Photovoltaic solar energy conversion for hydrogen production by alkaline water

electrolysis: Conceptual design and analysis

Rupsha Bhattacharyya*1, Apurva Misra2, Sandeep KC1
1

Heavy Water Division, Bhabha Atomic Research Centre, Trombay, Mumbai-400 085,
Maharashtra, India

Department of Chemical Engineering, Banasthali University, Rajasthan-304 022, India

*Email: rupshabhattacharyya1986@gmail.com, rupsha@barc.gov.in, Phone: +91-022-2559-2962

Abstract
The use of solar energy for electricity generation and use of this electricity for hydrogen
production by alkaline water electrolysis promises to be a truly sustainable scheme for the
postulated hydrogen economy. This work addresses the design of a standalone solar photovoltaic
(PV) energy system that meets the energy requirements of the electrolysis process, followed by
the performance analysis under different environmental conditions. Energy requirement for
electrolysis depends on the hydrogen production rate desired and the operating conditions of the
electrolysis cell and it has been predicted from an essentially thermodynamic analysis. Mean
solar irradiation data is estimated from location specific meteorological data. The current-voltage
output characteristics of the solar modules have been predicted as function of the solar
irradiation using the five parameter, single diode model of a solar panel and they have been
linked to the production rates of hydrogen. The module behaviour and its thermodynamic and
conversion efficiencies have also been predicted for actual operating conditions. Thus a step-bystep simplified approach for the preliminary PV power system design and analysis for an
electrolysis based hydrogen production unit has been presented in this study.
Keywords: Solar photovoltaics; alkaline water electrolysis; hydrogen production; exergy analysis;
conceptual design
1

1. Introduction
The use of hydrogen as an important energy carrier for the future has been widely proposed [1].
The hydrogen based energy system is not only an alternative to carbon based fossil fuels on
which we are primarily dependent for our energy requirements today, it is also expected to
become a necessity in the face of depleting fossil fuel reserves and increasing concerns about
environmental pollution arising from green house gas emission when using the conventional
fossil fuels. Apart from its use in the energy sector, hydrogen has wide spread applications as a
raw material in the chemical, petroleum and nuclear industries and most of the demand is
currently met by production of hydrogen from hydrocarbon resources [2]. Hydrogen production
from an abundantly available raw material like water and use of renewable energy resources like
solar energy for hydrogen production by alkaline water electrolysis are truly representative of a
possible environmental friendly and sustainable solution (albeit an initially expensive one even
with the technology currently available) that will at least partly meet the increasing global
demands for energy and ensure long term energy security [3].
Alkaline water electrolysis can be carried out in stacks of electrolysis cells by using an electrolyte
solution of around 30% potassium hydroxide in water. The current to be supplied depends on the
production rate of hydrogen desired. The produced hydrogen and oxygen gases are kept
separated in the cell through the use of a gas-impermeable but ion-conducting membrane like
asbestos or Zirfon and they are subsequently separated from the electrolyte in bubble columntype gas separators fitted with wire mesh demister pads. Thus very high purity hydrogen as well
as valuable by-product oxygen (having applications in manufacturing processes, chemical
process industries, oxidative waste treatment processes and medical requirements) can be
simultaneously produced and recovered in such a system [4, 5].
2

The only hindrance to large scale electrolytic hydrogen production is the prohibitive cost of
electrical power. Roof top mounted or ground located solar photovoltaic (PV) modules at the site
of hydrogen production can be used to generate the electricity required for the process, thus
allowing clean and power-grid independent operation and ultimately enormous reduction in
process operating costs, as solar energy is available free of cost and only an initial investment in
the solar PV modules, batteries and ancillary hardware would be required to set up the facility.
Fig. 1 shows a conceptual diagram of such a system.
The idea of coupling a clean renewable energy source like solar energy for generation of an
energy carrier like hydrogen using a nearly inexhaustible resource such as water has been
explored by several researchers in the past. Various advanced concepts have also been
examined recently. Amongst these, Ferrari et. al have studied the operation of a grid connected
alkaline water electrolyser and have then theoretically studied the economic consequences of
replacing grid power by solar PV based electricity, through their in-house software. They have
demonstrated the dependence of economic feasibility on scale of hydrogen production and have
arrived at an optimal size of the hydrogen plant that can justify the use of solar PV systems [6].
The combination of two or more sources of renewable energy, like wind energy and solar PV
system, to obtain hydrogen by water electrolysis has also been explored [7]. It has been shown
that wind energy is energetically more efficient as compared to solar PV system in that particular
case. This shows the dependence of such an analysis on the particular location considered and
justifies the need to do this exercise for any new location where a hydrogen production facility
needs to be established. Multiple renewable energy based hydrogen production processes,
including some where solar PV as well as solar thermal energy have been utilised, have been
analysed and compared by Dincer and Ratlamwala [8]. Energy and exergy efficiencies as well as
the sustainability indices have been estimated for these processes. Use of hydrogen derived from
different technologies like alkaline water electrolysis, Proton Exchange Membrane electrolysis as
well as high temperature steam electrolysis coupled with intermittent renewable energy sources
in the Power-to-Gas technology has been reviewed by Gtz et. al., with emphasis on process
technology as well as process economics [9]. A software package PSCAD/EMTDC has been used
3

by Park et. al. to model and simulate the production of hydrogen using solar PV modules and
water electrolysis using solid polymer electrolyte and results have been validated by actual
experiments [10]. Demonstrations of alkaline water electrolysis with solar PV energy with and
without maximum power point trackers were carried out by Ahmed and Shenawy [11]. Other
than alkaline water for hydrogen production, ammonium sulphite solution has also been used as
the electrolyte for pulsed electrolysis using PV based electrolysis [12]. A novel combined solar
and geothermal energy based system for both power and hydrogen production has been
proposed and analysed by Bicer and Dincer [13]. A system comprising of multiple methods of
harnessing solar energy and also producing hydrogen, including photocatalysts, solar PV
modules, thermal engines and chemical energy storage has been developed and assessed
comprehensively by Zamfirescu and Dincer [14]. Yet another integrated facility for hydrogen
production and its on-site combustion for power generation has been proposed by Islam et. al.
[15]. Such a system is expected to eliminate the high costs of hydrogen storage and
transportation to other sites. Waste heat recovery from the combustion engine exhaust is used to
run a chiller unit that allows the PV panels to be cooled, thereby enhancing the energy and
exergy efficiencies of the overall process. Control aspects of the solar PV based hydrogen
generation facility have been discussed by Ulleberg [16].
The present work integrates many of these ideas together and attempts to demonstrate the
basic analyses that need to be performed in order to conceptually design a solar PV based
hydrogen production facility from alkaline water electrolysis at a chosen geographical. The inhouse code developed for this is expected to be of assistance in gauging the techno-commercial
feasibility and the performance assessment of such a system, irrespective of the end use of the
hydrogen produced through it. The detailed design of the electrical system that will serve to
integrate the electrolyser plant with the PV based solar arrays has not been presented in this
work. The focus is on feasibility and performance analysis of the electrolyser system when
coupled with a renewable energy source.

In this work the design and parametric analysis of the performance of a standalone solar PV
system for alkaline water electrolysis based hydrogen production at Mumbai, India have been
considered. The effect of electrolysis cell operating conditions on the electrical energy
requirement of the process has been first established. It is followed by estimation of solar
irradiance at the site of the facility using region specific meteorological data and study of the
behaviour of a standard photovoltaic module at the proposed location of the production plant
under various environmental conditions. PV module electrical output has been correlated to the
possible monthly and annual average hydrogen production rate. Exergetic and conversion
efficiency of the PV system have also been calculated at different conditions of irradiance, which
in turn govern the module temperature and the power output. The major aim of the work is to
develop a simple methodology for sizing a PV system for hydrogen production by water
electrolysis, without the use of any specific proprietary software and to perform parametric
analysis of the systems hydrogen production capacity, given the characteristics of the PV
module and the location specific solar irradiation and meteorological data. These exercises are
important at the feasibility study stage and for preliminary conceptual design work.
2. Thermodynamic analysis of electrical power requirement for water electrolysis
A schematic diagram illustrating the electrolyser plant coupled with the solar energy system has
been shown in Fig. 2. The process involves electrolysis of a 30% potassium hydroxide solution in
demineralised water in compact electrolyzer stacks consisting of a number of cells (50
electrolysis cells have been considered in this work) connected in series. The electrolyte mixed
with hydrogen on one side and oxygen on the other are cooled and separated in dedicated gas
liquid separators and are then sent through catalytic recombiners or purifiers to remove trace
amounts of oxygen from hydrogen and vice versa. This ensures hydrogen purity of greater than
99.9% (v/v) at the plant battery limit. Demineralised water has to be replenished from time to
time to make up for its usage during electrolysis.

Fig. 1. Schematic diagram for solar photovoltaic system for hydrogen production by alkaline
water electrolysis
The electrical power required for electrolysis is calculated as the product of the current required
(which in turn depends on the desired hydrogen production rate, as governed by Faradays laws
of electrolysis) and the decomposition voltage required for the electrochemical reaction which is
governed by the pressure and temperature at which the electrolysis is to be carried out [17, 18].
The minimum thermodynamic cell potential required for water splitting can be estimated from
the free energy change of the reaction (H 2O H2 + 0.5O2) at a given temperature and pressure
as follows:

ET ,P =

G(T , P)
ne F

(1)

In actual operation there are several other electrical resistances to be overcome (arising out of
activation over-potentials for the electrode reactions, concentration polarization, transport
resistances, presence of gas bubbles) and hence a voltage drop greater than that predicted by
Eqn (1) has to be maintained across the cell. These resistances may be estimated from empirical
and theoretical considerations and they are functions of the current being used for the
electrolysis as well as the design of the cell [18]. The value of G(T,P) depends on the pressure
and temperature of electrolysis and can be estimated from thermodynamic calculations. In this
6

section, the method proposed by Onda et. al. [17] has been used to calculate G(T,P) and hence
ET,P for alkaline water electrolysis at different temperatures and pressures in the electrolytic cell.
The over-potentials are also estimated and added to the reversible cell potential to obtain the
theoretical cell voltage for water electrolysis in a typical zero-gap configuration of the
electrolysis cell [18]. Results are presented in Figs. 2-4. It is seen that the theoretical cell
potential decreases with increasing temperature at a given pressure, but increases with
increasing pressure at constant temperature. For high pressure electrolysis, it is assumed that
the alkaline water is pressurized to the required pressure before it is fed to the high pressure cell
for electrolysis. This allows for the production of the gases at higher pressures directly without
the need to compress them later for charging into storage tanks, though the electrolytic cell has
to be designed to withstand the higher pressures. So the total power required to be supplied by
the PV modules includes the electrical power necessary for the actual electrolysis as well as that
needed for electrolyte pumping. The calculated values of the power requirement are shown in
Fig. 5. Depending on the particular production rate, pressure and temperature condition chosen
for the electrolysis, the solar photovoltaic system needs to be designed to supply the necessary
total power. Not only the total power requirement, the current and voltage requirements have to
be individually considered for this particular application as the hydrogen production rate will
correspond to a particular current, which has to be supplied at a voltage just above the
theoretical minimum value. The design and operating parameters for the solar photovoltaic
system are shown in Table 1, along with values or range of values applicable to them.

1.3

Reversible cell voltage (Volt)

1.28

1.26

1.24
298 K
313 K
323 K
333 K
343 K
353 K

1.22

1.2

1.18

12
16
Pressure(bar)

20

24

28

Fig. 2. Reversible cell potential for alkaline water electrolysis at different temperatures and
pressures
2.6

Theoretical Cell Potential (Volt)

2.58
2.56
2.54
2.52
2.5

298 K
313 K
323 K
333 K
343 K
353 K

2.48
2.46
2.44

12
16
Pressure (bar)

20

24

28

Fig. 3. Theoretical cell potential for alkaline water electrolysis at different temperatures and
pressures

2.56

Theoretical Cell Potential (Volt)

2.6

2.55
2.54

2.55

2.53
2.52

2.5

2.51
2.5

2.45
360

2.49
340

320
Temperature(K)

30
20
300
280

10
Pressure(bar)

2.48
2.47
2.46

Fig. 4. Surface of theoretical cell potential as function of pressure and temperature for alkaline
water electrolysis
The diurnal variation of solar irradiance causes varying current-voltage output of the photovoltaic
modules, thus affecting the rate of hydrogen production. It has been assumed in this work that
the hydrogen production rate will follow this time-variant irradiance pattern and hence the total
plant capacity can be stated in terms of the cumulative amount of hydrogen produced during the
total working or sunshine hours per day at the facility location. Based on total power
requirement, number of working hours as well as the individual current and voltage requirement
for the electrolysis process, the total number of panels (based on operation at standard
conditions) to be installed has been determined [19]. Results are shown in Table 1. Owing to the
large value of current required a large number of PV modules have to be connected in parallel to
magnify the output current from the array. The voltage requirement poses no problem since the
output voltage per module is sufficiently high for water electrolysis under the entire range of
operating conditions considered.

Total power requirement (kWhr/Nm3)

6.15

6.1

6.05

6
298 K
313 K
323 K
333 K
343 K
353 K

5.95

5.9

5.85

12
16
Pressure(bar)

20

24

28

Fig. 5. Total power requirement for alkaline water electrolysis and electrolyte pumping at
different temperatures and pressures
3. Modeling hydrogen production using solar PV modules
3.1 Assumptions
The following assumptions have been made for the system design and analysis calculations:
a) Shadow-free operation of the solar PV modules has been assumed during the sunshine hours.
b) The solar panels considered in this analysis are designated as ELDORA VSP.60.260.03, whose
details are presented in Table 1 [20].
c) Fixed angle of tilt of all the panels has been assumed and maximum power point tracking
devices have been considered. Thus at given time, depending on the prevalent solar irradiance,
the current and voltage output from the solar modules correspond to those at the condition of
maximum power. This current, when fed to the electrolyser unit, governs the production rate of
hydrogen Solar panels would be roof top mounted with adequate support structures to ensure
the desired tilt angle.
d) All panels are assumed to face due south.
10

e) Operation of electrolysis plant takes place only within the sunshine hours and no batteries for
charge storage are assumed to be present. The system is a standalone PV system with no grid
connection.
f) Ambient conditions at the site of the facility are assumed to be represented by the average
values of day time temperature and wind speed as reported in Table 2.
g) The solar PV modules supply the electrical energy necessary for the electrolysis and for feed
liquid pumping. Additional power which would be required in practice for lighting, space cooling,
instruments, control systems etc have not been considered in this analysis.
h) It is assumed that adequate floor area is available at the site for setting up the modules for
the desired production rate of hydrogen.
Table 1
Design and Operating Parameters for the Solar PV System
Parameter
Nominal average hydrogen production rate
Pressure/Temperature
Electrical power requirement

Value/Range
10 Nm3/hr
5 bar (a)/70 deg C
6000 Whr/Nm3 * 10 Nm3/hr = 60000 W =
60 kW

Current requirement per cell for electrolysis at

480 A
nominal hydrogen production rate
Number of cells
50
Voltage requirement per cell
2.5 V
Total voltage requirement
2.5 V * 50 cells = 125 V
Panel mounting
Roof top mounted, pointing due south
Geographical coordinates
19.0379oN, 72.923oE (Mumbai, India)
Nominal hours of plant operation per day
8
Daily electricity consumption
480 kWhr
Average number of sunshine hours per day
10 (7:00 hrs to 17:00 hrs)
Average number of peak sunshine hours per day
6

0.2
Solar PV module data (Standard Test Conditions viz. m = 1.5, 25 oC, 1000 W m-2 solar
irradiance) [20]
Module type
Pmax
Vmp

ELDORA VSP.60.260.03
260 W
30.98 V
11

Imp
Voc
Isc
T
T
Number of solar cells in series per module
Dimensions of the modules
NOCT

Number of panels required in parallel

8.39 A
37.75 V
8.95 A
0.0049 A/oC
-0.096 V/oC
60
1.64 m (H) X 0.992 m (W) X 40 mm (thick)
45oC
20o
115 (accounting for non-constancy of
solar irradiation and hence output current
from each panel)
125 V/30 V per panel = 4 (assuming

Number of panels required in series

average

voltage

output

of

30

Total number of panels required

throughout sunshine hours)

460 (Total peak power output = 460

Area required for panel placement

panels * 260 W per panel = 120 kW)

1150 m2 (assuming 25% excess area
margin

for

accommodating

electrical

systems, prevention of shadow effect on

panel performance, maintenance work,
etc)

3.2 Estimation of solar irradiation at site

The hourly average solar irradiance data for the location of the PV system (Mumbai, India) have
been estimated by the method proposed by Tiwari and Dubey [21] for the entire year. The
atmospheric parameters (TR, , K1 and K2) have been selected based on the assumption that the
average climatic system (as defined by the clear, hazy or cloudy conditions of a day) at Mumbai
is of Type a from January to May, Type b in June, Type d from July to September and Type a from
October to December [21]. This is indicative of a coastal and tropical climate with no great
variation of diurnal temperature over the year. The governing equations for hourly solar
irradiation calculation (during the sunshine hours) at the site for each day of the year are as
follows:
12

a) Solar declination for any day of the year

=24.35 sin ( 284+ n )

360
degrees
365

(2)

b) Hour angle at any solar time

=( ST 12 )15 degrees

(3)

c) Angle of incidence at a given hour

cos i =( cos cos + sin sincos ) cos cos+cos sin sin sin + ( sin cos cos sin cos ) sin

(4)

d) Zenith/Azimuthal angle at the given hour

cos z=cos cos cos+ sin sin

(5)

e) Solar irradiance on a horizontal surface

I g=I N cos z +I d

(6)

I N =I ON exp [( m . . T R + ) ]

I ON =I s [1+ 0.033 cos

(7)

n
]
( 360
365 )

(8)

m=[ cos z + 0.15 ( 93.885 z )1.253 ]

=4.52910 m 9.6686510 m+0.108014

I d=K 1 ( I ON I N ) cos z + K 2

(9)

(10)

(11)

f) Solar irradiance on an inclined surface

13

I t =I b R b + I d R d + Rr (I b + I d )

(12)

cos i
cos z

(13)

Rd =0.5(1+cos )

(14)

Rr =0.5(1cos )

(15)

R b=

The calculated values of It for each day of an entire year are presented in Fig. 6. From the figure it
is apparent that the daily variation of solar irradiance follows a sinusoidal distribution from
sunrise to sunset hours. For any given month, the daily variation in solar irradiance pattern is not
very high and so for each hour of the day or for the entire day, a monthly average can be
estimated and it can be used as the representative value for the whole month. The months from
July to September show significant dip in the average irradiance received as these are the
monsoon months at the geographical location considered in this study and this time is mostly
characterized by overcast and low irradiance conditions.

14

Fig. 6. Hourly variation of total solar irradiance on inclined surfaces (20 degree inclination from
the horizontal) pointing due south and located at Mumbai, India for each day of a year
Table 2
Meteorological data for Mumbai, India
Average

Average
(oC) speed

wind Mean

day

Day

Month

temperature

number

January
February
March
April
May
June
July
August
Septemb

[22]
24.5
24.8
26.9
28.7
30.2
29.2
27.7
27.3
27.7

(m s-1) [22]
3.84
4.66
5.10
5.62
5.67
5.64
6.46
5.85
4.05

[19]
January 17/17
February 16/47
March 16/75
April 15/105
May 15/135
June 11/162
July 17/198
August 16/228
September 15/258

er
October
Novemb

28.7
28.0

3.48
3.41

October 15/288
November 14/318

er
Decemb

26.3

3.45

December 10/344

er

3.3 Governing equations for module behaviour

The electrical characteristics of the PV modules are assumed to be described by the well known,
5-parameter, single diode model [23]. The I-V characteristics of a solar cell can be represented as
[23]

[ (

I =I L I 0 exp

V + I Rs
V + I Rs
1 (
)
a
R sh

(16)

The five module parameters are IL, Io, Rs, Rsh and a. These are expressed as functions of solar
irradiance and temperature in terms of the reference parameter values [23, 24], as shown in
Equations (17-22). First, the reference values of these parameters at standard test conditions are
15

obtained by the simultaneous solution of five governing equations, using data from the
manufacturers datasheets. These equations are derived using a method proposed by Lun et. al.
which involves simplifying the V-I equation using a Taylor series expansion to ultimately
represent it as a cubic polynomial in current [23] and they are solved by the principle of
minimization of the sum of squares of residuals of the equations to obtain the reference
parameters. The details of the method are available in literature and are not repeated here.

a=ar

I L=

It
I t , ref

Tm
T ref

(17)

[ I Lr + T ( T m T ref ) ]

[ ] (( ) ( ))

T
Eg ,ref
E
I O =I m exp [
g
T ref
T ref
Tm

(18)

1.602071019
]
k

(19)

E g=E g ,ref [ 10.0002667 ( T mT ref ) ]

(20)

Tm
I
( 1 T ln t )
T ref
I t , ref

(21)

Rs =R sr

Rsh=R shr

I t ,ref
It

(22)

Once the reference parameters (shown in Table 3) are known, the values at any other irradiance
and temperature condition can be calculated. Then for a chosen value of the voltage V at a given
irradiance, the current I is calculated from the simplified, cubic polynomial representation of the
non-linear V-I characteristics equation [23]. Typical calculated V-I profiles and power-voltage
profiles at NOCT (45oC) for different values of irradiance are shown in Figs. 7 and 8. Similar
calculations can be performed to determine the entire V-I curve using the actual value of the
average solar irradiation at any time on any day of the year, as calculated in section 3.2.
16

Table 3
Intrinsic reference parameters for ELDORA VSP.60.260.03
ILr
8.8 A

IOr
4.5*10-13 A

Rsr
0.38

Rshr
4000

ar
1.233

Fig. 7. Current-voltage characteristics of ELDORA VSP.60.260.03 at NOCT as a function of solar

irradiance
250

900 W/m2

200
800 W/m2

Power (W)

150

600 W/m2
500 W/m2

100

400 W/m2

50

200 W/m2
100 W/m2

-50

10

15

20
Voltage (V)

17

25

30

35

40

Fig. 8. Module output power-voltage characteristics of ELDORA VSP.60.260.03 at NOCT as a

function of solar irradiance
3.4 Hydrogen production pattern at NOCT
The solar PV modules produce an output current that corresponds to the variation in solar
irradiance over the day. Since the amount of hydrogen produced is directly proportional to the
current supplied to the electrolysis cell, the hydrogen production rate also shows a similar
variation in the standalone PV system as there are no other provisions made for constant current
operation throughout the day. It has already been assumed that the modules will always be
made to operate at the current and voltage levels corresponding to the maximum power point
for the particular value of solar irradiance and temperature prevalent at a given time, by use of
suitable power point trackers. Operation at the maximum power point also ensures lower values
of the module temperature and hence higher power output from it. In this section, using the
maximum power point current corresponding to the prevailing irradiance level at a particular
time of the day, the cumulative hydrogen generation rate has been evaluated by Faradays first
law of electrolysis from the 50 electrolyser cells. Results are shown in Fig. 9. The nearly
sinusoidal variation in the production rate that corresponds with the sinusoidal diurnal variation
of solar irradiance is clearly indicated. The average production rate for any given day of the year
or an annual average production rate value can be evaluated from these results. Daily average
hydrogen production rate has been shown in Fig. 10. The annual average hydrogen production
rate at NOCT (using only peak current and voltage output data for sizing the PV array) from the
PV modules is determined to be 6.3 Nm 3 hr-1, which is about 37% less than the desired nominal
production rate of 10 Nm 3 hr-1. This is mainly due to the diurnal variation in the current output
from the PV array. Thus the number of PV panels to be connected has to be enhanced so that an
increased (albeit time-varying) current output can be obtained. By trial and error, it is determined
that 115 panels have to be connected in parallel and 4 panels have to be connected in series,
giving a total of 460 panels in the PV array, so that the yearly average production rate is about
10.5 Nm3 hr-1. This requires around 1150 m 2 of land area and the peak power output from the
18

plant becomes about 120 kW (which is about twice the required constant power output for 10
Nm3 hr-1 hydrogen production rate).
12

Hydrogen production rate (Nm 3/hr)

14

10

12
10

8
6

4
2

0
400
300
200
100
Day of the year

10

12
2

Sunshine hours

Fig. 9. Hourly variation of hydrogen production by solar electricity based electrolysis using
ELDORA VSP.60.260.03 modules for each day of the year under NOCT at Mumbai, India

4. Analysis of PV system performance

4.1 Effect of environmental conditions on module output and hydrogen production
The current-voltage characteristics and hence the power output from the PV module depends on
the temperature of the module and the amount of solar irradiance intercepted by it. Module
temperature at any time depends on the energy balance between incoming solar energy and its
conversion to electrical energy and the various losses from the modules. These losses in turn
depend upon the system as well as the ambient conditions at any particular time. Thus a steady
state thermal analysis of a solar module considering diurnal and seasonal variations of solar
irradiance, ambient temperature and prevalent wind speed enables us to calculate module
19

temperature and hence predict the actual current and voltage output obtainable from it under
given conditions [24, 25]. This is therefore an analysis of solar PV module performance based on
the first law of thermodynamics. The following simplifying assumptions have been made [24, 25,
26]:

Fig. 10. Daily average hydrogen production rate at NOCT using ELDORA VSP.60.260.03 modules
at Mumbai, India
a) The temperature of the module is assumed to be spatially uniform at any instant of time.
b) The thermal properties of the various layers of the module are taken to be constant and
independent of temperature.
c) The heat losses from the module are assumed to be due to convection (both natural and
forced convection due to wind flow across the faces of the module) and radiation. From the sunfacing (i.e. top) surface of the module, convective heat losses are assumed to be due to both free
and forced convection while from the bottom or roof facing surface, only free convection has
been assumed to be of significance.
d) The electrical power generated from the module corresponds to that at the maximum power
point for the actual solar irradiation being received by it.
20

e) Average ambient temperature and wind speed data are assumed to be constant for the entire
month at the values reported in Table 2. Wind speed data are for an altitude of 50 m from mean
sea level.
f) An hourly average value of the solar irradiation has been used to obtain the steady state
temperature of the module for each of the sunshine hours on any given day. By this method, the
diurnal variation of the module temperature can be tracked even with a steady state analysis,
hour to hour.
g) Monthly average solar irradiance on the module has been taken to be the actual irradiance on
the mean day of each month [19]. Day numbers corresponding to each mean day are presented
in Table 2.
4.1.1 Calculation methodology
The steady state energy balance for the PV module can be expressed as [25]

Q QlossQ power =0

(23)

From the calculated value of solar irradiance at any time intercepted by the module,

Q =A I t

(24)

The term Qpower is the maximum power output from the module corresponding to an irradiance of
It. It is module temperature and solar irradiance dependent, as shown in Fig. 8.
The term Qloss includes losses due to natural convection effects, forced convection due to the
prevailing wind speed and radiation losses from both the front and back faces of the module.
Thus [13],
4
Qloss= A h f ( T mT ) + A hb ( T mT ) + A f F fs ( T 4mT 4sky ) + A f F fr ( T 4m T roof
)+ A b F bs (T 4mT 4sky )+ A b F br (T 4mT 4roof

(25)

21

The natural convection heat transfer coefficient for the PV panel (which is an inclined flat plate)
is evaluated from the following equation [27]

N u L = 0.825+

where R a L=

g air ( T mT ) L3
air ai r

( (

N u L=

hn L
k air

)))

0.387 R a
0.492
1+
Pr

1
6
L

9 8
16 27

(26)

Pr

C p air air
k air

and

L=

A
2(H +W )

while the forced convection heat transfer coefficient is estimated as follows [24]:

h fr =

0.931air airC p air R e 0.5

2

(27)

LP r 3

where

L v wind air
air

The combined free and forced convection coefficient is calculated from the following expression
[27]
3
3
3
hc = hn + hfr

(28)

The thermodynamic and transport properties of air for the convection calculation are all obtained
from literature at the mean film temperature [28].
The view factors are calculated from Eqns (29) to (32) [24]

1+ cos
F fs =0.5

(29)

22

( )
1+ cos
Fbs =0.5

(30)

1cos
F fr =0.5

(31)

( )
1cos
Fbr =0.5

(32)

The temperature of the sky is given by [24]

1.5

T sky =0.0552T

(33)

while the roof temperature is obtained from [24]

T roof =T +aroof I g

(34)

Table 4
Data for the thermal analysis of the solar modules
Parameter
H
W
A (= H x W)

Value
1.64 m [20]
0.992 m [20]
1.627 m2

0.90 [29]

0.77 [29]

aroof

0.08 m2 K

W-1

[24]

The energy balance equation is an implicit equation in the module temperature T m and it has to
be solved by an iterative procedure due to its highly non-linear temperature dependent terms,
using monthly mean values of solar irradiance to obtain the minimum average and maximum
23

average temperature of the module for each month of the year. Results are presented in Fig. 11.
The minimum temperatures are found to occur at the sunset and sunrise hours since the least
amount of solar irradiance is received by the module at these hours while the maximum
temperature occurs at midday. The module undergoes nearly 25-30 oC temperature rise above
the average ambient temperature (as indicated by the height of the bars in Fig. 11) over the
course of a day when clear, cloudless weather conditions prevail, which in turn affects the
module performance and therefore hydrogen production rate significantly.

Fig. 11. Monthly average maximum and minimum temperatures of ELDORA VSP.60.260.03
modules at Mumbai, India

24

Fig. 12. Monthly average maximum and minimum hydrogen production rates using ELDORA
VSP.60.260.03 modules at Mumbai, India
4.2 Steady state exergy analysis of the solar PV system
The exergy analysis of the solar PV system provides a generalized method of quantifying its
efficiency based on both the first and second laws of thermodynamics. This analysis not only
accounts for the system properties but also considers the ambient parameters which affect the
thermodynamic performance of the modules.
The input exergy to the module is due to the solar irradiance input received by it and it can be
expressed as [30]

E x =I t A [1

4T T 4
+
]
3T s
Ts

( )

(35)

Exergy destruction due to heat loss from the module to the ambient is given by [30, 31, 32, 33]

E xt =Qloss [1

T
]
Tm

(36)

Electrical exergy is given by [31]

25

E x elec =V mp I mp

(37)

Thus exergy output from the system is given by [31]

E x out =E x elec E x t

(38)

Exergetic efficiency of the solar PV module under given conditions can therefore be defined as
[31, 32, 33]

=100

E x out
E x

(39)

The solar energy to electrical energy conversion efficiency, under the maximum power output
conditions is given by

=100

V mp I mp
It A

(40)

Calculated values of the exergetic and energy conversion efficiencies are shown in Fig. 13 for the
monthly minimum average and maximum average solar irradiance received by the PV module.
The energy conversion efficiency varies within the range of 14-16%, depending on the month
and the associated weather and irradiance conditions and it conforms to the efficiency (15.98 %)
under standard test conditions, quoted by the module manufacturer [20]. It is seen that during
midday when solar irradiance received is high, the module temperature is also high and hence
the power output (primarily due to lower voltage output at higher temperature) from the module
suffers degradation, thereby leading to lower conversion efficiencies. This is true for those
months when clear conditions prevail. Under cloudy conditions, the efficiency does not vary
much over the course of the day. Exergetic efficiency shows a minimum value at noon and the
maximum value at the beginning or end of the sunshine hours for any given day of a particular
month. Exergy efficiency is quite low at midday and highest during periods of low solar irradiance
and hence low module temperature. The heat losses or thermal exergy destruction are reduced
26

when the module and ambient temperature difference is low and this condition occurs at sunrise
and sunset hours, thereby enhancing the exergy efficiency over that prevailing at midday. It is
also observed that the highest exergetic efficiencies at any time of the day are obtained during
the monsoon months (7-9) which have cloudy weather conditions. This is because of the lowest
module temperatures attained. But the hydrogen production rates are also lowest in these
months owing to the lower current output from the modules.

Fig. 13. Monthly average conversion and exergetic efficiencies of ELDORA VSP.60.260.03
modules at midday and sunrise/sunset time in Mumbai, India
4.3 Economic analysis in the Indian scenario
A simplified cost analysis is performed in this section to establish the economic feasibility of
using solar PV based energy for hydrogen production by water electrolysis. The additional
investment necessary for implementing solar PV modules for water electrolysis consists of two
components the modules themselves and the cost of land area needed to accommodate the
modules. It has been previously determined that the total number of modules needed is 460 for
the average production rate of 10 Nm3 hr-1 and based on the area of each module, about 1150
m2 of free land area would be required (which is assumed to be separately procured for this
purpose). The land cost can be significantly reduced if the panels can be housed on existing
27

rooftops and available free land space. The average land cost in Mumbai, India at present is
about 108000 m-2 and the cost of a single solar panel producing a nominal power of 260 W is
about 12000. Thus the additional investment necessary is (460*12000+108000*1150)=
1.297*108. Assuming that the high purity electrolytic hydrogen and the by-product medical grade
oxygen can be sold at 250 Nm-3 and 245 Nm-3 respectively, the total earning per hour of
operation of the plant will be 3725. If the plant runs for 10 hours per day and for 300 days per
year, then the number of years of operation required to earn back the additional investment for
PV based solar energy is about 11.6 yrs. The initial fixed cost of the electrolytic plant, additional
maintenance charges, cost of gas cylinders etc. have not been taken into account here, only the
expenditure to be incurred if solar electricity is used instead of grid electricity. Even then, it is
seen that a practical payback period (which will definitely be significantly lower than the typical
20 year lifetime of the electrolysis plant and 25 year lifetime of the present day solar PV
modules) is realizable for the additional investment in a solar PV system. Improvement of the
conversion efficiency of the PV modules can bring down the cost of hydrogen production
significantly and further lower the payback period.
5. Results and conclusion
Use of solar energy for production of hydrogen by water electrolysis represents use of a
renewable energy resource for generating of a versatile and efficient energy carrier as well as a
raw material for various applications in the chemical process industries. Therefore this
technology has particularly important implications in the context of the proposed hydrogen
economy in the near future. The major problems in this scheme lie in the varying hydrogen
production rate with diurnal variations of solar irradiance and ambient temperature and the
enormous floor space and economic investment requirements for setting up the modules. This
work first establishes the energy requirements for alkaline water electrolysis for a given pressure
and temperature through a thermodynamic analysis. It is shown that for a nominal average
production rate of 10 Nm3 hr-1, at 70 deg C and 5 bar pressure, approximately 60 kW energy is
required. The site specific solar irradiation levels are next predicted for the entire year using
28

appropriate meteorological data. Irradiation levels vary from 70 W m -2 to 900 W m-2 depending on
the time of the day and year. For a chosen kind of PV module, detailed analysis has been done to
predict the ambient condition dependent current-voltage characteristics, using information from
manufacturers datasheets. Module performance has then been correlated with the hydrogen
production rates over the year, assuming operation at maximum power point under all
conditions. Hydrogen production closely follows the irradiance pattern at any given time. The
annual average production rate achievable under the actual ambient conditions is 10.5 Nm 3 hr-1,
which is about 5% higher than the nominal rate required. It is seen that while 60 kW is the
nominal power output required from the PV system, in practice a power plant of 120 kW output
rating will have to be set up in order to account for the variable rate of incoming solar radiation
and the consequent non-uniform hydrogen production, such that the annual average production
rate is close to the desired value. The calculated energy conversion (i.e. solar to electrical)
efficiency of the PV modules is found to lie within 14 to 16%. Exergetic efficiency ranges from 8
to 16%, depending on time of the year and time of the day. Actual heat losses from the module
are likely to be less since the wind velocities actually prevalent may not in practice, be as high as
the values reported for an elevation of 50 m as all the panels may not be mounted on a roof top
50 m high. This may lead to higher efficiencies and improved hydrogen production. The effect of
shadows of surrounding structures on the I-V characteristics of the solar modules has not been
explicitly taken into account in this study as it depends on specific features of the location of the
facility and its surroundings.
Owing to the high current requirement of the process, a large number of PV panels will have to
be connected in a series-parallel array configuration. Given the high cost of the solar panels and
the large floor area requirement (and hence land cost) to set them up for electricity production,
the scheme may not always be economically feasible as of now on a large scale but it is a clean
and sustainable solution using no carbonaceous raw material for hydrogen generation. Remote
locations with abundant sunshine and water may be the most suitable locations for setting up
facilities such as these. The high purity of hydrogen obtained by electrolytic methods implies that
this technology can be used to produce small quantities of hydrogen for specialized applications
29

in food, pharmaceuticals and other fine chemical industries [34]. For the scale of operation
considered in this work, the additional investment can be recovered in about 12 years of
operation.
This work primarily aims to systematically demonstrate the preliminary calculations and analyses
to be performed in order to establish the feasibility and minimum requirements for a facility that
produces solar energy based hydrogen by water electrolysis. An integrated in-house code for the
specific case of hydrogen production by alkaline water electrolysis using solar PV energy has
been developed for such feasibility studies. Using location specific and PV module specific data,
similar calculations and analyses for a basic conceptual design can be performed for any other
region using the same code. Optimization studies have not been explicitly carried out, the major
aim of this work being the correlation between a particular PV module type, the energy
requirements for water electrolysis and the design and analysis of the PV array for the energy
required for electrolytic hydrogen production. The long term benefits of including solar PV
electricity in a hybrid energy system can be clearly demonstrated by a thorough economic
analysis that allows us to predict the hydrogen production cost from water electrolysis and this
problem will be addressed for the Indian scenario in another study in the near future. Use of
other renewable like wind energy to generate hydrogen is also potentially very beneficial and will
therefore be analysed similarly.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial,
or not-for-profit sectors.
Conflict of Interest
The authors declare no conflict of interest.
Acknowledgements

30

The authors are thankful to Dr Sadhana Mohan and Shri Kalyan Bhanja, Heavy Water Division,
Bhabha Atomic Research Centre for an introduction to the field of electrolytic production of
hydrogen and the discussions on possibility of using solar PV modules as the source of electrical
energy required for the process.

Nomenclature
a

Lumped ideality factor for the PV module under any condition

ar

(dimensionless)
Lumped ideality factor for the PV module under reference condition

aroof
A
Cpair
Eg
Eg.ref
ET,P
Exelec
Exin
Exout
Ext
F
Fbr
Fbs
Ffr
Ffs
g
H
hb

(dimensionless)
Solar irradiance absorptivity coefficient of the roof (m 2 W-1 K-1)
Area of the module exposed to solar radiation (m2)
Specific heat capacity of air (J kg-1 K-1)
Solar PV material band gap energy (eV)
Band gap energy for silicon (1.121 eV)
Reversible cell potential for water electrolysis at T and P (V)
Electrical exergy rate (W)
Solar input exergy rate (W)
Output exergy rate (W)
Thermal exergy destruction rate (W)
Faradays constant (96485 C mol-1)
Radiation view factor for back of module and roof (dimensionless)
Radiation view factor for back of module and sky (dimensionless)
Radiation view factor for front of module and roof (dimensionless)
Radiation view factor for front of module and sky (dimensionless)
Acceleration due to gravity (9.81 m s-2)
Length of the PV module (m)
Convective heat transfer coefficient for back surface of module (Wm -2 K1

hc
hf

)
Combined free and forced convection heat transfer coefficient (Wm -2 K-1)
Combined convective heat transfer coefficient for front surface of

hfr
hn
I
Ib
Id
Ig
IL
ILr
Imp
IN

module (Wm-2 K-1)

Forced convection heat transfer coefficient (Wm-2 K-1)
Free convection heat transfer coefficient (Wm-2 K-1)
Module output current under any operating condition (A)
Beam or direct radiation received by a surface (W m -2)
Diffuse radiation received by a surface (W m-2)
Global irradiance on a horizontal surface (W m-2)
Photo current in the PV module at any condition (A)
Photo current in the PV module at reference condition (A)
Module current at maximum power point (A)
Direct beam irradiance on a surface perpendicular to the direct beam (W
m-2)
31

ION

Extraterrestrial radiation measured on a plane normal to the radiation

IO

(W m-2)
Saturation current in the PV module under reverse bias at any condition

IOr

(A)
Saturation current in the PV module under reverse bias at reference

IS
Isc
It
It,ref
L
k
kair
K1, K2

condition (A)
Extra-terrestrial solar constant (1367 W m-2)
Module short circuit current (A)
Total radiation received by a surface (W m-2)
Total radiation received by a surface under reference conditions (W m -2)
Characteristic dimension for the solar module (m)
Boltzmann constant (1.38*10-23 J K-1)
Thermal conductivity of air (W m-1 K-1)
Lumped atmospheric parameters for diffuse radiation (dimensionless, W

m
n
ne
ni
NOCT
NuL

m-2)
Air Mass (dimensionless)
Day number in a year (dimensionless)
Number of electrons transferred during electrolysis (dimensionless)
Ideality factor (dimensionless)
Nominal operating cell temperature (C)
Free convection Nusselt number for heat transfer from a flat, inclined

P
Pmax
Pr
q
Qin
Qloss
Qpower
RaL
Rb
Rd
Re
Rr
Rs
Rsh
Rshr
Rsr
ST
T
Tm
Tref
Troof
TR
Ts
Tsky
T
vwind
V
Vmp
Voc

plate (Wm-2 K-1)

Electrolysis pressure (Pa)
Maximum power output from solar module (W)
Prandtl number (dimensionless)
Electronic charge (1.602*10-19 C)
Thermal energy input to the PV module (W)
Thermal energy loss from the PV module (W)
Electrical power output from the PV module (W)
Rayleigh number for free convection (dimensionless)
Conversion factor for beam radiation (dimensionless)
Conversion factor for diffuse radiation (dimensionless)
Reynolds number (dimensionless)
Conversion factor for reflected radiation (dimensionless)
Series resistance at any condition ()
Shunt resistance at any condition ()
Shunt resistance at reference condition ()
Series resistance at reference condition ()
Solar time (hours)
Electrolysis temperature (K)
Module temperature during operation (K)
Reference temperature of module (K)
Temperature of the roof (K)
Linke turbidity factor (dimensionless)
Sun temperature (5780 K)
Temperature of the sky (K)
Ambient temperature at the site (K)
Wind velocity (m s-1)
Module output voltage under any operating condition (V)
Module voltage at maximum power (V)
Module voltage at open circuit (A)
32

Width of the PV module (m)

Greek symbols

Lumped atmospheric parameter for beam radiation at any location

air
T

air
T

GT,P

(dimensionless)
Thermal diffusivity of air (m2 s-1)
Temperature coefficient of current of the solar module (A oC-1)
Angle of tilt of the solar PV panels (degree)
Coefficient of volumetric expansion of air (K -1)
Temperature coefficient of voltage of the solar module (V oC-1)
Solar declination angle (degree)
Free energy change in water electrolysis process (J mol -1 hydrogen)
Integrated Rayleigh scattering optical thickness of the atmosphere

b
f

air
air

air

i
z

(dimensionless)
Emissivity of the back surface of the PV module (dimensionless)
Emissivity of the front surface of the PV module (dimensionless)
Solar azimuth angle (degree)
Conversion efficiency (dimensionless)
Viscosity of air (Pa s)
Kinematic viscosity of air (m2 s-1)
Ground reflection coefficient (dimensionless)
Density of air (kg m-3)
Stefan-Boltzmann constant (W m-2 K-4)
Angle of incidence of the sunrays (degree)
Solar zenith angle (degree)
Longitude of the site (degree)
Exergetic efficiency (dimensionless)
Hour angle (degree)

References
[1] Sherif S A, Barbir F, Veziroglu N. Towards a Hydrogen Economy. The Elec Journal
2005;18(6):62-76.
[2] Ramchandran R, Menon R K. An Overview of Industrial Uses of Hydrogen. Int J Hydrogen
Energy 1998;23(7):593-7.
[3] Pagliaro M, Konstandopoulos A G. Solar Hydrogen: Fuel of the Future. Cambridge: RSC
Publishing; 2012.
[4] Stojic D Lj, Marceta L C, Sovilj S P, Miljanic S S. Hydrogen generation from water
electrolysispossibilities of energy saving. J Power Sources 2003;118:315-9.

33

[5] Zeng K, Zhang D. Recent progress in alkaline water electrolysis for hydrogen production and
applications. Prog in Ener Combus Sci 2010; 36:307-26.
[6] Ferrari M L, Rivarolo M, Massardo A F. Hydrogen production system from photovoltaic panels:
experimental characterization and size optimization. Energy Conv and Mgmt 2016;116:194-202.
[7] Khalilnejad A, Riahy G H. A hybrid wind-PV system performance investigation for the purpose
of maximum hydrogen production and storage using advanced alkaline electrolyzer. Energy Conv
and Mgmt 2014;80:398-406.
[8] Dincer I, Ratlamwala T A H. Development of novel renewable energy based hydrogen
production systems: A comparative study. Energy Conv and Mgmt 2013;72:77-87.
[9] Gtz M, Lefebvre J, Mrs M et. al. Renewable Power-to-Gas: A technological and economic
review. Renew Energ 2016;85:1371-90.
[10] Park M, Lee D, Yu I. PSCAD/EMTDC modeling and simulation of solar-powered hydrogen
production system. Renew Energ 2006;31:2342-55.
[11] Ahmad G E, Shenawy E T E. Optimized photovoltiac system for hydrogen production. Renew
Energ 2006;31:1043-54.
[12] Huang C. Solar hydrogen production via pulse electrolysis of aqueous ammonium sulfite
solution. Sol Energy 2013;91:394-401.
[13] Bicer Y, Dincer I. Development of a new solar and geothermal based combined system for
hydrogen production. Sol Energy 2016;127:269-84.
[14] Zamfirescu C, Dincer I. Assessment of a new integrated solar energy system for hydrogen
production. Sol Energy 2014;107:700-13.
[15] Islam S, Dincer I, Yilbas B S. System development for solar energy-based hydrogen
production and on-site combustion in HCCI engine for power generation. Sol Energy
2016;136:65-77.
34

[16] Ulleberg . The importance of control strategies in PVhydrogen systems. Sol Energy
2004;76:323-9.
[17] Onda K, Kyakuno T, Hattori K, Ito K. Prediction of production power for high-pressure
hydrogen by high-pressure water electrolysis. J Power Sources 2004;132:64-70.
[18] Roy A, Watson S, Infield D. Comparison of electrical energy efficiency of atmospheric and
high-pressure electrolysers. Int J Hydrogen Energy 2006;31:1964-79.
[19] Solanki C S. Solar Photovoltaics - Fundamentals, Technologies and Applications, 3 rd ed. Delhi:
PHI Learning Pvt. Ltd.; 2015.
[20]Eldora Ultima Silver Series Technical Data
Available

at

http://www.vikramsolar.com/products/products_details-Mg==-Mw==-

Eldora_Ultima_Silver (Last accessed on 24.08.2016)

[21] Tiwari G N, Dubey S. Fundamentals of Photovoltaic Modules and Their Applications.
Cambridge: RSC Publishing; 2010.
[22] Wind Speed in Mumbai, Maharashtra, India
Available at http://www.synergyenviron.com/tools/wind_data.asp?loc=Mumbai%2CMaharashtra
%2CIndia (Last accessed on 22.08.2016)
[23] Lun S, Du C, Guo T, Wang S, Sang J, Li J. A new explicit IV model of a solar cell based on
Taylors series expansion. Sol Energy 2013;94:221-32.
[24] Lobera D T, Valkealahti S. Dynamic thermal model of solar PV systems under varying
climatic conditions. Sol Energy 2013;93:183-94.
[25] Armstrong S, Hurley W G. A thermal model for photovoltaic panels under varying
atmospheric conditions. App Therm Energy 2010;30:1488-95.

35

[26] Lee Y, Tay A A O. Finite Element Thermal Analysis of a Solar Photovoltaic Module. Energy
Proc 2012;15:413-20.
[27] Incropera F P, DeWitt D P, Bergmann T L, Lavine A S. Fundamentals of Heat and Mass
Transfer, 6th ed. New Jersey: John Wiley and Sons, Inc; 2007.
[28] Yaws C L. The Yaws Handbook of Physical Properties for Hydrocarbons and Chemicals, 2 nd ed.
Oxford: Gulf Professional Publishing; 2015.
[29] Giovannetti F, Fste S, Ehrmann N, Rockendorf G. High transmittance, low emissivity glass
covers for flat plate collectors: Applications and performance. Energy Proc 2012;30:106-15.
[30] Sudhakar K, Srivastava T. Energy and exergy analysis of 36 W solar photovoltaic module. Int
J Amb Energy 2014;35(1):51-7.
[31] Pandey A K, Pant P C, Sastry O S, Kumar A, Tyagi S K. Energy and Exergy Performance
Evaluation of a Typical Solar Photovoltaic Module. Therm Sci 2015;19(S2):S625-36.
[32] Joshi A S, Dincer I, Reddy B V. Exergetic assessment of solar hydrogen production methods.
Int J Hydrogen Energy 2010;35:4901-8.
[33] Oruc M E, Desai A V, Kenis P J A, Nuzzo R J. Comprehensive energy analysis of a photovoltaic
thermal water electrolyzer. App Energy 2016;164:294-302.
[34] da Rosa A V. Fundamentals of Renewable Energy Processes, 3 rd ed. Oxford: Elsevier
Academic Press; 2013.

36