Advances in Engineering Research, volume 198

International Seminar of Science and Applied Technology (ISSAT 2020)

Energy Management Control Based on Standalone

Photovoltaic Battery and Supercapacitor Hybrid
Energy Storage System Using PI Controller
Soedibyo1,* Avian Lukman Setya Budi1* Sjamsjul Anam1 Mochamad Ashari1
Adi Soeprijanto1
1
Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Jawa Timur 60111,
Indonesia
*
Corresponding author. Email: dibyosoe@gmail.com, avianlukmans@gmail.com

ABSTRACT
Standalone photovoltaic system is a system that does not connect with main grid. So, it is necessary to have a backup
supply, considering day and night cycle, where there are no solar radiation at night. Backup supply is often be the grid
in the on-grid system, but in an off-grid system, backup supply has to be energy storage system, as in this paper, storage
system contains battery and supercapacitor. As stated, standalone photovoltaic with hybrid energy storage system must
have its own energy management system. The proposed method in the paper is to use PI controller as the energy
management system control. The examined aspect is the voltage and power stability in both battery and supercapacitor.
The result compared with the stability characteristics from uncontrolled energy management system. From the result,
the PI controlled system is proven to give the better result due to compensation of the load power surplus during the
maximum irradiance and increase the battery input by 162.261 W or 69.836% more power to the battery. The load has
been optimized, in both maximum irradiance by the decreasing of 178.309 W and half irradiance by the increasing of
15.166 W.

Keywords: Battery, Photovoltaic, Supercapacitor, Stability, PID

1. INTRODUCTION 1.1. Literature Review

Renewable energy is energy generated from sources that are There is research behind this research, including stand-alone
not depleted or can be replenished in human life. The most photovoltaic systems, batteries, supercapacitors, and
common examples include wind, solar, geothermal, evaluation studies, conclusions, and recommendations for
biomass, and hydropower. This is different from non- previous energy management systems.
renewable sources such as fossil fuels.
Although renewable energy is often considered a new
technology, harnessing the forces of nature has long been 1.1.1. Standalone Photovoltaic System
used for heating, transportation, lighting, and more. But over
the last 500 years, humans have increasingly turned to The concept of solar power generation is a simple concept
cheaper and dirtier sources of energy such as coal and to convert sunlight into electrical energy. Sunlight is a form
crushed gas. of energy from natural resources. Solar natural resources
From the related field, it is generally known that stand-alone have been widely used to supply electric power to
photovoltaic systems must be equipped with energy storage communication satellites via solar cells. These solar cells
systems. There are various types of energy storage today, can produce an unlimited amount of electrical energy which
includes using multiple energy stores with parallel is drawn directly from the sun, with no rotating parts and
connections. However, energy management systems for requires no fuel. Thus, the solar system is often said to be
multiple energy storage systems in parallel should function clean and environmentally friendly [1].
properly for better stability and continuity. A standalone photovoltaic system is a system that is not
connected to the main grid. Therefore, it is necessary to have
a backup supply, given the day and night cycle, where solar
radiation is at 0 W / m2 at night. The backup supply is often
networked in a grid system, but in an off-grid system the

Copyright © 2020 The Authors. Published by Atlantis Press B.V.

This is an open access article distributed under the CC BY-NC 4.0 license -http://creativecommons.org/licenses/by-nc/4.0/. 1
 Advances in Engineering Research, volume 198

backup supply must be an energy storage system, such as a 1.1.2.1. Battery.

battery, hydro pump storage, heat storage, or supercapacitor.
The voltages and currents for these standalone systems are Batteries are one of the most widely used energy storage
not sufficient for many applications, so generally a number devices in the power sector, especially in stand-alone
of solar cells are arranged in series to form a solar module. photovoltaic systems. Each battery consists of a Positive
One solar module typically consists of 28-36 solar cells, and Terminal (Cathode) and a Negative Terminal (Anode), as
produces a total DC voltage of 12 V under standard well as an electrolyte that functions as a distributor. The
irradiation conditions (Air Mass 1.5). Solar modules can be electric current output from the battery is direct current
combined in parallel or in series to increase the total voltage (DC).
and output current according to the power required for a Lead Acid battery type is a type of battery suitable for solar
particular application. panel systems. This can be seen clearly because by using
Lead Acid type batteries, users can take advantage of the
electrical energy stored in the battery (discharge) when the
1.1.2. Energy Storage System solar panels are not exposed to sunlight. Conversely, when
there is sun, the battery will be charged (charged) by solar
The energy storage system is essential for the operation of panels [2].
the power system. They ensure continuity of energy supply VRLA AGM or VRLA Gel Deep Cycle batteries are the
and increase system reliability. Energy storage systems most suitable and most widely used batteries for solar panel
come in all shapes and sizes. The size, cost, and scalability systems, both SHS (Solar Home System), SPSL (Solar
of energy storage systems depend largely on the form of Street Lighting), Solar Power Pumps, Solar Street Lights,
energy stored. Energy can be stored as potential, kinetic, PV (Solar Power Generation) Power.
chemical, electromagnetic, thermal, etc. Some forms of The reason for using VRLA AGM and VRLA Gel Deep
energy storage are more suitable for small scale systems and Cycle batteries is because they have charge cycle resistance,
some are used only for large scale storage systems. For use resistance, spillage / leakage, and are maintenance free.
example, a chemical battery is perfect for small systems It is the most preferred option for energy storage of stand-
ranging from watches and computers to building backup alone photovoltaic systems.
systems, but is still expensive when considering the From previous research work, it was chosen to use Li-ion
megawatt scale. Pumped hydropower storage, on the other batteries because there was no need for further research on
hand, which stores a large amount of energy in the form of VRLA batteries because of their better performance.
water potential energy, can only be found in large power
systems [3].
Examples of chemical energy storage systems include 1.1.2.2. Supercapacitor.
batteries, flow batteries, and fuel cells. Mechanical (kinetic
and potential) energy storage systems include pumped Supercapacitors are capacitors whose capacitance values far
storage hydropower, flywheel, and pressurized gas storage exceed other capacitors (but with a lower voltage limit), and
systems. Thermal energy can be stored as molten salt and is can be thought of as an intermediary between (ordinary)
also mainly used for large scale systems. Magnetic energy electrolytic capacitors and rechargeable batteries.
can be stored in superconducting magnet storage systems, Supercapacitor is considered as storage of electrical system
which is still a relatively new and expensive technology. because it has two working frequencies. When the
Using an energy storage system (ESS) is proposed and is supercapacitor is given a low working frequency, it acts as
one of the most appropriate solutions in this area. This new an ordinary capacitor (having capacitance characteristics).
category allows engineers to manage the power system
However, if the supercapacitor is given a stable frequency
optimally.
of 50 Hz to 60 Hz, it works with resistive characteristics.
Generally, the ESS operation is categorized as follows:
Super capacitors can store 10 to 100 times more per cubic
 The charging period: This process is applicable
using the network electrical energy, during the off- charge than electrolytic capacitors, can accept and distribute
peak intervals when the electrical energy is charges faster than batteries, and have better tolerance to
available at lower prices. usage cycles than rechargeable batteries.
 The discharging period: In times of peak the stored Unlike ordinary capacitors which use solid dielectrics,
energy in an ESS is used. It should be mentioned supercapacitors use a double layer of electrostatic
that in this period the network electrical energy has capacitance and electrochemical pseudo capacitance, both
a higher price and use of distributed generators of which contribute to the total capacitance of the
(DGs) is more economical. Accordingly, supercapacitor, with some differences as stated below:
application of an ESS system is mainly explainable • Electrostatic double-layer capacitors (EDLCs) use
for reducing or even eliminating the uncertainties carbon as electrodes or the like with the amount of
of renewable DG. electrostatic double-layer capacitance greater than
It is worth mentioning that the most commonly used method the amount of electrochemical pseudo-capacitance,
in ESS is based on DC type, so the use of this system is more resulting in the separation of Helmholtz double-
related to power electronic devices to connect to the national layer charges on the surface of the conductive and
power grid. electrolyte electrodes. The amount of charge
Generally, various ESSs can be provided in terms of separation is several ångström (0.3-0.8 nm),
technology, location, capacity, demand and investment smaller than an ordinary capacitor.
costs.

2
 Advances in Engineering Research, volume 198

• Electrochemical pseudo-capacitors use metal can produce power at the MPP point. The Duty Cycle will
oxides or electrodes made from conductive control switching on the converter.
polymers with higher amounts of electrochemical Power input is obtained through the conversion of solar and
pseudo-capacitance coupled with double-layer using photovoltaics. Then, the voltage and current will be
capacitance. Pseudo-capacitance is achieved by read at the sensor of each photovoltaic. MPPT will operate
Faradays electron traction by redox, intercalation, according to its algorithm to get maximum power. In the
and absorption by the electrode surface [4]. MPPT algorithm a duty cycle will be generated which will
Hybrid capacitors, such as lithium-ion capacitors, use control switching on the multi-input converter. When the
electrodes with different characteristics: one emphasizes power produced by and PV is greater than the load power
electrostatic capacitance while the other emphasizes (Pload), then the surplus energy will be used by the converter
electrochemical capacitance. and does not produce power.

1.1.2.3. Energy Storage Management System.

ESMS (Energy Storage Management System) maximizes
the value of energy storage by predicting and controlling the
flexible capacity of any storage device, anywhere on the
network. It increases ROI by taking advantage of all energy
storage use cases and providing a new revenue stream. The
most preferred method of dividing the input power into
storage is by filtering the first sequence into the battery and
the other into the supercapacitor. Previous research on first-
order filters has uncontrolled scheme for the energy
management system.

Figure 1 Bidirectional converter for the PV-connected

1.2. Our Contribution
storage system
This paper presents some improvements based on the
When the power produced by PV is equal to the load power
control system in the energy management system. The
(Pload), the energy will not produce power. When the power
aspect that acted as the parameter is the voltage and power
produced by PV is less than the load power (Pload), then the
stability. The result without control and with PI controller is
lack of energy will be compensated by the supercapacitor by
compared. More noise and oscillation, high starting
producing electricity. This hybrid system will still be able to
overshoot or undershoot will be categorized as less
operate in a state where one input voltage cannot provide
contribution to system stability.
voltage. Therefore, a hybrid energy storage system using the
MPPT P&O algorithm is expected to improve the efficiency
of the power generated so that the output power in this
1.3. Paper Structure system becomes more optimal [7].
The rest of the paper is organized as follows. Section 2
introduces the methodology used in this paper, which
include Maximum Power Point Tracker (MPPT), input 2.2. Perturb and Observe Algorithm
model and the PI tuning. Section 3 presents a simulation
The Perturb and Observe (P&O) algorithm is often used in
results and comparison analysis. Finally, Section 4
concludes the paper and presents direction for future searching for Maximum Power Point (MPP) in PV. This
research. method can be used for various types of characteristics of
PV and does not require information about the
characteristics of energy storage systems. However, the
2. METHOD AND METHODOLOGY P&O method has the disadvantage that there is an oscillation
in steady state conditions caused by constant duty cycle
changes.
2.1. Maximum Power Point Tracker
MPPT is a method for determining the operating point that
produces maximum power. The point that produces
maximum power is called the Maximum Power Point
(MPP). MPPT works by increasing or decreasing operating
voltage so that maximum power is obtained [6]. The MPPT
working scheme is by measuring current and voltage from
photovoltaics. Currents and voltages will be processed in the
MPPT algorithm and a duty cycle will be generated which

3
 Advances in Engineering Research, volume 198

represents sudden cloud in Surabaya, Indonesia sample

area). The temperature is given constant at the value of
25 °C. The simulation is only 2 seconds due to simulation
computational load limit. The input also had given the
model so load power input could be examined [5].

Figure 2 Perturb and Observe Algorithm based

Maximum Power Point Tracker
Perturb and observe algorithm is also called hill climbing
method. In this algorithm the MPPT method is based on the Figure 4 Input model for photovoltaic input
calculation of output power. In this algorithm the
PV Modules used in the simulation is the model from
perturbation leads to change in the output power of system.
MATLAB Simulink Waaree Energies WU-120. The load
If perturbation increases towards the maximum power point
model in the system is the same for both simulation, 500 W
then voltage must be increased. And if perturbation with unity power factor, for easier simulation tracking since
decreases away from the maximum power point then the PV and the load is in the same direct current (DC) bus.
voltage must be decreased. With this the duty cycle is also
changed and this process continues until the maximum
power point has reached [6].

Figure 5 PV-modules parameter

The supercapacitor used is electric double layer capacitors

(EDLC) type with parameter as shown in figure 6.

Figure 3 Flowchart of Perturb and Observe Algorithm

Figure 3 shows the flowchart of perturb and observe
algorithm. First, we determine the value of V(n) and I(n).
Then after finding power, we check the slope dP/dV at three
different conditions.
If slope dP/dV= 0 at MPP Figure 6 Supercapacitor parameter
If slope dP/dV>0, at left of MPP
If slope dP/dV<0, at right of MPP The battery used is Lithium-ion type with parameter as
shown in figure 7. Initial state of charge is set in 50%.

2.3. Input Model

Data used in this simulation is divided into 2 phases, the
0-1 s (first phase) contains the normal condition (using
Surabaya, Indonesia sample at maximum (1000 W/m2)
during sunny day). The 1-2 s (second phase) contains the
condition where the cloud and/or pollutant in terms of
dust pollutant cover up the panel so the irradiance is half
of the maximum (50% pollutant equals to 500 W/m2

4
 Advances in Engineering Research, volume 198

Figure 10 Battery State of Charge (%)

From the simulation result, the battery shows there is not
much contribution during the 500 W/m2.

Figure 7 Battery parameter

Figure 11 Supercapacitor Voltage (V) and Current (A)

3. SIMULATION RESULT

3.1. Uncontrolled System

This is the result from uncontrolled system, from previous
research, it is known for its inability to control the power Figure 12 Supercapacitor State of Charge (%)
load. Simulation is run in MATLAB Simulink, including the
default system simulation. From the simulation result, the supercapacitor shows there
is not much contribution because the slope in the irradiance
change is almost the same. Besides, the supercapacitor had
given non-first order input, which is very small comparing
to the first order input in real power.

Figure 8 Load Power Stability (W)

From the simulation result, it is proven the uncontrolled
system could not maintain the stability in different
irradiance input due to direct input power from PV to the Figure 13 Battery and Supercapacitor Power
load in the same bus.
Comparison (W)
From the simulation result, the supercapacitor acted as the
change compensator, while the battery acted as the main
energy storage supply. While at full irradiance, the battery
is charging 232.344 W and the supercapacitor use 3.806 W,
while at half irradiance, because the input is less than load
power demand (484.239 W < 500 W). The battery is
discharging 14.048 W, and supercapacitor is 3.335 W to the
system. The total load power demand of 500 W is satisfied
with the input of 484.239 W + 14.048 W + 3.335 W =
501.622 W. In the 1000 W/m2 simulation, load power is
Figure 9 Battery Voltage (V) and Current (A) 180.595 W more than the demand power.

3.2. PI-controlled System

This is the result from uncontrolled system, from previous
research, it is known for its inability to control the power

5
 Advances in Engineering Research, volume 198

load. Simulation is run in MATLAB Simulink, including the

default system simulation.

Figure 14 Load Power Stability (W) Figure 17 Supercapacitor Voltage (V) and Current (A)
From the simulation result, the PI-controller could maintain
the stability in different irradiance input, increasing the
power efficiency by decreasing 36.119% power into the
load.

Figure 18 Supercapacitor State of Charge (%)

From the simulation result, the supercapacitor shows there
is not much contribution because the slope in the irradiance
change is almost the same. Besides, the supercapacitor had
given non-first order input, which is very small comparing
to the first order input in real power.

Figure 15 Battery Voltage (V) and Current (A)

Figure 19 Battery and Supercapacitor Power

Comparison (W)
Figure 16 Battery State of Charge (%)
From the simulation result, it is clear that PI-controlled
From the simulation result, the battery shows there is more system has the better result due to compensation of the load
contribution during the 500 W/m2 than the uncontrolled power surplus during the 1000 W/m2 and increase the
system. battery input by 162.261 W or 69.836% more power to the
battery. The supercapacitor has not affected significantly
due to its contribution as secondary energy storage.

Table 1 Uncontrolled and PI-controlled system power input comparison result.

Load Power Battery Power SC Power

Control 1000
500 W/m2 1000 W/m2 500 W/m2 1000 W/m2 500 W/m2
W/m2
680.595 -232.344
No Control 484.239 W 14.048 W -3.806 W 3.335 W
W W
502.286 -394.083
PI Controller 499.073 W 68.811 W -0.628 W 1.609 W
W W
The battery power and SC power after being 500 W/m2 could not contribute much due to direct input from PV to the load in the same bus.
Meanwhile, the load power is still could not stable in decreasing of irradiance input. PI-controller system compensate the inability of
maintain the load power.

4. CONCLUSION 69.836% more power to the battery. The load has been
optimized, in both 1000 W/m2 by the decreasing of
From the result, it is obtained that PI-controlled 178.309 W and 500 W/m2 by the increasing of 15.166 W.
standalone system has the better result due to Hopefully this research can help another research in the
compensation of the load power surplus during the 1000 renewable energy and storage system fields.
W/m2 and increase the battery input by 162.261 W or

6
 Advances in Engineering Research, volume 198

ACKNOWLEDGMENT P1562/D1.1, March 2018 , vol., no., pp.1-33, 10

March 2018.
This work was supported by Institut Teknologi Sepuluh
Nopember, especially Department of Electrical
Engineering Energy Conversion Laboratory and Power
System Simulation Laboratory.

REFERENCES
[1] W. D. P. Vallejos, "Standalone photovoltaic system,
using a single stage boost DC/AC power inverter
controlled by a double loop control," 2017 IEEE
PES Innovative Smart Grid Technologies
Conference - Latin America (ISGT Latin America),
Quito, 2017, pp. 1-6, doi: 10.1109/ISGT-
LA.2017.8126705.
[2] "IEEE Recommended Practice for Installation and
Maintenance of Lead-Acid Batteries for
Photovoltaic (PV) Systems," in IEEE Std 937-2019
(Revision of IEEE Std 937-2007) , vol., no., pp.1-
25, 28 Feb. 2020, doi:
10.1109/IEEESTD.2020.9057793.
[3] R. Mars, B. Bouzidi and A. Yangui, "On the
comparison of two hybrid energy storage systems:
Parallel connection of battery-RC ultracapacitor/-
multibranch ultracapacitor circuits," 2016 Eleventh
International Conference on Ecological Vehicles
and Renewable Energies (EVER), Monte Carlo,
2016, pp. 1-8, doi: 10.1109/EVER.2016.7476418.
[4] A. Harrouz, M. Abbes, I. Colak and K. Kayisli,
"Smart grid and renewable energy in Algeria," 2017
IEEE 6th International Conference on Renewable
Energy Research and Applications (ICRERA), San
Diego, CA, 2017, pp. 1166-1171, doi:
10.1109/ICRERA.2017.8191237.
[5] M. A. Jirjees, D. A. Al-Nimma and M. S. M. Al-
Hafidh, "Voltage Stability Enhancement based on
Voltage Stability Indices Using FACTS
Controllers," 2018 International Conference on
Engineering Technology and their Applications
(IICETA), Al-Najaf, 2018, pp. 141-145, doi:
10.1109/IICETA.2018.8458094.
[6] S. Phiouthonekham and S. Chaitusaney,
"Mitigating impact of partial shading on
photovoltaic array configuration by using
rearrangement," 2015 12th International
Conference on Electrical Engineering/Electronics,
Computer, Telecommunications and Information
Technology (ECTI-CON), Hua Hin, 2015, pp. 1-6,
doi: 10.1109/ECTICon.2015.7207052.
[7] "IEEE Draft Guide for Array and Battery Sizing in
Stand-Alone Photovoltaic (PV) Systems," in IEEE