Designing and Simulation of a DC Microgrid in

PSCAD
M.H.F Ahamed*, U.D.S.D Dissanayake, H.M.P De Silva, H.R.C.G.P kumara, N.W.A Lidula, Member, IEEE.
Department of Electrical Engineering,University of Moratuwa,
Katubedda, Sri Lanka.
*fahamedi91@gmail.com

Abstract— DC and AC-DC hybrid microgrids are evolving Energy management and control of a DC microgrid based on
technologies used in telecommunication industry concerning its an intermittent source is very critical. A well-functioning energy
reliability, safety and efficiency in supplying power. This paper management and control scheme will ensure reliable operation of
presents a DC Microgrid system designed for telecommunication the microgrid even under various disturbances. Introducing an
power supply system, and three possible modes of operations are energy storage element for the microgrid allows operation
discussed. The DC microgrid comprises of a solar PV array as the flexibility while making the control scheme complex. The
distributed energy source, a battery bank as the energy storage purpose of the energy management and control is to maintain the
element and the utility grid. The solar characteristics are verified
DC bus voltage within acceptable limits and to protect the battery
using manufacture specification. The irradiance data used for solar
from over charging and deep discharging. The state of charge
PV is actual data, which was measured in a typical sunny day. A
proper Energy Management and Control is proposed for the
(SOC) of battery is the signal that indicates how much further the
reliable operation of the DC microgrid. The simulated results show battery can be discharged within the safe limit. For the battery
that the energy management and control works as expected. control circuits SOC is given as one of the indicator to decide
charging or discharging of the battery [2].
Keywords—DC microgrid; Distributed resource; DC-DC buck Plenty of research has been carried out regarding control of
converter; Energy storage; Energy management and control. DC microgrids. In [3] a battery and super capacitor based control
scheme is introduced. A super capacitor is capable of giving
I. INTRODUCTION immediate responses for rapid changes in power, but due to its
With the increased concerns on environmental factors and limited capacity it cannot handle a long term power requirement.
increasing demand for electricity, renewable energy sources have Compared to super capacitors battery shows a slow response but
captured a spectacular place in the line of energy sources. can meet a long term power requirement. In [3], the control
Microgrids enable to integrate distributed energy sources offering algorithm is designed such that the battery converter will be
higher reliability. Microgrids can be AC, DC or hybrid those are switched ON after the super capacitor converter is switched ON
defined according to the bus voltage. if there is a power exchange. Passivity based control is
introduced in [4]. In [5] the authors describe two possible modes
The debate AC verses DC is historical and as old as the of operation of a diesel generator in a solar PV and wind turbine
power systems itself [1]. DC microgrids offer plenty of based system. The diesel generator is operated in continuous
advantages when compared to an AC microgrid. Most of the operation mode and ON/OFF mode as two different cases and the
renewable sources like solar PV can be integrated to a DC cost of operation of each mode is compared.
microgrid with reduced conversion stages. Energy storage
elements will require a single DC-DC converter when connected Energy management and control of a DC microgrid based on
to a DC microgrid. Therefore, compared to an AC microgrid the a solar PV and battery is discussed in this paper. Three possible
overall efficiency will be increased when using a DC microgrid. modes of operation including grid connected mode of operation
In addition to this the key advantage of LV DC microgrid is that is presented. The designed system can be further expanded by
low risk of dangerous electric shock. Most of the modern integrating wind turbines, super capacitor, fuel cells and fly
electronic loads, LED lighting, telecommunication equipment, wheels etc. and, it can be used to analyze various power system
electric vehicles (EV) and HVAC systems are operated on DC scenarios including transients and protection.
power. The next section provides an overview of the designed system
Most of the commercial buildings, data centers, and simulation models. Section III describes the energy
telecommunication and broadcasting stations, space crafts, ships, management and control scheme. Simulation results and
rural electrification systems and EV charging centers use DC discussions are presented in section IV. Finally the conclusions
microgrids. DC microgrid is an emerging concept, but still are given in Section V.
standards and regulations are not documented, which is an
obstacle for the development of DC microgrid technologies.

978-1-4673-8848-1/16/$31.00 ©2016 IEEE 1

 II. OVERVIEW OF THE MICROGRID 10 2
200 W/m
The layout of the designed DC Microgrid system is shown in 8

C u rren t (A )
Fig. 1, which was simulated in PSCAD/EMTDC which is a 400 W/m2
6
power system simulation tool. The modelled DC microgrid was 600 W/m
2
4
designed based on an existing telecommunication DC-AC hybrid 800 W/m2
microgrid system. This system was taken as the reference system 2
1000 W/m2
to select ratings of system components such as PV rating, battery 0
capacity, size of the load etc. As per the reference system the bus 0 10 20 30 40
Voltage (V)
voltage is 48 V DC. A voltage variation of +/- 10% is acceptable
(a)
for healthy operation of telecommunication equipment.

(b)
Fig. 1. Layout of the microgrid. Fig. 2. (a) Solar PV simulated curve, (b) Solar PV characteristics curves by
the manufacture [6].
A. PV array
There is an in-built library model for solar PV available in 1200
PSCAD/EMTDC, which is used in this simulation. The model 1000
Irradiance (W/m2)

has to be configured by specifying the irradiance data and the cell 800
temperature. Series and parallel combination of solar cells and 600
modules can be customized according to the voltage and power 400
requirement. The PV array used in this simulation is based on
200
“ReneSola” manufacture, which is the PV module manufacture
for our reference DC-AC hybrid microgrid. The solar PV 0
6.30 am 9.00 am 12.00 noon 3.00 pm 6.00 pm
characteristic curves are validated with the characteristic curves Time of Day
provided by the manufacture for different irradiances, which are Fig. 3. Variation of irradiance.
shown in Figs. 2 (a) and (b). The PV array parameters, which
were used in the simulations, are shown in Table 1. These data The irradiance input for the solar PV is an actual variation,
were referred from the manufacture specifications. which was measured in Hambanthota, Sri Lanka. The sampling
rate at which the irradiance was measured is 30 s. Fig. 3 shows
TABLE 1 SOLAR PV RATINGS PROVIDED BY THE MANUFACTUER the variation of irradiance for an 11 hours and 30 minutes form
6.30 AM to 6.00 PM for a typical sunny day. Irradiance
No. of modules connected in series per array 6
undergoes fluctuations due to disturbances like clouds covering
No. of module strings in parallel per array 4 the sun etc. The maximum irradiance recorded is around 1100
No. of cells connected in series per module 60 W/m2.
No. of cell strings in parallel per module 1
Reference irradiation (W/m2) 1000 B. Battery
Reference cell temperature(̊ C) 30 There is an in-built library model available for battery in
Effective area per cell (m2) 0.02435 PSCAD/EMTDC version 4.6, which is based on Shepherd
model. The battery capacity is 600 Ah. For the healthy operation
of the battery the over-charging limit is set to 90% and

2
discharging limit is set to 30% of its capacity. Because of this harnessing the maximum power from the solar PV. As shown in
provision it can be shown that the battery can supply the loads Fig. 4(b), the PV output voltage and current are sent through a
for around 9 hours without interruption. Charge and discharge first order low pass filter to filter the higher frequencies and
rates of batteries are governed by C rates. It is recommended by harmonics. The filtered signals are then fed to the MPPT model,
the battery manufacture to charge the battery at rate of 0.8C to which generates a reference voltage called Vmpp. The method
prolong battery life and to use around 1C discharging rate or used in this model is incremental conductance, an algorithm
higher. During a fault or an abnormal situation for the healthy based on the slope of the curve shown in Fig.4(c). The slope at
operation of the battery a current limiting mechanism should be maximum power point is zero, positive on right and negative on
implemented. the left [8]. Then, the solar output voltage (VPV) is forced to
follow Vmpp by firing the IGBT in the buck converter, which is
III. ENERGY MANAGEMENT AND CONTROL shown in Fig.4(a). The PWM pulses are generated by comparing
The distributed resource used in this study is solar PV, which the PI controller output with a triangular carrier signal. It can be
is intermittent in nature. An intelligent energy management and seen that even under varying atmospheric conditions the solar
proper control scheme is required for stable operation of the output voltage is very closely following Vmpp.
microgrid and maximum utilization of solar power. The
implemented control scheme for each component is as follows:
• Control of DC-DC buck converter is to harness the
maximum power from solar PV.
• Control of battery charging and discharging circuits is to
maintain the DC link voltage within the acceptable limit.
• Control of the grid tied rectifier is to supply the required
power to the system in the absence of solar PV and
(a)
battery power.
The main indicator for charging and discharging of the
battery is battery SOC. To protect the battery from overcharging
and deep discharging, battery SOC is maintained within 30% and PPV
90%. These numbers can be freely chosen according to the
design frame work without changing the control algorithm. The
following actions are taken according to the detected SOC VMPP
VPV
conditions [7]. (b) (c)
• Charge battery if; Fig. 4. (a) Solar PV array with DC-DC buck converter (b) Maximum power
o SOC < 90% and PPV > PLoad point voltage generating model in PSCAD. (c) P-V characteristics curve of solar
module.
• Discharge battery if;
o SOC > 30% and PPV < PLoad
B. Battery Chrging and discheging Control
• Load shedding if;
Bidirectional converter is used for battery control. However,
o SOC < 30%, PPV < PLoad and utility grid is not
in this simulation, an isolated DC-DC converter is used for
available.
battery charging and a buck converter is used for battery
• Disconnect solar PV if; discharging. This topology is chosen to simplify the control of
o SOC > 90% and PPV > PLoad the battery converter. In the battery charger circuit an isolation
transformer is used to provide galvanic isolation between the
• Grid connection if;
o SOC < 30% and PPV < PLoad input and the output.

• Disconnect the grid if; C. Grid Connected Rectifier Control

o 30 % < SOC and PPV > 0 or The grid connected rectifier used for the simulation is shown
o SOC > 90%
in Fig. 5. The rectification is done by the six pulse IGBT
converter and the voltage control is done by buck converter.
A. Maximum Power Point Tracking (MPPT)
The modeled Solar PV array with the DC-DC buck converter The feedback of the bus voltage is compared with the
is shown in Fig. 4(a). Maximum power point tracking is done for reference voltage of 48 V DC to produce an error signal, which is
fed to the PI controller.

3
Fig. 5. Grid connected rectifier model in PSCAD.

8
Then the PI controller output signal is compared with a saw (a)
tooth wave of frequency 5 kHz and amplitude 1 to produce the 6

Ppv (kW)
PWM signals to fire IGBT in the buck converter.
3
The control system will monitor the solar PV power and
battery SOC. If solar PV power is less than load power and 0
0 50 100 150 200
battery SOC is less than 30% the grid rectifier will be Time (s)

automatically connected to serve the load and charge the battery. (b)

Irradiance (W/m )
2
1000
During other operation periods it can be seen that the solar output
voltage is closely following Vmpp. This will ensure that 500
maximum power is harnessed from solar PV.
0
0 50 100 150 200
IV. SIMULATIONS AND RESULTS 60
Time (s)
(c)
56
Volatge (V)

The simulations are mainly focused on control and energy 52

management of PV and battery during islanded mode and grid 48
connected mode when the microgrid is importing power from the 44

utility grid. The simulation run times were chosen as to 40

0 50 100 150 200
Time (s)
demonstrate the battery SOC dynamics. The simulations are done
under three scenarios: (d)

• Scenario (1) Islanding mode of operation of the DC

microgrid. At point of operation battery reaches its
maximum charging limit.
• Scenario (2) At point of operation in islanding mode
battery reaches its minimum charging limit and load (e) 250
Vmpp
shedding is done to protect the battery. 200
Vpv
Voltage (V)

• Scenario (3) Transition of the microgrid from islanding 150

mode to grid connected mode. 100

Fig. 6 shows the simulated results for Scenario 1. In this 50

0 50 100
Time (s)
150 200

simulation, for the convenience of visualization and to reduce (f)

simulation time, 30 s in real time was chosen as 1s in simulation
time. Therefore, total simulation time of 200 s indicates 1 hour
and 40 minutes of real time. According to the time scale the
battery capacity was de-rated.

Fig. 6. (a) Irradiance input (b) Solar PV power (c) DC bus voltage (d) Battery
state of charge (e) VMPP and VPV (f) Load power for scenario 1

4
 At the begging, the battery state of charge is less than 90%. This SOC value can be chosen arbitrarily to reduce the
With the increase of solar power, the battery will get charged and number of ON/OFF operations of the breaker and to have higher
the SOC of battery reaches 90% as shown by point 1 in Fig. 6 utilization of solar power. It can be seen from Fig. 6 (e) that PV
(d). If the battery is overcharged it will reduce the battery life output voltage raises to around 225 V during the open circuited
time. Since the grid is not connected PV array is disconnected period. Fig. 6 (f) shows the load power variation. Even during the
allowing the battery to discharge. When the SOC reaches 80% dynamic operations, the load is supplied without interruptions at
the PV array is connected back to system, which is shown by 2 kW at 48 V by the microgrid system.
point 2.
(a) 1000 (a)
Irradiance (W/m2 )

1500

Irradiance (W/m2)
1000
500
500

0 0
0 50 100 150 200 0 24 48 72 96
Time (s) Time (s)

(b) 5 (b) 8
4

Ppv (kW)
6
Ppv (kW)

3
4
2
1 2
0 0
0 50 100 150 200 0 24 48 72 96
Time (s) Time (s)

(c) 60 (c) 64
56
Voltage (V)

Voltage (V)
56
52 48
48 40
44 32
40 24
0 50 100 150 200 0 24 48 72 96
Time (s) Time (s)

(d) (d)

(e) 200 (e)

Voltage (V)

100

0
0 50 100 150 200
Time (s)

(f) 100
Grid Current (A)

(f) 2.3
80
2
Pload (kW)

60
1.5 40
20
0.9 0
0 50 100 150 200 0 24 48 72 96
Time (s) Time (s)

Fig. 7. (a) Irradiance input (b) Solar PV power (c) DC bus voltage (d) Battery Fig. 8. (a) Irradiance input (b) Solar PV power (c) DC bus voltage (d) Battery
state of charge (e) VMPP and VPV (f) Load power for scenario 2 state of charge (e) Load power (f) Grid current at DC interface for scenario3.

5
 Fig. 7 shows the simulated results for scenario 2. At the Acknowledgment
beginning, battery SOC is at 75%. Solar power is not enough to
The authors would like to thank Dr. Dharshan Muthumuni,
feed the load in this case and therefore, battery gets discharged
Managing director of HVDC research center Manitoba Canada,
and reaches its minimum discharging limit of 30% at point 3,
for his continuous support in doing this project. Authors would
which can be seen from Fig. 7 (d). Therefore, the Energy like to show their gratitude to Eng. Kasun Wijethunga who
Management System (EMS) will automatically disconnect the supported us in collecting data from Dialog GSM, Sri Lanka.
noncritical load of 1.1 kW to protect the battery from deep
discharge. In this scenario, solar PV power is more than enough
to serve the critical load of 900 W, without further draining the
battery.
References
[1] Elsayed, A. Mohamed and O. Mohammed, "DC microgrids and
As seen in Fig.7 (d) the excess solar power is utilized to distribution systems: An overview", Electric Power Systems Research, vol.
charge the battery. If the solar power is not enough to serve the 119, pp. 407-417, 2015.
critical load of 900 W, the EMS will disconnect the total load and [2] Lifshitz and G. Weiss, "Optimal energy management for grid-connected
storage systems", Optimal Control Applications and Methods, vol. 36, no. 4,
allow the battery to charge with the available solar power. After pp. 447-462, 2014.
battery SOC reaching 30% the EMS will reconnect the non- [3] J. Lee, H. Kim, B. Han, Y. Jeong, H. Yang and H. Cha, ‘‘DC Micro-
critical loads, which is shown in Fig.7 (f). As it can be seen in grid Operational Analysis with Detailed Simulation Model for Distributed
Fig.7 (e) the PV voltage is closely following Vmpp, which Generation’’ Journal of Power Electronics, vol. 11, no. 3, pp. 350-359,
ensures maximum power point operation of solar PV throughout 2011.
the Scenario 2. [4] Y. Gu, W. Li and X. He, "Passivity-Based Control of DC Microgrid for
Self-Disciplined Stabilization", IEEE Trans. Power Syst., vol. 30, no. 5, pp.
The simulated results for scenario 3 are shown in Fig. 8. This 2623-2632, 2015.
simulation was done for 24 hours. For simulation purpose the [5] K. Kusakana, "Optimal scheduled power flow for distributed
photovoltaic/wind/diesel generators with battery storage system", IET
simulation time is scaled down to 96 s which represent 4 s of Renewable Power Generation, vol. 9, no. 8, pp. 916-924, 2015.
simulation time equals to one hour in real time. The battery [6] "RENESOLA UNITED STATES", Renesola.com, 2016. [Online]. Available:
capacity is de-rated according to the simulation time. Simulation http://www.renesola.com/product/Solar. [Accessed: 05- Apr- 2016].
time begins at 6.30 AM of the day. During the evening time solar [7] R. Zamora and A. Srivastava, "Energy Management and Control
power gradually decreases and ultimately becomes zero. algorithms for Integration of Energy storage Within Microgrid", Renewable
and Sustainable Energy Reviews, vol. 14, no. 7, pp. 2009-2018, 2010.
Therefore, battery discharges to its minimum SOC limit of 30%,
[8] X. Han and Q. Liu, "Modeling and Simulation of Grid-Connected
which is shown in Fig.8 (d). To maintain continuous power Photovoltaic System Based on PSCAD", AMR, vol. 986-987, pp. 367-370,
supply to the load the EMS will automatically connect the DC 2014.
microgrid to the utility grid. As shown in Fig.8 (c) the DC bus
voltage is undergone a transient due to breaker operation.
Consequently, the battery gets charged and load is supplied by
the grid. According to the simulation results, it can be ensured
that even when the solar power is not available the load can be
served by operating the system in grid connected mode.

V. CONCLUSION.
In this paper, typical DC microgrid for a telecommunication
system is modeled and three different scenarios were simulated.
Proper Energy Management and Control algorithm is proposed
for the DC microgrid. The major concern is given to maintain the
DC bus voltage at a desired value and to protect the battery. The
simulation results indicate that the proposed control algorithm
ensures required continuous operation for a telecommunication
system at a loss of grid under intermittency of solar PV.
Therefore, the algorithm will be tested experimentally in a real
system as the next phase of the project. For further studies,
current limiting mechanisms can be designed for battery, rectifier
and solar PV interconnecting circuits. Also transient studies can
be carried out for various disturbances and fault conditions.