IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO.

10, OCTOBER 2020 6217

A Benchmark Test System for

Networked Microgrids
Mahamad Nabab Alam , Member, IEEE, Saikat Chakrabarti , Senior Member, IEEE,
and Xiaodong Liang , Senior Member, IEEE

Abstract—The coordinated operation of multiple micro- supply loads within a clearly defined electrical boundary. MGs
grids (MGs) enables high penetration of locally available can sell excess power to distribution grids in a grid-connected
distributed energy resources. It enhances the reliability and mode or operate autonomously in an islanded mode. In an MG,
resiliency of the power network and reduces the cost of
energy. Although networked MGs have attracted significant nonrenewable DGs include diesel engines (DEs), microturbines
research interests, validation of various studies is difficult (MTs), fuel cells (FCs), and combined heat and power (CHP)
because there is no benchmark test system available for plants, and renewable DGs include solar photovoltaic (PV) pan-
such systems. A benchmark test system can be used to els and wind turbines (WTs) [2]. These DGs may be dispatchable
validate static and dynamic studies related to the network-
or nondispatchable. The power generation from PV panels and
ing of multiple MGs, such as optimal power flow, energy
management, control, stability, and protection. To fill in WTs are usually nondispatchable in nature, whereas the power
this research gap, a benchmark test system for networked output from CHPs, MTs, FCs, and DEs are fully dispatchable. In
MGs is proposed in this article, where four independent the islanded mode of an MG, a dispatchable DG with sufficient
MGs are interconnected and coordinated. Required data, capacity is required to maintain power balance and avoid load
such as line parameters, load data, and power generating shedding [3].
sources, have been prepared for each MG considered in the
system. To provide a general test platform, typical datasets The increased frequency and severity of weather-related
are made as close to practical MGs as possible. Parameters events and catastrophic man-made incidents, such as cyber-
used to evaluate reliability indices and resiliency measures attack, equipment failure, technical errors, can create wide-
of the system are given for the entire test system. Future spread blackouts, cause huge financial losses and discontinuity
potential studies, which can be tested on the proposed of power supply to the consumers [4]. Networked MGs enhance
benchmark test system, are discussed.
resiliency of power system infrastructure [5], provide more
Index Terms—Benchmark, distributed generation, eco- reliable and economic power supply to consumers [6], and
nomic scheduling, load flow data, microgrids (MGs), offer competitive ancillary services [7], and thus, have attracted
networked microgrids.
significant research interest recently. During emergencies, co-
ordinated operation of networked MGs can continue supplying
I. INTRODUCTION power to the critical loads for a longer period of time, and help
restoring the main grid power by providing black-start support
ETWORKING of multiple microgrids (MGs) is emerging
N as one of the promising alternative to improve resiliency
and reliability of power systems. It provides suitable electric in-
to conventional power stations [4], [5].
Since networked MGs have emerged as one important aspect
of modern smart grids, many research and development centers
frastructure to utilize cost-effective and environmental-friendly
all over the globe are involved in conducting miscellaneous stud-
electric power generated from distributed energy resources
ies in this area [8]–[10]. The reported research in the literature
(DERs) [1]. MGs are defined as entities to integrate clusters of
on networked MGs can be categorized as follows:
renewable and non-renewable distributed generations (DGs) to
1) system planning, considering reliability and cost evalua-
tion [11]–[15];
Manuscript received August 28, 2019; accepted February 23, 2020.
Date of publication February 27, 2020; date of current version June 22,
2) voltage and frequency control [16]–[18];
2020. This work was supported by the Department of Science and Tech- 3) power flow and power sharing through advanced control
nology, Ministry of Science and Technology, Government of India, under and communication techniques [19]–[21];
the Project DST/EE/2018174. Paper no. TII-19-3941. (Corresponding
author: Mahamad Nabab Alam.)
4) economic dispatch through optimization, energy manage-
Mahamad Nabab Alam and Saikat Chakrabarti are with the Depart- ment and scheduling [22]–[25];
ment of Electrical Engineering, Indian Institute of Technology, Kanpur 5) service restoration [26], [27];
208016, India (e-mail: itsmnalam@gmail.com; saikatc@iitk.ac.in).
Xiaodong Liang is with the Department of Electrical and Computer En-
6) stability enhancement [28].
gineering College of Engineering, University of Saskatchewan, Saska- Initially, the focus is on analyzing economic benefits, relia-
toon, SK S7N 5A9, Canada (e-mail: xil659@mail.usask.ca). bility improvement, and resiliency against system failures. The
Color versions of one or more of the figures in this article are available
online at http://ieeexplore.ieee.org.
development of a master controller for seamless integration of
Digital Object Identifier 10.1109/TII.2020.2976893 different types of MGs, efficient energy management systems,

1551-3203 © 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6218 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

and adaptive protection scheme further advance the networked II. NEEDS OF A BENCHMARK TEST SYSTEM FOR
MGs research. The ancillary services in a competitive energy NETWORKED MICROGRIDS
market and emergency demand response are the additional
A. Purpose
emerging areas. However, these studies have been reported
on diverse test systems of different sizes, and it is difficult to Networked MG is a new and highly explored research area,
compare the performances of one approach with the other. and can provide the best electrical infrastructure to utilize DERs
All existing test systems considered on networked MGs in and offer ancillary services in competitive energy market. Sev-
the literature can be broadly divided into two categories: 1) eral universities and research laboratories are working toward the
benchmark by dividing an existing test system into multiple MGs development of suitable technologies for seamless integration of
[6], [12], [15], [17], [27] and 2) benchmark by adding MGs to multiple MGs. However, since there is no benchmark test system
an existing test system [11], [20], [22], [23], [26], [28]. In the currently available in this area, it becomes difficult to analyze
first category, the networked topology remains almost the same superiority of one study over the other proposed for the planning
as the original test system, and only the DERs are placed into the and seamless operation of such systems. To fill in this research
system, which cannot represent a real networked MG system. gap, a benchmark test system for networked MGs is proposed
In the second category, the topological and complete data of in this article.
the added MGs are missing, which makes them difficult to be The purpose of the benchmark test system is to provide a
used for validation of different studies other than the considered suitable test platform for performing various simulation studies,
one. Hence, none of these test systems can be considered as a such as power flow, control, stability, protection, and energy
generalized benchmark test system. A few test systems do not management for networked MGs. Additionally, comparative
belong to the above categories [13], [16], [25]. They are small, analysis of algorithms developed for these studies can also be
very simplified, and only suitable for a specific use, which makes performed on this test system. It can provide full dataset to
them hard to be accepted as a benchmark for networked MGs. the researchers to validate their research works. The test cases
None of the existing test systems give reliability and resiliency should be as per the latest IEEE Standard 1547–2018 [35].
quantification data. They are mostly suitable for static simulation
studies, while the dynamic simulation data have not been given
adequately for networked MGs. Therefore, there is an urgent B. Benefit
need of a benchmark test system for networked MGs, so that a In the proposed benchmark test system, four MGs of different
comparative study can be performed to identify and validate the sizes and topologies have been considered. There is a total of
most effective approaches for any applications. five point of common couplings (PCCs) to interconnect among
In the literature, several benchmarks are available for MGs the four MGs. The PCCs can be switched “ON/OFF” as per
[29]. The CIGRE has also developed MG benchmarks for dif- the requirement. The presented benchmark gives hourly load
ferent voltage levels [30]. These benchmarks have been selected flow data for one year. It also gives the data of DERs, energy
by designers as general standard networks to test upon [31]. The storage, curtailable and non-curtailable loads, failure rate, and
studies on MGs, such as analyzing the impact of DERs on power repair time for each component considered. These data can be
flow, voltage profile, system stability, power quality, energy used to test control and stability studies of networked MGs. The
management, overall system reliability, control, and protection, evaluation of reliability and resiliency indices are one of the
can be validated [32]–[34]. However, it is difficult to validate major advantages of this benchmark. Studies related to demand
studies on networked MGs because there is no benchmark test side management and energy storage planning can be tested on
system available for such systems. this benchmark test system.
In this article, a benchmark test system for networked MGs
with four MGs of different sizes is developed. The data for the
entire system have been chosen as per the IEEE Standard 1547- III. NETWORKED MICROGRID SYSTEM
2018 [35] and the guidelines adopted in [36]. This proposed
benchmark test system provides a simulation platform to validate A. Microgrids
several studies including power flow, reliability, and resiliency An MG is a single control entity powered by DGs to serve
analysis for a networked MG system. It is suitable for both static local loads with the ability to interconnect with the main grid.
and dynamic simulation studies of networked MGs. Individual MGs may be of ac, dc, or hybrid ac–dc type depending
The rest of this article is organized as follows: In Section II, on the local power generation and the demand. From ownership
purpose and benefits of the proposed benchmark test system point of view, an MG may belong to utility, community, or private
are provided; in Section III, the topology of the test system party [16]. The operation scheduling of these MGs may be
is described; various system data for the load flow analysis, diverse depending on their own energy planning and economic
dynamic simulation, and evaluating reliability indices are given rules. Load and generation balance must be maintained during
in Sections IV, V, and VI, respectively; the information related to all the operating conditions to keep frequency and voltage of
quantifying resiliency of the system are discussed in Section VII; MGs within the operating limits. In a grid-connected mode, this
the potential test cases using the proposed benchmark test system balance is handled by the grid, which works as the slack bus
are given in Section VIII. Section IX concludes this article. for the MG. In an islanded mode, dispatchable DG with enough

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6219

TABLE I
POSSIBLE CONNECTION STATUS OF PCCs IN NETWORKED MGs

Fig. 1. Block diagram of the benchmark test system for SCPUIC.

capacity must be present to provide the load and generation

balance [3].
In the presented system, all individual MGs are considered
to be of ac type and equipped with synchronous generator (SG)
based dispatchable DG to maintained full load of the MG during
autonomous operation [3]. Voltage level of each MG is the
same, and thus, the tie-line does not require any tie-converter.
Additionally, energy storage system (ESS) is placed in each MG
to provide full load backup for at least two hours.

B. Networked Microgrids Configurations

1) Single Common Point Utility Interconnection: The block
diagram of the proposed benchmark test system considering
single common point utility interconnection (SCPUIC) is shown
in Fig. 1. This networked MG system consists of four MGs,
which can be interconnected using five PCC tie-cables. MG1 is
connected to the utility grid directly. MG2 and MG3 can receive
the power either through MG1 or directly from the utility grid if
MG1 is out of service. MG4 receives power from MG2 and/or
MG3.
The total possible ways of networked MG connections may
be large. Each PCC has one operation status: either “ON” or
“OFF.” Therefore, the total number of possible configurations of
networked MGs depends on the status of PCCs, which can be
defined as follows:
TN = 2NPCC (1)
Fig. 2. Block diagram of the benchmark test system for MCPUIC.
where NPCC is the number of PCCs and TN is the total number
of possible configurations. Not every possible configuration
is feasible during the operation due to restraints, such as the MGs are possible in this system. Table I provides the possible
control, protection, and stability of the network. connections of the system, and connection types of MGs are
Mainly four types of integration among the MGs can be mentioned in the “Remarks” column in the table.
identified, which are as follows: 2) Multiple Common Point Utility Interconnection: The block
1) one or more islanded MGs; diagram of the proposed benchmark test system considering
2) two or more islanded but interconnected MGs; multiple common point of utility interconnection (MCPUIC)
3) radially connected MGs; is shown in Fig. 2. In this configuration, MGs can be intercon-
4) meshed connected MGs. nected to the distribution feeders at different nodes through a
In this benchmark test system, a total of 32 interconnections common point of interconnection. In this type of configuration,
among the four MGs are possible. The utility grid may or may microgrids can be interconnected to the utility grid through two
not be connected to any of the three MGs (MG1, MG2, and different feeders. All possible topologies mentioned in Table I
MG3) and two or more MGs may also get connected at the are also valid for this configuration. Additionally, this configu-
same time, and thus, the connection of utility grid can be of eight ration allows the benchmark to create well-explored topologies,
types. Therefore, a total of 8 × 32 interconnections among the as discussed in [5] and [9], as follows:

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6220 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

conductor are considered in the study. It is to be noted that the

parameters of MGs have been carefully selected following the
IEEE Standard 1547–2018 [35].
The four MGs considered in the benchmark test system are
adopted from the different benchmark test systems of different
sizes. However, their parameters are selected to fit for MGs and
the networking. MG1 is very similar to the 6-bus test system in
[38]. MG2 and MG3 are adopted from the IEEE 33- and 69-bus
test systems in [39] and [40], respectively. MG4 is the distri-
bution section of the IEEE 14-bus test system [41]. Parameters
of these adopted test systems have been redefined by keeping
network topologies close to the existing test systems. Also,
load and generation data are considered as per the microgrid
requirement. While developing the benchmark, guidelines from
the IEEE Test Feeder Working Group of the Distribution System
Analysis Subcommittee [42] have been followed.
The first bus of each MG (101, 201, 301, and 401) serves as the
slack bus of that MG and is equipped with a conventional SG to
provide load balancing and reactive power support for the MG to
operate properly in an islanded mode [3]. In the grid-connected
mode, the slack bus of MG1 (101) serves as the slack bus for
the entire system. Load balance and V/f control of the system is
performed through this bus.
MG1 is a meshed power distribution network consisting of 6
buses and 11 underground cables. Three standby SGs each of
5000 kVA connected to bus 101 are available to supply power
during emergencies. Three PV panels are installed at buses 2, 3,
and 4. This MG can share power with MG2 through PCC1 and
MG3 through PCC2. MG2 is a radial power distribution network
with 9 buses and 8 underground cables. Three standby SGs are
available to power the network through bus 201. Additionally,
three PV panels (at buses 202, 203, and 204) and two WT sets
(at buses 206 and 209) are always available to contribute gener-
Fig. 3. Topologies for MCPUIC configuration. (a) Serial MGs on ated power to the network. MG3 is a radial power distribution
feeder 1. (b) Parallel MGs on feeder 1. (c) Interconnected MGs on network with 18 buses and 17 underground cables. Bus 301 has
feeders 1 and 2. three standby SGs to power the network during emergencies or in
islanded mode. Additionally, six PV panels are installed at buses
303, 304, 305, 306, 307, and 315, and two WTs are installed at
1) serial MGs on a single distribution feeder [see Fig. 3(a)];
buses 314 and 317. MG2 and MG3 can exchange power with
2) parallel MGs on a single distribution feeder [see
all the remaining MGs through the PCCs shown in Fig. 4. In
Fig. 3(b)];
case of MG1 failure, MG2 or MG3 can be connected to the
3) interconnected MGs on multiple distribution feeders [see
utility grid directly for the smooth operation of the remaining
Fig. 3(c)].
system. MG4 is a meshed network with seven bus and eight
Thus, all types of possible topologies can be obtained with this
underground cables, three PV are installed at buses 405, 406, and
MCPUIC. Researchers can use any of the topologies of SCPUIC
407. Two standby SGs connected to bus 401 are available. MG4
and MCPUIC as per their needs.
can share power with MG2 and MG3. These standby SGs are
of equal capacity and have the self-starting capability. Various
C. System Data loads of the entire system can be categorized as normal, priority,
The electrical single-line diagram of the developed bench- or critical loads. The networked MGs are assumed to be under
marked test system with detailed buses, cables, loads, and gen- supervisory control and data acquisition (SCADA) system for
eration is shown in Fig. 4. The five PCCs are clearly marked in monitoring, control, and protection.
the figure. Each bus in the system is numbered using a three-digit The networked MG system is designed to operate on
number. All lines are three-core underground cables in the four 11 kV three-phase underground cables. The utility feeder is
MGs. Interconnecting tie-lines are also similar underground 33 kV three-phase single-circuit overhead lines. There are three
cables. More information about components of the benchmark 20 MVA parallel step-down transformers to power the sys-
test system can be found in Table II. Note: These parameters are tem. Each transformer is equipped with on-load tap changers
fixed as per Havells Catalogue 2016 [37]. Cables made of copper (OLTCs), which maintain the voltage of bus 101 at 11 kV.

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6221

Fig. 4. Single-line diagram of the benchmark networked MG system.

TABLE II TABLE III

DETAILS OF MGs OF THE BENCHMARK TEST SYSTEM LINE DATA OF THE NETWORKED MG SYSTEM

Critical loads include health care centers, online data storage

rooms, SCADA system, security control centers, etc. These
loads must be supplied continuously. Also, change in these loads
are less as compared to the normal loads. Therefore, such loads
can be supplied efficiently through separate transformers.

IV. DATA FOR LOAD FLOW ANALYSIS

Load flow analysis of electric power distribution networks is
a preliminarily and important study [43]. Line and bus data are
required to perform load flow or state-estimation of a system. To
perform a load flow analysis, adequate data must be provided.
The line data such as resistance, reactance, normal current
rating, and cable length of the networked MG system are given
in Table III. Interconnection cable, i.e., tie-cable data are given
in Table IV. Details about the cable considered, such as cross-
sectional area and number of cables used in each line is given
in Table V. More information about the cable can be found in
[37]. The long-time (24 h) and short-time (15 min) emergency
ratings of each cable are 116% and 120% of the normal rating,

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6222 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

TABLE IV TABLE VI
PCC/TIE-CABLES DATA OF THE NETWORKED MG SYSTEM LOAD DATA OF THE NETWORKED MG SYSTEM

TABLE V
CABLE SIZE AND PLACEMENT DETAILS IN THE SYSTEM [37]

respectively. The short-circuit current rating (SCCR) for 1 s is

given in the last column of these tables.
For the MCPUIC configuration, each node of the feeders is
equally spaced. The distance between two nodes, and the first
node and the utility substation is one km. The underground
double cables with 400 sq. mm cross section are used in the
two feeders. It is assumed that the distance between the MGs
and concerned nodes of feeders for interconnection is small.
In this system, a one-year long dataset has been given. The
complete bus data of the networked MG system are provided in
Tables VI–XIII. The total load and critical load of the annual
peak at each bus, and bus load percentages in terms of the total
system load are given in Table VI. The total load of the system
is 30.802 MW and 6.374 MVAR, out of which 6.16 MW and
1.27 MVAR are critical loads. Normal and critical loads are
supplied through separate distribution transformers (DTs). All Base voltage: 11 kV; Specified voltage at all buses: 1.0 p.u.
the loads are under direct control of the SCADA system at the
control center of each MG. Bus type mentioned in the table for TABLE VII
WEEKLY LOAD DATA IN PERCENTAGE OF ANNUAL PEAK LOAD
each bus represents the following information: 1) Bus type 1:
Slack bus; 2) Bus type 2: Generator or DG bus; 3) Bus type 3:
Load bus.
Hourly load data at each bus of the system can be obtained
using Tables VI–IX. The weekly load data in terms of the
percentage of annual peak load is given in Table VII. The daily
load data of any week can be obtained based on weekly peak
load data (see Table VIII). Hourly load data of the system can be
obtained for any day of a year (see Table IX). The annual peak
load of the system occurs in between 7 and 8 P.M. on Thursday
of the 12th week of the year. The loading pattern on each bus
during this peak hour is as given in Table VI.
The details of DERs installed at several locations of MGs are
given in Table X, whereas the details of SGs are discussed in
Table XI. PV panels have been installed in each MG to share the The capacity of inverter/converter installed at various buses are
majority of loads during the day. The excess power is stored in the same as those mentioned in column 3 of Table X.
ESS installed with each PV system to contribute some portion of Hourly power generation from PV panels and WTs for the
power during the night. The standby SGs have enough capacity entire year is shown in Tables XII and XIII, respectively. It
to handle all loads of each MG individually, whenever is needed. is assumed that the PV power generation remains constant for

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6223

TABLE VIII TABLE XI

DAILY LOAD DATA IN PERCENTAGE OF WEEKLY PEAK LOAD STANDBY SYNCHRONOUS GENERATOR DATA IN MGs OF THE SYSTEM

the mentioned hour and the same pattern is repeated for the
TABLE IX mentioned weeks. Similarly, the wind power generation during
HOURLY LOAD DATA IN PERCENTAGE OF DAILY PEAK LOAD the mentioned hours of the weeks is assumed to be fixed. All
four MGs are geographically located near to one another. The
percentage power contributions from PV panels and WTs are in
proportion to their installed capacities in all MGs. The deficit
power at any time is handled by the utility grid directly. In case
of no power supply from the utility grid or as per the energy
management strategy, one or more standby SGs can be started
to contribute power requirement. The system operator of net-
worked MGs is authorized to schedule generation level of each
SG to contribute power so that load is supplied economically at
any particular hour.
The details of ESS installed at various buses are given in
Table XIV. These ESSs are Lithium-ion batteries. With 80%
depth of discharge, the installed battery systems can supply
each MG for more than one hour during the peak load. The
charging and discharging efficiencies of the ESSs including
inverter/converter sections are 0.95 and 0.92, respectively [44].
As stated earlier, normal and critical loads are supplied through
separate DTs. Modeling of loads and generation of buses are
assumed as shown in Fig. 5. All the DTs used in MG1 and
at bus 406 of MG4 are of 500 kVA and 11 kV/415 V rating,
whereas the rest of the DTs are of 250 kVA and 11 kV/415 V
rating. The kVA rating of the converter transformer has been
adopted so that it can supply the maximum power as mentioned
in Table XIV. All the SGs are generating power at 11 kV, and
thus, can be connected directly to the allocated bus in MGs.
TABLE X Similarly, all WTs are assumed to supply power to the system
INSTALLED CAPACITY OF PV AND WT IN MGs OF THE SYSTEM
at 11 kV. The WT generators are asynchronous ac generators,
which generate power at 690 V and stepped up to 11 kV using
a step-up transformer. The PV modules are arranged in a way
to produce power at 415 V dc at all the locations in MGs. The
battery banks are arranged in the same way to charge or discharge
at 415 V. The power converter for PV modules converts dc into ac
at the same voltage, which is then stepped up at 11 kV to integrate
to the concerned PV bus. The data for step-up transformer are
given in Table XV. The per-unit (p.u.) values mentioned in these
tables are on the nameplate rating of the equipment. The kVA
rating of the transformer is the same as the WT rating. Also, the
p.u. parameters of converter transformer and DTs are considered
the same on their nameplate rating as mentioned in Table XV.
The vector connection of all the transformers is considered as
Dyn-11. The DERs have low voltage ride through capability as
mentioned in the latest IEEE standard [35].
Initially, the state-of-charge (SoC) of the batteries are assumed
to be the lowest, i.e., 20% of their full capacity. Charging and

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6224 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

TABLE XII
AVAILABILITY OF SOLAR POWER IN PERCENTAGE OF THE
INSTALLED CAPACITY

TABLE XIII
AVAILABILITY OF WIND POWER IN PERCENTAGE OF THE INSTALLED CAPACITY

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6225

TABLE XIV
ENERGY STORAGE CAPACITY

Fig. 6. Load flow solution to the three considered cases.

TABLE XVI
PARAMETERS OF SYNCHRONOUS GENERATOR

1) Case 1: Fully islanded MGs without utility grid, hours

8–9 A.M., total load 90%, PCCs open, ESSs idle, PV and
Fig. 5. Bus modeling in networked MG system. WTs operating at unity power factor.
2) Case 2: Networked MGs without utility grid, hours 10–11
TABLE XV A.M., total load 95%, PCC 1, 2, and 3 are closed while
DATA FOR TRANSFORMER
others kept open, ESSs idle, PV and WTs are operating
at unity power factor.
3) Case 3: Networked MGs with utility grid, hours 7–8 P.M.,
peak load condition, PCCs closed, ESSs contributing 20%
of its maximum supplying capacity, no PV, WTs operating
at unity power factor.
The load flow solution is obtained using Newton–Raphson
discharging scheduling can be set by the users. One possible
method in the above cases. The magnitudes of bus voltage (|V|)
scheduling can be explained as follows: charging the batteries
and their corresponding angles (θ) are shown in Fig. 6. In Case 1,
is the first priority of PV panel, so once the SoC level of
two standby SGs units of each MG must be operating to maintain
batteries reaches 90%, PV panels can start supplying loads
power balance. In Case 2, PV panels and WTs are contributing
through converter transformers. This scheduling philosophy
significant portion of the loads and only two standby SGs of MG1
allows sufficient capacity available in the ESS to handle an
need to operate to maintain power balance of the entire system.
emergency situation properly. Another charging option of the
Case 3 is very different than previous two cases. This is the hour
batteries is through the main grid, which is used as a precau-
when annual peak load is there. Here, utility grid is maintaining
tionary measure of a coming emergency situation. In this test
power balance, while WTs and ESSs are contributing a little
system, proper scheduling of the ESSs is left for the users to
portion of the loads. All the standby SGs are idle and also there
decide. This flexibility can provide the researchers with the
is no power supply from PV panels.
option for better scheduling of charging/discharging cycle of
ESSs. Also, preparing suitable database to store the data is left
V. DATA FOR DYNAMIC SIMULATION STUDIES
to the researchers so that they can use the information as per their
needs. The required data for dynamic simulation studies, such as
The load flow solution to three cases are included in this transient and subtransient of SG and induction generator of WT
section. These cases are on 4th day and 12th week of the year. data, are given in Tables XVI and XVII. The zero-sequence
The details of these cases are given as follows. impedance for cables can be considered thrice of the actual

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6226 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

TABLE XVII The number of customers connected to each bus in the net-
PARAMETERS OF INDUCTION GENERATOR
worked MGs is given in Table XIX. Customers of normal loads
and priority loads are connected to the networked MGs through
different DTs as shown in Fig. 5. The total number of DTs
connected to each bus for supplying normal and priority loads
are included in Table XIX. The number of customers connected
to each DT at a particular bus can be assumed nearly equal as
per Table XIX.
TABLE XVIII The other reliability data related to terminal substation equip-
FAILURE RATE AND REPAIR DURATION OF CABLES IN MGS ment are adopted as follows [50].

A. Circuit Breaker
1) Failure rate = 0.0066 f/yr.
2) Outage duration = 36 h.

B. Distribution Transformer
1) Failure rate = 0.02 f/yr.
2) Outage duration = 192 h.

C. Bus Section
1) Failure rate = 0.22 f/yr.
2) Outage duration = 10 h.
The reliability data of tie-cables are not included because these
are not the part of a particular MG. The reliability measure of
the proposed benchmark test system is analyzed by assuming
seven outage incidents given in Table XX. The first column of
the table shows the time at which the interruption occurs. It is
indicated as week of the year (W), day of the week (D) and hour
of the day (H). It is assumed that before the mentioned incidents,
impedance [45] and the negative-sequence impedance for the
PCC1, PCC2, and PCC3 were “ON,” while the other two PCCs
cables can be assumed equal to the actual impedance. These
were “OFF.” The actual outage duration of cables B01 and C08
datasets can be used to test the performance of networked
are 72 and 24 h, respectively, as mentioned in Table XVIII.
MGs under topological switching and widely changing load
However, the services in the affected areas were restored quickly
and generation. These details are very useful for doing stability
by networking the healthy MGs to supply the affected areas.
studies, short-circuit analysis, and protection planning.
Loads of buses 309–310 were restored in 10 min and that of buses
202–209 were restored in 30 min. Thus, autonomous operation
VI. DATA FOR RELIABILITY EVALUATION of MGs takes longer time to restore supply as compared to
To evaluate reliability of a power distribution network, relia- networked MGs.
bility indices need to be calculated as discussed in the IEEE stan- By using the information given in Table XX, three reliability
dard 1366–2012 [48]. Commonly used reliability indices include indices, viz., SAIFI, SAIDI, and ENS, are calculated. It is
the system average interruption frequency index (SAIFI), system assumed that the load remains constant throughout the outages as
average interruption duration index (SAIDI), customer average mentioned in this table. For autonomous MG system, the values
interruption duration index (CAIDI), the energy not supplied of SAIFI, SAIDI, and ENS are 0.34, 9.45 h, and 215316 kWh, re-
(ENS), and average energy not supplied [49]. spectively. On the other hand, for networked MG system, SAIDI
To enable the proposed benchmark test system for reliability and ENS reduced to 3.6 h and 71011 kWh, respectively. SAIFI
evaluation, the following data are prepared. Failure rate corre- remains the same in the both the cases. Significant reduction
sponding to permanent outages and repair duration of all the in SAIDI and ENS indicates that networking of multiple MGs
cables used in the four MGs are given in Table XVIII. The improves reliability of the system significantly.
transient outages are rare in distribution systems, where only Recently, some new reliability indices especially defined for
underground cables are used. The outage rate of a given cable islanded MGs [51] can also be used to analyze the reliability
is calculated as follows [50]: improvement with the given information.

P = 0.02 × L (2) VII. DATA FOR RESILIENCY EVALUATION

where L is the length of the cable in kilometer and P is the The resiliency of a power system can be defined as the ability
number of outages/yr of the cable [50]. to recover rapidly from major disruptions caused by extreme

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6227

TABLE XIX
NUMBER OF CUSTOMERS AND DTs IN THE NETWORKED MG SYSTEM

TABLE XX
INTERRUPTION DATA IN A CALENDAR YEAR

Fig. 7. System performance during extreme event in networked MGs

[4].

events through their ability to prepare adequately and respond where the quantification of performance loss against extreme
comprehensively [4]. Thus, a resilient power system will be able event is expressed as follows:
to withstand disruptions pertaining to severe weather (such as  t4  
1 P0 − P (t)
hurricanes, earthquakes, tsunamis). It may include some other loss = dt. (4)
severe conditions, such as climate changes, catastrophic man- t 4 − t 1 t1 P (t)
made incidents (such as malicious attack, network operator mis- From (3) and (4), it is observed that the degree of resiliency
steps), and a combination of such incidents [52], [53]. ranges from 0 to .  indicates that the system is perfectly
Fig. 7 shows the typical performance curve of the system resilient, whereas 0 indicates no resiliency. Perfect resiliency
in the wake of an extreme event with respect to time [4]. The means no performance degradation, while 0 resiliency means an
performance of the system (i.e., system power supply) is P0 at immediate collapse of the system performance after the extreme
time t1 when an extreme event occurs. After a period of time event.
at t2 , the active power supply of the system rapidly declines An extreme event is assumed to occur in the calendar year,
to its lowest level Pmin . Restoration work started at t3 to bring which causes a wide-spread outage in the networked MG system.
back the performance of the system from the lowest level to the This extreme event is assumed to damage the transmission line
normal active power level, which is reached completely at t4 . of the utility grid. The power supply from the utility grid stops
The system resiliency is defined as the reciprocal of loss of because of this damage at 10:20 P.M. on the 2nd day of the 16th
system performance as given in the following equation [4]: week of the year. Following the failure of the main supply, PCCs
1 are opened and MGs islanded from each other. In a short while,
resiliency = (3) standby SGs of MG2, MG3, and MG4 started to handle the
loss

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6228 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

Fig. 8. Timeline of the events occurred in the networked MG system on the 2nd day of the 16th week of the year.

TABLE XXI
POTENTIAL STUDIES SUITABLE ON THE PROPOSED BENCHMARK SYSTEM

Fig. 9. Performance curve of the system during severe condition

mentioned in Fig. 8.

full load of their individual MG. However, the standby SGs of

MG1 failed to start, and its ESSs also got some power quality
issues. As a result, the power supply of MG1 got interrupted.
The remaining three MGs supply their full load independently
with their internal standby SGs and ESSs. The timeline of all
the event and process performed following utility grid failure
are shown in Fig. 8.
After two hours, an attempt was started to restore power
to MG1. The normal load of each MG was curtailed while
the critical load continued to get supply. Initially, PCC3 was
closed and MG2 and MG4 were interconnected. After that,
MG3 was connected to this interconnected MGs (MG2 + MG4)
through PCC4. The V/f control is provided by MG2 in this
interconnection. Finally, PCC1 was closed and supply to MG1
was restored only for the critical loads. After that, ESSs of
MG1 were enabled to share the loads. Now, the networked MG
system works properly but supplying only critical loads. As a
precautionary measure, only critical load was allowed to get
supply and the normal load remained curtailed. After six hours,
the standby SGs of MG1 were started. Now, the V/f control was grid service restoration) of the entire system was achieved
taken over by MG1 from MG2 and attempt was started to restore because of networking multiple MGs. The hypothetical se-
supply to the normal load. First, the normal load of MG4 was quence of events described above may be used to test numerous
connected. After that, the normal load of MG3, MG2, and finally control and restoration strategies to explore resiliency of the
MG1 was restored. Thus, the power supply to the entire system system.
was restored at 6:15 A.M. using internal sources only. Finally, Fig. 9 shows the performance curve of the system during the
the power supply of the utility grid was restored at 2:30 P.M. extreme event described in Fig. 8. It shows the load curve ex-
This information can be used to calculate resiliency of the pected without outage and power supplied during the outage with
networked MGs. An early restoration (much before the main the system restoration considering autonomous and networked

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
ALAM et al.: BENCHMARK TEST SYSTEM FOR NETWORKED MICROGRIDS 6229

operation of MGs. Resiliency measure of the system using (3) [14] S. Mojtahedzadeh, S. N. Ravadanegh, and M. R. Haghifam, “Optimal mul-
and (4) for the autonomous and networked MGs are found to be tiple microgrids based forming of greenfield distribution network under
uncertainty,” IET Renewable Power Gener., vol. 11, no. 7, pp. 1059–1068,
0.207 and 0.756, respectively. Here, resiliency index is improved 2017.
significantly because of networked MGs operation. [15] K. P. Schneider et al., “Enabling resiliency operations across multiple
microgrids with grid friendly appliance controllers,” IEEE Trans. Smart
Grid, vol. 9, no. 5, pp. 4755–4764, Sep. 2018.
VIII. POTENTIAL STUDIES ON NETWORKED MICROGRIDS [16] R. Zamora and A. K. Srivastava, “Multi-layer architecture for voltage and
frequency control in networked microgrids,” IEEE Trans. Smart Grid,
Table XXI gives the detail of some potential studies suitable vol. 9, no. 3, pp. 2076–2085, May 2017.
to test on the proposed benchmark test system. [17] M. S. Golsorkhi, D. J. Hill, and H. R. Karshenas, “Distributed voltage
control and power management of networked microgrids,” IEEE J. Emerg.
Sel. Topics Power Electron., vol. 6, no. 4, pp. 1892–1902, Dec. 2017.
IX. CONCLUSION [18] M. H. Cintuglu and O. A. Mohammed, “Behavior modeling and
auction architecture of networked microgrids for frequency sup-
In this article, a benchmark test system for networked MGs port,” IEEE Trans. Ind. Informat., vol. 13, no. 4, pp. 1772–1782,
Aug. 2017.
was proposed and developed. Four independent MGs of dif- [19] A. Ouammi, H. Dagdougui, and R. Sacile, “Optimal control of power flows
ferent sizes and topological configurations were considered. A and energy local storages in a network of microgrids modeled as a system
complete dataset required to perform load flow analysis under of systems,” IEEE Trans. Control Syst. Techn., vol. 23, no. 1, pp. 128–138,
Jan. 2015.
different topologies, loading level, and power generations was [20] F. Zhang, H. Zhao, and M. Hong, “Operation of networked microgrids in a
given. Data related to quantifying the reliability and resiliency of distribution system,” CSEE J. Power Energy Syst., vol. 1, no. 4, pp. 12–21,
the networked MG system were also included for the benchmark Dec. 2015.
[21] P. Tian, X. Xiao, K. Wang, and R. Ding, “A hierarchical energy manage-
test system. It can be used as a simulation platform to validate ment system based on hierarchical optimization for microgrid community
several studies on a network MG system including load flow, economic operation,” IEEE Trans. Smart Grid, vol. 7, no. 5, pp. 2230–
reliability, and resiliency analysis. 2241, Sep. 2016.
[22] Z. Wang, B. Chen, J. Wang, M. M. Begovic, and C. Chen, “Coordinated
Additional data related to economic dispatch, demand-side energy management of networked microgrids in distribution systems,”
management, protection coordination analysis, control and sta- IEEE Trans. Smart Grid, vol. 6, no. 1, pp. 45–53, Jan. 2015.
bility studies, etc. will be reported in the extended work. [23] X. Hu and T. Liu, “Co-optimisation for distribution networks with multimi-
crogrids based on a two-stage optimisation model with dynamic electricity
pricing,” IET Gener. Transmiss. Distrib., vol. 11, no. 9, pp. 2251–2259,
REFERENCES 2017.
[24] W. Y. Chiu, H. Sun, and H. V. Poor, “A multiobjective approach to
[1] A. R. Malekpour and A. Pahwa, “Stochastic networked microgrid energy multimicrogrid system design,” IEEE Trans. Smart Grid, vol. 6, no. 5,
management with correlated wind generators,” IEEE Trans. Power Syst., pp. 2263–2272, Sep. 2015.
vol. 32, no. 5, pp. 3681–3693, Sep. 2017. [25] A. M. Jadhav and N. R. Patne, “Priority-based energy scheduling in a smart
[2] N. Nikmehr and S. N. Ravadanegh, “Optimal power dispatch of multimi- distributed network with multiple microgrids,” IEEE Trans. Ind. Informat.,
crogrids at future smart distribution grids,” IEEE Trans. Smart Grid, vol. 6, vol. 13, no. 6, pp. 3134–3143, Dec. 2017.
no. 4, pp. 1648–1657, Jul. 2015. [26] Z. Wang and J. Wang, “Service restoration based on AMI and networked
[3] H. Lotfi and A. Khodaei, “AC versus DC microgrid planning,” IEEE Trans. MGs under extreme weather events,” IET Gener., Transmiss. Distrib.,
Smart Grid, vol. 8, no. 1, pp. 296–304, Jan. 2017. vol. 11, no. 2, pp. 401–408, 2017.
[4] Z. Li, M. Shahidehpour, F. Aminifar, A. Alabdulwahab, and Y. Al-Turki, [27] A. Arif and Z. Wang, “Networked microgrids for service restoration in
“Networked microgrids for enhancing the power system resilience,” Proc. resilient distribution systems,” IET Gener., Transmiss. Distrib., vol. 11,
IEEE, vol. 105, no. 7, pp. 1289–1310, Jul. 2017. no. 14, pp. 3612–3619, 2017.
[5] M. N. Alam, S. Chakrabarti, and A. Ghosh, “Networked microgrids: State- [28] Y. Li, P. Zhang, and P. B. Luh, “Formal analysis of networked microgrids
of-the-art and future perspectives,” IEEE Trans. Ind. Informat., vol. 15, dynamics,” IEEE Trans. Power Syst., vol. 33, no. 3, pp. 3418–3427,
no. 3, pp. 1238–1250, Mar. 2019. May 2018.
[6] F. S. Gazijahani and J. Salehi, “Stochastic multi-objective framework for [29] H. Farhangi and G. Joos, “Microgrid benchmarks,” in Microgrid Planning
optimal dynamic planning of interconnected microgrids,” IET Renewable and Design: A Concise Guide, 1st ed. Hoboken, NJ, USA: Wiley, 2019,
Power Gener., vol. 11, no. 14, pp. 1749–1759, 2017. pp. 25–36.
[7] M. M. Esfahani, A. Hariri, and O. A. Mohammed, “A multi-agent based [30] “CIGRE task force C6.04.02: Benchmark systems for network integration
game-theoretic and optimization approach for market operation of multi- of renewable and distributed energy resources,” CIGRE, Paris, France,
microgrid systems,” IEEE Trans. Ind. Informat., vol. 15, no. 1, pp. 280– Apr. 2014.
292, Jan. 2019. [31] P. Kotsampopoulos et al., “A benchmark system for hardware-in-the-loop
[8] M. Shahidehpour, Z. Li, S. Bahramirad, Z. Li, and W. Tian, “Networked testing of distributed energy resources,” IEEE Power Energy Technol. Syst.
microgrids: Exploring the possibilities of the IIT-Bronzeville grid,” IEEE J., vol. 5, no. 3, pp. 94–103, Sep. 2018.
Power Energy Mag., vol. 15, no. 4, pp. 63–71, Jul. 2017. [32] D. E. Olivares, C. A. Cañizares, and M. Kazerani, “A centralized energy
[9] G. Liu et al., “Networked microgrids scoping study,” Oak Ridge Nat. Lab., management system for isolated microgrids,” IEEE Trans. Smart Grid,
Oak Ridge, TN, USA, Tech. Rep. - ORNL/TM-2016/294, Oct. 2016. vol. 5, no. 4, pp. 1864–1875, Jul. 2014.
[10] S. N. Backhaus et al., “Networked microgrids scoping study,” Los [33] B. V. Solanki, A. Raghurajan, K. Bhattacharya, and C. A. Cañizares,
Alamos Nat. Lab., Los Alamos, NM, USA, Tech. Rep. LA-UR-16-27399, “Including smart loads for optimal demand response in integrated energy
Jan. 2016. management systems for isolated microgrids,” IEEE Trans. Smart Grid,
[11] Z. Wang, B. Chen, J. Wang, and J. kim, “Decentralized energy management vol. 8, no. 4, pp. 1739–1748, Jul. 2017.
system for networked microgrids in grid-connected and islanded modes,” [34] P. S. Sauter, B. V. Solanki, C. A. Cañizares, K. Bhattacharya, and
IEEE Trans. Smart Grid, vol. 7, no. 2, pp. 1097–1105, Mar. 2016. S. Hohmann, “Electric thermal storage system impact on northern commu-
[12] W. Yuan, J. Wang, F. Qiu, C. Chen, C. Kang, and B. Zeng, “Robust nities’ microgrids,” IEEE Trans. Smart Grid, vol. 10, no. 1, pp. 852–863,
optimization-based resilient distribution network planning against nat- Jan. 2019.
ural disasters,” IEEE Trans. Smart Grid, vol. 7, no. 6, pp. 2817–2826, [35] IEEE Standard for Interconnection and Interoperability of Distributed
Nov. 2016. Energy Resources With Associated Electric Power Systems Interfaces,
[13] Z. Wang, B. Chen, J. Wang, and C. Chen, “Networked microgrids for self- IEEE Std 1547-2018, Feb. 2018.
healing power systems,” IEEE Trans. Smart Grid, vol. 7, no. 1, pp. 310– [36] C. Grigg et al., “The IEEE reliability test system-1996,” IEEE Trans. Power
319, Jan. 2016. Syst., vol. 14, no. 3, pp. 1010–1020, Aug. 1999.

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.
6230 IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, VOL. 16, NO. 10, OCTOBER 2020

[37] “LT/HT power & control cables,” Havells, Catalogue, 2016. [Online]. Mahamad Nabab Alam (Member, IEEE) re-
Available: https://www.havells.com/content/dam/havells/brouchers/ ceived the Ph.D. degree in electrical engineer-
Industrial%20Cable/Cable%20Catalogue-2016.pdf ing from the Indian Institute of Technology Roor-
[38] M. N. Alam, B. Das, and V. Pant, “An interior-point method-based kee, Roorkee, India, in 2018.
protection coordination scheme for directional overcurrent relays in He is currently a Postdoctoral Fellow with the
meshed networks,” Elect. Power Energy Syst., vol. 81, pp. 153–164, Department of Electrical Engineering, Indian In-
2016. stitute of Technology Kanpur, Kanpur, India. His
[39] M. F. Baran and F. F. Wu, “Network reconfiguration in distribution systems current research interests include active distri-
for loss reduction and load balancing,” IEEE Trans. Power Del., vol. 4, bution system protection, economic operation of
no. 2, pp. 1401–1407, Apr. 1989. power systems, and networked microgrids.
[40] J. S. Savier and D. Das, “Impact of network reconfiguration on loss
allocation of radial distribution systems,” IEEE Trans. Power Del., vol. 22,
Saikat Chakrabarti (Senior Member, IEEE) re-
no. 4, pp. 2473–2480, Oct. 2007.
ceived the Ph.D. degree in electrical engineer-
[41] M. Ojaghi and V. Mohammadi, “Use of clustering to reduce the number of
ing from the Memorial University of Newfound-
different setting groups for adaptive coordination of overcurrent relays,” land, St. Johns, NF, Canada, in 2006.
IEEE Trans. Power Del., vol. 33, no. 3, pp. 1204–1212, Jun. 2018.
He is currently a Professor with the Depart-
[42] K. P. Schneider et al., “Analytic considerations and design basis for the
ment of Electrical Engineering, Indian Institute
IEEE distribution test feeders,” IEEE Trans. Power Syst., vol. 33, no. 3,
of Technology Kanpur, Kanpur, India. His re-
pp. 3181–3188, May 2018.
search interests include power system state es-
[43] J. J. Grainger and W. D. Stevenson, Power System Analysis. New York,
timation, dynamics and stability, smart grid, and
NY, USA: McGraw-Hill, 1994.
microgrid.
[44] B. Zakeri and S. Syri, “Electrical energy storage systems: A comparative
life cycle cost analysis,” Renewable Sustain. Energy Rev., vol. 42, pp. 569–
596, 2015. Xiaodong Liang (Senior Member, IEEE) was
[45] N. D. Tleis, Power Systems Modelling and Fault Analysis Theory and born in Lingyuan, Liaoning, China. She re-
Practice. Amsterdam, The Netherlands: Elsevier, 2008. ceived the B.Eng. and M.Eng. degrees from
[46] P. Kundur, Power System Stability and Control, 1st ed. New York, NY, Shenyang Polytechnic University, Shenyang,
USA: McGraw-Hill, 1994. China, in 1992 and 1995, respectively, the M.Sc.
[47] G. Abad, J. Lopez, M. A. Rodrıguez, L. Marroyo, and G. Iwanski, degree from the University of Saskatchewan,
Doubly Fed Induction Machine Modeling and Control for Wind Energy Saskatoon, SK, Canada, in 2004, and the Ph.D.
Generation. Hoboken, NJ, USA: Wiley, 2011. degree from the University of Alberta, Edmon-
[48] IEEE Guide for Electric Power Distribution Reliability Indices, IEEE Std. ton, AB, Canada, in 2013, all in electrical
1366–2012 (Revision of IEEE Std. 1366–2003), May 2012. engineering.
[49] R. Billinton and R. N. Allan, Reliability Evaluation of Power Systems. From 1995 to 1999, she served as a Lecturer
New York, NY, USA: Springer, 1996. with Northeastern University, Shenyang, China. In October 2001, she
[50] R. Billinton et al., “A reliability test system for educational purposes-basic joined Schlumberger in Edmonton, and was promoted to be a Principal
data,” IEEE Trans. Power Syst., vol. 4, no. 3, pp. 1238–1244, Aug. 1989. Power Systems Engineer in 2009 by this world’s leading oil field service
[51] S. Wang, Z. Li, L. Wu, M. Shahidehpour, and Z. Li, “New metrics for as- company. After serving Schlumberger for almost 12 years, from 2013
sessing the reliability and economics of microgrids in distribution system,” to 2019, she was with Washington State University in Vancouver, United
IEEE Trans. Power Syst., vol. 28, no. 3, pp. 2852–2861, Aug. 2013. States, and Memorial University of Newfoundland in St. John’s, Canada,
[52] A. Gholami, T. Shekari, F. Aminifar, and M. Shahidehpour, “Microgrid as an Assistant and later Associate Professor. In July 2019, she joined
scheduling with uncertainty: The quest for resilience,” IEEE Trans. Smart the Department of Electrical and Computer Engineering, College of
Grid, vol. 7, no. 6, pp. 2849–2858, Nov. 2016. Engineering, University of Saskatchewan, where she is currently an
[53] T. Ding, Y. Lin, Z. Bie, and C. Chen, “A resilient microgrid formation Associate Professor. Her research interests include power systems,
strategy for load restoration considering master-slave distributed genera- renewable energy, and electric machines.
tors and topology reconfiguration,” Appl. Energy, vol. 199, pp. 205–216, Dr. Liang is a registered Professional Engineer in the province of
2017. Saskatchewan, Canada.

Authorized licensed use limited to: UNIVERSITY PUTRA MALAYSIA. Downloaded on March 15,2022 at 08:28:28 UTC from IEEE Xplore. Restrictions apply.