2009 Third Asia International Conference on Modelling & Simulation

Modeling and Simulation of Photovoltaic Water Pumping System

Taufik, Akihiro Oi Makbul Anwari Mohammad Taufik

Electrical Engineering Department Faculty of Electrical Engineering Department of Physics
Cal Poly State University Universiti Teknologi Malaysia Universitas Padjadjaran
San Luis Obispo, CA 93407 USA 81310 UTM Skudai, Malaysia Bandung 40132, Indonesia
taufik@calpoly.edu makbul@ieee.org mhtaufik@yahoo.com

Abstract—This paper presents modeling of a simple but

efficient photovoltaic water pumping system. The system
employs maximum power point tracker (MPPT). Two
MPPT algorithms, subsystems and control methods will be
modeled and simulated using Matlab and Simulink with
actual irradiance data. Simulation results will be presented
in terms of performance parameters such as total energy
produced and total volume of water pumped per day. The
results further validate that MPPT can significantly improve
efficiency and performance of PV water pumping system
compared to the system without MPPT.
Figure 1. Block diagram of Proposed System
Keywords—Water Pumping System, Renewable Energy
Sources, Photovoltaic System Maximum power point tracker (MPPT) is now prevalent
in grid-tied PV power systems and is becoming more
popular in stand-alone systems. Analog controllers have
I. INTRODUCTION traditionally performed control of MPPT. However, the use
of digital controllers is rapidly increasing because of their
Water resources are essential for satisfying human advantages over analog controllers. First, digital controllers
needs, protecting health, and ensuring food production, are capable of implementing advanced algorithm with
energy and the restoration of ecosystems, as well as for relative ease. It is far easier to code the equation, x = y × z,
social and economic development and for sustainable than to design an analog circuit to do the same [3]. Digital
development [1]. However, according to UN World controllers are immune to time and temperature drifts
Water Development Report in 2003, it has been estimated because they work in discrete, outside the linear operation.
that two billion people are affected by water shortages in As a result, they offer long-term stability. They are also
over forty countries, and 1.1 billion do not have sufficient insensitive to component tolerances since they implement
drinking water [2]. There is a great and urgent need to algorithm in software, where gains and parameters are
supply environmentally sound technology for the consistent and reproducible [3]. They allow reduction of
provision of drinking water. Remote water pumping parts count since they can handle various tasks in a single
systems are a key component in meeting this need. It will chip. Therefore, a method of digital control was selected for
also be the first stage of the purification and desalination the MPPT in the proposed system. A model of MPPT is
plants to produce potable water. In this paper, a simple developed on the premise that implementation in a
but efficient photovoltaic water pumping system is microcontroller or a DSP will be performed, and its
presented. It provides theoretical studies and modeling of algorithm is readily transferable.
photovoltaics (PV) and comparative study on the use of Two types of pumps are commonly used for PV water
maximum power point tracker (MPPT). pumping applications: positive displacement and
centrifugal [4]. Positive displacement types are used in
II. THE PROPOSED SYSTEM
low-volume pumps [5] and cost-effective. Centrifugal
The water pumping system is a stand-alone 150 W pumps have relatively high efficiency [4] and are capable of
system without backup batteries as shown in Figure 1. pumping a high volume of water [5]. However, the cost and
The system consists of a single PV module, a maximum complexity of these systems will be significantly higher. In
power point tracker (MPPT), and a DC water pump. general, DC motors are preferred because they are highly
efficient and can be directly coupled with a PV module.
Brushed types are less expensive and more common
although brushes need to be replaced periodically [4]. The

978-0-7695-3648-4/09 $25.00 © 2009 IEEE 497

DOI 10.1109/AMS.2009.85
water pump selected was a submersible solar pump which where: Go is the nominal value of irradiance, which is
is a diaphragm-type positive displacement pump equipped normally 1KW/m2. The reverse saturation current of diode
with a brushed permanent magnet DC motor and designed (Io) at the reference temperature (Tref) is given below with
for use in stand-alone water delivery systems. It operates the diode ideality factor added:
with a low voltage (12~30V DC), and its power
requirement is as little as 35W [5]. I
Io = sc
qVoc / nkT
(4)
III. MODELING OF PV MODULE (e − 1)
The PV module chosen is a 72 multi-crystalline silicon
solar cells in series able to provide 150W of maximum The reverse saturation current (Io) is temperature dependant
power [1]. The study done in [6] shows the electric model and the Io at a given temperature (T) is calculated by the
of the PV with moderate complexity, shown in Figure 2, following equation:
gives very accurate results. It consists of a current source
3
(Isc), a diode (D), and a series resistance (Rs). The effect − q⋅ E g ⎛ 1 1 ⎞
⎛ T ⎞n ⎜ −
n⋅k ⎜⎝ T Tref
⎟
⎟
of parallel resistance (Rp) is very small in a single I o |T = I o |Tref ⋅⎜ ⎟ ⋅e ⎠ (5)
module, thus it is not included. To make a better model, it ⎜T ⎟
⎝ ref ⎠
must also include temperature effects on the short-circuit
current (Isc) and the reverse saturation current of diode The diode ideality factor (n) is unknown and must be
(Io). It uses a single diode with the diode ideality factor estimated. It takes a value between one and two. However,
(n) set to achieve the best I-V curve match. the value of n=1 (for ideal diode) is used until the more
Rs accurate value is estimated later by curve fitting. The effect
of varying the ideality factor is simulated in Matlab as
+ shown in Figure 3. The series resistance (Rs) of the PV
Isc module has a large impact on the slope of the I-V curve near
D Load
the open-circuit voltage (Voc), as shown in Figure 4. Hence,
the value of Rs is calculated by evaluating the slope of the I-
-
V curve at the Voc. The equation for Rs was derived and
yields:
Figure 2. Equivalent circuit used in the MATLAB simulations
dI nkT q (6)
Rs = − − ⎛ V + I ⋅ Rs ⎞
The following equation describes current-voltage dV q⎜ ⎟
relationship of the PV cell [6]. Io ⋅e ⎝ nkT ⎠

For open circuit voltage, V=Voc (also I=0):

⎡ q ⎛⎜ V +nkT
I ⋅ Rs ⎞
⎟ ⎤
I = I sc − I o ⎢e ⎝ ⎠
− 1⎥ (1)
⎢⎣ ⎥⎦ dV nkT q (7)
Rs = − − qVoc
dI Voc
Io ⋅ e nkT
where: I is the cell current = the module current, V is the
cell voltage = {module voltage} ÷ {# of cells in series}, T 5

is the cell temperature in Kelvin. The short-circuit current 4.5

n=1

(Isc) was first calculated at a given cell temperature (T):

4 n=2

[
I sc |T = I sc |Tref ⋅ 1 + a (T − Tref )] (2) 3.5
Module Current (A)

where: Isc at Tref is given in the datasheet (measured under 2.5

irradiance of 1000 W/m2), Tref is the reference temperature 2

of PV cell in Kelvin (K), usually 298K (25oC), and a is 1.5

the temperature coefficient of Isc in percent change per
1
degree temperature also given in the datasheet. The short-
circuit current (Isc) is proportional to the amount of 0.5

irradiance. Isc at a given irradiance (G) is: 0

0 5 10 15 20 25 30 35 40 45
Module Voltage (V)

⎛G ⎞ (3)
I sc |G = ⎜⎜ ⎟⎟ I sc |Go Figure 3. Effect of diode ideally factors
⎝ Go ⎠

498
 5

4.5

3.5 Rs=0
Module Current (A)

2.5
Rs=5 mOhm
2
Rs=10 mOhm

1.5 Rs=15 mOhm

0.5

0
0 5 10 15 20 25 30 35 40 45
Module Voltage (V)

Figure 4. Effect of series resistances

Figure 6. Simulated I-V curve of PV module
Finally, the equation of I-V characteristics is solved. It
is, however, complex because the solution of current is
recursive by inclusion of a series resistance in the model.
The Newton’s method was chosen for rapid convergence
of the answer as shown below:

⎡ q ⎛⎜ V + I n ⋅ Rs ⎞⎟ ⎤
I sc − I n − I o ⎢e ⎝ nkT ⎠ − 1⎥
I n +1 = In − ⎣⎢ ⎥⎦ (8)
⎛ V + I n ⋅ Rs ⎞
⎛ q ⋅ R s ⎞ q ⎜⎝ nkT ⎟⎠
−1− Io ⎜ ⎟e
⎝ nkT ⎠

Figure 5 shows the I-V characteristics of various module

temperatures simulated by MATLAB model and provides
good correspondence to data from the manufacturer’s
Figure 7. I-V and P-V relationships of PV module
datasheet. Figure 6 depicts a point on the I-V curve which
yields the maximum power. This point is called a
maximum power point (MPP), at which the module IV. DC MOTOR MODEL
operates with the maximum efficiency and produces the Many PV water pumping systems employ DC motors
maximum output power. It is possible to visualize the
because they could be directly coupled with PV arrays and
location of the MPP by fitting the largest possible
make a system very simple. Among different types of DC
rectangle inside of the I-V curve as illustrated in Figure 7,
motors, a permanent magnet DC (PMDC) motor is
and its area equal to the output power which is the
preferred in PV systems because it can provide higher
product of voltage and current. It reveals that the amount starting torque. Figure 8 shows an electrical model of a
of power produced by the PV module varies greatly PMDC motor. When the motor is turning, it produces Back
depending on its operating condition.
EMF (E) which is a voltage proportional to the angular
speed of the rotor. From the equivalent circuit, the DC
voltage equation for the armature circuit is:

V = I ⋅ Ra + K ⋅ ω (9)

O0C

Figure 5. I-V curves for various temperatures Figure 8. Electrical model of permanent magnet DC motor

499
Figure 9 shows a major problem with a direct coupled V. SIMULATION RESULTS
PV-motor setup in efficiency because of mismatching of The location of the MPP in the I–V plane is not known
operating points. For this example the water pumping beforehand and always changes dynamically depending on
system would not start operating until irradiance reaches irradiance and temperature. Therefore, the MPP needs to be
at 400W/m2. Once it starts to run, it requires as little as located by tracking algorithm. There are a number of
200W/m2 of irradiance to maintain the minimum methods that have been proposed. Among different
operation. It means that it cannot utilize a fair amount of algorithms are the Perturb & Observe and Incremental
insolation in the morning. Conductance (incCond) discussed here.
On a sunny day, the irradiance level changes gradually
since there is no influence of cloud. MPP tracking is
supposed to be easy. As shown in Figure 11, both
algorithms locate and maintain the PV operating point very
close to the MPPs without much difference in their
performance.
On a cloudy day, the irradiance level changes rapidly
because of passing clouds. MPP tracking is supposed to be
challenging. Figure 12 shows the trace of PV operating
points for (a) P&O algorithm and (b) incCond algorithm.
For both algorithms, the deviations of operating points from
the MPPs are obvious when compared to the results of a
sunny day. A close inspection of Figure 12 reveals that the
P&O algorithm has slightly larger deviations overall and
some erratic behaviors (such as the large deviation pointed
Figure 9. I-V curves with varying irradiance and a DC motor
by the red arrow). The simulation results show the
efficiency of 99.3% for the P&O algorithm and 99.4% for
Also, when the motor is operated under the locked
the incCond algorithm. The experimental results show
condition for long time, it may result in shortening of the
96.5% & 97.0%, respectively, for a partly cloudy day.
life of the motor due to input electrical energy converted
to heat rather than to mechanical output [5]. To overcome
this problem, a MPPT called a linear current booster
(LCB) may be used. The MPPT maintains the input
voltage and current of LCB at the MPP of PV module.
As shown in Figure 10, the power produced at the
MPP is relatively low-current and high-voltage which is
opposite of those required by the pump motor. The LCB
shifts this relationship around and convert into high-
current and low-voltage power which satisfies the pump
motor characteristics. For example, as shown in Figure
11, the LCB could start the pump motor as little as
50W/m2 of irradiance.

Figure 10. PV I-V curves with constant power lines (dotted) and a DC Figure 11. MPP tracking on a sunny day
motor I-V curve

500
 (a) PV Power vs. Voltage

Module Power (W)

Module Voltage (V)
(b) PV Current vs. Voltage

Module Current (A)

Module Voltage (V)

Figure 12. MPP tracking on a cloudy day
Figure 14. MPPT simulations with the DC pump motor load (20 to
1000W/m2, 25oC)
The permanent magnet DC motor with a constant field
was modeled and simulated with SIMULINK as
Figure 15 shows the I-V curve produced by the
illustrated in Figure 13. The parameters of DC machine
SIMULINK simulation. The result was then compared with
that correspond to the actual pump motor are unknown,
the direct-coupled PV water pumping system without
thus they are chosen by modification of the default values MPPT. The results as depicted in Figure 15 showed that the
and estimation from other references [10] [11]. Figure 14 PV water pumping system without MPPT has poor
(a) and (b) show that the trace of operating point staying
efficiency because of mismatching between the PV module
close to the MPPs throughout the simulation.
and the DC pump motor load. On the other hand, it shows
that the system with MPPT can utilize more than 99% of
PV capacity. More specifically, an efficiency increase of
35% was achieved compared to the system without MPPT.
Figure 15 further emphasizes the severe disadvantage of
the direct-coupled PV water pumping system because the
pump stays idle for nearly two more hours in the morning
while the same system with MPPT is already pumping
water. Similarly, it goes idle nearly two hours earlier than
the system with MPPT in the afternoon. The flow rate of
water is also lower throughout the operating period. The
total volume of water pumped for the 12-hour period is also
calculated for both systems. The results show that MPPT
offers significant performance improvement. It enables to
pump up to 87% more water than the system without
Figure 13. Simulink model for DC Motor MPPT.

501
 [7]. Hussein, K. H., I. Muta, T. Hoshino, & M. Osakada “Maximum
Photovoltaic Power Tracking: an Algorithm for Rapidly Changing
Atmospheric Conditions” IEE Proceedings – Generation,
Transmission and Distribution – v. 142 January 1995, page 59-64
[8]. Hohm, D. P. & M. E. Ropp “Comparative Study of Maximum Power
Point Tracking Algorithms” Progress in Photovoltaics: Research
and Applications November 2002, page 47-62
[9]. MathWorks Inc. Embedded Target for the TI TMS320C2000™ DSP
Platform For Use with Real-Time Workshop® User’s Guide Version
1 2005
[10]. Castañer, Luis & Santiago Silvestre Modelling Photovoltaic Systems,
Using PSpice John Wiley & Sons Ltd, 2002
[11]. Sharaf A. M., Abdulla Ismail, R. A. El-Khatib & S. I. Abu-Azab “A
Photovoltaic Utilization System with Bang-Bang Self-Adjusting
Maximum Energy Tracking Controller” International Journal of
Energy Research, Volume 22, Issue 12 December 1998, page 1091 -
1098

Figure 15. Flow rates of PV water pumps

VI. CONCLUSION
This paper presents a simple but efficient photovoltaic
water pumping system. It models each component and
simulates the system using MATLAB and SIMULINK.
Simulation of the developed model of the system shows
that the PV model using the equivalent circuit in moderate
complexity provides good matching with the real PV
module. Further, simulations were also performed to
compare two MPPT algorithms using actual irradiance
data in the two different weather conditions. The incCond
algorithm shows narrowly but better performance in terms
of efficiency compared to the P&O algorithm under the
cloudy weather condition. However, due to cost
constraint, the P&O control method was chosen.
Simulation with SIMULINK to model a DC pump motor
which was then incorporated into MATLAB verifies
functionality and benefits of MPPT. Lastly, simulations of
the model were then compared against the system without
MPPT in terms of total energy produced and total volume
of water pumped a day. The results validate the benefits
of the MPPT which can significantly increase the
efficiency of energy production from PV and the
performance of the PV water pumping system compared
to the system without MPPT.

REFERENCES
[1]. UNEP “Water Policy and Strategy” (viewed on
www.unep.org/dpdl/water/ , August 2005)
[2]. UNESCO The UN World Water Development Report, 2003
(viewed on www.unesco.org/water/wwap/wwdr/, August 2005)
[3]. Texas Instruments “Converting Analog Controllers to Smart
Controllers with TMS320C2000 DSPs” Application Report, June
2004 (downloaded from dspvillage.ti.com/)
[4]. Rashid, Muhammad H. Editor-in-Chief Power Electronics
Handbook Academic Press, 2001
[5]. Kyocera Solar Inc. Solar Water Pump Applications Guide 2001
(downloaded from www.kyocerasolar.com)
[6]. Thompson, Marry A. Reverse-Osmosis Desalination of Seawater
Powered by Photovoltaics Without Batteries Doctoral Thesis,
Loughborough University, 2003

502