Integration and Control of an Off-grid Hybrid

wind/PV Generation System for Rural

Applications
Tariq Kamal'I, Murat Karabacak 1, Syed Zulqadar Hassan 2, Hui Li 2, Ali Arsalan 3,4, Rajprasad Kumar Rajkumar4
lDepartment of Electrical and Electronics Engineering, Sakarya University, Turkey
2School of Electrical Engineering, Chongqing University, Chongqing, China
3Department of Electronics Engineering, Karachi Institute of Economics and Technology, Karachi, Pakistan
4Faculty of Engineering, The University of Nottingham Malaysia Campus, Jalan Broga, 43500, Semenyih, Malaysia
[email protected], [email protected], [email protected], [email protected],
[email protected]

Abstract- This study focuses on a stand-alone hybrid system resources. Even though, Malaysia ranks second in the
combining photovoltaic/wind with super-capacitor/battery production of oil and natural gas in Southeast Asia region in
storage for rural residential applications. Various DC/DC order to sustain the natural resources and address environmental
converters are used to control the power flow to the load and issues. The Malaysian government has assured that it will cut
Maximum Power Point Tracker (MPPT) is used for maximum
down its carbon emissions by 40 % of the levels recorded in
power extraction from the photovoltaic/wind systems. In the
proposed architecture, the Photovoltaic (PV) has the priority to 2005 by the year 2020 [6]. The total population of Malaysia was
meet the required load. A wind turbine system is integrated in the estimated to be around 29.45 million in the year 2012 with
system to support PV during demand. Both super-capacitor and around 70% living in cities [7]. The rest of the people live in the
battery are used as an energy storage system during surplus power rural areas. To provide electricity from grid to these people is
and/or backup device during demand. The proposed system sometimes difficult and not feasible, because, of the large
operates in such a way to maintain the State of Charge (SoC) of distance from the grid and landscape limitations and difficulties.
battery above 80%, which consequently improves the battery life The only source of electricity for these people is through
and its performance. The proposed system uses real time solar, generators burning petrol or diesel which is causing global
wind data and a load of five rural households available at the
warming. The solution to this problem is to install off grid RE
University of Nottingham, Malaysia campus. MATLAB
simulation results conclude that the proposed system can sources that can provide clean energy. Although, many sources
compensate power fluctuation and meet the required load without of renewable energy are available in Malaysia but the scope of
any grid connections. the project is limited to wind, solar energy hybrid supported by
storage devices battery and supercapacitor bank. The annual
Index Terms- pv, Wind, Hybrid storage, Battery life, stand- average irradiance in Malaysia is about 1.643 kWhlm 2 [6]. This
alone hybrid system amount is quite favorable for the production of electricity from
solar PV. Hence the main energy source in the proposed system
I. INTRODUCTION
has been taken as PV. In Malaysia, the average annual wind
The depleting of fossil fuel resources and a rapidly speed is less than 2 mls. The wind speeds are not uniform
increasing global warming rate has brought a new challenge to throughout the whole region and have monthly variations.
fulfil the world's energy demand and to be environmentally However, the wind energy can be store in battery or super-
friendly at the same time. The most viable solution that the capacitor. Therefore, wind system is also the part of the
world is trying to implement is tapping into renewable energy. proposed system. Furthermore, to keep the operation of system
Fossil fuels once utilized cannot be reused, while renewable for 24 hours with utility, a hybrid storage system (battery and
energy (RE) resources such as wind, sunlight, tides, rain are super-capacitor) is connected in the proposed structure.
naturally renewed, thus can be sustained for eternity [1]. Several PV or wind or PV/wind based hybrid system are
Renewable energy also does not produce harmful emissions like discussed in the literature. For example, in [8], the researchers
C02 which are the leading cause of global warming. An have developed PV/wind/battery and PV/wind/diesel/battery
estimated 19.1 % of the global final energy consumption was hybrid systems for rural electrification in Malaysia. PV and wind
provided by RE in the year 2013, and increased even more in based generation with hydrogen/battery storage is the objective
2014 [2]. Beside several benefits of RE, they are highly of [9], [10]. In [11], a MATLAB simulation based solar PV
intermittent and sensitive to climatic conditions, which system is integrated with the supercapacitor and directly
compelled RE to integrate various energy sources [3], [4]. connected to the grid. The system contains only the
Malaysia is considered a growing economy and its supercapacitor as storage device and uses it to stabilize the
economy in the March quarter of2016 increased by 4.2 % year- output power. The author has proposed two different controllers
on-year [5]. Growing economy translates into increasing power for the control of power flow through the system. The first
demand for growth, hence the dependence on oil and gas controller is simple Proportional-Integral (PI) based the second
*Corresponding Author: [email protected]

978-1-5090-3310-2/ 17/$3 \.00 ©2017 IEEE

controller is a hybrid of fuzzy logic and PI controller called as II. DESCRIPTION OF PROPOSED HYBRID SYSTEM
fuzzy tuned PI. Standalone/grid-tied wind with battery/ Figure 1 shows the structure of proposed off-grid system
Supercapacitor system is proposed in [12]. Similarly, Control consists of two main buses i.e., an AC bus and a DC bus in the
and simple energy management method for a PV/supercapacitor proposed design. A 20 kW wind turbine and PV with their
/battery system is simulated in [13]. In most studies, the authors respective converters and MPPT controllers are connected to the
used average load and virtual generated weather patterns. DC bus. The energy storage elements, supercapacitor and
This work provides PV /wind power generation with hybrid battery are also connected with DC bus via the bi-directional
storage systems, e.g., Super-capacitor (SC) and battery. The converters. The AC bus is interfaced with the DC bus through
proposed system is designed for a load of five rural households. the hysteresis current controlled inverter. The AC bus provides
Real time weather data have been used in the proposed structure three phase power to the load connected. The control of power
which was measured at the University of Nottingham, Malaysia flow between the load and the energy sources is regulated by the
campus. Both PV and wind system are controlled through Power Control Unit (PCU). The PCU makes decision on the
MPPT techniques. Various non-isolated power converters are actual power of the wind (Pwr), SC (Psc), PV (Ppv), Battery
used to control the power follow between different energy (PBB), Load (PL), State of Charge (SoC) of SC (Sse) and Battery
sources as well as load which ensure 24 hours power flow with (SBB). Based on these parameters, reference signals for the
high ratability without a grid connection. It is important to switches are generated and the respective powers are regulated.
mention the proposed system is controlled in such way that the III. CONTROL OF SYSTEM COMPONENTS
SoC of battery should maintain 80%, which improves the life
and performance of battery. This work is arranged as follows: In the proposed system, the PV has tracked for maximum power
though incremental conductance MPPT technique. The output
after the introductory part, system description has provided in
ofPV is controlled through Proportional Differential Integrator
Section II. Control of system is covered in Section III. Section
(PID) embedded boost converter. The PID tries to minimize the
IV focuses on the operation of proposed algorithm. Simulation MPPT error "e" . This error is important to control the duty cycle
results and conclusion are scrutinized in Section V and Section of PV boost converter as shown figure 2. Similarly, to track
6, respectively. wind for MPP an Optimal Torque (OT) based controller is

r-----------,
I -- Power Li ne I
DCfDC I I

So lar Array Conve rter : ~_ :~~o~ :i~e_:

~
....---"'--..., PCU

Bidirectiona l D C
SC Bank 10 DC convener

Fig I : Complete structure of proposed hybrid system

2
modelled in MATLAB. The OT algorithm regulates the diminish the error in order to achieve the desired active and the
Permanent Magnet Synchronous Generator (PMSG) torque reactive powers.
according to maximum power reference torque. It can be
achieved when y reaches Yopt. The optimal torque at a given
speed is written as;
Residential
(1) Load

-~
I
I
I
I
I

Fig 5: 3-phase DC/AC inverter control

IV. PCU OF PROPOSED HYBRID SYSTEM

Fig 2: Control system for PY array To prevent the system from blackouts and keep the system
stable, it is required to design a supervisory control unit which
OTBascd Voltage manages the overall power flow and also generates the
MPPT Controller appropriate references to the local controller. In this regard, the
PCU is developed in the proposed system. The PCU decides the
power flow depends on several inputs such as; PWT, Pse, Ppv,
Fig 3: Control scheme for WT system PBB, PL, Sse and SBB. The power balance equation for proposed
system can be written as;
where Copt is a function of yopt . The OT based MPPT produces
references in terms of voltages. Depend on voltage error, an (2)
appropriate duty cycle (DwT) is generated. The overall control
scheme of Wind Turbine (WT) system is presented in figure 3. Using (2), the excess power generated by wind and PV can be
Super-capacitor and battery are controlled on the basis of their expressed as;
charging/discharging capabilities. The bi-directional flow ofSC (3)
and battery needs buck-boost converter. The buck-boost
converter has controlled via PID controller. The control system The excess power is consumed by SC and battery. Similarly, the
for super-capacitor/battery is shown in Figure 4. equation for power shortage created by wind and PV can also be
expressed as;
IsdlBB
(4)

Fig 4: Charging/Discharging controller for SC/battery

Figure 5 describes the basic control strategy of the proposed

three phase inverter connecting the DC bus voltage of the
hybrid system. The inverter operates on power conversion from
the DC bus power to load. The current controlled inverter is
controlled to provide sinusoidal current after filtering, which is
fed to the utility line to distribute the required power. It also
stabilizes the DC bus voltage. The inverter increases the output
power when the DC bus voltage is increased and vice versa.
The proposed control technique uses PI controllers and
hysteresis current control pulse width modulation technique to
generate suitable gate signals for driving the controllable
switches of the inverter. The inputs to PI controllers are the
errors calculated between the measure and actual values of the
active and the reactive powers. The PI controllers try to Fig 6: Flow pattern of PF A

3
Based on power equations, the Power Flow Algorithm (PFA) is
designed. The PF A is designed with the help of various logic
gates in Simulink simpower system toolbox and embedded in
PCu. The complete flow pattern of PF A is shown in figure 6
and explored below;
(b)
1) Initially, the input parameters of the PCU are monitored and
stored in local memory.
2) The PCU will check the equation (2) condition, if it is
satisfied, then goto 1 else goto 3.
3) PCU will check the power status depending upon condition
PL>PPV+PWT. If this condition is true, then it represents

t3:~
power shortage in system and goto 4 otherwise it will
represent excess power in system and goto 6.
4) Check weather Tsc>20%, if true, then apply power shortage (/) 2.5

sc
reference to SC; P = PL - (PWT + Ppv) and goto 1 else
'0
c:
5: 20 4 8 12 16 20 24
goto 5. Time (Hrs)
5) Apply power shortage reference to battery; P~B = PL - Fig 7: Metrological data (a) Irradiance (b) Temperature (c) Wind Speed
(PWT + Ppv ) and goto 1. The output powers of different energy sources and load are
6) Check weather T sc>90%, if true, then apply excess power shown in figure 8. At t=0-6.5 Hrs, the load demand starts from
reference to battery; P~B = PL - (PWT + Ppv ) and goto 1 6 kW and increases gradually till 13 kW. In this region, the PV
else goto 7. output power is zero due to night hours. Wind system provides
an average power of 1 kW due to very low wind speed.
7) Apply excess power reference to SC; P sc= P L - (PWT + Therefore, the remaining load burden is applied on SC and
Ppv ) and goto 1. battery. According to PMU priorities, the battery will only be
used when SC does not able to provide any power. So, in this
V. SIMULATION RESULTS A ND DISCUSSION
case, the SC successfully fulfills the remaining load demand. At
The proposed system is designed to operate on real world t=6.5-7.5 Hrs, the load demand increases to 14.5 kW. On the
meteorological conditions. The wind speed and solar irradiance other side, the PV system starts generating power due to sunrise
are recorded for every 5 minutes and halfhour. In this study, the as shown in figure 8. In this region, the need ofSC becomes less
real domestic load of five households near University of and at the final stages its output power become zero. The battery
Nottingham, Malaysia campus is considered. The recorded is used as load compensator which absorb/deliver any excess
irradiance, temperature and wind speed for a typical summer day power and power shortage generated by other energy sources.
are shown in figure 7. To verify the effectiveness of proposed At t=6.5-17 Hrs, the load demand increases to a peak load of
system, all the recorded data along with mathematical model of 20.3 kW, but PV system generates enough power to overcome
different energy sources are modelled in MA TLAB/Simulink. all the peaks in this region . The PV system also generates some
In this research, a WT of20 kW, PV array of20 kW, SC of 500 excess power which is used for SC charging (negative power).
F and battery of 5 x200Ah are used for simulation purpose. At t=17-18 Hrs, the PV power start decreasing.
4

~X~10-r--r-~--~~~~~r==r==~~==~~===c==r==r~~~---r--r--r--r-~--~~
2.5 r

l----- P p,J .......... P BaU - - P SG ----- P L - - P W;nd I

2- I, .- 1-== :::: ,..- :== --_ I I--
__ !,--;:-~-- ---- --d,I
---- -i,
1.5 1-- - '1--- --t:

~ 1M- ---- J
----F rL ',:
'--,
~o
~ L-t==t--
' .... ~
----t-==., \,.....

F.
0.5 I

(1
--- --'.........l;:: ........ ......... ........ ...................................... ·......I....·F· ........ ---- ----
1
' Lh- r-~r--I--J..-.r\...-Jj,..J
.0.50n --~----'-----"----------"---5'--------6~----'-----'-
8 ----'-9--~
10----'1-
1 - -1'=-
2 - -1:':-
3- -1-'-4---:':
15:---:":
16:----:'
17=----:1'::-
8 -------:1"::-
9 -------:2=0- -2
::':1- -=22:---:.
23=----::24
lime (Hrs)
Fig 8: Output powers of di fferent energy sources

4
Therefore, due to low wind speed, the SC is needed to overcome Figure 10 shows the battery and SC parameters, the battery SoC
the power shortage. SC is providing maximum power of 14.5 is maintained above 80% throughout the day. The SC current
kW. At t=18-24 Hrs, the load attains its second peak of21 kW. varies according to the weather conditions and the load. During
Due to low wind speed and no radiations, the entire system only t=0-5.5 & 18-24 Hrs, the solar irradiation is zero the SC current
relies on Sc. The SC is providing maximum power of20.2 kW, is positive and the supercapacitor is discharging. When the sun
while keeping the system operational and prevent it from de- rises, the supercapacitor goes into charging mode and the current
synchronization. becomes negative.
The real and reactive powers for residential load are shown in VI. CONCLUSION
figure 9. In figure 9, the blue solid line represents the actual
power while the dotted red line shows the reference power In this work, a windlPV ISC/battery hybrid power system
demand. At t=0-0.1 Hrs, the actual load fluctuates very fast, was modeled and simulated for a stand-alone
however, the proposed system tracks the load demand very application. Wind and PV were utilized as primary power
quickly. The zoomed figures in figure 9 represents the load sources of the system while the SC/battery were employed as
tracking performance of proposed system for a step change of storage system during excess power and backup during
2/ 1.5 kW (real/reactive power). It is clearly revealed from demand. Special care was considered for long life of battery.
zoomed figures, that the proposed system tracked the step To this aim, the proposed algorithm was operated in such way
change very quickly. to maintain the SoC of battery above 80%. A detailed
, (a) simulation test-bed was established which consents designing
x 10
and analyzing any wind/PV ISC/battery hybrid system with
different power levels and parameters. The dynamic
[ 1.5 performance ofthe proposed test-bed was tested under real load
~ and weather patterns at the University of Nottingham, Malaysia
"- campus. Simulation results obtained indicates the feasibility of
the proposed system.
0 . 5 "-----~---~---~--~---~---------'

o
ACKNOWLEDGMENT
,
x 10 The authors wish to thank Turkish Government Scholarship
1.6 entry 2015-2016 for fmancial support, and University of
1.4
Nottingham, Malaysia campus for providing real time data
[ 1.2
during the completion of this research work.
~ 1
~ 0.8
REFERENCES
0.6
0.4 L -_ _
[1] M. Almaktar, H. Abdul Rahman, M. Y. Hassan, and 1. Saeh,
_ _ _~ _ _ _~ _ _~ _ _ _~ _ _~
"Artificial neural network-based photovoltaic module temperature
~

estimation for tropical climate o f Malaysia and its impact on

l ::'
photovoltaic system energy yield," Prog. Photovoltaics Res. App/. ,
Fig 9: AC Load actual and reference power (a) Real Power (b) Reactive Power vol. 23, no. 3, pp. 302- 318, 2015.
(a) [2] J. L. Sawin et al., " Renewables 2016 Global Status Report. Key
findings. A Record Breaking Year for Renewable Energy: New

' ~
_ I ----- SOC Balt - - SOC sc Installations, Policy Targets, Investment and Jobs. Mainstreaming
renewables: guidance for policy makers. "

~ ': -------------------------:-------~----------------j
110 r
[3] P. Balamurugan, S. Ashok, and 1. L. Jose, "Optimal operation of
biomass/wind/PV hybrid energy system for rural areas," Tnt. 1. Green
Energy, vol. 6, no. I , pp. 104---116, 2009.
o 4 8 12 16 20 24 [4] M. K. Deshmukh and S. S. Deshmukh, "Modeling of hybrid
(b) renewable energy systems," Renew. Sustain. Energy Rev. , vol. 12,
no.l , pp.235- 249, 2008.
~ · '~ G=-I ' ~
200 ls~ 1 ' [5] "Malaysia GDP Annual Growth Rate 2000-2016 I Data I Chart I
Calendar," 2016. [Online]. Available:

j ': t-------~ --- [6]

hup://www.tradingeconomics.comlmalaysialgdp-growth-annual.
F. M. Hossain, M. Hasanuzzaman, N. A. Rahim, and H. W. Ping,
"Impact of renewable energy on rural electrification in Malaysia: a
-100 t.----~~--~---~---~---~--
---'

o 4 8 12 16 20 24 review," Clean Technol. En viron. Policy, vol. 17, no. 4,pp. 859- 871 ,
(c) 2015.
200· ~--~--~---~--~---~--~ [7] S. E. Hosseini and M. Abdul Wahid, 'T he role of renewable and

i :: ~:::=::_-
_ _ ::~
sustainable energy in the energy mix of Malaysia: a review," Int. 1.
Energy Res. , vol. 38, no. 14, pp. 1769- 1792, 2014.
[8] M. Fadaeenejad, M. A. M. Radzi , M. Z. A. AbKadir, and H. Hizam,
t-____________ ____________ "Assessment of hybrid renewable power sources for rural
electrification in Malaysia," Renew. Sustain. E nergy Rev. , vol. 30,
o 4 8 12 16 20 24 pp.299- 305, 2014.
Time (Hrs) [9] S. Z. Hassan, S. Mumtaz, T. Kamal , and L. Khan, "Performance of
grid-integrated photovoltaic/fuel celli electrolyzer/battery hybrid
Fig 10: Battery and SC parameters (a) SoC (%) (b) Current (A) (c) Voltage (V) power system," in 2015 Power G eneration Systems and Renewable

5
 Energy Technologies, PGSRET 2015, 2015 , pp. 1- 8.
[10] T. Kamal, S. Z. Hassan, H. Li, S. Mumtaz, and 1. Khan, " Energy
management and control of grid-connected wind/fuel cellibattery
Hybrid Renewable Energy System," in 2016 International
Conference on fntelligent Systems Engineering (ICISE) , 2016, pp.
161- 166.
[II] O. O. Mengi and T. H. Altas, " An off-grid PY/SC green energy
production system," in Smart Grid Congress and Fair (lCSG), 2015
3rd international fstanbul, 2015, pp. 1- 5.
[12] S. Z. Hassan, H.Li , T. Kamal, and M. Awais, "Stand-alone/grid-tied
wind power system with baUerylsupercapacitor hybrid energy
storage," in 2015 international Conference on Emerging
Technologies, ICET 201 5, 2015, pp. 1- 6.
[13] S. Z. Hassan, H. Li, T. Kamal, S. Mumtaz, L. Khan, and 1. Ullah,
"Control and Energy Management Scheme for a PY/SC/BaUery
Hybrid Renwable Power System," Sci. Int. , vol. 28, no. 2, pp. 955-
964, 2016.