Real-Time Simulation of a Microgrid Control System using MODBUS Communication

Abstract

In recent years, microgrids have become increasingly common in power systems around the world. Microgrids have complex control and operational requirements, thus, a dedicated communication network is often a requisite for its functionality. Due to economic and logistical concerns, it is also desirable to keep the communication system as simple as possible. Among contemporary system automation protocols, the Modbus protocol is generally the simplest one to implement. This paper demonstrates the applicability of Modbus TCP communication in facilitating the control system of a detailed microgrid simulated in a real-time power system simulator. The simulated microgrid exchanges voltage, frequency, power measurements, breaker statuses as well as power set points of generators with a central control location via Modbus protocol, where a control system continuously monitors the microgrid. Control centre issues commands depending on the operating conditions of the microgrid such as to switch the microgrid controls from the grid connection mode to the islanded mode, change the protection relay settings groups and activate a load-shedding scheme. Implementation of relevant microgrid controls and Modbus communications are presented using an example microgrid case with the results obtained.

Full text

1
Real-Time Simulation of a Microgrid Control System
using MODBUS Communication
Sachintha Kariyawasam, Dinesh Rangana Gurusinghe*, and Dean S. Ouellette
RTDS Technologies Inc., Winnipeg, Canada
(*Corresponding author: dinesh@rtds.com)
ABSTRACT
In recent years, microgrids have become increasingly common in power systems around the world.
Microgrids have complex control and operational requirements, thus, a dedicated communication
network is often a requisite for its functionality. Due to economic and logistical concerns, it is also
desirable to keep the communication system as simple as possible. Among contemporary system
automation protocols, the Modbus protocol is generally the simplest one to implement. This paper
demonstrates the applicability of Modbus TCP communication in facilitating the control system of a
detailed microgrid simulated in a real-time power system simulator. The simulated microgrid exchanges
voltage, frequency, power measurements, breaker statuses as well as power set points of generators
with a central control location via Modbus protocol, where a control system continuously monitors the
microgrid. Control centre issues commands depending on the operating conditions of the microgrid such
as to switch the microgrid controls from the grid connection mode to the islanded mode, change the
protection relay settings groups and activate a load-shedding scheme. Implementation of relevant
microgrid controls and Modbus communications are presented using an example microgrid case with
the results obtained.
Keywords – Microgrids, Distributed Energy Resources (DER), Islanding, Modbus Communication,
Hardware-In-the-Loop (HIL) Simulation, Real-Time Simulator
1. INTRODUCTION
A microgrid is a portion of a distribution network and is usually designed to operate in parallel with the
power grid or autonomously as a power island [1]. In contrast to traditional distribution networks, a
microgrid is generally a self-sustained entity with distributed energy resources (DERs) catering for its
energy consumption. Integration of small-scale, renewable energy resources such as solar-PV, wind
power and mini-hydro power is a key attribute of microgrids [2]. Modern microgrids are also equipped
with energy storage devices such as battery banks. When properly managed, microgrids improve the
overall power system performance in addition to providing significant economic and environmental
benefits. However, safe and reliable operation of a microgrid in grid connected and islanded operation
is a challenging task due to issues like islanding detection, protection coordination with bidirectional
power flow, system stability concerns, and intermittent nature of renewable energy resources [3]-[6].
With their increasing popularity, safe, reliable and cost effective tools are necessary to study challenges
in microgrid implementation, operation and control. Real-time hardware-in-the-loop (HIL) simulations
are one of the best approaches in this regard, as the simulated system can replicate the actual microgrid
system in detail [7], while allowing the interfacing of external device to the simulation. Real-time power
system simulators can interface protection and control devices as well as power sources and
communication equipment to a modelled microgrid. This enables systematical evaluation of
performances of individual devices as well as the system as a whole.

2
The remainder of this paper is organized as follows. Section 2 explains the importance of a
communication system in a microgrid. It further discusses why Modbus protocol is worth investigating
as a suitable candidate for system automation in a microgrid. Section 3 provides information regarding
the test microgrid system and its communication interface using Modbus protocol. In Section 4, test
results are presented for several cases, where control actions are performed using the implemented
Modbus communication interface. Section 5 follows with conclusions emphasizing the main findings of
this paper.
2. INFORMATION EXCHANGE IN A MICROGID
Unlike conventional distribution systems, microgrids undergo frequent changes in operating conditions,
enforced by energy sources as well as loads. Therefore, a microgrid has additional control and
operational requirements than a conventional system, which demands real-time information exchange.
2.1. Requirement for a Dedicated Communication System in a Microgrid
As explained above, a microgrid must operate within acceptable limits in both grid connected and
islanded modes. When switching to the islanded mode of operation, a microgrid should take appropriate
control actions to maintain its local frequency. This usually involves changing the operation mode of one
of its generators from droop control mode to isochronous frequency control mode, upon detection of the
islanding [8], [9]. In addition, islanding often causes variations in power flow and voltage profiles within
the microgrid. Changes to the microgrid topology may be required to remedy some of these issues.
Furthermore, it is necessary to maintain power balance in the islanded mode by changing the reference
power set points of DERs, connecting/disconnecting reactive power sources such as capacitor banks
and enforce load shedding if load demand exceeds the available generation [3], [4].
In a power system protection perspective, islanding might cause complications to protection
coordination of the system [5], [6]. In general, fault levels in a microgrid are lower in islanded mode than
they are in grid-connected mode, as the fault current contributions from the main grid are now absent.
Fault levels become further unpredictable due to the intermittent nature of renewable type DERs, which
are often an integral part of modern microgrids [10]. Moreover, DERs, particularly those with full-
frequency converters such as solar-PV and type-4 wind generation, pose unique challenges to power
system protection [11]. These complicated operating conditions demand protection relays to vary their
protection settings and configurations, depending on circumstances.
Operation of modern microgrids are also affected by economic considerations, which demand available
energy resources to be dispatched optimally, without violating system operating limits [12]. This
generally involves harnessing the maximum amount of power available from renewable type DERs. The
intermittent nature of the power output from these type of sources essentially means that the power set
points of other generation must be varied dynamically.
It is the role of the microgrid control system to manage all of the above-mentioned control and protection
considerations. A microgrid control system can be implemented as a central unit or a distributed system,
but both architectures require a microgrid to have a dedicated communication network. In addition,
monitoring and supervision of system data is an integral part of the control system as well.
2.2. Automation of Microgrid Communication
There are a number of existing protocols suitable for automation of information exchange in a microgrid
such as IEC 61850, DNP3, IEC 60870-5-104 and Modbus [13]. Although less refined than its newer
counterparts, Modbus protocol is often the simplest and the least expensive option among them. When

3
configuring a comparatively smaller system such as a microgrid, using Modbus can be more convenient
to use than a protocol like IEC 61850. However, IEC 61850, with object oriented data modelling and a
rigorous engineering process, has distinct advantages when the system under consideration is large
[14]. Low bandwidth requirements of Modbus is also ideal for a microgrid, where laying out a complex
communication network is often not feasible. Moreover, Modbus protocol provides sufficient
interoperability and robustness and, remains a commonly supported protocol by relays and other
intelligent electronic devices (IEDs).
Modbus is a protocol with a client/server architecture, originally designed with serial communication for
system automation to facilitate communication between a master station and a Remote Terminal Unit
(RTU) [15]. A Modbus master (the client) interfaces with one or multiple slave stations (the server) using
a request/reply routine. In general, a Modbus transaction is always initiated by the master and the slaves
transmit data only as a reply. Currently, Modbus protocol has few implementation variants and supports
both serial and internet protocol (IP) communication [16]. In this paper, Modbus TCP protocol is
proposed for data communication and automation of a microgrid control system.
3. MICROGRID CASE AND TEST SETUP
The test system used in this paper is a modified version of CIGRE C6.04.02 benchmark North American
medium voltage distribution network [17]. The topology of the microgrid structure is shown in Fig. 1.
Table I Information exchanged between Modbus slaves at each location and the master at control centre
Location
Input/output Data Points
Monitoring (read only)
Control (read-write)
Discrete Inputs
Input Registers
Coils
Holding registers
Bus 1 (B1)
S1 status
Scap status
L1 status
Current through S1
B1 voltage
B1 frequency
S1
Scap (1 to 3)
R2 activation
Bus 2 (B2)
L2 status
BRK26 status
B2 voltage
B2 frequency
BRK26
R1 Setting group
R3 activation
Bus 3 (B3)
L3 status
S3 status
BRKPV status
B3 voltage
B3 frequency
PV power output
S3
BRKPV
R4 activation
Bus 4 (B4) and
Bus 5 (B5)
L4, L5 status
S2 status
BRKWind status
B4 voltage
B4 frequency
Wind power output
S2
BRKWind
R5 activation
Bus 6 (B6)
L6 status
B6 voltage
B6 frequency
Bus 7 (B7)
L7 status
BRKDie status
B7 voltage
B7 frequency
DG power output
BRKDie
DG Mode Control
DG Power set-point

4
The microgrid is connected to a 138 kV main grid through a 25 MVA, 138/13.2 kV Δ-Y transformer with
8% leakage impedance [18]. There are three DERs in the microgrid including a 1.74 MW PV system
connected to bus B3 and a 2.0 MW doubly-fed induction generator (DFIG) wind turbine system
connected to bus B5. In addition, a 5.5 MVA diesel generator is connected to bus B7. A switched
capacitors bank rated at 500 kvar is connected at bus B1 to provide reactive power support. The
microgrid is interconnected using a circuit breaker S1. Circuit breakers S2 and S3 are kept open to
maintain a radial network in grid connected mode. S2 will be closed following an islanding to increase
the system reliability. The total active and reactive loading of the microgrid are 7.39 MW and 2.936 Mvar,
respectively [18].
GRID
1200 MVA
M1
T/f
138/13.2 kV
8%
S1
Capacitor
Bank
L1
7.2 km
Scap
B1
1.3 km
0.6 km
L2
L3
2.3 km
PV
S3
0.5 km
1.4 km
L7
Diesel S2
L4
Wind
L5
B5
1.7 km
B6
L6
B4
B3
B2
B7
MODBUS
Slave
MODBUS
Slave
MODBUS
Slave
MODBUS
Slave
MODBUS
Slave
MODBUS
Slave
MODBUS
Master
R1
(50/51)
Line 26
R2
(81U)
R3
(81U)
R4
(81U)
R5
(81U)
BRKWind
BRKPV
BRKDie
BRK26
Control Centre
Fig. 1 Topology of the Microgrid with MODBUS Slaves

5
It is assumed that each bus shown in the microgrid has a Modbus slave capable of responding to
requests sent by the Modbus master at the central control centre. Locations of the Modbus
communication devices are also shown in Fig. 1. Information exchanged between Modbus slaves and
the master at the control centre are provided in Table I.
In an actual microgrid system, each of the six locations as illustrated in Fig. 1 must have Modbus slave
devices in order to complete the microgrid communication system. In this real-time simulation setup,
one Modbus slave is sufficient to interface with the remote Modbus master as all the data points are
internally generated and therefore, available inside the simulation itself. The Modbus slave interface in
the real-time simulator has the capability to accommodate hundreds of data points of each type [19]. If
required, however, this test setup can be modified without difficulty to accommodate independent slaves
by adding multiple network interfaces cards.
A schematic of main connections of the microgrid test setup is shown in Fig. 2. The microgrid given in
in Fig. 1 and all its controls are modelled inside the main simulation. The main simulation uses an
electromagnetic transient type solution algorithm and runs on a dedicated hardware platform in real-
time. Data generated inside the simulation are mapped to the Modbus slave using an internal points
mapping file attached to the Modbus component in the simulation. Combination of the Modbus
component and the network interfaces card of the simulator works as a Modbus slave, and any properly
configured Modbus master can connect to it. All communications are carried out using Modbus TCP
protocol.
GTNET MODBUS
Component
Internal Data
Mapping
Network Interface card
(GTNET Card)
Network
MODBUS (TCP/IP)
Remote MODBUS Master
(RSCAD runtime MODBUS
master interface)
MODBUS
(TCP/IP)
External Device
Main Simulation
Real-Time Simulator
Fig. 2 Connection Setup between Devices

6
Runtime scripting feature of the real-time simulator supports Modbus master commands, enabling
implementation of the Modbus master interface for the microgrid control centre in the workstation
computer that runs the simulation. As depicted in Fig. 2, the remote Modbus master resides external to
the real-time simulator in the workstation computer and implemented using a runtime script file.
4. Test Results
The Modbus master is continuously requesting measurements and statuses values from the slaves as
indicated in Table I. In turn, the control centre issues appropriate commands, when changes in operating
conditions of the microgrid are detected. The following sub-sections present few such scenarios with
the results obtained.
4.1. Islanding Detection
In this microgrid test setup, an islanding event is detected based on the status of the circuit breaker at
the point of common coupling, S1. The microgrid control system continuously monitors the status of the
S1 circuit breaker via Modbus. When the control system detects the opening of S1, it issues a Modbus
command to the diesel generator to switch into isochronous frequency control mode from droop control.
The diesel generator is then responsible for maintaining the local frequency while the microgrid is in the
islanded mode. Similarly, the control system commands the diesel generator to switch back to droop
control mode, once after the microgrid has reconnected to the main grid.
4.2. Switching Relay Setting Groups
As explained in Section 2, islanding of the microgrid causes changes in the system that demands
adjustments in protection settings. One such case is presented in this sub-section with results.
In grid connected mode (and with both S2 and S3 open), phase A current of Line26 measured at bus B2
is around 93 A (in this case, the load currents are unbalanced). During an A to ground fault on bus B6,
phase A of Line 26 will observe a steady state fault current around 1400 A at bus B2. In contrast, when
the microgrid has islanded, the same line would carry a load current of 75 A and observe a steady state
fault current of only 171 A. The marked decrease of fault current magnitude in the islanded mode is
understandable due to the absence of the fault current contributions from the main grid. Here, the fault
is mainly fed by a solar PV system, which is known to produce weak fault current contributions [11].
Overcurrent protection is often the preferred protection method used in distribution level, including
microgrids and in order to counter such a variation in fault levels, different relay settings are necessary
[20]. The most convenient way to accommodate this is to have two settings groups in the relay and
switch between them depending on the microgrid operation mode.
In order to demonstrate this phenomenon, overcurrent protection is applied to Line26 as indicated in
Fig. 1. Here, the relay R1 at bus B2 carries two setting groups and assumed to have Modbus slave
capabilities. The control centre commands the relay to switch between setting groups using Modbus
communication. Fig. 3 below presents the fault currents and the corresponding trip signals from the relay
R1, for an A-G fault at bus B6. One can observe the variation of fault currents from grid connected mode
to islanded mode. Accurate settings allow prompt operation of relays, enabling the microgrid to return
to its stable operation following a fault, without enforcing load shedding.

7
(a) Grid Connected Mode
(b) Islanded Mode
Fig. 3 Fault Currents and Trip Signals
4.3. Adjusting the Power Reference of the Diesel Generator
Economics involved with energy production demand that irrespective of the mode of operation of the
microgrid, both the solar PV system and the wind plant harness the maximum available amount of
energy from their respective sources. In islanded mode, the diesel generator must cater for the rest of
the load by its own, while maintaining the load-generation balance of the microgrid (and consequently,
the frequency as well). In grid-connected mode, however, the diesel generator is operating on a droop
with a specified active power reference. This allows the diesel generator to adjust its power output in an
optimal manner. In this study, active power imported from the main grid is kept constant at a low value
when possible, with the idea of keeping the microgrid a self-sustaining system.
-2.0
-1.0
0.0
1.0
2.0
3.0
Fault Current - Phase A (kA)
-0.2
0.3
0.8
1.3
0.0 0.1 0.1 0.2 0.2 0.3 0.3 0.4 0.4 0.5 0.5
Trip Signal
Time (s)
-2.0
-1.0
0.0
1.0
2.0
3.0
Fault Current - Phase A (kA)
-0.2
0.3
0.8
1.3
0.0 0.1 0.1 0.2 0.2 0.3 0.3 0.4 0.4 0.5 0.5
Trip Signal
Time (s)

8
Plots shown in Fig. 4 illustrates how the microgrid control centre adjust the active power reference set
point of the diesel generator, upon detecting an increase of the wind power output. The controller
constantly monitors the power outputs of the wind and solar systems and adjusts the power reference
set point of the diesel generator accordingly. In Fig. 4, it can be seen that as the power output of the
wind plant increases, there is a temporary decrease of the active power imported from the grid until the
diesel generator settles at its new power reference set point. A similar behavior can be observed when
the power output of the wind plant decreases. In addition, there is no discernable change in either the
voltage magnitude or the frequency at the point of common coupling due to this control action.
Fig. 4 Responses of Wind Output, Diesel Output and Power Imported to the Microgrid with Wind Speed Variation
4.4. Load shedding scheme
An under frequency load shedding scheme is implemented in this microgrid system to arrest a significant
drop in system frequency. Its function is vitally important to maintain the load-generation balance when
the microgrid has islanded. The microgrid control centre enables the under frequency function at
different locations of the microgrid using Modbus commands, following an islanding. In this setup, loads
connected to buses B1, B2, B3 and B4 are presumed to be non-critical and will be dispensed with by the
load-shedding scheme in an emergency. The loads are shed by under frequency relays at respective
buses in response to local frequency measurements. The relays’ algorithm has pickup thresholds for
both under frequency and rate of change of frequency. In Fig. 5, the operation of the load-shedding
8.0
9.0
10.0
11.0
12.0
13.0
Wind Speed (m/s)
0.0
1.0
2.0
3.0
PWind (MW)
2.5
3.0
3.5
4.0
4.5
PDiesel (MW)
0.60
0.62
0.64
0.66
0.68
0.70
010 20 30 40 50 60
PGrid (MW)
Time (s)

9
scheme is provided for a sudden decrease of wind power output. In this case, microgrid frequency
recovered back to the nominal range by disconnecting L1 and L2 only (loads connected at buses 1 and
2, respectively).
Fig. 5 Operation of the Load-Shedding Scheme due to a Frequency Drop caused by
a Loss of Wind Power Output
0.0
1.0
2.0
3.0
PWind (MW)
59.0
59.4
59.8
60.2
Freq. at B1(Hz)
59.0
59.4
59.8
60.2
Freq. at B2(Hz)
59.0
59.4
59.8
60.2
Freq. at B3(Hz)
59.0
59.4
59.8
60.2
Freq. at B4(Hz)
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
0246810 12 14 16 18 20
Load Shedding Signals
Time (s)
L1 L2 L3 L4

10
5. CONCLUSIONS
This paper has presented the use of the Modbus TCP protocol for operational and control applications
of a microgrid. Several microgrid control actions were performed on a detailed microgrid model
simulated in an EMT type real-time simulator. All information was exchanged using real Modbus TCP
communication between a Modbus master and a slave. Results presented in this paper substantiates
that Modbus TCP protocol is well suited for carrying out communication requirements of a smaller
system like a microgrid. Furthermore, the results indicate that when implemented on modern hardware
platforms, Modbus TCP protocol is responsive enough to execute fast and complex control operations
such as dynamically changing the power reference of a generator.
In addition, this paper demonstrates the necessity of a dedicated communication system for modern
microgrid systems with complex operational and control requirements. This paper further highlights the
advantages of using a real-time simulator for microgrid studies, where the users can not only model the
system in detail, but also interface external protection and control devices using real communication.
6. REFERENCES
[1] F. Katiraei, M. R. Iravani, and P. F. W. P. Lehn, “Micro-grid autonomous operation during and
subsequent to islanding process,” IEEE Transactions on Power Delivery, vol. 20, no. 1, pp. 248-
257, Jan. 2005.
[2] M. Ahmed, U. Amin, S. Aftab, and Z. Ahmed, “Integration of renewable energy resources in
microgrid,” Energy and Power Engineering, 7, pp. 12-29, 2015.
[3] A. A. Salam, A. Mohamed and M. A. Hannan, “Technical challenges on microgrids,” ARPN Journal
of Engineering and Applied Sciences, vol. 3, no. 6, Dec 2008.
[4] D. E. Olivares et al., "Trends in Microgrid Control," in IEEE Transactions on Smart Grid, vol. 5, no.
4, pp. 1905-1919, July 2014.
[5] G. Buigues, A. Dyśko, V. Valverde, I. Zamora, and E. Fernández, “Microgrid Protection: Technical
challenges and existing techniques,” Renewable Energy and Power Quality Journal, vol.1, no.11,
Mar. 2013.
[6] L. Che, M. E. Khodayar and M. Shahidehpour, "Adaptive Protection System for Microgrids:
Protection practices of a functional microgrid system," in IEEE Electrification Magazine, vol. 2, no.
1, pp. 66-80, Mar. 2014.
[7] P. Forsyth and R. Kuffel, “Utility applications of a RTDS simulator,” in Proc. International Power
Engineering Conference, Singapore, pp. 112-117, Dec. 2007.
[8] N. W. A. Lidula, “Determination of requirements for smooth operating mode transition and
development of a fast islanding detection technique for microgrid,” Ph.D. Thesis, Dept. of Electrical
& Computer Engineering, University of Manitoba, Canada, 2012.
[9] D. R. Gurusinghe and D. Ouellette, “Real-time implementation of PMU based islanding detection
schemes for microgrids,” in Proc. of 7th International Conference on Advanced Power System
Automation and Protection (APAP2017), Jeju, Republic of Korea, Oct. 2017.
[10] T. S. Ustun, C. Ozansoy, and A. Zayegh, “Modeling of a Centralized microgrid protection system
and distributed energy resources according to IEC 61850-7-420,” IEEE Transactions on Power
Systems, vol. 27, no. 3, pp. 1560-1567, Aug. 2012.
[11] J. Keller and B. Kroposki, “Understanding fault characteristics of inverter-based distributed energy
resources”, National Renewable Energy Laboratory, Golden, CO, USA, Technical Report REL/TP-
550–46698, Jan. 2010.
[12] V. A. Papaspiliotopoulos, G. N. Korres, V A. Kleftakis, and N. D. Hatziargyriou, “Hardware-in-the-
loop design and optimal setting of adaptive protection schemes for distribution systems with
distributed generation”, IEEE Transactions on Power Delivery, vol. 32, no. 1, pp. 393-400, Feb.
2017.

11
[13] A. Bani-Ahmed, L. Weber, A. Nasiri, and H. Hosseini, "Microgrid communications: State of the art
and future trends," International Conference on Renewable Energy Research and Application
(ICRERA), Milwaukee, WI, USA, pp. 780-785, Oct. 2014.
[14] S. Mohagheghi, J. Stoupis and Z. Wang, "Communication protocols and networks for power
systems-current status and future trends," IEEE/PES Power Systems Conference and Exposition,
Seattle, WA, USA, pp. 1-9, Mar. 2009.
[15] Modbus Organization, “MODBUS application protocol specification V1.1b3,” Apr. 2012. [Online].
Available: http://www.modbus.org/docs/Modbus_Application_Protocol_V1_1b3.pdf.
[16] Modbus Organization, “Modbus messaging on TCP/IP implementation guide V1.0b,” Oct. 2006.
[Online]. Available: http://www.modbus.org/docs/Modbus_Messaging_Implementation_Guide_V1
_0b.pdf.
[17] Benchmark Modeling and Simulation for Analysis, Design, and Validation of Distributed Energy
Systems, CIGRE C6.04.02 Task Force, Sep. 2006.
[18] O. Nzimako, “Real-time simulation of a microgrid system with distributed energy resources,” M.Sc.
Thesis, Dept. of Electrical & Computer Engineering, University of Manitoba, Canada, 2015.
[19] RTDS Controls library manual, RTDS Technologies Inc., Winnipeg, MB, Canada , pp. 7I.1-7I.12,
2017.
[20] V. Telukunta, J. Pradhan, A. Agrawal, M. Singh and S. G. Srivani, "Protection challenges under
bulk penetration of renewable energy resources in power systems: A review," in CSEE Journal of
Power and Energy Systems, vol. 3, no. 4, pp. 365-379, Dec. 2017.