A Benchtop DC Microgrid for Renewable Energy

Sources Integration
Diego A. Aponte Roa, MS Jeffrey Borres Martinez Xavier Collazo Fernandez
Department of Electrical and Department of Electrical and Department of Electrical and
Computer Engineering Computer Engineering Computer Engineering
University of Turabo University of Turabo University of Turabo
Gurabo, Puerto Rico Gurabo, Puerto Rico Gurabo, Puerto Rico
aponted1@[Link] jborres2 @[Link] xcollazo4@[Link]

Wayne W. Weaver, Ph.D

Department of Electrical and
Computer Engineering
Michigan Tech. University
Michigan, USA
wwweaver@[Link]

Abstract After decades of the traditional Distributed Grid,

—

have the same frequency, it only need to be of the same

the microgrid are emerging as the new order on modern smart voltage value. There is no standard for the voltage bus of DC
grids. However, due to the excessive use of DC to AC converters houses, but ones of the typical values are 24V and 48V.
to allows the integration of renewable energy based on DC, as
photovoltaic systems, researchers are now focus in studying the What it is necessary is to have a control system to stabilize
feasibility of implementing DC distribution on a given the DC DC converters to make the microgrid operation
-

application, by involving DC microgrids in the future smart optimized. There are many controls that can been applied, but
grids. These DC microgrids have control system advantages and the droop control is the most used. In these systems, it is
their easy integration with IoT technology, makes it a real implemented a “virtual” impedance “R droop” that with a
alternative in applications as future smart homes. This paper controllable variable voltage (because of the converters), it
presents the simulation of a DC Microgrid and the designing and creates a closed-loop feedback controller that could make the
preliminary implementation of a benchtop DC microgrid. system share a load between the sources available [ 9-10].
Keywords —

Distributed Systems; DC Microgrid; Droop This paper presents a DC microgrid system derivation and
Control; DC DC Converter; MATLAB/Simulink
-

a simulation in MATLAB/Simulink environment using the

microg rid model shown in Fig. 1. The model includes two
I. INTRODUCTION sources and one load that could be anything (ac, dc, voltage
Microgrids is a set of distributed generation (DG) , levels, power). All elements are connected by a common DC
controlled to operate either in Islanded mode or connected to bus. Simulations were performed without a control load
the grid [1]. These grids are view like the solution for the sharing and with a droop control implemented to the system to
environmental problems by increasing the number of demonstrate a desired load sharing between the DC-DC
renewable sources integration on the distributed generation, converters for a specific case of study. Preliminary results of a
which allow to add energy storage elements and new DC microgrid test bench is presented as well.
technologies [2 ]. There are different types of microgrids
depending which control system will be developed [3]. The
majority of the microgrids are AC, since is easier to integrate
them in the grid, but there are also AC /DC hybrid [4 ] and
purely DC microgrid. Some advantage of DC microgrids is
that the distribution is more efficient than AC due to most of
the loads on building and house are DC, meaning that it will
not require an AC-DC converter which create losses. In DC
grid is easier to implement a control system due to the
elimination of frequency controls as well [5]. In overall,
compared with AC systems, DC systems have advantages of
higher reliability and efficiency, simpler control and natural
interface increasing the number of DC renewable generations,
energy storage and electronic loads [6].
Fig. 1. DC Microgrid Model.
DC microgrids supply power to the DC bus loads by
interconnected sources such as: fuel cell power plants, PVs,
micro wind turbine and battery storage systems, operating in
-

II. DC M ICROGRID S IMULATIONS

parallel. The DC -DC converters used in DC microgrids are to To investigate how the droop control operates, a DC
interface some renewable sources to the bus or load, since the
Microgrid model was performed in Simulink using the model
renewable energy resources operate normally at low voltage
[7 -10]. In comparison of AC source, the DC doesn’t require to

978-1-7281-0554-3/19/$31.00 ©2019 IEEE

0079
scheme showed in Fig. 1. System is assuming to works in
continuous mode. Table I shows selected parameters:

TABLE I. PARAMETERS SELECTED

B. Simulink Model
Vbus 24V
V1 12V To obtain load sharing between DC-DC converters, it is
L1 1mH necessary to adjust both converter’s duty cycle. The values
R1 100mΩ obtained are: D1 = 0.499917 and D2 = 0.831675. Fig. 3 shows
V2 20V the DC microgrid model implemented in MATLAB/Simulink
(each equation was implemented with a specific color to
L2 2mH
facilitate their tracking). A script was performed as well
R2 200mΩ
(outputs can be obtained in the Scope or by using the script).
CLoad 2000μF
RLoad 12Ω

From the DC microgrid model presented in Fig. 1, we can

identify two important facts:

i. Common bus voltage:

ii. Sum of bus currents:

A. Converter Model
Assuming that both sources are boost converters, the
corresponding converter model is presented in Fig. 2.

Fig. 3. DC Microgrid in MATLAB/Simulink.

C. Simulink Model Response

The converters dynamics is shown in Fig. 4. The system
reaches its steady-state in 20ms. Each boost converter supplies
current to the load.

Fig. 2. DC Microgrid Topology.

Assuming each switch act in complement (q1 = 1 – q2), the

average-mode state equation is:

Where D is the duty cycle. Two algebraic expression are

founded: Fig. 4. Inductor currents response for each converter.

0080
Fig. 5. Total current supplied by each converter to the bus

A control system for dynamic (transient of the system)

was not implemented yet. Fig. 5 and Fig. 6 shows the current
and voltage available in the bus. As expected, the steady-state
bus current was around 2A while the bus voltage has reached Fig. 7. DC Microgrid under droop control
the desired value of 24V.
To integrate the droop control, it is necessary to have a
virtual resistance Rdroop in front of a nominal reference voltage
Vref. Having this, the controller will then adjust the voltage to
the desired value. To reach that value a PI controller was
implemented for each source (see Fig. 8):

Fig. 6. Bus Dynamics Response

D. Droop Control
The power supplies and boost converters can be lumped as
equivalent voltage sources. To model the dynamic microgrid
with both sources under droop control, it is necessary the Fig. 8. Closed-loop feed-back controller
measures of both output current and output voltage in each
converter (see Fig. 7). Some important facts are identified: Following equation are obtained:

From the model, the source currents are:

Fig. 9 shows the PI controller implemented in Simulink

while Fig. 10 show the system response.

0081
 III. PROPOSED DC MICROGRID TEST BENCH

Fig. 12 shows a diagram of the proposed DC Microgrid

Test Bench. In the test bench a micro-wind turbine (MWT) of
400W will be used as a source. This turbine has been tested
with a wind speed around 7.5m/s. An electronic load was
connected and adjusted from 3W to 12W, which indicates that
the turbine could give enough power to small devices like a
led and cellphone charger. In addition, two PV of 24V/100W
are going to be used with MPPT and batteries of 12V and 7
A/h will be connected as well.

Fig. 9. Closed-loop PI controller in Simulink

Fig. 12. Proposed DC Microgrid Test Bench

The PVs were connected in parallel giving 20V, and 40V

when it was in series. To ensure that the MPPT was working
and the PV could charge the battery while having a load, a
circuit was implemented with the PVs connected in Parallel
while using a buck-boost set on 24V.
IV. PROPOSED DC MICROGRID TEST BENCH RESULTS
Fig. 10. System Response with PI Controller
Table II presents the data collected from the PV test
If all the sources have implemented the closed-loop droop performed in a sunny day. Results demonstrated that the PVs
control effectively, then the system will behave like the below used are going to fit as planned for the test bench that will be
circuit shown in Fig. 11. implemented.

TABLE II. PV GENERATION ON DIFFERENT LOADS

Voltage (V) Current (A) Power (W)

11.63 0.086 1
23.70 0.211 5
23.81 0.420 10
23.81 0.630 15
23.81 0.840 20
23.81 1.050 25
24 1.25 30
23.81 1.68 40
23.70 2.11 50
23.62 2.54 60

For the implementation of the bench, the sources were

already mentioned: two PV panels on parallel and a micro
wind turbine of 400W generator. The PV output was
connected to the MPPT which was supplying to the battery
and to the buck-boost to reach the DC bus bar desired voltage.
Fig. 11. Droop control in the proposed Microgrid In a similar form, the wind turbine output was first rectified,
and then connected as input to the buck-boost converter
leading to the bus bar. The electronic load was connected to

0082
the DC bar to be work as a load. Each source had a multimeter integration of a battery, two renewable energy sources
to reading the current flow. Equipment used is presented in combined in a DC bus bar of 24V, a dummy load generated by
Fig. 13 while results are summarized in Table III. an electronic load.

Droop control strategy was simulated in

MATLAB/Simulink and seems to be an easy way to control
the proposed DC Microgrid. The closed-loop feed-back
controller allows the DC bus to reach the desired steady-state
value in 1 second.
VI. FUTURE WORK
The designed droop control will be implemented on the
test bench to verify the experimental results. If results are
positive, the DC microgrid will be set as the electrical system
of a small green house available in our institution. In addition,
a Bluetooth device with sensors will be integrated to the
system to show a life stream data about currents flow, voltage,
and power consumption to facilitate the data analysis and
Fig. 13. Equipment used to complete the test bench. calculate system efficiency. The final goal is to create a DC
microgrid smart house. This smart house is expected to have a
circuit simulating loads to recreate the human behavior. This
TABLE III. RESULTS OF THE DC MICROGRID TEST BENCH will allow us to study more deeper the efficiency and
performance of load-sharing control schemes applying to DC
PV Current MWT Total Current Power
grids on a real-life scenario. Also, a comparison between AC
(A) Current (A) (A) (W)
and DC microgrids will be done to validate the assumption of
0 0.044 0.044 1 DC microgrids will being more efficient.
0.094 0.121 0.215 5
0.211 0.200 0.411 10 ACKNOWLEDGMENT
0.320 0.310 0.630 15 The authors gratefully acknowledge the support of this
0.444 0.397 0.841 20 work by the José Domingo Pérez school of Engineering at the
0.570 0.47 1.044 25 Universidad del Turabo, the Puerto Rico Energy Center
0.690 0.560 1.25 30 (PREC), and its Consortium for Integrating Energy System in
0.865 0.590 1.455 35 Engineering and Science Education (CIESESE), an initiative
funded by the U.S. Department of Energy (DE-NA0003330).
Results show how the sources were contributing to supply
the require power as expected. Typically, the micro wind REFERENCES
turbine can supply around 12W to the electronic load with a [1] H. Shanvanfar, S. Malek, “Photovoltaic Microgrids Control by the
7m/s wind speed applied. By implementing the PV to the dc Cooperative Control of Multi-Agent Systems”, 30th International Power
bus bar, the system was able to sustain a 35W load as well as System Conference (PSC), 2015.
provide charge to the 12V storage element. Nevertheless, after [2] T. Dragicev, X. Lu, J. Vasquez, and J. Guerrero, “DC Microgrids-Part I:
A Review of Control Strategies and Stabilization Technique”, IEEE
passing the 35W the turbine shutdowns demonstrate the Trans. Power Electronics, vol 31, no. 7, pp. 4876-4891, July. 2016.
importance of having a control system. This test validates the [3] J. Paniagua, E. Unamuno, and J. Andoni, “Experimental Test Bench for
flexibility and feasibility of the system by combining two Testing DC Microgrid Control Strategies”, Mondragon Universiteit,
power sources to sustain a common DC load. Mondragon, Spain, 2017.
[4] J. Paniagua and E. Unamuno, “Hybrid ac/dc microgrids—Part I: Review
and classification of topologies”, Renewable and Sustainable Energy,
V. CONCLUSION Reviews 52, 2015.
After decades of the traditional DG a new technology with [5] E. Rodríguez, X. Su, M. Savaghebi, J. Vasquez, M. Han, and J.
Guerrero, “Intelligent DC Microgrid Living Laboratories - A Chinese-
intelligent interconnected grid using Microgrids are being Danish Cooperation Project”, State Key Laboratory of Alternate
watching as a feasible solution. Many cities and islands are Electrical Power System with Renewable Energy Sources, Electric
already working their electricity with Microgrids, but the Power University, Beijing, China, 2015.
majority of these, if not all, are based on AC current, making [6] F. Li, Z. Lin, Z. Qian, and J. Wu, “Active DC bus signaling control
method for coordinating multiple energy storage devices in DC
the system more inefficient because of the ac to dc or dc to ac microgrid” IEEE Second International Conference on DC Microgrids
converters through all the system connection. (ICDCM), Nuremburg, Germany, Jun. 2015.
[7] T. Pavlovic, T. Bjazic, and Z. Ban, “Simplified Averaged Models of
Our study is to create a DC microgrid with the main DC-DC Power Converters Suitable for Controller Design and Microgrid
Simulation,” IEEE Trans. Power Electronics, vol 28, no. 7, pp. 3266-
objective of eliminates a big part of these converters and 3275, Jul. 2013.
increase the system efficiency by the implementation of a less
complex control system than a traditional one. Results
presented in this paper was obtained in the proposed test
bench and demonstrates the feasibility of these scheme by the

0083
[8] P. Chaudhari, M. Khadse, V. Jadhav, P. Patil, K. Daware, M. Rajwade, [11] S. Augustine, M. Mishra, and N. Narasamma, “Proportional Droop
and S. Sharma, “Novel Control Strategy for Dynamic Load Sharing Index Algorithm for Load Sharing in DC Microgrid”, IEEE
Between DC-DC Converters for DC Microgrid”, in Proc. IEEE Signal International Conference on Power Electronics, Drives and Energy
Processing, Informatics, Communication and Energy Systems Systems (PEDES), 2014.
(SPICES), 2015. [12] R. Ferreira, H. Braga, A. Ferreira, and P. Barbosa, “Analysis of Voltage
[9] M. Abdollahi, and L. Wang, “Control of DC-DC Bidirectional Droop Cotrol Method for dc Microgrids with Simulink: Modeling and
Converters for Interfacing Batteries in Microgrid”, IEEE International Simulation”, IEEE International Conference on Industry Applications
Conference on Power Systems Conference and Exposition (PSCE), (INDUSCON), 2012.
2011.
[10] Z. Fusheng, and R. Naayagi, “Power Converters for DC Microgrids –
Modelling and Simulation” IEEE Innovative Smart Grid Technologies -
Asia (ISGT Asia), Singapore, Singapore, May. 2018.

0084