Journal of Power Sources 298 (2015) 150e157

Contents lists available at ScienceDirect

Journal of Power Sources

journal homepage: www.elsevier.com/locate/jpowsour

Measuring the state of charge of the electrolyte solution in a vanadium

redox flow battery using a four-pole cell device
Kittima Ngamsai, Amornchai Arpornwichanop*
Computational Process Engineering Research Unit, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok 10330,
Thailand

h i g h l i g h t s

The four-pole cell device is proposed to measure electrolyte imbalance in a VRB.

The method is sufficiently precise to be applied in real-time mode.

The effect of free ions in electrolyte solution on the cell voltage is discussed.

The effect of electrolyte compositions on the performance of VRB system is described.

a r t i c l e i n f o a b s t r a c t

Article history: The decrease in the efficiency and capacity of a vanadium redox flow battery (VRB) caused by an elec-
Received 25 April 2015 trolyte imbalance is an important impediment to its long-term operation. Knowing the state of charge
Received in revised form (SOC) of an electrolyte solution can quantify the level of the electrolyte imbalance in the VRB. In this
5 August 2015
study, a four-pole cell device is devised and employed to predict the SOC. The proposed method directly
Accepted 8 August 2015
measures the ionic resistance of the electrolyte solution and is sufficiently precise to be applied in real-
Available online xxx
time mode. Experimental studies on the effects of the operating current on the four-pole cell and the
concentrations of vanadium and sulfuric acid in the electrolyte solution are carried out. The results show
Keywords:
Vanadium redox flow battery
that the four-pole cell method can be utilized to measure the electrolyte SOC. The concentrations of
State of charge vanadium and sulfuric acid in the electrolyte solution affect the ionic resistance of the solution.
Electrolyte imbalance Regarding the capacity and efficiency of the VRB system, the results indicate that the electrical charge is
Four-pole cell device determined from the concentration of vanadium and that the cell voltage depends on the concentration
of sulfuric acid in the electrolyte solution. The decreased vanadium concentration and increased sulfuric
acid concentration improves the cell voltage efficiency.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction VRB is used together with renewable energy systems to compen-

A vanadium redox flow battery (VRB) is an energy storage
technology that is based on the electrochemical principle by which
electrical energy can be stored as a form of chemical energy and
then re-generated when the electricity is demanded [1e3]. The key
advantage of the VRB is that its power and energy can be designed
separately [4e6], leading to flexibility in scale up. Furthermore, the
empty energy of the electrolyte solution can be replaced by the full
energy of the electrolyte solution, which is a simple way to charge
the VRB system [7,8]. Consequently, the VRB system has received
much attention and is found in many applications. At present, the
* Corresponding author.
E-mail address: Amornchai.a@chula.ac.th (A. Arpornwichanop).
sate for the intermittence of the energy supply [9e11].
The VRB consists of two key components, an electrochemical
cell (for energy conversion) and an electrolyte solution (for energy
storage). When the VRB system needs to be used for energy storage
or generation, an electrolyte solution is fed into the electrochemical
cell and then the redox reaction (reduction and oxidation reactions)
of vanadium occurs in the cell. The cell has two half-cells (positive
and negative) that are separated by an ion exchange membrane,
which is utilized to prevent cross mixing of the electrolyte between
the half-cells [12e14]; however, ions can be transported across the
membrane to balance the circuit during the charging or discharging
process. Because vanadium is a transition element and has four
oxidation states, it can be employed as both positive and negative
electrolytes [15e17].

http://dx.doi.org/10.1016/j.jpowsour.2015.08.026
0378-7753/© 2015 Elsevier B.V. All rights reserved.
 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157
An electrolyte solution imbalance is considered a major problem
of the VRB [18e20] that is caused by side reactions and vanadium
ion and water transfers across the membrane [21e23]. This prob-
lem affects the energy capacity and efficiency of the VRB system.
Research groups have attempted to solve the problem through
several approaches; for example, new membranes were synthe-
sized to minimize the transfer of vanadium ions and waters across
the membranes [24e27], the operation of the VRB was optimized
[28e31] and the electrochemical cell and electrolyte tank were
designed to decrease the side reactions [32]. Alternatively, the
concept of electrolyte system management is proposed to maintain
the level of the electrolyte imbalance in an acceptable range. For
this approach, the determination of the state of charge (SOC) on
each side of the VRB electrolyte is necessary. When the real-time
measurement of the SOC can be performed, the electrolyte in the
VRB system can be automatically carried out, leading to high sys-
tem effectiveness.
In general, an open circuit voltage cell (OCV cell) is applied to the
measurement of the total system SOC, indicating the amount of
energy remaining in the VRB system. As the SOC on each side cannot
be identified by this method, the level of the electrolyte imbalance
cannot be determined. To detect the SOC on each electrolyte side, a
potentiometric titration method with potassium permanganate as
the titrant is the most popular method [18]; however, this method
requires time and is thus, unsuitable for real-time applications. The
measurement of the electrolyte solution absorbance can also be
utilized to measure the SOC. As vanadium is a transition element, the
color of the electrolyte solution changes with the vanadium oxida-
tion state [18]. Because the positive electrolyte is opaque due to the
high concentration of vanadium, this method cannot identify the
SOC on the positive side. The transmission spectrum can also be
utilized to determine the SOC based on the shape of the spectrum
from many wavelengths instead of the absorbance at a single
wavelength [20]. Although this method can be employed to detect
the SOC in both the positive and negative electrolyte solutions, a
large database of the standard spectra is needed for comparison with
the sample spectrum. The measurement of the conductivity of the
electrolyte solution is another potential method [18], but the cali-
bration of the probe is necessary to make a precise measurement.
Additionally, inserting a reference electrode is another way of
knowing the state of charge at each separate electrode [33,34].
However, the reference electrode should be employed in additional
cell in order to avoid the operating cell disturbance and the mea-
surement of SOC should be done in the condition of no current flow.
In this study, the measurement of the SOC in the electrolyte on
each side of the VRB using a newly developed four-pole cell with
simplicity and versatility is investigated. The two outer poles are
used to pass a constant current into the electrolyte, and the two
inner poles are used to measure the potential difference. Based on
this approach, the polarization effect is eliminated, and the voltage
of the four-pole cell represents the actual voltage that results from
the characteristics of the electrolyte solution [35]. The effects of key
operating parameters, such as the constant current passed through
the four-pole cell and the concentrations of vanadium and sulfuric
acid in the electrolyte solution on the capacity and efficiency of the
VRB system, are investigated.
2. Experimental
The experiment is divided into two parts, (1) measurement of
the SOC in the electrolyte solution on each side of the VRB using a
four-pole cell and (2) investigation of the effects of the electrolyte
compositions on the capacity and performance of the VRB system.
151
2.1. Measurement of the SOC in the electrolyte solution on each side
of the VRB system by a four-pole cell
2.1.1. Preparation of electrolyte solution
Vanadium pentoxide (V2O5) powder was dissolved in sulfuric
acid solutions of varying concentrations. Then, an electrolysis cell
was employed to adjust the oxidation state of vanadium to þ3.5
(50% V(III) ions and 50% V(IV) ions) [36].
2.1.2. The four-pole cell
The four-pole cell is the device that was utilized to measure the
voltage difference across the electrolyte solution when a constant
current was applied. There are four poles (or electrodes) with
different distances between them, as shown in Fig. 1. The poles
were fabricated of conductive carbon with a thickness of 2.5 mm
and a width of 7 mm. The four-pole cell was connected at the inlet
of the electrolyte solution before passing through the VRB cell. After
that, a power supply was employed to supply a constant current at
the two outer poles. A multi-meter was utilized to measure the
voltage difference between the two inner poles.
2.1.3. Experimental method
The VRB system with the four-pole cell was set up as shown in
Fig. 1. Then, 230 mL of an electrolyte solution was added to each side
of the cell using the pumps. A VRB cell with an effective area of
120 cm2 was used to change the SOC through a charging process
with a constant current of 80 mA cm2. Meanwhile, the OCV, the
voltage of the cell and the voltage of the four-pole cell were recorded
every 3 s using a Wisco data logger (model DL2200) until the OCV
was unchanged, indicating that it was fully charged. Then, the same
method was utilized to test the electrolyte solution on the other side.
Three variables, including the constant current through the
four-pole cell and the concentrations of vanadium and sulfuric acid
in the electrolyte solution, were investigated. The operating tem-
perature, the test conditions and the parameters of the test device
such as the flow rate of the electrolyte solution, the VRB cell, the
constant current to charge for the change of the SOC, and the four-
pole cell device were the controlled variables. To investigate the
effects of the three variables on the four-pole cell operation, con-
stant currents of 10 mA, 30 mA and 50 mA were employed through
the four-pole cell. Vanadium concentrations of 1.0 mol L1,
1.5 mol L1 and 2.0 mol L1 with a sulfuric acid concentration of
4.0 mol L1 and sulfuric acid concentrations of 3 mol L1, 4 mol L1
and 5 mol L1 with a vanadium concentration of 1.5 mol L1 were
also tested. The concentrations were selected based on the pa-
rameters of practical electrolyte solutions to verify that the four-
pole cell method is applicable to real systems. The calculated
charging time and OCV were utilized to represent the SOC.
In the electrolyte solution, the major components are vanadium
salt and sulfuric acid, which comprise vanadium, bisulfate and
sulfate ions and protons. These ions can affect the performance of
the VRB system, so the effect of the electrolyte composition on the
capacity and efficiency of the VRB system is also studied. It is noted
that the variations in pH occurring during operation, e.g. water
evaluation or hydrogen evolution, would affect the measured OCV.
Hydrogen evolution at the negative electrode can lower the proton
activity and lead to the precipitation of vanadium species.
2.2. Effect of electrolyte composition on the capacity and efficiency
of the VRB system
The VRB system without the four-pole cell was set up as shown
in Fig. 1. A single cell with an effective area of 120 cm2 was utilized.
GFD 4.6 felts (SGL) with thermal activation were employed as the
fiber electrodes. An APS-4 Selemion membrane was used. Then, the
152 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157
Fig. 1. The schematic of a VRB system with the four-pole cell set-up and the attributes of the four-pole cell.
same electrolyte solutions as above were fed to the cell by two
peristaltic pumps. 230 mL of electrolyte solution was employed on
each side, and a constant current of 80 mA cm2 was applied to
charge the VRB cell. The OCV and the voltage of the cell were
recorded every 3 s using a Wisco (model DL2200) data logger until
the OCV was unchanged or until the negative electrolyte was pur-
ple and the positive electrolyte was yellow, indicating that the cell
was fully charged. To study the effect of the vanadium concentra-
tion in the electrolyte solution, vanadium concentrations of
1.0 mol L1, 1.5 mol L1 and 2.0 mol L1 with a sulfuric acid con-
centration of 4.0 mol L1 were investigated. Similarly, to study the
effect of the sulfuric acid concentration in the electrolyte solution,
sulfuric acid concentrations of 3 mol L1, 4 mol L1 and 5 mol L1
with a vanadium concentration of 1.5 mol L1 were also investi-
gated. The calculated charging time and the OCV were utilized to
represent the SOC.
3. Results and discussion
3.1. Measurement of the SOC in the electrolyte solution on each side
of the VRB system using the four-pole cell
Three variables that affect the correlation between the voltage
of the four-pole cell and the SOC on each side are investigated. The
results are shown as follows.
3.1.1. Effect of constant current through the four-pole cell
Three values of constant currents through the four-pole cell,
10 mA, 30 mA and 50 mA, were investigated. The results, as illus-
trated in Fig. 2, show that the correlations between the voltage of
the four-pole cell and the SOC on each side for four-pole cell cur-
rents of 10 mA and 30 mA are linear. However, for 50 mA, the
correlations are quite curved because the measurements are
interfered by side reactions that can occur at sufficiently high
supply voltages. The optimum constant current for the four-pole

.
2VOSO4 þ e þ ða þ 1ÞHþ þ SO2 4 ða  1ÞHþ þ H2 SO4 aHþ þ HSO
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
free ion
cell needs to be known as when the constant current is too low,
the slope is also low, so the SOC cannot be determined, and when
the constant current is too high, side reactions can occur, leading to
imprecise measurements due to interference. Moreover, the supply
current poles (the two outer poles) of the four-pole cell can also be
damaged by oxygen evolution. Increasing the distance between the
measured poles (the two inner poles) and increasing the effective
area of the supply current poles are two ways to reduce the effect of
side reactions. Consequently, a constant current of 30 mA
(14.46 mA cm2) is the optimum for this four-pole cell.
According to the results in Fig. 2, when the same constant cur-
rent was applied on both sides (negative and positive) of the four-
pole cell, the voltage of the negative side is higher than that of
positive side. This illustrates that the resistance of the negative
electrolyte is higher than that of the positive electrolyte because
when the oxidation of the vanadium ions in an electrolyte changes,
the free ions in the solution (sulfate, bisulfate and proton) also
change, as will be explained below. However, the voltage of the
four-pole cell decreases when the free ions in the electrolyte so-
lution increase. Likewise, the voltage increases when the free ions
decrease. The results further reveal that when the SOC is equal to 1
on the negative side, it is equal to 0 on the positive side due to the
connection of the four-pole cell, as can be explained by Eqs. (1)e(4).
The operating process of the VRB can be divided into two regions,
the conditioning region and the operating region. In the condi-
tioning region, the reactions shown in Eqs. (1) and (2) occur. It is
noted that the initial electrolyte solution is prepared by 50% VO2þ
and 50% V3þ (called V3.5þ). The conditioning region is set by
adjusting the negative electrolyte (V3þ) to 100% and the positive
electrolyte (VO2þ) to 100%. As a result, the electrolyte can be used
for charge and discharge processes, which is called the operating
region.
Conditioning region
Negative:
. 
4 ⇔V2 ðSO4 Þ3 þ 2H2 O (1)
 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157 153

01.40 01.40
Negative 10 mA Positive 10 mA
01.20 30 mA 01.20 30 mA
50 mA 50 mA
Voltage of four-pole cell (V)

Voltage of four-pole cell (V)

01.00 01.00

0.80 0.80

0.60 0.60

0.40 0.40

0.20 0.20

00.00 00.00
00.00 0.20 0.40 0.60 0.80 01.00 00.00 0.20 0.40 0.60 0.80 01.00

SOC SOC

Fig. 2. Correlation between the voltage of the four-pole cell and the SOC at three constant currents for the four-pole cell, 10 mA, 30 mA and 50 mA, in an electrolyte solution with a
vanadium concentration of 1.5 mol L1 and a sulfuric acid concentration of 3 mol L1.

voltage between negative and positive sides is by an amount equal

to the voltage of the four-pole cell, causing a change of one
Positive: oxidation state of vanadium in the electrolyte solution. Then, the
VRB system will be in the operating region, in which the reactions
of Eqs. (3) and (4) occur.
V2 ðSO4 Þ3 þ2H2 O⇔2VOSO4 þe

. .  Operating region
þ ða þ 1ÞHþ þSO2 4 ða  1ÞHþ þ H2 SO4 aHþ þHSO 4
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl} Negative:
free ion
(2)

. . 
V2 ðSO4 Þ3 þe ⇔2VSO4 þ ða þ 1ÞHþ þ 2SO2 4 ða  1ÞHþ þ H2 SO4 þ SO2 4 aHþ þ HSO 4 þ SO4
2
(3)
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
free ion

Positive:

. . 
2VOSO4 þ2H2 O⇔ðVO2 Þ2 SO4 þe þ ða þ 1ÞHþ þ SO2 4 ða  1ÞHþ þ H2 SO4 aHþ þ HSO 4 (4)
|fflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflfflffl}
free ion

In the conditioning region, the vanadium oxidation state of þ3.5 In the operating region, the free ions increase on both sides, so
is utilized as the initial electrolyte solution. During the charging the voltages of the four-pole cell decrease at the same rate, as
process, the oxidation state changes from V(IV) to V(III) in the shown in Fig. 3. Therefore, in the fully charged state (where the SOC
negative half-cell, as shown in Eq. (1), which causes the free ions in of the negative side is equal to 1 and the SOC of the positive side is
the electrolyte solution to slightly decrease, and so the voltage of also equal to 1), the numbers of free ions on the negative and
the four-pole cell slightly increases. On the other hand, the oxida- positive sides of electrolyte solution are equally increased from the
tion state of vanadium changes from V(III) to V(IV) in the positive values at the conditioning-operating region interface. These
half-cell, as shown in Eq. (2), so the free ions in the electrolyte numbers are equal to the numbers of free ions generated from a
solution slightly increase, leading to a slight decrease in the voltage change of one oxidation state of vanadium in the electrolyte solu-
of the four-pole cell (Fig. 3.). Considering Fig. 3 at the conditioning- tion. Consequently, the voltage of the four-pole cell on the negative
operating region interface (where the SOCs of the negative and side when the SOC is equal to 1 can be connected to that of positive
positive sides are equal to 0), the voltage of the four-pole cell is side when the SOC is equal to 0. This is beneficial for the electrolyte
higher on the negative side than that of positive. This different sample, as it can also indicate whether it is a negative or positive
154 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157
Fig. 3. Correlation between the SOC and the voltage of a four-pole cell in the condi-
tioning and operating regions using a constant current for the four-pole cell of 30 mA
and an electrolyte with a vanadium concentration of 1.5 mol L1 and a sulfuric acid
concentration of 3.0 mol L1.
electrolyte solution.
3.1.2. Effect of concentration of vanadium
To study the effect of the concentration of vanadium, electrolyte
solutions with vanadium concentrations of 1.0 mol L1, 1.5 mol L1
and 2.0 mol L1 are investigated. The results show that the voltage
of the four-pole cell is highest when the electrolyte solution uses a
vanadium concentration of 2.0 mol L1 at a sulfuric acid concen-
tration of 4 mol L1, when the resistance of the electrolyte solution
is greatest (Fig. 4.). Furthermore, the correlation of the 2.0 mol L1
vanadium concentration is quite curved as because the voltage
supplied is high, side reactions occur that interfere with the mea-
surement, although it is still measurable.
To study the performance of the VRB system, the voltage effi-
ciency in the charging process (Vch_eff) is considered. The efficiency
is calculated from the average Vch_eff in the whole region of VRB
operation (SOC ¼ 0 to SOC ¼ 1), which is equal to the OCV divided
by the voltage of cell at the same SOC (Vch_eff ¼ (Vocv/Vcell)*100). The
experimental data show that the voltage efficiencies in the
charging process are equal to 94.36%, 93.39% and 93.00% for elec-
trolytes with vanadium concentrations of 1.0 mol L1, 1.5 mol L1
and 2.0 mol L1, respectively. The configuration of the electro-
chemical cell, the operating conditions and the charging current are
01.00
Negative
1.0 mol L
1.5 mol L
0.80 2.0 mol L
Voltage of four-pole cell (V)
0.60
0.40
0.20
00.00
00.00 0.20 0.40 0.60 0.80
SOC
Fig. 4. Effect of the vanadium concentration on the correlation between the voltage of the four-pole cell and the SOC for electrolyte solutions with vanadium concentrations of
1.0 mol L1, 1.5 mol L1 and 2.0 mol L1, a sulfuric acid concentration of 4 mol L1, and a constant current across the four-pole cell of 30 mA.
tightly controlled. The results indicate that the electrolyte with a
higher concentration of vanadium has a lower voltage efficiency.
The difference in the ionic resistance of the electrolyte solution is
the reason for the varying efficiencies. Therefore, an electrolyte
solution with a higher concentration of vanadium has a higher ionic
resistance than that with a lower concentration of vanadium when
the sulfuric acid concentration is the same.
3.1.3. Effect of the sulfuric acid concentration
The sulfuric acid in the electrolyte solution affects the resistance
of the electrolyte, as can be observed in Fig. 5. The highest voltage of
the four-pole cell occurs when the electrolyte solution contains
3 mol L1 sulfuric acid and 1.5 mol L1 vanadium. When the elec-
trolyte solution with a low concentration of sulfuric acid is
employed, the voltage of the four-pole cell is higher than that with
a high concentration of sulfuric acid at the same SOC. In addition,
the results indicate that the voltage of the four-pole cell is dispro-
portionate to the concentration of sulfuric acid because the corre-
lation between the conductivity of a solution and its concentration
of sulfuric acid forms an asymmetrical inverse parabola, caused by
the self-obstruction of ions. Therefore, the highest conductivity is
measured at a sulfuric acid concentration of 5e6 mol L1,
depending on the important factor of temperature [37].
The voltage efficiency in the charging process can be calculated
using the same method applied in the study of the vanadium
concentration effect. The results reveal that the voltage efficiency of
electrolyte solutions with sulfuric acid concentrations of 3 mol L1,
4 mol L1 and 5 mol L1 and identical vanadium concentrations of
1.5 mol L1 are 92.90%, 93.39% and 93.66%, respectively, when the
same configuration of the electrochemical cell, operating condi-
tions and charging current are employed. Consequently, the voltage
efficiency of an electrolyte solution with a high concentration of
sulfuric acid is higher than that of an electrolyte solution with a low
concentration of sulfuric acid due to its lower ionic resistance.
A higher voltage efficiency is obtained in an electrochemical cell
when the ionic resistance of the electrolyte solution is lower.
Therefore, utilizing an electrolyte with a low vanadium concen-
tration and high sulfuric acid concentration can increase the effi-
ciency of the VRB system. However, an electrolyte with a high
sulfuric acid concentration increases the chance of the precipitation
of the electrolyte solution and the crystallization of vanadium ions.
Moreover, the viscosity of the electrolyte solution will increase,
leads to a greater pumping power requirement. Additionally, the
01.00
Positive 1.0 mol L
1.5 mol L
0.80 2.0 mol L
Voltage of four-pole cell (V)
0.60
0.40
0.20
00.00
01.00 00.00 0.20 0.40 0.60 0.80 01.00
SOC
 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157 155

0.80 0.80
Negative Positive 3 mol L
3 mol L
0.70 0.70 4 mol L
4 mol L
5 mol L 5 mol L
Voltage of four-pole cell (V)

Voltage of four-pole cell (V)

0.60 0.60

0.50 0.50

0.40 0.40

0.30 0.30

0.20 0.20

0.10 0.10

00.00 00.00
00.00 0.20 0.40 0.60 0.80 01.00 00.00 0.20 0.40 0.60 0.80 01.00

SOC SOC

Fig. 5. Effect of the sulfuric acid concentration on the correlation between the voltage of the four-pole cell and the SOC for electrolyte solutions with sulfuric acid concentrations of
3 mol L1, 4 mol L1 and 5 mol L1, a vanadium concentration of 1.5 mol L1, and a constant current through the four-pole cell of 30 mA.

Fig. 6. The difference between the four-pole cell and the two-pole cell.

excessive free ions in the electrolyte solution may lead to an elec-
trolyte imbalance, which may lead the redox reaction in the posi-
tive half-cell to be shifted back to V(IV) instead of V(V) in the
charging process, although the details of this process require
further investigation. Finally, the utilization of an electrolyte solu-
tion with a concentration of sulfuric acid higher than that which
has the highest conductivity is unbeneficial.
The advantages of the four-pole cell method include the ability
to take measurements in real-time and its simplicity. Furthermore,
the method is precise because only the characteristics of the elec-
trolyte solution are measured, and the polarization effect does not
generate interference, as opposed to the two-pole cell method, in
which the measured voltage is a function of the electrical resistance
of all connections and the electrolyte solution. However, some
conductivity meters still utilize the two-pole cell method, so
interference that impacts the measurement leading to error is still
an issue. The differences between the four-pole cell and two-pole
cell are illustrated in Fig. 6. It is noted that the two-pole cell con-
ductivity can be acceptable for the determination of the electrolyte
conductivity in case where electroactive species are not present. In
the case of VRB, the use of two-pole is highly arguable since Fara-
daic reactions are occurring and thus, the kinetic overpotential has
a large contribution to the overall overpotential.
Fig. 7 demonstrates the relation of the measured SOC and
electrolyte imbalance of the VRB. To identify the electrolyte
imbalance, the SOCs of positive (SOCp) and negative (SOCn) are
measured and compared. When SOCp is equal to SOCn, the elec-
trolyte solution is balanced; otherwise it becomes unbalanced. An
imbalance of the positive side is occurred when SOCp is higher than
SOCn, whereas that of the negative side is observed when SOCn is
higher than SOCp. Both types of the imbalance result in a reduction
of SOC useable region as shown in Fig. 7. It is noted that for the
electrolyte balance condition, the SOC useable region of both the
electrolytes sides is in a range of 0 and 1.
According to the results of the experiments, the capacity and
performance of the VRB system will be different when the com-
ponents of the electrolyte solution are different. Therefore, the ef-
fect of the electrolyte composition on the capacity and efficiency of
the VRB system are investigated in the next section.
3.2. Effect of electrolyte composition on capacity and efficiency
The capacity is the energy that can be stored in a VRB system.
Generally, the unit of capacity is kilowatt*hour (kW$h), which can
156 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157

(a) 1.80

1.60

1.40

OCV (V)
1.20

1.0 mol L
1.5 mol L
1.00 2.0 mol L
Nernst Eq.

0.00
00.00 0.20 0.40 0.60 0.80 01.00

SOC

Fig. 7. Electrolyte imbalance at the positive and negative sides of VRB. (b) 1.80

also be considered as the product of the electrical charge (ampere 1.60

second) and voltage. The electrical charge is determined by the
volume and concentration of the electrolyte solution [38e40], with
one mole of vanadium in an electrolyte solution having an electrical 1.40
OCV (V)

charge of 96,485 A s (Faraday's constant). Consequently, an elec-

trolyte solution with a greater concentration of vanadium can store
more energy. However, the solubility of vanadium salt in a sulfuric 1.20
acid solution is another factor that needs to be taken into account.
Many research groups have tried to improve upon the energy 3 mol L
4 mol L
density of a VRB, which is quite low at the present [41e43]. 1.00
5 mol L
Obtaining the highest capacity electrolyte solution is an objective of Nernst Eq.
considerable research. The operating temperature and concentra- 0.00
tion of the sulfuric acid solution are other important factors that 00.00 0.20 0.40 0.60 0.80 01.00

need to be considered, as both factors impact the solubility of va- SOC

nadium. Moreover, an electrolyte solution with a higher concen-
tration of sulfuric acid can have a lower ionic resistance, although
side reactions are also more likely to occur.
To consider the voltage of the VRB system that influences the
capacity, the correlation between the OCV and SOC that is described
by the Nernst equation is investigated. The operating temperature
and the concentrations of vanadium and sulfuric acid affect the
voltage of the VRB system. The operating temperature is assumed
to be constant. The concentration of vanadium is considered in
terms of the SOC, and the concentration of sulfuric acid is consid-
ered in terms of the proton concentration. Therefore, the effects of
the concentrations of vanadium and sulfuric acid in an electrolyte
solution on the capacity of the VRB system are investigated.
The results indicate that the concentration of vanadium in an
electrolyte solution negligibly affects the correlation between the
OCV and SOC, as shown in Fig. 8(a). This shows that the alteration of
the vanadium concentration in an electrolyte solution does not
affect the voltage of the VRB system in term of OCV but impacts the
solution's conductivity, leading to a change in the voltage of the
VRB system when the current is applied, as well as the efficiency of
the VRB cell.
The results show that the alteration of the sulfuric acid con-
centration affects the correlation between the OCV and SOC, as
shown in Fig. 8(b). The correlation for the electrolyte solution with
a concentration of sulfuric acid of 3 mol L1 is the lowest line,
indicating that the capacity of this solution is the lowest. The cor-
relations of the solutions with concentration of 4 mol L1 and
5 mol L1 are close together, indicating that the addition of sulfuric
acid beyond this point does not increase the capacity of the VRB
system. This composition corresponds to the highest conductivity
Fig. 8. Correlation between the OCV and SOC of electrolyte solutions with (a) vana-
dium concentrations of 1.0 mol L1, 1.5 mol L1 and 2.0 mol L1 and a sulfuric acid
concentration of 4 mol L1 and (b) sulfuric acid concentrations of 3 mol L1, 4 mol L1
and 5 mol L1 and a vanadium concentration of 1.5 mol L1.
composition, as explained above. It is noted that the OCVs predicted
by Nernst equation at different SOCs are also given in Fig. 8(a) and
(b) to demonstrate the difference between the experimental data
and theoretical results.
Consequently, the capacity of the VRB system can be increased
by increasing the concentration of vanadium in the electrolyte so-
lution, which increases the electrical charge of the system, or
increasing the concentration of sulfuric acid, which increases the
cell voltage. However, the solubility of vanadium salt in the elec-
trolyte solution, the conductivity and stability of the solution, the
required power of the pumps and the risk of side reactions must
also be taken into account.
4. Conclusions
In this study, the measurement of the SOC using the voltage of a
four-pole cell with a constant current applied to the electrolyte
solution is investigated. In the developed method, the ionic resis-
tance (or conductivity) of the electrolyte solution varies when the
operating temperature and the composition of the solution are
altered. The change in the oxidation state of vanadium (or the
change of the SOC) causes a change in the free ions in the elec-
trolyte solution, leading to a change in the voltage of the four-pole
cell. The effect of the constant current applied through the four-
 K. Ngamsai, A. Arpornwichanop / Journal of Power Sources 298 (2015) 150e157
pole cell and the concentrations of vanadium and sulfuric acid in
the electrolyte solution on the voltage of the four-pole cell and the
SOC in negative and positive electrolyte solutions are investigated.
The results indicate that when the applied constant current for the
four-pole cell is too low, the SOC cannot be determined because the
difference in the four-pole cell voltage for each SOC is quite low, but
when the applied constant current through the four-pole cell is too
high, the measurement of the SOC experiences interference, lead-
ing to inaccurate measurements. From the experimental data, the
optimal constant current for the four-pole cell is approximately
14 mA cm2. To study the effect of the vanadium and sulfuric acid
concentrations, concentrations in the practical region were inves-
tigated, with the results showing that the four-pole cell method can
be utilized in the whole region of practical electrolyte solutions. The
composition of an electrolyte solution affects both the capacity and
efficiency of a VRB system. The results show that the capacity de-
pends on the concentration of vanadium, while the concentration
of sulfuric acid affects the voltage of the VRB cell. Decreasing the
vanadium concentration or increasing the sulfuric acid concentra-
tion can increase the efficiency of the VRB system owing to the
decrease in ionic resistance (or increase in conductivity) of the
electrolyte solution.
Acknowledgments
Financial support provided by the 90th Anniversary of Chula-
longkorn University Fund (Ratchadaphiseksomphot Endowment
Fund), Cellennium (Thailand) Co., Ltd. and the Thailand Research
Fund is gratefully acknowledged.
References
[1] J. Baker, Energy Policy 36 (2008) 4368e4373.
[2] B. Christian, R. Alfred, in: IEEE International Conference on Sustainable Energy
Technologies, ICSET 2008, Singapore, 2008, pp. 696e701.
[3] K.C. Divya, J. Østergaard, Electric Power Syst. Res. 79 (2009) 511e520.
[4] L. Xuanli, L. Zhengzhong, X. Jingyu, W. Zenghua, Z. Wentao, C. Liquan,
Q. Xinping, J. Phys. Chem. B 109 (2005) 20310e20314.
[5] Z. Mai, H. Zhang, X. Li, S. Xiao, H. Zhang, J. Power Sources 196 (2011)
5737e5741.
[6] A.Z. Weber, M.M. Mench, J.P. Meyers, P.N. Ross, J.T. Gostick, Q. Liu, J. Appl.
Electrochem. 41 (2011) 1137e1164.
[7] C. Ponce de Leo
n, A. Frías-Ferrer, J. Gonz

nto, F.C. Walsh,
alez-García, D.A. Sza
J. Power Sources 160 (2006) 716e732.
[8] A.A. Shah, M.J. Watt-Smith, F.C. Walsh, Electrochim. Acta 53 (2008)
8087e8100.
[9] J.-G. Kim, S.-H. Lee, S.-I. Choi, C.-S. Jin, J.-C. Kim, C.-H. Ryu, G.-J. Hwang, J. Ind.
157
Eng. Chem. 16 (2010) 756e762.
[10] C. Sun, J. Chen, H. Zhang, X. Han, Q. Luo, J. Power Sources 195 (2010) 890e897.
[11] B. Caglar, P. Fischer, P. Kauranen, M. Karttunen, P. Elsner, J. Power Sources 256
(2014) 88e95.
[12] T. Mohammadi, M. Skyllas-Kazacos, J. Membr. Sci. 107 (1995) 35e45.
[13] B. Tian, C.W. Yan, F.H. Wang, J. Membr. Sci. 234 (2004) 51e54.
[14] Q. Luo, H. Zhang, J. Chen, D. You, C. Sun, Y. Zhang, J. Membr. Sci. 325 (2008)
553e558.
[15] K.-L. Huang, X.-g. Li, S.-q. Liu, N. Tan, L.-q. Chen, Renew. Energy 33 (2008)
186e192.
[16] D. Chen, S. Wang, M. Xiao, D. Han, Y. Meng, J. Power Sources 195 (2010)
7701e7708.
[17] C. Flox, J. Rubio-García, M. Skoumal, T. Andreu, J.R. Morante, Carbon 60 (2013)
280e288.
[18] M. Skyllas-Kazacos, M. Kazacos, J. Power Sources 196 (2011) 8822e8827.
[19] A. Tang, J. Bao, M. Skyllas-Kazacos, J. Power Sources 196 (2011) 10737e10747.
[20] L. Liu, J. Xi, Z. Wu, W. Zhang, H. Zhou, W. Li, X. Qiu, J. Appl. Electrochem 42
(2012) 1025e1031.
[21] M. Skyllas-Kazacos, L. Goh, J. Membr. Sci. 399e400 (2012) 43e48.
[22] S. Rudolph, U. Schro €der, I.M. Bayanov, J. Electroanal. Chem. 703 (2013) 29e37.
[23] A.H. Whitehead, M. Harrer, J. Power Sources 230 (2013) 271e276.
[24] J. Qiu, M. Zhai, J. Chen, Y. Wang, J. Peng, L. Xu, J. Li, G. Wei, J. Membr. Sci. 342
(2009) 215e220.
[25] X. Teng, Y. Zhao, J. Xi, Z. Wu, X. Qiu, L. Chen, J. Membr. Sci. 341 (2009)
149e154.
[26] F. Wang, J.M. Sylvia, M.M. Jacob, D. Peramunage, J. Power Sources 242 (2013)
575e580.
[27] K. Jihoon, J. Jae-Deok, K. Seung-Yeop, Eletrochem. Commun. 38 (2014) 6870.
[28] H. Al-Fetlawi, A.A. Shah, F.C. Walsh, Electrochim. Acta 55 (2010) 3192e3205.
[29] X. Ma, H. Zhang, C. Sun, Y. Zou, T. Zhang, J. Power Sources 203 (2012)
153e158.
[30] M.J. Watt-Smith, P. Ridley, R.G.A. Wills, A.A. Shah, F.C. Walsh, J. Chem. Technol.
Biotechnol. 88 (2013) 126e138.
[31] A. Tang, J. Bao, M. Skyllas-Kazacos, J. Power Sources 248 (2014) 154e162.
[32] J. Escudero-Gonz
alez, P.A. Lo
pez-Jime
nez, J. Power Sources 251 (2014)
243e253.
[33] D. Aaron, C.N. Sun, M. Bright, A.B. Papandrew, M.M. Mench, T.A. Zawodzinski,
ECS Electrochem. Lett. 2 (2013) A29eA31.
[34] E. Ventosa, M. Skoumal, F.J. V
azquez, C. Flox, J.R. Morante, J. Power Sources
271 (2014) 556e560.
[35] Chemical Sensor and Biosensor Group, Conductivity Theory and Practice,
University of Crete, Greece, 2010. http://www.analytical-chemistry.uoc.gr/
files/items/6/618/agwgimometria_2.pdf (January 27, 2014).
[36] T. Sukkar, M. Skyllas-Kazacos, J. Membr. Sci. 222 (2003) 235e247.
[37] DKL Engineering, Inc, Conductivity of Sulfuric Acid, 2005e2011. http://www.
sulphuric-acid.com/techmanual/Properties/properties_acid_cond.htm
(January 27, 2014).
[38] M. Rychcik, M. Skyllas-Kazacos, J. Power Sources 22 (1988) 59e67.
[39] M. Gattrell, J. Qian, C. Stewart, P. Graham, B. MacDougall, Electrochim. Acta 51
(2005) 395e407.
[40] P. Qian, H. Zhang, J. Chen, Y. Wen, Q. Luo, Z. Liu, D. You, B. Yi, J. Power Sources
175 (2008) 613e620.
[41] M. Kazacos, M. Cheng, M. Skyllas-Kazacos, J. Appl. Electrochem. 20 (1990)
463e467.
[42] G. Oriji, Y. Katayama, T. Miura, J. Power Sources 139 (2005) 321e324.
[43] F. Rahman, M. Skyllas-Kazacos, J. Power Sources 189 (2009) 1212e1219.