Paper f2
Paper f2
Abstract - This paper presents the modelling and simulation system, and load is presented in [2], a larger scale project of a
of an 80kW AC microgrid network in MATLAB/Simulink DC-based microgrid system with residential unit integration is
environment. The network comprises a 50 kW photovoltaic also presented in [3].
system, a 10 kW fuel cell system, and a 20 kW battery energy Due to its scalable configuration, the network is adaptable
storage system (BESS). The model is simulated under four to integrate various types of renewable energy sources.
operating conditions: (i) grid-connected mode, (ii) islanded Depending on the selected location, most studies and
mode (iii) islanded mode with excess power generation from simulation of the microgrid network utilise the available
Distributed Energy Resources (DERs) and (iv) islanded mode renewables resources in the area to maximise the power
with limited power generation from DERs. Simulation results generation in the local community. As an example, the
of the microgrid during various operating conditions were integration of wind energy conversion system with microgrid
observed to investigate the charging and discharging and utility in [4] was conducted specifically for wind-powered
characteristics of the battery and the total harmonic distortion systems as the country was found to have a suitable average
of the power network during grid-connected mode and wind speed in operating wind turbines system and a potential
islanded modes. site for an offshore power generation in the future. By
accessing the environmental factor, suitable renewable
Keywords—Microgrid, Photovoltaic, Fuel cell, Battery energy technology can be selected to ensure system stability such as
storage system, grid forming controller, grid feeding controller, a photovoltaic system [5] where the location receives longer
maximum power point tracking (MPPT). sunlight hour creating a suitable environment to adopt the
photovoltaic system. In the meantime, a fuel cell-powered
I. INTRODUCTION
microgrid was selected in [6] as the system requires higher
Due to the rapid growth in the energy industries and the power stability and cost was not a limitation.
economies, the demand for power increases rapidly, which
causes fossil fuels to undergo excessive depletion. As the With the reduced complexity, microgrid allows more
demand for power increases, the rate of fossil fuel flexibility in terms of scheduling and control of the overall
consumption increases leading to an increase in carbon network in real-time conditions. To maximise the usages of
dioxide (CO2) production which has hazardous consequences the system, optimization techniques have been developed and
on the earth's atmosphere causing damage to the environment implemented to various parts of the microgrid network such
and gives rise to global warming. These issues have led the as energy management for the end-user and power
researchers to tap into clean energy technologies, such as solar management for renewable energy technologies.
energy, wind energy, fuel cell, and biomass, etc. Therefore, With the rapid expansion of the electrical network, more
renewable energy technologies have become one of the best AC microgrid is adopted increasing the overall complexity of
alternatives to conventional power sources. However, due to the control and reliability of the network [7] To ensure the
the dynamic nature of renewable energy sources, their seamless transfer between modes of operation, the control
integration to the existing power grid poses numerous technique is often used to reduce the transient of the loads
challenges such as power instability and harmonic, etc. Hence, during the transition. One example of the commonly used
an alternative system configuration in the form of a microgrid technique is master-slave control [8] the technique operates
was initiated. in two different modes under different conditions. During
grid-connected mode, the DERs functions as a state feedback
The microgrid is defined as an independent low voltage
control loop to the inverter under grid-feeding while during
(LV) local distribution network that can be operated in either
islanded mode the DERs functions as a voltage control loop.
grid-connected mode or islanded mode. With this advantage,
With the above technique, the smooth transition between
the network is often adopted in a remote area where the power
modes can be achieved by changing the control algorithms
grid is inaccessible or for a small community that consumes
under a different mode of operation.
power at a smaller scale such as a residential microgrid [1].
With relevant literature documenting research on DC To minimise the current and transient time during the
microgrid with various distributed energy resources (DERs) transition period, the inverter current and voltage have to be
such as PV array, battery energy storage system, fuel cell controlled. In [9], a seamless interchange method for a DC
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
microgrid during grid-connected mode and islanded mode B. Photovoltaic
with a suitable PI controller is presented. A droop controlled The solar cell utilizes solar irradiance and converts them
technique was used in [10] to control the inverter to ensure into direct current (DC) through the process of photovoltaic
seamless transfer between the modes of operation. Whereas effect. The performance of the system depends highly on two
standalone network transitioning to grid mode is often based factors, irradiance and temperature. To generate the desired
on Phase-locked-loop (PLL). output power, PV cells with a voltage level of 48.2 V and a
AC microgrid configuration is important due to its power rating of 230.04 W are connected in series and parallel
effective power transfer during long-distance power to form a module or an array. For the system requirement of
transmission. With the global environmental concern and the 50 kW, a total of 273 modules are formed using 13 parallel
rising energy demand, together with the advancement of string and 21 series-connected modules per string.
renewable energy technologiesMetropolitan cities are driven
to adopt more renewables energy technologies, thus TABLE I SPECIFICATION OF PHOTOVOLTAIC ARRAY
alternative power generation network is required to fulfill the
ever-changing energy demand. Thus, the industry needed Parameters Value
more research and a solution to get an extensive overview of Parallel String 13
how the microgrid network can be incorporated into the Series connected modules per 21
existing power network and to further verify the reliability string
and sustainability of the solution for various countries. Maximum power, Pmax (W) 230.04
This paper provides the performance analysis of an AC Open circuit voltage, Voc (V) 48.2
microgrid network with multiple DERs such as photovoltaic, Short circuit current, Isc (A) 6.05
fuel cell, and battery energy storage system (BESS) using
MATLAB/Simulink environment. In particular, the overall
performance of the microgrid during various operational To ensure the system output the maximum power at
modes is investigated. The battery state of charge (SOC) different solar irradiation levels, the Maximum power point
during the charging and discharging cycle will be observed tracking device is implemented with Perturb and Observe
along with the analysis of Total harmonic distortion (THD) (P&O) control algorithms. The algorithms regulate the PV
in the power network. array operating voltage by observing the output power of the
array. The algorithms observe the output power of the PV
II. MODELLING O F A C M I C R O G R I D array and obtain the differences in voltage. Through the
comparison, the algorithm predicts the next maximum power
A. AC Microgrid Network point voltage, enabling the command to increase or decrease
Microgrid network has been emerging all over the world the operating voltage of the PV array to ensure maximum
due to its ability to operate without any external power source power point can be achieved.
and is considered to be a self-sustainable system. Singapore's
recent initiative to adopt more renewables energy
technologies such as the photovoltaic system to address
environmental concerns. Microgrid network can potentially
be integrated into the existing electrical network. In this
paper, the simulation is based on a three-phase system at 80
kW, 400 Vrms, and 50 Hz system. The simulation model
consists of the following renewables energy resources: (i) 50
kW photovoltaic (ii) 10 kW fuel cell and (iii) 20 kW battery
storage system.
Figure 1 Block diagram of the 80kW AC microgrid Figure 2 Flow chart of P&O MPPT algorithms [11]
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
C. Fuel Cell TABLE II SPECIFICATION OF FUEL CELL BOOST
CONVERTER
Fuel cells are electrochemical devices that convert
chemical energy into electrical energy. To supply 10kW of Parameters Value
power to the network two stacks of Proton-Exchange Switching Frequency (kHz) 10
Membrane Fuel Cell (PEMFC) with a voltage level of 45 Vdc Input voltage, Vin (V) 64
and power rating of 6 kW are connected in series to deliver the 402
Output voltage, Vout (V)
desired output power. A DC-DC converter is implemented to
step up the voltage generated by the fuel cell system to a Inductor, L (^H) 8.985
higher voltage according to the specifications listed in Table Capacitor, C (^F) 554.8675
II. The duty cycle of the converter was calculated based on the Duty cycle (D) 0.8875
DC input voltage and the desired output voltage. The duty
cycle obtained is crucial in maintaining the desired output
voltage. D. Battery Energy Storage System
More renewable energy technologies are deployed into the
Viin(min) x n energy industry however, these renewables technologies rely
Duty cycle = 1 (1)
Vn on a dynamic variable. Thus, the level of reliability in
Vin(min) = Minimum input voltage delivering power is unstable. Besides, these technologies are
n = Efficiency of the converter incapable of storing the generated power or excess energy of
its own. Therefore, energy is being transformed into another
Vout = Desired output voltage unusable form of energy such as heat, resulting in power loss.
To resolve the issue, deep cycle batteries are used to store
The inductor selection depends on the DC input voltage, the energy in terms of electrical energy.
desired output voltage, minimum switching frequency of the
converter, and the estimated inductor ripple (2). The inductor TABLE III SPECIFICATION OF LITHIUM-ION BATTERY
current ripple was assumed to be 1% while the minimum
switching frequency was set to 10 kHz. Parameters Value
Number of battery 9
Vin x (Vout - Vin) Nominal voltage (V) 48
Lmin = (2) Rated capacity (Ah) 450
AL x f s x V out
Lmin = Minimum inductor value State of charge, SOC (%) 80
fs = Minimum switching frequency of converter Battery response time (ms) 10
AL = Inductor ripple current Nominal discharge current (A) 195.65
The output capacitor of the converter was selected based on With its characteristic to charge and discharge electrical
the relationship of the maximum output current of the circuit, energy, a battery storage system is commonly used in a
duty cycle, switching frequency, and the desired output standalone network that is self -sustainable. The battery can
voltage ripple (3). In this calculation, the output voltage perform as both load and source in the network in different
ripple was assumed to be 1% while the duty cycle value was conditions. making the system desirables to be adopted in
based on equation (1) various electrical networks. In this paper, a total of 9 lithium-
ion battery with a nominal voltage of 48 V and rated capacity
lout(max) X
D
of 450Ah was used to supply 20kW of power to the network
Cout(min)= (3) To facilitate the charging and discharging of the battery, a
fS X AVout
bi-directional DC-DC boost converter was implemented. The
Cout(min) = Minimum output capacitance different operation cycle depends on the power available on
Iout(max) = Maximum output current in the circuit the network.
D = Duty cycle calculated in equation (1)
TABLE IV SPECIFICATION OF BI-DIRECTIONAL DC-DC
CONVERTER
Equation (2) is the minimum inductance for the boost
converter to operate at Continuous conduction mode (CCM). Parameters Value
Thus, the inductor used must be equal or higher than the Switching Frequency (kHz) 10
calculated value. Desired output voltage ripple (%) 1
The capacitor value is selected based on equation (3), the Inductor ripple current (%) 1
selected capacitor value must equal to or higher than the Inductor resistance (Q) 0.05
obtained value to ensure the output voltage ripple is within Inductor, L (mH) 0.576
the operating range. The parameter of the DC-DC boost
Capacitor resistance (Q) 0.195
converter is shown in Table II.
Capacitor, C (^F) 500
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
Figure 3 Simulink model of AC microgrid
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
Time [s)
Figure 4 Voltage waveform of grid and PV during grid feeding
Figure 8 (a) battery voltage (b) battery current during grid-
connected mode
B. Islanded mode
The microgrid is transitioning from grid-connected
mode to islanded mode from t=5sec to t=10sec. After the time
t=4.5sec the main grid is isolated from the rest of the network.
Thus, the loads connected to the AC bus bar must be powered
by the respective DERs. A total of three different AC loads
with a total demand of 75kW is connected to the bus bar
starting from t=4.5sec to t=10sec. Figure 10 indicates that the
voltage generated by the DERs experiences a slight decrease
Time (s) x10 & in amplitude during the transition period before the network
Figure 6 Real power injection tothe AC bus bar during grid- enters a complete islanding mode at time t=5.5sec. During the
connected mode transition from grid-connected mode to islanded mode, the
current waveform from the grid in Figure 11 experiences
ripple before the main grid becomes completely isolated from
the rest of the network.
As the network has a grid-forming controller that controls
the voltage and current output of the inverter when the main
grid has been isolated. The inverter output voltage was able
to regain its amplitude to meet the reference value and
function as the power source as shown in Figure 12. The real
power output from the grid in Figure 13 indicates that the grid
was isolated from the network from time t=5.5sec. Since the
total power generated from both fuel cell and photovoltaic is
Figure 7 Battery state of charge percentage during grid-connected lower than the power demand by loads, the battery energy
mode storage system is activated and enters a discharging cycle at
time t=5.5sec as indicated in Figure 14. The islanded mode
lasted to time t=10sec where all the loads are supplied by the
DERs.
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
4,95 5 5.05 5.1 5.15 5.2 5,25 5.3
Time (s)
Figure 10 Voltage waveform of grid and PV during transitioning Figure 14 Battery state of charge percentage during the transition
mode from grid-connected mode to islanded mode from grid-connected mode to islanded mode
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
Figure 19 indicates the power delivered to loads during the D. Total harmonic distoriton (THD)
simulation, the waveforms show no disruption of power at
any point. Thus, the result above has verified that the network
can operate at islanded mode and the battery is capable to
discharge power to meet the demand of loads when power
generated by the DERs is limited.
P o w e r f r o m D E R s - Grid c o n n e c t e d m o d e
i (FC)
f0.7F
§ 0.6
s
- 0.5
1.5 2 2.5 3 3.5 4 4.5 5
Time(s) x103
Figure 21 THD analysis of voltage and current from fuel cell
inverter without any filter
Time (s)
Figure 17 Power generated by each DERs during islanded mode A passive filter was implemented at the output of each
inverter to minimise the harmonics present at the output to
ensure a smooth power transfer between the source and the
load. The total harmonic distortion of the respective system
is kept to be below 5% as required by the IEEE standards.
I V . CONCLUSION
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.
REFERENCES
[1] M. Laour, F. Akel, D. Bendib and M. Chikh, "Residential Microgrid
Load Management and Optimal Control in grid Connected and Islanded
Mode," in IEEE, Rabat, Morocco, Morocco, 2018
[2] J. Mirez, "A modeling and simulation of optimized interconnection
between DC microgrids with novel strategies of voltage, power and
control," in IEEE, Nuremburg, Germany, 2017.
[3] T. Sakagami, A. Werth, M. Tokoro, Y. Asai and D. Kawamo,
"Performance of a DC-based microgrid system in Okinawa," in IEEE,
Palermo, Italy, 2015.
[4] Y. Fan, M. Goyal, A. Ghosh and F. Shahnia, "Integration of Wind Energy
Conversion System with Microgrid and Utility," in IEEE, Perth, WA,
Australia, 2014.
[5] A. Khamis, A. Mohamed, H. Shareef, A. Ayob and M. S. M. Aras,
"Modelling and Simulation of a Single Phase Grid Connected Using
Photovoltaic and Battery Based Power Generation," in IEEE, Manchester,
UK, 2013.
[6] M. INCI, "Design and Modeling of Single Phase Grid Connected Fuel
Cell System," in IEEE, Elazig, Turkey, Turkey, 2019.
[7] Y. W. M. C. L. Ying Wu, "AC Microgrid Small-Signal Modeling:
Hierarchical Control Structure Challenges and Solutions," in IEEE, China,
2019.
[8] H. Xu, X. Zhang, F. Liu, D. Zhu, R. Shi, H. Ni and W. Cao,
"Synchronization strategy of microgrid from islanded to grid-connected
mode seamless transfer," in IEEE, Xi'an, China, 2013.
[9] V. Kleftakis, D. Lagos, C. Papadimitriou and N. D. Hatziargyriou,
"Seamless Transition Between Interconnected and Islanded Operation of DC
Microgrids," in IEEE Transactions on Smart Grid, vol. 10, no. 1, pp. 248-
256, Jan. 2019
[10] S.-H. Hu, C.-Y. Kuo and T.-L. Lee, "Design of virtual inductance for
droop-controlled inverter with seamless transition between islanded and
grid-connected operations," in IEEE, Raleigh, NC, USA, 2012.
[11] R. R. R. M. Natarajan Pandiarajan, "Application of Circuit Model for
Photovoltaic Energy Conversion System," in ResearchGate, India, 2012.
Authorized licensed use limited to: Carleton University. Downloaded on May 27,2021 at 16:06:11 UTC from IEEE Xplore. Restrictions apply.