solar radiation into electrical energy to meet the to increasing demand. In this paper we present control of electrical power supplied by Photovoltaic cell fed to single phase induction motor which can be used for water pumping applications. . The overall configuration of this system gives photovoltaic system is given, dynamic models for the Z-source inverter, single phase induction motor and neural network based Maximum power point tracking. Simulation results are presented to show the overall system performance.

Keywords: photovoltaic (PV) module, Impedance, Source Inverter(Z-SI).neural network, Single Phase Induction Motor I. INTRODUCTION Recently, as the fossil fuel exhaustion and environmental pollution are aggravated, the concer of the development of alternative energy systems, which are renewable and pollution free, has been increased continuously. Among them the photovoltaic (PV) power generation systems standout as an important solution because they produce electric power without inducing environmental pollution, by directly transforming solar irradiation into electricity. The main drawbacks of PV systems are high fabrication cost and low energyconversion efficiency, which are partly caused by their nonlinear and temperature dependent V–I and P–I characteristics. To overcome these drawbacks, three essential approaches can be followed:  Improving manufacturing processes of solar arrays: many research efforts have been performed with respect to materials and manufacturing of PV arrays.  Controlling the insolation input to PV arrays: the input solar energy is maximized using sun-tracking solar collectors.  Utilization of output electric power of solar arrays: the main reasons for the low electrical efficiency are the nonlinear variations of output voltage and current with solar radiation levels, operating temperature, and load current. To overcome these problems, the maximum power operating point of the PV system (at a given condition) is tracked using online or offline algorithms and the system operating point is forced toward this optimal condition.

Fig 1. Block Diagram of Photovoltaic Energy Conversion System For Water Pumping Application

Many MPPT techniques have been proposed, analyzed, and implemented. They can be categorized as: A. Look-up table method -- The nonlinear and timevarying nature of pv cells and their great dependency on radiation and temperature levels as well as degradation (aging, dirt) effects, make it difficult to record and store all possible system conditions. B. Perturbation and observation (P&O) method-Measured cell characteristics (current, power) are employed along with an online search algorithm to compute the corresponding maximum power point independent of insolation, temperature, or degradation levels. C. Computational method -- The nonlinear V–I characteristics of PV panel is modeled using mathematical equations or numerical approximations. Based on the modeled V–I characteristics, the corresponding maximum power points are computed for different load conditions as a function of cell open-circuit voltages or cell short-circuit currents. II. MODELLING OF PV MODULE The most commonly used one – diode equivalent circuit as Since the shunt resistance Rsh is neglected. This simplified circuit is modeling of a PV-cell.

model for PV-cell is shown in figure (2). large, it is normally used in this paper for

P.P. Kiran Kumar Reddy is a P.G Student and J. Nagarjuna Reddy is working as Asst.Professor, both are from Dept. EEE, RGMCET, Nandyal, Emails: [email protected], [email protected]

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

23

Tcref= Reference temperature (25˚C issued in this study); µIsc= temperature coefficient of the short-circuit current (A/˚C). From the above equation for light current it can be observed that IL is a function of both temperature and irradiance. Both ILrefand µIsccan be obtained from manufacturer data sheet. B. Saturation Current (IO) (a)

(b)

Fig.2. One-diode equivalent circuit model for a PV cell. (a) Five parameters model (b) Simplified four parameters model

The non-linear of Vpv-Ipv and P-V curves are correspondingly drawn as shown below:

I =I

exp

∝

1− (4)

Where Ioref= saturation current at the reference condition (A); egap= band gap of the material1. 17eV for Simaterials); Ns = Number of cell sin series of a PV module; q = Charge of an electron (1.60217733×10-19 C); αref= The value of α at reference condition. Ioref can be calculated as: I

Fig.3. VI & P-V characteristics of a PV cell

=I

exp −

V ∝

(5)

Where Vocref= the open circuit voltage of the PV module at reference condition (V). C. Calculation of α

From figure (1.b) the relation between the output Vpv and the output current Ipv can be expressed as: I I

= I −I

= I −I

exp

(1) −1

∝

∝=

∝

(6)

The value of αref can be calculated as:

(2)

∝

Where IL= Light current; Io= Saturation current; Rs= Series Resistance; α = Thermal voltage timing completion factor. The above four parameters are need to be determined to obtain the I-V characteristics of PVmodule. Thus, this model can be termed as Fourparameter model. The equations for determining the four parameters are given below:

(7)

Where Vmpref= maximum power point voltage at the reference condition (V); Impref =maximum power point current at the reference condition (A); Iscref= short circuit current at the reference condition (A).

A. Light Current (IL) D. Series Resistance (RS) I =

I

+μ

(T − T

)

(3)

Where G= irradiance (W/m2); Gref = Referenceiradiance(1000W/m2 issued in this study); ILref= light current at the reference condition (1000W/m2 and 25˚C); Tc= PV cell temperature (˚C);

Some manufacturers provide the value of Rs.If not provided,the following equation can be used to estimate its value: ∝

=

(8)

RS is taken as a constant in the model of this study.

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

E. Thermal Model of PV From equations (1) to (7), it can be noted that the temperature plays an important role in the PV performance. Therefore, it is necessary to have a thermal model for a PV cell / module. In this study, a lumped thermal model is developed for the PV module. The temperature of the PV module varies with surrounding temperature, irradiance, and its output current and voltage, and can be written as: ( − ) = − − (9) = the overall heat capacity per unit area of the PV cell CPV / module [J/ (˚C-m2)]; Kinpv= Transmittance – absorption product of PV cells; Kloss= overall heat loss coefficient [W/ (˚C-m2)]; Ta = ambient temperature (˚C); A = effective area of the PV cell / module (m2) Maximum Power Point Tracking Of Pv Cell Using Neural Networks The block diagram for identifying the optimal operating point using neural networks is shown in fig(4).

24

1 … (10) − ( ) 1+ The term Ii(k) is the input signal given to the neuron I at the Kth sampling. The input Ii (k) is given by the weighted sum from the previous nodes as follows: ( )=

( )=

( ) ( ) … (11)

In the above equation, Wij is the connection weight from the neuron j to the neuron i and Oj (k) is the output from neuron j.The process of determining connection weights is referred to as training process[10]-[12]. In the training process, we need a set of input-output patterns for the neural network. The computations are performed offline during the training process. With the training patterns, the connection weights Wij recursively until the best fit is achieved for the input-output patterns in the training data. A commonly used approach is the generalized delta rule, where the sum of the squared error described below is minimized during the training process. ( )− ( )

=

… (12)

Where N is the total number of training patterns. T(k) is the target output from the output node and O(k) is the computed one. For all the training patterns, the error function E is evaluated, and the connection weights are updated to minimize the error function E. III OPERATING PRINCIPLE OF ZSI

Fig.4 Block Diagram for the identification of optimal operating point The configuration of 3-layer feed-forward neural network is shown in fig (5). The network has 3 layers with 3 neurons in input, 4 neurons in hidden, and 1neuron in output layers [8].

Fig.5. Configuration of a Neural Network The neuron in the input layer gets input solar irradiance (G) and cell temperature (Tc). These signals are directly passed to the neurons in the hidden layer. The neuron in the output layer provides the identified maximum Imp. For each neuron in the hidden and the output layer, the output Oi(k) is given as follows:

The operation of a Z-source inverter can be demonstrated with the help of a simplified circuit diagram, depicted in Fig.4 wherein a Z-source network is interposed between the dc-link and the semiconductor switches of the inverter. The Z-source network employs two inductors (L1 and L2) and two capacitors (C1 and C2) that are connected in ‘X’ shape .As compared to conventional VSI and CSI, a Z-source inverter has three modes of operation – active state, zero state and an additional shoot-through state. This additional shootthrough state provides the boost feature to the Z-source inverter. Active state occurs when any two switches from the upper and lower phase legs (but not from the same phase leg at a time) become ON. During zero state, all the two switches either from the upper leg or all from lower leg remain ON. Shoot through state occurs when any two switches of the same phase leg become ON. Since L1 and L2 are identical so as C1 and C2, it is obvious that Zsource network is symmetrical and hence



=

=

(13)

=

=

(14)

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

25

Fig. 7 shows the equivalent circuit during shootthrough mode of operation. During this mode, the effective dc-link voltage becomes zero since both the switches of same phase leg remains ON. At the same time, the diode becomes reversed biased. Hence L1 and C1 form a parallel combination with each other and similarly L2 and C2 form a parallel combination with each other. Considering Fig. 7 it can be noted that during this mode, the inductor voltage and effective dc-link voltage can be written as

Fig .6.Equivalent circuit of ZSI viewed from equivalent dclink

=

(17)

=0

(18)

Combining (13) and (15), the average voltage across the Inductor over a switching period is determined. As the average voltage across an inductor at steady state over a switching period is zero, this leads to

(

=

)

(19)

=

Fig.7. Circuit configuration of ZSI During non-shoot

(20)

Wherein TS is switching period, T1 is the time duration for the occurrence of any one from the eight non shoot-through states and T0 is the time interval for shootthrough state. Combining (16) and (18) the average value of effective dc-link voltage, Vi over a switching period, can be obtained as (

through state

)

( )∗

(21)

Shoot-through duty ratio can be defined as =

(22)

From equation (21) VC can be expressed in terms of d as .

Fig 8. Circuit configuration of ZSI During shoot through state

Fig. 6 represents ZSI during active state as viewed from effective dc-link side, wherein Viis the effective dc-link voltage of the inverter. So, the circuit configuration during zero state remains almost the same as that of Fig. 2, the only difference being that Ii assumes zero value during this state. Therefore zero state and active state are termed as non shoot-through states. During non shoot-through mode of operation, inductor voltage and the effective dc-link voltage can be expressed as = =

− −

(15) =2

−

(16)

=

(23)

Peak value of the effective dc-link voltage can be expressed as =2

−

=

.

=

.

(24)

IV. INDUCTION MOTOR MODEL The equivalent circuit of the induction motor based on double revolving field theory is shown in where ‘a’ is the turns ratio of the auxiliary to main winding Rlm, Xlm are the resistance and reactance of the main winding, Rla, Xla are the resistance and reactance of the auxiliary winding, Rc, Xc are the equivalent series resistance and reactance of the capacitor, Rf, Xf are the forward equivalent series resistance and leakage reactance of the rotor referred to the main winding, Rb, Xb are the backward equivalent series resistance and leakage

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

26

reactance of the rotor referred to the main winding, Im, Ia, I are the main, auxiliary and motor currents, respectively, Efm, Ebm the self-induced voltages in auxiliary winding by its forward and backward fluxes,respectively, Efa/a, Eba /a are the mutually induced voltages in main winding by its forward and backward fluxes of the auxiliary winding, respectively.

Where ωs is the synchronous speed in rad/sec. The mechanical power developed is given by = (1 − )

(36)

WhereS is the per unit slip. The output power is =

Fig. 9.Equivalent circuit of Induction Motor

From fig. 1, the following equations are written =

+ =(

+ +

)

− +

+



+

+



=

+

=

=

=

=

=

(

(27) )

+

(28)

+ (

+

(29) )

(30)

Substituting from equations (27)-(30) into equations (25) and (26) gives = =

+

+

−

− +(

+

− +

(31) +

) (32)

The solution of equations (31) and (32) gives the main and auxiliary windingcurrents under any operating conditions. Hence, the total motor current is obtained as = + (33) The net amount of power transferred across the air gap is obtained as =( +

) − ) sin( −

+2 )

Where, Prot is the rotational losses. The voltage equations (28) and (29) constitute the steady state mode of the single-phase induction motor. The solution of these equations under any operating conditions gives the main and auxiliary winding currents. Hence, all the performance characteristics of the motor at particular load point can be calculated .It should be noted that particular load point means a given value for the applied voltage and motor speed.

V. SIMULATION RESULTS

−

where, =

(37)

(25)

(26)

=

−

+ (34)

Where θm and θa are the phase angles of the main and auxiliary winding currents,respectively.The electro mechanical torque developed is = / (35)

Based on the mathematical equations discussed before, a dynamic model for a PV module consisting of 153 cells in series has been developed using MATLAB/Simulink. The input quantities (solar irradiance G and the ambient temperature Ta) together with manufacturer data are used to calculate the four parameters. Then, based on equation (1), the output voltage is obtained numerically. The thermal model is used to estimate the PV cell temperature. The two output quantities (PV output voltage Vpv and the PV cell temperature Tc), and the load current Ipv, are fed back to participate in the calculations. The model parameters used in the simulation are given in Table I. Table I THE PV MODEL PARAMETERS ISCref(ILref)

2.664A

αref

5.472

RS

1.324Ω

VOCref

87.72V

VMPref

70.731V

IMPref

2.448A

Gref

1000w/m2

Tcref

25°c

Cpv

5*104J/(0c-m2)

A

1.5m2

Kinpv

0.9

Kloss

30W/(0c-m2)

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

27

A. Model Performance The model Ipv-Vpv characteristic curves under different irradiances are given in Figure (10) at 25˚C. It is noted from the figure that the higher is the irradiance, the larger are the short-circuit current (Isc) and the opencircuit voltage (Voc). And, obviously, the larger will be the maximum power (P), shown in Figure (11). B. Training of a Neural Network The training of a neural network consists of solar irradiance and cell temperature as the input patterns. The target pattern is given by measured Imp for training the neural network. The Imp is calculated for different values of irradiance and cell temperature w.r.t above modeled PV module. This calculated Imp values are given as a training data to the neural network. Fig(15), shows the convergence of error during training process. During the training process, the convergence error is taken as 0.01.

Fig.12. P-Vpv characteristics for constant Tc and Varying G

C. Simple boost PWM (SB-PWM) Past research implemented this method in single phase Zsource converter. This work used the same concept as of Figure 10 and implemented it in three-phase ZSI. For control of shoot-through duty ratio, two straight lines equal to, or greater than, the peak value of the reference signal, are used. When the carrier signal is greater than the upper straight line or lower than the bottom straight line, the circuit turns into shoot-through state ; else, it operates like a traditional carrier-based PWM.

Fig 13. The Output of VI characteristic

Fig 14. The Output of PV characteristic Fig.10 Simple boost PWM control A – upper straight line B – carrier signal C – reference signals D – lower straight line

200 150

V o l t a g e ( v o l ts )

100 50 0 -50

-100 -150 -200 0

0.05

0.1

0.15

0.2

0.25 Time (s)

0.3

0.35

0.4

0.45

0.5

Fig 15. Output Voltage waveform of ZSI

Fig.11 Vpv-Ipv characteristics for constant Tc and Varying G

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

28

phase inverter and single phase induction motor. We can conclude that this work will be a contribution to the analysis of the photovoltaic pumping system with regards to the results of simulation of the model. This paper also investigates a photovoltaic powered AC motor drive for water pumping application and rural electricity. A number of experimental PV powered DC motor drives for water pumping are already in use, however such schemes find limited applications due to high cost and maintenance problems commonly associated with DC machines. The experimental setup and its analysis may be considered as the future work. REFERENCES Fig.16 Generation of pulses for S1&S4(Switches)

Fig.17 Convergence error during training process.

Fig 18.Simulation outputs for speed, armature current and torque

CONCLUSION The work simulated in this paper examines the possibility of utilizing a PV cell to supply a single-phase induction motor through a single-phase ZSI. The system explained here is a PV system for water pumping, using a single

[1]. S.Yuvarajan, Dachuan Yu and ShanguangXu,“A novel power converter for photo voltaicapplications” Elsevier Journal of Power Sources, June-2004. [2]. M. Buresch: Photovoltaic Energy Systems Design and Installation,McGraw-Hill, New York, 1983. [3]. K. Benlarbi, L. Mokrani, M.S. Nait-Said: A fuzzy global efficiency optimization of a photovoltaic water pumping system, Solar Energy, 77 (2004) 203–216. [4]. F. Z. Peng, "Z-source inverter", IEEE Transactions on IndustryApplications, vol. 39, pp. 504-510, MarApr 2003. [5]. Rabi BJ and Arumugam R, ‘‘Harmonics study and comparison of zsourceinverter with traditional inverters’’, American Journal of appliedscience, vol2, no.10, pp.1418-1426. [6]. Bindeshwar Singh, S. P. Singh, J. Singh, and MohdNaim, “Performanceevaluation of three phase induction motor drive fed from z-sourceinverter”, International Journal on Computer Science and Engineerin(IJCSE). [7]. AtulKushwaha, Mohd. Arif Khan, AtifIqbal and Zakir Husain, “ZSourceInverter Simulation and Harmonic Study”, Global Journal ofAdvanced Engineering Technologies-Vol1-Issue1-2012. [8]. B.Y. Husodo, M. Anwari, and S.M. Ayob, “Analysis and Simulations ofZ-Source Inverter Control Methods”, IEEE Transactions on IndustryApplications, vol. 42, pp. 770 – 778, MayJun 2006. [9]. NejibHamrouni, MoncefJraidi, AdneneCherif Theoretical and experimentalanalysis of the behavior of a photovoltaic pumping system Science direct, SolarEnergy 83 (2009); 1335–1344 . [10]. Odeh I, Yohanis YG, Norton B.Influence of pumping head, insolation and PVarray size on PVwater pumping system performance .Solar Energy 80 (2006);51–64. [11]. A Betka, A. Moussi Performance optimization of a photovoltaic inductionmotor pumping system Science direct, Renewable Energy 29 (2004) 2167– 2181. [12]. Nang Saw YuzanaKyaing, WunnaSwe Design Considerations of PV WaterPumping and Rural Electricity System (2011) in Lower Myanmar. WorldAcademy of Science, Engineering and Technology 75, 2011.

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.

P.P. Kiran Kumar Reddy and J. Nagarjuna Reddy

29

J.Nagarjuna Reddy was born in 1985 in India. He received the B.Tech from Dr.Paul Raj Engg. College in 2005 and Post graduated from Jawaharlal Nehru Technological University (JNTU), Hyderabad, In 2006-2008. He is currently working as a assistant professor in the department of electrical and electronics engineering in RGM college of engineering and technology, Nandyal, Andhra Pradesh, India. He has Two years of teaching experience. His main areas of research include Electrical Drives & Renewable energy source.(E-mail: [email protected]). P.P. Kiran Kumar Reddy was bornin Kurnool India. He received the B. Tech (Electrical and Electronics Engineering) degree from the Jawaharlal Nehru Technological University, Hyderabad in 2011 and perusing the M.Tech (Power Electronics) from Jawaharlal Nehru Technological University, Anantapur. His area of interest in the field of power electronic converters andElectricDrives.(E-mail: [email protected]).

International Journal of Emerging Trends in Electrical and Electronics (IJETEE – ISSN: 2320-9569)

Vol. 10, Issue. 2, Mar-2014.