International Journal of Engineering, Management & Sciences (IJEMS)

ISSN-2348 3733, Volume-2, Issue-5, May 2015

A New Control Scheme for Battery-Supercapacitor

Hybrid Energy Storage System for Standalone
Photovoltaic Application
Akanksha Thakur, Prof. L. M. Saini

 Np, Ns Number of PV cells in series and parallel

Abstract In this paper, a battery-supercapacitor (SC) respectively.
hybrid Energy Storage System (ESS) is employed in a E, Eo Battery no load voltage and battery constant
standalone photovoltaic (PV) system to maintain continuity in voltage respectively.
the supply. The battery ESS is characterized by high energy r Internal resistance of battery
density, low power density, degradation due to frequent and
i Battery current
partial charge/discharge cycles. By incorporating SC ESS which
has high charge/discharge rates, a system having both high
SOC State of Charge of battery
energy and power capabilities can be designed. In the presented A, B Battery constant parameters
control scheme, high frequency current reference along with Q Capacity of battery (Ah)
battery error current is utilized as the reference for the SC ESS V_ref , Vout System reference voltage and system output
resulting in an improved voltage profile as compared to that of voltage respectively.
the conventional methods. Also, an improvement in the life span, I_ref Total reference current generated
reduction in size and cost of the battery and better Ib_ref, Isc_ref Reference current generated for battery and
generation-demand power balance can be obtained. The control supercapacitor respectively.
scheme is validated using Matlab/Simulink models under
Ihfc_ref, Ib_err High frequency current reference and
varying conditions of irradiation and load.
battery error current respectively
Index Terms battery, bidirectional dc-dc converter, hybrid Pb_uncomp Uncompensated battery power
energy storage system, standalone photovoltaic system, Kp_pv, Ki_pv PI controller gains for boost converter
supercapacitor. control
Kp_b, Ki_b PI controller gains for bidirectional dc-dc
NOMENCLATURE converter1 control
Kp_sc, Ki_sc PI controller gains for bidirectional dc-dc
Vpv, Vb, Vsc Terminal voltage across the PV module,
converter2 control
battery and supercapacitor respectively.
fsw Switching frequency of the converters used
Ipv, Ib, Isc Terminal current of the PV module, battery
and supercapacitor respectively.
Voc, Isc Open circuit voltage and short circuit
I. INTRODUCTION
current of the PV module respectively.
Ppv Instantaneous power delivered by PV panel The low emission and low maintenance benefits of solar
Pmpp Maximum power delivered by the PV photovoltaic (PV) generation are offset by high cost and the
module intermittent nature of solar energy. Fortunately, the increased
Lpv, Lb, Lsc Filter inductance of PV panel, battery and activity in the PV market has led to decreasing costs due to
SC respectively. increased production of solar panels and improvements in the
SW1-SW5 Control switches using IGBT technology technology. The intermittency issue for standalone PV
C Filter Capacitance systems can potentially be addressed by large-scale
R System Resistive load interconnections with complementary energy sources such as
Iph Photo-current or light generated current wind or natural gas operated generation units, however the
Irs Reverse saturation current of diode most self-contained approach is to add storage to the system,
Q Charge of an electron thus creating a hybrid PV-storage system that can balance the
a Diode ideality constant
PV generation with the load. Though there are many available
K Boltzmann s constant
storage technologies including supercapacitors, pumped
Tc, T Absolute temperature of PV cell and
hydro, compressed air, superconducting magnetic and
operating temperature respectively.
thermal storage, the most cost effective and mature
technology on a small scale is battery storage. Lead-acid
Manuscript received February 20, 2015. batteries remain for the moment the most cost effective
Akanksha Thakur, Electrical Department, National Institute of solution, as well as the most mature with respect to recycling
Technology, Kurukshetra, India,
Prof. L. M. Saini, Electrical Department, National Institute of at the end-of-life. Batteries possess high energy density but
Technology, Kurukshetra, India, low power density which means it has low charge/discharge

25 www.alliedjournals.com
 A New Control Scheme for Battery- Supercapacitor Hybrid Energy Storage System for Standalone Photovoltaic
Application

rates [6]. The supercapacitors are a new form of energy

storage which stores energy by means of static charge. When
compared to batteries, SC possesses high power density but II. THE SYSTEM STRUCTURE
low energy density which means it has high charge/discharge
rates [7]. In Table I performance of the battery and the Ppv
supercapacitor is compared [5]. When only battery is PV Panel Boost converter

employed as storage, then it has to be oversized to take care of

the peak load demand which may be required for a very short SW1

duration of time. An ideal ESS in a standalone PV system MPPT

controller
should have both high energy and power capacities to handle
situations such as solar irradiation changes and load step
changes. Thus, the objective is to harness the advantages of Pb
Bidirectional
both the storage systems to design a hybrid energy storage Battery dc-dc
converter1
Load

system with high power and energy density. By utilizing a

battery-supercapacitor hybrid ESS the following merits can SW2, SW3
be achieved: i) longevity of battery life ii) reduction in battery Power

size and hence the cost iii) reduction in battery stress and iv)
management
Controller

improvement in balance between generation and load SW4, SW5

demand. Various control strategies have been reported in Psc

Bidirectional

literature [1] [11], for coordinating power sharing between Supercapacitor dc-dc
converter2

battery and super capacitor. N. Mendis et al. [3] has addressed

the advantages of including a SC to a battery storage system in
a wind based hybrid remote area power supply (RAPS)
Fig. 1. Block diagram for the PV/battery/supercapacitor hybrid system
system. An energy management algorithm (EMA) has been structure
established between the battery storage system and SC to
operate both the storage systems in a designated manner. A. Overall circuit description and modelling
Authors in [6] have utilized hybrid ESS for microgrid Lpv
applications. Amine Lahyani et al. [10] have presented an
Diode
analysis on the reduction in battery stress with the help of SC. Ipv
Authors in [11], have shown that hybrid ESS reduces the PV SW1 R Load
Vpv C Vout
battery cost and enhances the overall system ef ciency. The Panel

fundamental idea of all the control strategies reported so far is

that battery provides low frequency component of power and
SC provides high frequency component of power
momentarily. In this paper, a new control strategy is
demonstrated for hybrid ESS consisting of PV, battery, SC
and load. The presented method is based on decoupling of SW2 SW4
low and high frequency current components. It utilizes the Lb Lsc
error current of battery in addition to the high frequency
current reference to control the SC while the rest of the Ib Isc

current is used as reference to control the battery ESS. This Battery Vb

SW3 SW5 Super-
Vsc capacitor
paper is organized as follows. In Section II system structure
and its modelling is given. The proposed control structure is
designed and discussed in Section III. Simulation results are
reported in Section IV.
Fig. 2. Typical configuration of standalone PV system using hybrid energy
TABLE I. Comparison of lead-acid battery and supercapacitor
storage
Lead acid battery Supercapacitor
A standalone PV system consisting of battery and SC
Specific energy density 10 - 100 1 - 10 arrangement is shown in Fig. 1. The PV panel is connected to
(Wh/kg) the load using a boost converter. The function of the boost
Specific power density <1000 >10000 converter used here is to extract the maximum power from PV
(W/kg) panel using incremental-conductance maximum power point
Life cycle 103 cycles 106-108 cycles tracking (MPPT) algorithm. Hybrid ESS is connected to the
load using bi-directional DC/DC converters. Hybrid ESS is
Charge/discharge 70 85% 85 98%
Efficiency
used to maintain the constant DC voltage (Vout) even if there is
a mismatch between generation and demand. When the
Charge time 1 5 hr. 0.3 30 sec. demand is more than generation, Vout drops from its reference
Discharge time 0.3 3 hr. 0.3 30 sec. value, consequently Hybrid ESS will discharge to provide the
surplus demand. Similarly, when the demand is less than the
generation, Vout increases from its reference value,

26 www.alliedjournals.com
 International Journal of Engineering, Management & Sciences (IJEMS)
ISSN-2348 3733, Volume-2, Issue-5, May 2015

consequently Hybrid ESS will be charged to absorb the

surplus power. Buck-boost converter is used as a
bi-directional converter to facilitate the bi-directional power
ow between the load and Hybrid ESS. In this paper, the total
load is represented as DC load with resistance R. In Fig. 2 Vpv,
Vb and Vsc are PV panel, battery and SC voltages respectively,
Ipv, Ib and Isc are PV panel, battery and SC current
respectively, Lpv, Lb and Lsc are PV panel, battery and SC lter
inductance respectively, Vout is output voltage, C is lter
capacitance and R is load resistance. SW1, SW2, SW3, SW4
and SW5 are control switches.

B. PV Module
The photovoltaic (PV) cell is basically a p-n junction
fabricated in a thin wafer of semiconductor. The solar energy
is directly converted to electricity through photovoltaic effect.
The equivalent circuit of PV module is as shown in Fig. 3. PV
cell exhibits a nonlinear P-V and I-V characteristics which
vary with cell temperature and solar irradiance. The P-V and
I-V characteristics for an irradiation of 1000 W/m2 are given
in Fig. 4,5 respectively.

Fig. 4. P-V characteristics of the photovoltaic module

Fig. 3. Equivalent circuit of photovoltaic module

The nonlinear voltage-current characteristic equation of PV

module is given in [12] as,
Vp I pv R s
q(  )
Ns Np
N pV pv / N s  I pv Rs
I pv  N p I ph  N p I rs {e akT  1} 
Rsh
(1)
Various parameters can be obtained from speci cations gi ven
in [13].

C. MPPT Controller
The switching of the boost converter is controlled by the
MPPT controller which utilizes the incremental-conductance
algorithm for its operation. The flowchart of the algorithm is Fig. 5. I-V characteristics of the photovoltaic module
shown in Fig. 6. The method exploits the assumption that the The above equation could be written as
ratio of change in output conductance is equal to the negative
output conductance. We have, Ipv/ Vpv = - Ipv/Vpv (5)
Ppv = Vpv Ipv (2)
Applying the chain rule for the derivative of products yields to The MPPT regulates the PWM control signal of the dc-dc
boost converter until the condition,
Ppv/ Vpv = [ (VpvIpv)]/ Vpv (3)
( Ipv/ Vpv) + (Ipv/Vpv) = 0 (6)
At maximum power point, as
is satisfied.
Ppv/ Vpv = 0 (4)

27 www.alliedjournals.com
 A New Control Scheme for Battery- Supercapacitor Hybrid Energy Storage System for Standalone Photovoltaic
Application

Start

PV array voltage and

current sampling

Calculate
di= i-io
dv= v-vo

no yes
dv= 0?

yes yes
di/dv= -i/v? di= 0?
Fig. 8. Model of the battery
no no


 B idt


KQ
E  Eo   Ae (7)
yes yes
di/dv> -i/v? di> 0?
Q  idt
no no

Vb  E  Ri (8)
D= D- dD D= D+ dD D= D+ dD D= D- dD

 idt ) *100
vo = v
1
% SOC  (1  (9)
io= i

Q
End

The battery model characteristics for discharging currents of

Fig. 6. Flowchart for incremental-conductance algorithm
1.5 and 3A is as shown in Fig. 9. Other parameters can be
obtained using [14].
Fig. 7 shows the PV power obtained for an irradiation of
520W/m2 when MPPT algorithm is enabled at t= 0.11 s. This
leads to an increment in the PV power is from 251.25W to
488.65W.

Fig. 9. Discharge characteristics of the battery model

A. Supercapacitor
The SC is an emerging technology in the eld of energy
storage systems. The SC resembles a regular capacitor with
the exception that it offers very high capacitance in a small
package. Energy storage is by means of static charge rather
than of an electro-chemical process that is inherent to the
Fig. 7. PV power obtained for a constant irradiation of 520W/m2 when
MPPT algorithm is enabled at t= 0.11 s battery. The charge characteristics of the model is shown in
Fig. 10.
D. Battery
Lead acid batteries are the most commonly used batteries in a
PV system. It is modelled using a simple controlled voltage
source in series with a constant resistance. The battery is
modelled as a nonlinear voltage source as shown in the Fig. 8
using the following equations [14],

28 www.alliedjournals.com
 International Journal of Engineering, Management & Sciences (IJEMS)
ISSN-2348 3733, Volume-2, Issue-5, May 2015

able to track the Ib_ref instantly.

Therefore, the uncompensated battery power is given as
Pb_uncomp = (Ihfc_ref +Ib_err)*Vb (12)
This uncompensated battery power is to be compensated by
SC. Therefore, the reference current of SC is taken as
Isc_ref = Pb_uncomp/Vsc (13)
Isc_ref is compared with the actual SC current (Isc), and the
error is fed to the PI controller. The PI controller generates the
required duty ratios. These duty ratios are given to the PWM
generator to generate switching pulses corresponding to SC
switches (SW4, SW5).

IV. SIMULATION RESULTS

The presented control strategy is validated using
Matlab/Simulink for the following cases: i) variation in solar
irradiation and ii) variation in load demand. The objective is
to maintain the output voltage at Vref = 50 V. The nominal
parameters of the system are given in Table II. The initial
Fig. 10. Charge characteristics of the supercapacitor State of Charge (SOC) of the battery is set at 50%. The PI
controller gains of the outer voltage loop, inner current loop
of battery and inner current loop of SC are Kp_pv = 1.477, Ki_pv
III. THE CONTROL STRUCTURE = 3077, Kp_b = 0.043, Ki_b = 0.65, Kp _sc = 0.41 and Ki _sc =
14800 respectively.

TABLE II. Parameters used in simulation

System Specifications
Component
PV array Voc= 37.5V, Isc= 33.2A, Pmpp= 882.7W
Battery Lead acid- 24V, 14Ah
SC 29F, 32V
Ib_err
DC-DC Converters Lpv= 0.352mH, Lb= 0.3mH, Lsc= 0.355mH,
C= 390uF, Vout= 50V, R= 5 , f sw= 16kHz

Fig. 11. New Control scheme

Fig. 11 shows the block diagram of the proposed control

strategy. The objective of this algorithm is to reduce the stress
on battery and hence increase the lifetime of battery. In this
algorithm, the DC output voltage (Vout) is compared with
reference voltage (Vref), and the error is fed to the
proportional integral (PI) controller. The PI controller
generates the total current required (Iref) from hybrid energy
storage system. Iref is separated into low frequency component
and high frequency component as,
Ib_ref = lowpassfilter(Iref) (10)
Ihfc_ref = Iref Ib_ref (11)
The low frequency component of current is given as the
Fig. 12. Simulation results for an increase in solar irradiation showing PV
reference current to battery. Ib_ref is compared with the actual power, output voltage and load power using the proposed control scheme
battery current (Ib), and the error (Ib_err) is given to the PI
controller. The PI controller generates the duty ratios. These
duty ratios are fed to the PWM generator to generate
switching pulses corresponding to battery switches (SW2,
SW3). Due to the slow dynamics of the battery, Ib may not be

29 www.alliedjournals.com
 A New Control Scheme for Battery- Supercapacitor Hybrid Energy Storage System for Standalone Photovoltaic
Application

Fig. 14. Simulation results for a decrease in solar irradiation showing PV

power, output voltage and load power using the proposed control scheme

Fig. 15. Simulation results for a decrease in solar irradiation showing the
output voltage using the conventional control scheme

 Decrease in irradiation:
In the study, initially maximum PV power extracted
with irradiation 550W/m2 and T= 25oC is 388W. At
t= 1s, the irradiation is decreased to 450W/m2 . At
this value of irradiance, the power supplied by the
Fig. 13. Simulation results for an increase in solar irradiation showing the PV module is 315W , but the load demand is 502W.
output voltage using the conventional control scheme To maintain the voltage at its reference value, the
surplus demand is to be supplied by the storage
A. Variation in solar irradiation
system. Fig. 14 shows the waveforms of PV power,
 Increase in irradiation: output voltage and the load power using the
In the study, initially maximum PV power extracted proposed control scheme. Using the conventional
with irradiation 450W/m2 and T= 25oC is 315W. At control scheme [6], the output voltage waveform
t= 1s, the irradiation is increased to 650W/m2 . At obtained for the same conditions is shown in Fig. 15.
this value of irradiance, the power supplied by the It is observed that during the transient, the voltage is
PV module is 465W , but the load demand is 502W. suddenly decreased to 45V. It can be seen that the
To maintain Vout at its reference value, the surplus percentage decrease in voltage is reduced from 10%
supply is to be absorbed by the storage system. Fig. to 0% using the new method. Unlike the
12 shows the waveforms of PV power, output conventional methods, the sudden change in the
voltage and the load power using the proposed voltage due to the slow dynamics of the battery is
control scheme. Using the conventional control removed by utilizing the battery error current.
scheme [6], the output voltage waveform obtained
for the same conditions is shown in Fig. 13. It is B. Variation in load demand
observed that during the transient, the voltage is  Increase in load power:
suddenly increased to 58.1V. It can be seen that the In the study, initially PV panel is operated at
percentage increase in voltage is reduced from constant irradiation of 550W/m2, T= 25oC and R=
16.2% to 0% using the new method. Unlike the 5 . At t= 1s, the load demand is suddenly increased
conventional methods, the sudden change in the by decreasing R to 2.7 . At this instant, the power
voltage due to the slow dynamics of the battery is supplied by the PV module is 390W, but the load
removed by utilizing the battery error current. demand is 920W. To maintain Vout at its reference
value, the surplus demand is to be supplied by the
storage system. Fig. 16 shows the waveforms of PV
power, output voltage and the load power using the
proposed control scheme. Using the conventional
control scheme [6], the output voltage waveform
obtained for the same conditions is shown in Fig. 17.
It is observed that during the transient, the voltage is
suddenly decreased to 44V. It can be seen that the
percentage decrease in voltage is reduced from 12%
to 0% using the new method. Unlike the
conventional methods, the sudden change in the
voltage due to the slow dynamics of the battery is
removed by utilizing the battery error current.

30 www.alliedjournals.com
 International Journal of Engineering, Management & Sciences (IJEMS)
ISSN-2348 3733, Volume-2, Issue-5, May 2015

Fig. 16. Simulation results for an increase in load demand showing load
power, output voltage and PV power using the proposed control scheme

Fig. 18. Simulation results for a decrease in load demand showing load
power, output voltage and PV power using the proposed control scheme

Fig. 17. Simulation results for an increase in load demand showing the
output voltage using the conventional control scheme

 Decrease in load power:

In the study, initially PV panel is operated at
constant irradiation of 550W/m2 , T= 25oC and R=
3.3 . At t= 1s, the load demand is suddenly
decreased by increasing R to 10 . At this instant,
the power supplied by the PV module is 390W, but
the load demand is 250W. To maintain Vout at its
reference value, the surplus supply is to be absorbed
by the storage system. Fig. 18 shows the waveforms Fig. 19. Simulation results for a decrease in load demand showing the output
voltage using the conventional control scheme
of PV power, output voltage and the load power
using the proposed control scheme. Using the
conventional control scheme [6], the output voltage
waveform obtained for the same conditions is shown
in Fig. 19. It is observed that during the transient, the
voltage is suddenly increased to 55.9V. It can be
seen that the percentage increase in voltage is
reduced from 11.8% to 0% using the new method.
Unlike the conventional methods, the sudden change
in the voltage due to the slow dynamics of the battery
is removed by utilizing the battery error current.

31 www.alliedjournals.com
 A New Control Scheme for Battery- Supercapacitor Hybrid Energy Storage System for Standalone Photovoltaic
Application

V. CONCLUSION
In this paper, a new control strategy for coordinating the
power sharing between battery and SC hybrid system has been
established. The proposed method is based on extraction of
low frequency and high frequency component of power. It
utilizes error current of battery to control the SC unlike the
conventional methods. The low frequency power component
was managed by battery storage system while the high
frequency component along with the error current of the
battery was managed by SC. The simulation results show that
the proposed method is able to maintain the constant DC
output voltage during change in PV generation and load
demand. Hence, power balance between the generation and
demand is achieved.

REFERENCES
[1] A. M. Gee, F. V. P. Robinson, and R. W. Dunn (2013). "Analysis of
battery lifetime extension in a small-scale wind-energy system using
supercapacitors", IEEE Trans. Energy Convers., 28(1), 24-33.
[2] R. Karki and Billinton (2001). "Reliability/cost implications of pv and
wind energy utilization in small isolated power systems", IEEE Trans.
Energy Convers., 16(4), 368-373.
[3] N. Mendis, K. M. Muttaqi, and S. Perera (2012). "Active power
management of a supercapacitor-battery hybrid energy storage system
for standalone operation of dfig based wind turbines", in Proc. IEEE
Industry Applications Society Annual Meeting (lAS), 1-8.
[4] W. Kellogg, M. H. Nehrir, G. Venkataramanan, and V. Gerez (1998).
"Generation unit sizing and cost analysis for stand-alone wind,
photovoltaic, and hybrid wind/pv systems , IEEE Trans. Energy
Convers., 13(1), 70-75.
[5] M. Glavin, P. Chan, S. Armstrong, and W. Hurley (2008). "A
stand-alone photovoltaic supercapacitor battery hybrid energy storage
system", in Proc. 13th IEEE Power Electronics and Motion Control
Conference, 1688-1695.
[6] R. Sathish Kumar, K. Sathish Kumar, and M. K. Mishra (2012).
"Dynamic energy management of micro grids using battery super
capacitor combined storage", in Proc. Annual IEEE India Conference
(INDICON), 1078-1083.
[7] Z. Guoju, T. Xi sheng, and Q. Zhiping (2010). "Research on battery
supercapacitor hybrid storage and its application in microgrid", in
Proc. IEEE Power and Energy Engineering Coriference, 1-4.
[8] S. Vazquez, S. M. Lukic, E. Galvan, L. G. Franquelo, and J. M.
Carrasco (2010). "Energy storage systems for transport and grid
applications", IEEE Trans. Ind. Electron, 57(12), 3881-3895.
[9] W. Li and G. Joos (2008). "A power electronic interface for a battery
supercapacitor hybrid energy storage system for wind applications", in
Proc.IEEE Power Electronics Specialists Conference, 1-4.
[10] A. Lahyani, P. Venet, A. Guermazi, and A. Troudi (2013).
"Battery/supercapacitors combination in uninterruptible power supply
(ups)", IEEE Trans. Power Electron, 28(4), 1509-1522.
[11] L. Wei, G. Joos, and B. L. (2010). "Real-time simulation of a wind
turbine generator coupled with a battery supercapacitor energy storage
system", IEEE Trans. Power Electron, 57(4), 1137-1145.
[12] H.-L. Tsai, C.-S. Tu, and Y-J. Su (2008). "Development of generalized
photovoltaic model using matlab/simulink", in Proc. the World
Congress on Engineering and Computer Science, 846-851.
[13] F. M. Gonzalez-Longmatt (2005). "Model of photo voltaic module in
matlab", in Proc. Do Congreso Iberoamericano De Estudiantes De
Ingenieria Electrica, Electronica Y Computacion (II CIBELEC), 1- 5.
[14] O. Tremblay, L.-A. Dessaint, and A.-I. Dekkiche (2007). "A generic
battery model for the dynamic simulation of hybrid electric vehicles",
in IEEE Vehicle Power and Propulsion Conference, VPPC, 284-289.

32 www.alliedjournals.com