Received December 25, 2017, accepted January 28, 2018, date of publication February 5, 2018, date of current version

April 18, 2018.

Digital Object Identifier 10.1109/ACCESS.2018.2800279

Communication Modeling of Solar Home System

and Smart Meter in Smart Grids
S. M. SUHAIL HUSSAIN 1 , (Student Member, IEEE), ASHOK TAK2 ,
TAHA SELIM USTUN3 , (Member, IEEE), AND IKBAL ALI1 , (Senior Member, IEEE)
1 Department of Electrical Engineering, Jamia Millia Islamia, New Delhi 110025, India
2 Load Dispatch Center, Electrical T&D, Tata Steel Ltd., Mumbai 400001, India
3 Fukushima Renewable Energy Institute, AIST, Koriyama 963-0298, Japan

Corresponding author: S. M. Suhail Hussain (suhail@ieee.org)

ABSTRACT The future energy networks are envisioned to be green and clean with high penetration of
renewable energy-based generators. The most promising type is solar energy which has immense potential
around the globe. Solar home systems (SHSs) with rooftop solar panels are proliferating in urban cities
as well as in distant rural areas. Possible interaction of SHS with utility grid will result in dynamic power
flow which is a huge challenge for power utility authorities and consumers. The smart meters (SMs) are
being deployed to make this possible through bi-directional energy and information exchange. In order
to address this need, this paper develops the communication models of SHS and SM based on the IEC
61850 standard. These models provide standardized approach to these technologies and facilitate a series of
functionalities, such as power flow control, demand response, and other ancillary services, using configured
message exchange. The detailed models, their use cases and the messages are studied in detail. Furthermore,
extensive simulations are run with riverbed modeler to investigate the dynamic performance. IEC 61850-
based models of SHS and SM are implemented, message frames are developed according to use cases, and the
functionalities mentioned earlier are run as scenarios. Finally, the performances of different communication
technologies have been analyzed to estimate their adequacy for smart grid implementations.

INDEX TERMS Solar home system, smart meter, IEC 61850, communication infrastructure, smart pricing.

I. INTRODUCTION is a step towards Automatic Metering Infrastructure (AMI).

The electric networks are now transforming towards a clean High SM proliferation is an essential step for achieving ade-
power grid as more renewable energy resources are inte- quate communication in smart grids and can be leveraged to
grated. The integration of these distributed energy resources mitigate challenges such as power flow and protection [5].
(DERs) is challenging, considering the centralized nature of Considering the number of SMs and SHSs that may be
conventional grid and lack of reliability due to their inter- deployed in a grid and the possibility of them being from
mittency [1]. These DERs can be Photo Voltaic (PV) sys- different vendors, it is clear that a common language has to
tems, wind generators or Solar Home Systems (SHS) in rural be established for seamless and interoperable operation [6].
microgrids or urban residences. To facilitate standard communication in the power networks,
SHS is a small energy system including local rooftop various approaches have been developed [7], [8]. However,
PV generation with energy storage systems and loads [2]. most of these approaches present challenges of feasibil-
This concept is becoming popular with high deployment of ity, flexibility and interoperability. The most popular stan-
PV panels to residences with purpose of having clean, green dard with Object-Oriented and interoperability design is IEC
and independent energy supply [3]. When connected to the 61850 which is being deployed widely in the power utility
utility grid, SHS may have bi-directional power flow to sup- automation communication networks [9]. For communica-
ply or sink energy to/from the grid. Various novel challenges tion utilization in microgrids, IEC Working Group (WG 17)
are introduced by this behavior and they can be mitigated published extension part IEC 61850-7-420, which is very
with the implementation of effective communication and inclusive for integrating DERs into communication infras-
coordination [4]. tructure of power systems [10]. Different DERs and con-
Smart Meters (SMs) are used to monitor flow of energy trollable loads have been modeled with logical devices and
between a household and the utility grid. Their deployment logical nodes (LN) of IEC 61850 standards in [11]–[14].

2169-3536
2018 IEEE. Translations and content mining are permitted for academic research only.
VOLUME 6, 2018 Personal use is also permitted, but republication/redistribution requires IEEE permission. 16985
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

IEC 61850 communication models have been developed for

multi-level management [15], energy management automa-
tion [12], Volt-VAR optimization [16] and hybrid agent archi-
tecture for automation [17] of microgrids/smart grids. IEC
61850 based PMU communication models for wide area
communication have also been studied in [18], [19]. Recently,
IEC 61850 based communication models for DSTATCOM
has been reported in [20]. Previous work in [21] and [22]
addresses the modeling of Electric Vehicles (EVs) with IEC
61850 and development of system architecture, data set
design and communication for implementation of V2G sys-
FIGURE 1. Block Diagram of SHS.
tem with IEC 61850. As it can be seen, IEC 61850 is emerging
as one of the most promising solutions for communication
standardization in power utility automation domain.
stand-alone as well as grid connected mode depending upon
However, until now, the IEC 61850 standard does not cover
availability of local generation and the electricity price.
all entities that may be present in a smart grid such as SHS
Smart grids utilizes Information and Communication
and SM. Feuerhahn et al. [23] reported that IEC 61850 based
Technologies (ICT) to monitor and control power flows
MMS protocols are most suitable for SM communication
with an objective to make the power grid more resilient,
networks. However, there is little work on SHS and SM
efficient and cost effective. The SHSs in smart grid may
communication modeling according IEC 61850 in literature.
contribute in intelligent load shedding processes, Volt/VAR
Liu et al. [24] and Vyatkin et al. [25] have proposed some
control or demand response which can help in achieving the
preliminary models for SM, yet these models do not take
objectives of smart grid. Distribution System Operator (DSO)
smart pricing into effect and are inadequate for full smart grid
generates the schedules and control signals for all the DER
implementation.
systems in the distribution system. However, small DER sys-
In order to address this knowledge gap, this paper presents
tems (such as SHS) cannot directly exchange the energy with
IEC 61850 based modeling of SHS and SM to facilitate their
the grid and participate in electricity markets. Hence, the fig-
integration to power systems and to ensure interoperability
ure of the aggregator is introduced, which is widely accepted
among different devices. In this fashion, a more holistic
for enabling the participation of DERs in the grid energy
modeling can be achieved and different components that may
exchanges and electricity markets. SHS can participate in grid
be present in a grid can communicate in a standard manner.
energy exchanges by being the part of the aggregations. This
Since this removes the barrier of integrating equipment from
can be in three ways:
different manufacturers, it is a solid step towards plug-and-
play (PnP) in smart grids. Furthermore, the performance of 1. SHS can be part of the virtual power plants (VPP)
the proposed SHS communication model is evaluated for dispatched by DSO.
different communication technologies in terms of End to 2. SHS can participate in demand response methods
End (ETE) delay for different messages exchanged. (curtailing loads)
The rest of the paper is organized as follows: Section II 3. SHS via aggregators can participate in providing ancil-
introduces the SHS operation and related use cases in smart lary services.
grids. Section III details the developed IEC 61850 based
models of SHS and SM. Section IV presents different func- A. SHS AS ENERGY RESOURCE
tionalities using the developed models. Section V shows the Based on the energy consumption pattern and energy genera-
message modeling, simulations and discussions on perfor- tion profile, the SHS owner registers the technical capabilities
mance/validation using Riverbed Modeler tool [26]. Finally, of SHS system, such as power rating and location, with the
Section VI presents the conclusions. aggregator. The aggregator filters and selects the SHS out
of the pool that can serve the demand of the DSO. These
II. SHS OPERATION AND ITS USE CASES clustered SHS are offered as VPPs.
IN SMART GRIDS
SHSs are emerging in urban cities as well as in remote rural B. SHS FOR DEMAND RESPONSE
areas. With the focus on extracting solar energy, installation SHS can participate in demand response either individu-
of rooftop PV panels is trending around the globe [27]–[29]. ally or via an aggregator as per the regulator or utility policies.
A SHS consists of PV panels as generation, batteries as stor- In case the dynamic prices are made available at household
age devices, and local appliances including mobile charging, level; when the energy price is high, SHS may reduce the load
lightning bulbs and telecommunication devices as loads. PV consumption by taking off the non-critical load or shift the
panels are connected to power infrastructure through inverter load to other time periods when the price of energy is low.
which acts as node for SM to grid, residential loads and Sometimes, based on prognosis from grid topology,
storage devices as shown in Fig. 1. The SHS can operate in demand history, forecasted demand and grid capacity,

16986 VOLUME 6, 2018

S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

TABLE 1. Description of SHCT Logical Node.

FIGURE 2. SHS Modeling with IEC 61850.

the DSO identifies the areas where congestion or power short-

age or Volt/VAR fluctuations are likely to occur. DSO will,
then, issue a request for providing the energy to be diverted
to the concerned aggregators. In such cases, SHS may bid
with their aggregators to curtail a certain amount of load. The
aggregator clusters the load reduction bids of SHS and other
DER service bids, if any, and negotiates with DSO. For such
cases, the SHS systems must be ready to curtail their loads on
a short notice period.

C. SHS PARTICIPATING IN ANCILLARY SERVICES

Aggregators coordinate SHS to provide ancillary services
such as frequency support or energy balancing in real-time.
To participate in ancillary services, SHS must register their
power capacity within which they can be dispatched in real-
time to the aggregators. Depending upon the ancillary service
market rules, capacities may be required an hour to a week
ahead of the delivery of services. Whenever there is request
for ancillary services from DSO, aggregators bid to provide
these services. In turn, the aggregator’s issues ancillary ser-
vices profiles to all the SHS systems that has committed to
provide the ancillary services.

III. MODELING SHS AND SM WITH IEC 61850-7-420 FOR

SMART GRID COMMUNICATION INFRASTRUCTURE of SHS such as PV generations, storage batteries and loads.
A. SHS MODELING Fig. 2 shows the overview of all LNs associated with SHS.
Interaction of utility grid and SHS poses challenges for Making use of already developed LNs in IEC 61850-7-420,
efficient and automated interaction. Hence, communication PV System, Solar Home Controller, circuit breaker and mea-
infrastructure for SHS has to be designed for its integration surement devices have been modeled. It is contribution of this
in power grid with dynamic power flows. The consumer manuscript to combine these LNs towards modeling an SHS.
end behavior, demand side management and power supply The novel modeling for representing SHS functionalities as
economics can be enhanced significantly by applying com- discussed above is done by developing a new logical node
munication configuration of SHS. class, SHCT, as shown in Table 1 and discussed below.
IEC 61850-7-420 standard contains the LNs for dif- PV system is modeled with DVPM and DPVC
ferent DER systems communication modeling. However, LNs [12]. The interconnecting inverter/rectifier is modeled
it does not cover system-view and specifications about var- by ZINV/ZRCT LNs respectively. Through the DCCT the
ious DER components and their interaction with grid with SHS registers its technical capabilities to the aggregator.
specified functions in application view [8]. Same standard, The DSCH receives the energy service schedules for SHS
i.e. IEC 61850-7-420, is used for modeling all components from the aggregator. DSCC is used to control or set the

VOLUME 6, 2018 16987

 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

energy or ancillary service schedule in the SHS. The DC loads

are modeled by ZDCL, whereas the controllable loads are
modeled by CNLO [12]. The storage device and its controller
in SHS are represented by LNs ZBAT and ZBCT. SHS is
connected to grid via a switch and circuit breaker, modeled
by CSWI and XCBR LNs. The SHC block in SHS is SHS
controller to be used for controlling various LNs to make
suitable decision by user or utility grid.
SHC is a group of LNs which are used for interaction of
user and utility to control the bi-direction power flow and
economic schemes for energy consumption and generation.
The LNs in SHC are IHMI for SHS user interface and ITCI
for control and communication interface which is connected FIGURE 3. Virtualization of SHS with proposed LNs.

to DRCS and DRCT corresponding to DER controller status

i.e. PV panel status and its controller respectively.
The newly developed SHCT LN includes all the neces-
sary parameters to control SHS’s operation and functional-
ities. The SHS to Grid or Grid to SHS connection time is
monitored by Data Objects (DOs) ‘SH2GStart’, ‘SH2GEnd’,
‘G2SHStart’ and ‘G2SHEnd’ of SHCT. ‘SHReady’ DO
denotes the time interval when SHS is ready to deliver
power into grid. Similarly, ‘GridReady’ is time duration when
grid is ready for injection of power from remote distributed
resources.
‘IAlim’ and ‘IVlim’ are parameters for input current and
voltage limits, respectively, whereas ‘OAlim’ and ‘OVlim’ are
for output current and voltage limits for SH2G operation,
respectively. ‘ConnCount’ tracks the number of times grid FIGURE 4. SM Modeling with IEC 61850.

connected for recording history of stand-alone or grid con-

nected operations. ‘SH2GStatus’ is set to ‘‘True’’ if power
is delivered from SHS to grid and ‘G2SHStatus’ is set to considered to be the back-bone of distribution systems in
‘‘True’’ otherwise. ‘EconStatus’ presents whether SHS is future power grids. With the proliferation of local generation,
connected to grid for economic power transfer during non- e.g. PV panels, SMs need to deal with bi-directional power
peak and peak times. ‘SH2GEnable’ is set to ‘‘True’’ if SHS flow and they need to be controlled by user and utility through
is connected to grid and ‘‘False’’ if it is operating in islanded communication system. In order to facilitate their integration
mode, while ‘SH2GSwitch’ changes its operation from power into power systems, SMs should also have a standard mod-
sourcing (‘‘True’’) to power sinking (‘‘False’’). The switching eling based on IEC 61850. To that end, preliminary models
of different operation can be done by SHS user or, in case of of LN classes for specific functions of SM were developed
demand side management, by the utility grid using commu- in [24]. The detailed communication modeling of SM using
nication configuration. The interaction of SHS with the grid IEC 61850 standard is presented as shown in Fig. 4.
depends on the status of the latter. Hence, SHS interacts with The SM contains several LNs corresponding to different
the grid using bi-lateral agreement. Fig. 3 shows virtualiza- functions. The functions of SMs includes power flow con-
tion of SHS and SH2G technology by IEC 61850-7-420 based trol, measurement data acquisition and its processing, fault
modeling. detection and protection, human–machine interface (HMI)
The measured values section has items of delivered and smart pricing.
(‘DelPower’) and received power (‘RecPower’) as well as The power flow control section is responsible for track-
pricing for energy generated by SHS (‘SHPrice’) and energy ing dynamic power flow using controller interface (ITCI)
purchased from the grid (‘GridPrice’). These measured val- connected to SHS user (IHMI). The SHS can be switched
ues play an important role in power flow control and smart between grid connected to stand-alone mode using switch
pricing for realization of smart grid at house hold level. SMs controller (CSWI), physical switch (XSWI) and tripping
are used for this purpose to perform smart economics and they circuit breaker (XCBR). Measurement and data acquisition
should also be modeled in a standard fashion. section uses MMXU and MMTR LNs to measure voltages,
currents, powers and energy. Other specific LNs include
B. SM MODELLING MSQI for sequence and inter-harmonics measurement and
SMs play a vital role in smart grid operation with multiple MHAI for harmonics and inter-harmonics measurement. The
operations they perform, including energy pricing. These are protection and fault detection is done with help of RDER

16988 VOLUME 6, 2018

S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

TABLE 2. Description OF LGRP logical node. role in power flow control case using section data parameters
i.e. settings, status information of SHS and control methods.
The bi-directional power flow is controlled by parameters in
controls section of SHCT LN.
There are two possible operating modes, i.e. islanded mode
and grid connected mode, and these modes are dependent
upon local generation value. There are three possible cases,
considering local generations: (1) Sufficient generation only
for SHS, (2) Generation scarcity and (3) Over-generation to
be fed to grid.
If SHS has sufficient local generation, it will operate in
islanded mode with ‘SH2GEnable’ = False and status infor-
mation as ‘SH2GStatus’ = False as well as ‘G2SHStatus’ =
False. While in islanded mode, SHS simply acts as isolated
microgrid with no exchange with power grid and local gen-
eration is used for storage and local loads.
However, if SHS needs to import power from the grid,
TABLE 3. Description of LSHP logical node.
the status of grid is checked by ‘GridReady’ and if
‘GridReady’ = 1, then the control parameter G2SHEnable
is toggled to ‘‘True’’. The connection nature is deter-
mined either economic or immediate connection. Hence,
‘G2SHStatus’ = True, ‘SH2GStatus’ = False and timer
starts with connection i.e. ‘G2SHStart’ = 1 for allowed
time for particular connection. Also, ‘ConnCount’ is incre-
mented with each new connection. The power received from
grid is measured using parameter ‘RecPower’. ‘SH2GEnd’
is enabled after allowed time ends. And connection is ended
with ‘SH2GStatus’ = False.
When SHS has more generation, this can be exported to
the grid. Normally, the SHS owner registers the technical
capabilities of the SHS with the aggregator. This is done by
sending the information in DCCT LN of SHS to aggrega-
tor. The aggregator forms a VPP with the clusters of SHS
and other DERs, and presents it for economic dispatch. The
LN, used for disturbance recording and processing. If any aggregator receives the energy dispatch schedule from the
ambiguity detected, then alarm handling controller (CALH) DSO, in turn the aggregator sends the energy dispatch sched-
warns the system to switch operation. For smart pricing func- ules to the all components of VPP. The SHS receives the
tionality between SHS and utility grid, two new LNs LGRP energy schedules from aggregator in the DO ‘SchdTyp’ of LN
and LSHP are developed in this paper, which are shown DSCH. And the schedule is set in SHS by enabling the DO
in Table 2 and Table 3, respectively. ‘ActWSchd’ to ‘‘True’’ in DSCC LN. Next, if ‘SHReady’ =
The newly developed LGRP LSHP LNs contains the DOs True and ‘SH2GStatus’ = False, SHS is connected to grid
for the stored and current prices of utility grid and SHSs in with ‘SH2GEnable’ = True and ‘SH2GSwitch’ = True. The
the ‘StoredPrice’ and ‘CurrentPrice’, respectively. The time ‘SH2GStart’ is set ‘‘True’’ for allowed time.
period within which the current pricing signal must be applied The output current and voltage limits (‘OAlim’, ‘OVlim’)
is denoted by DO ‘WinTms’, while the DO ‘CrntTms’ denotes are checked before supplying power. The output deliv-
the timeout period of current pricing signal. ered power (‘DelPower’) is measured for duration between
‘SH2GStart’ = True and ‘SH2GEnd’ = True. When
IV. PROPOSED FUNCTIONALITIES FOR SHS AND SM ‘SH2GEnd’ = True, the connection is ended using SHC and
Based on SHS, SM models and operation case scenarios power flow control section of SM.
developed in previous sections, the communication mapping Data messages mapping for power flow control is as
is presented in this section. follows:
Consider a SHS with a SM installed for communication
A. SHS as Energy Resource and pricing of bi-directional power flow. The operation of
Bi-directional power flow between SHS and grid can be SHS for different scenarios, explained above, is used for
controlled by the house owner and grid via communication mapping the data messages used in SHCT LN and SM model.
infrastructure. The modelled SHCT LN plays an important Power flow control functionalities uses specific LNs block of

VOLUME 6, 2018 16989

 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

SM model and parameters of status information & control B. DEMAND RESPONSE THROUGH SHS AND SM
sections of SHCT LN. The aggregator or utility, through Utility Control Desk (UCD),
In an SHS, the bi-directional power can flow through SM. notifies the dynamic energy price of the grid to the SHS own-
Incoming power from grid to SHS (Pgrid) and available ers via the SM LN LGRP. The ‘CurrentPrice’ DO contains
power which can be sent to grid (Pshs) can be tracked by LNs the dynamic grid price. When the grid price is at higher side,
and its parameters. If SHS operator wants to check the various the SHS controller can reduce the load by controlling the
power values, it can be controlled using settings as follows direct controllable loads (DCL). DCL (e.g. air conditioners,
thermostats etc.) can be operated at less than rated capacity
Pgrid = MCPU → SHS1.IHMI
during emergency or peak load conditions. The operating
→ SHCT.(OAlim∗ OVlim) mode of DCL can be set by issuing a setting command at
Pgen = MCPU → SHS1.IHMI → SHS1.ITCI DO ‘DCLMode’ in the CNLO LN of controllable loads.
→ SHS1.DRCS => SHS1.DVPM The setting for implementing the demand response is per-
formed as follows:
Pload = MCPU → SHS1.IHMI → SHS1.ITCI
Initially the SHS is set to the demand response mode by
→ SHS1.DRCS → SHS1.CNLO setting ‘SH2GMode’ to 2 as follows,
Pstore = MCPU → SHS1.IHMI → SHS1.ITCI
MCPU → SHS.IHMI → SHCT.(SH2GMode = 2)
→ SHS1.DRCS → SHS1.ZBAT
The current price of grid is taken in SHS1 as follows,
The available power which can be feed to grid is,
MCPU → SHS1.DMSC → SHS1.SM1.LGRP
Pshs = Pgen – PLoad + Pstore → SHS1.SM1.LGRP.CurrentPrice
The SH2G operation includes enabling and disabling the If the current price is more than the threshold, the DCLs
grid connection using control parameters such as: are issued command to reduce their power consumption by
(a) For Enabling SH2G: setting the mode of DCL as follows:
There would be excess power with SHS which is to be fed MCPU → SHS1.CNLO → SHS1.CNLO1.DCLMode
to grid, therefore, Pshs is positive & ‘SHReady’ = ‘‘True’’.
→ SHS1.CNLO1.DCLMode.(0, 1, 2, 3)
If SHCT.SH2GStatus = ‘‘False’’, then
MCPU → SHS.IHMI → SHCT.(SH2GEnable = True) C. ANCILLARY SERVICES THROUGH SHS AND SM
Based on the expected energy production and consumption,
And if SHCT.G2SHStatus = ‘‘True’’, every SHS registers its power capacity within which it can
be dispatched in real-time. When there is a need for ancil-
MCPU → SHS.IHMI → SHCT.(SH2GSwitch = True)
lary services in the grid, the aggregators coordinate SHSs to
And the ‘SH2GMode’ is set to 1 (i.e. energy resource) provide ancillary services in real-time. The aggregators send
the ancillary services profiles such as contingency reserve
MCPU → SHS.IHMI → SHCT.(SH2GMode = 1) ‘‘spinning’’, contingency reserve supplemental, emergency
reserve, energy balancing, reactive power support, emergency
The energy service schedule received from the aggregator
islanding to the SHS controller. These profiles are set in the
is set by the following mapping:
‘SchdTyp’ of LN DSCH of the SHS controller. The commu-
Aggregator → SHS.IHMI → DSCH.(SchdTyp = 1) nication mappings for setting the ancillary services are as
MCPU → SHS.IHMI → DSCC.(ActWSchd = True) follows:
Initially the SHS is set into the ancillary services mode by
(b) For Disabling SH2G : setting ‘SH2GMode’ to 3 as follows,
SHS would be in need of grid power, and accepts the grid MCPU → SHS.IHMI → SHCT.(SH2GMode = 3)
power (Pgrid).
If SHCT.SH2GStatus = ‘‘True’’, then The energy service schedule received from the aggregator
is set by the following mapping:
MCPU → SHS.IHMI → SHCT.(SH2GEnable = False)
Aggregator → SHS.IHMI → DSCH.(SchdTyp = x)
And if SHCT.G2SHStatus = ‘‘False’’, then, MCPU → SHS.IHMI → DSCC.(ActWSchd = True)
MCPU → SHS.IHMI → SHCT.(SH2GEnable = False)
V. SIMULATION OF PROPOSED COMMUNICATION
And, power transfers are measured with MMXU node in INFRASTRUCTURE IN SHS NETWORKS
SM using settings time parameters i.e. ‘SH2GStart’, when SH Using new IEC 61850 based communication models of SHS
is connected to grid; ‘SH2GEnd’, when SH is removed from and SMs, different wired and wireless communication tech-
grid; similar for ‘G2SH’ operation etc. nologies are simulated using Riverbed Modeler. Different

16990 VOLUME 6, 2018

S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

FIGURE 5. Communication Network Architecture of Ad-hoc SHS microgrid.

sizes of ad-hoc SHS microgrids (such as 20 SHS, 50 SHS TABLE 4. Messages exchanged between SHS IEDs.
and 100 SHS based microgrids) are modeled and ETE delay
performance of different messages are investigated for differ-
ent communication technologies such as LAN, WiFi (IEEE
802.11n/g) and WiMAX. Figure 5 shows the communication
network architecture of the ad-hoc SHS microgrid. Each solar
home consists of SHS IED and SM IED. 5 SHSs are grouped
as a cluster. All the SHS and SM IEDs in a cluster are con-
nected to a receiver (access point) in WiFi/WiMAX networks
(or to a common switch in Ethernet based LAN networks).
The cluster access points are connected to base station access
point as shown in Fig. 5. The base station access point further
connects to the UCD via a wide area network, since the SHS
ad-hoc microgrids and UCD can be geographically distantly
located. The messages exchanged between SMs, SHSs and
the UCD are in form of IEC 61850 based GOOSE, Sample
Value (SV) and status update messages.
The measured data is cyclic in nature, i.e. each SM samples
the values of voltage and current at 4000/4800 (for 50/60 Hz)
A. TRAFFIC MODELING BETWEEN DIFFERENT IEDs samples per second. Status update messages can be classi-
The description of different messages communicated among fied as 2 types, i.e. the periodic updates and event based
various SHSs, for use cases discussed in section IV is updates. Each SHS and SM regularly reports its status to the
given in Table 4. As seen, their Application Protocol Data aggregator every 0.1 second in the form of periodic status
Unit (APDU) sizes, destination and source IEDs are also updates. In wake of any event, status update messages are
given. exchanged between SHSs and aggregators which are event

VOLUME 6, 2018 16991

 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

based status update messages. The events may occur at any

time, therefore, the traffic of event based updates are modeled
with suitable stochastic processes. Hence the event based
status update messages are modeled with negative exponen-
tial distribution function given by
F (t) = λe−λt , t>0 (1)
where (1/λ) is the average time interval between the two
consecutive packets. The value of λ is set to 10. Similarly
the GOOSE messages are also event triggered and bursty
in nature, therefore, these messages are also modeled with
Weibull distribution function. The burst data of GOOSE mes-
sages is transmitted for a short period of time. Weibull distri- FIGURE 6. Frame formats of different messages. (a) Frame format of
GOOSE APDU. (b) Frame format of Sample value APDU. (c) Frame format
bution function is suitable to represent the nature of GOOSE of Status update APDU. (d) Frame format of pricing signal APDU.
messages as it is a heavy tailed function normally used to
model fixed rate in ON period and ON/OFF period lengths,
TABLE 5. Parameters of wireless technologies.
when producing self-similar traffic. The Weibull distribution
function is given as
n o
β α
t
−
F (t) = 1 − e , t>0 (2)
For simulation, the values of β and α are consid-
ered as 0.1 and 0.9, respectively. It is assumed that the
UCD sends the pricing signals to the SMs for every
5 seconds.

B. MESSAGE FORMATS AND SIZES

The size of different messages exchanged is calculated by
adding suitable protocol overheads and data payload to
each message. The standard Ethernet frame has a protocol the time required to process the packet header by nodes
overhead of 24 bytes (7 preamble, 1 SFD, 6 destination (i.e. IEDs, routers and switches). The propagation delay is
address, 6 source address, 2 type, 4 CRC), whereas the WiFi time required to transfer/propagate a packet on the physical
(IEEE 802.11) and WiMAX (IEEE 802.16) have protocol medium from one end to the other. The queuing delay is
overheads of 34 bytes and 9 bytes, respectively. The size and the time a packet waits in the buffer queue of a node. The
details of different fields of GOOSE, SV, status update and queuing and processing delays constitute the major portion of
pricing signal APDUs are given in Fig. 6. SV and GOOSE ETE delay.
messages are mapped directly onto Ethernet layer hence do It is assumed that the ad-hoc microgrids are spread over an
not contain TCP/IP layers, while MMS type messages have area of 0.5 km by 0.5 km and the UCD is located at a dis-
TCP, IP headers. tance of 25 kms. Hence, in Riverbed Modeler the simulations
The size of data present in each message is calculated project scenario with SHS microgrids spanning over 0.5 km
by adding the sizes of required data objects of LNs for a by 0.5 km were simulated. The average distance between
particular application. The data (in IEC 61850) is normally SHS/SM IED and access points was about 200-250 m. The
in TLV (tag, length, value) ASN.1 BER type format. The size UCD was connected to the base station access point over
of ‘‘value’’ field for different data objects is either boolean a wide area network with three hops. The simulations are
(1 byte), INT32 (4 bytes) or string (8 bytes). carried out for Ethernet based LAN networks employing
100 Mbps and 1 Gbps links. Table 5 summarizes the different
C. SIMULATION RESULTS AND INFERENCES parameters of the wireless technologies used in the simulation
The proposed SHS based microgrids of different sizes are of SHS based microgrid networks. The simulations were
simulated in Riverbed Modeler to investigate the commu- performed and ETE delays of different messages summarized
nication infrastructure performance. In order to evaluate the in Table 4 for different communication technologies were
effectiveness of the proposed communication, the ETE delay collected.
performance and packet loss under different background For all SHS microgrid networks with varying sizes, ETE
traffic are studied. The ETE delay is the time required to delays for status update and pricing signal messages obtained
send a packet from the source IED to the destination IED. for different communication infrastructures are tabulated
The ETE delay is sum of the processing delays, propaga- in Table 6. The ETE time delays of GOOSE messages, which
tion delays and queuing delays. The processing delay is are time critical, for different communication technologies

16992 VOLUME 6, 2018

S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

TABLE 6. ETE delay of status update and pricing signals for different SHS
networks.

FIGURE 7. Comparison of Sample Values (SVs) delays for different

communication technologies in 100 SHS microgrid networks.

is presented in Table 9. As the GOOSE messages are brust

type of messages and has negligible impact with respect to
packet loss, hence GOOSE messages are omitted for packet
loss analysis. The SV message the has highest packet loss
TABLE 7. ETE delay of GOOSE messages for different SHS networks.
since it has the highest sampling rates. High sampling rates
results in high data traffic which inturn causes congestion
in network which leads to packet drop or packet loss. From
Table 9 it can be observed that SVs over WiFi 802.11G
and 802.11n communication technologies has highest packet
loss per second. It can be observed from the results that the
wireless technologies (WiFi 802.11n, 802.11g and WiMAX)
are more prone to packet loss due to poor SNR and external
interferences. The packet loss analysis for pricing signal and
(LAN, WiFi and WiMAX in different SHS microgrids) is status update messages is of parmount importance. Since the
tabulated in Table 7. The ETE delays for GOOSE mes- pricing signal and staus update messages use TCP in the
sages for different communication technologies and different transport layer, every packet that is lost has to be retransmit-
sizes of SHS are within the limits, i.e. 3 ms, as speci- ted. Hence, the packet loss in these type of messages results in
fied in the standards. The ETE delays for SV messages retranmission of message and thereby considerably incresing
in 100 SHS networks is shown in Fig. 7. The ETE delays the ETE delay.
for SV and GOOSE messages are relatively lower with From the results it is evident that LAN has low ETE
respect to pricing signal and status update messages due delays and packet loss compared to WiFi and WiMAX net-
to the fact that these messages are exchanged between the works. LAN is the best option for low-latency communication
SM and SHS, of a particular smart home system, which among SHSs as well as between SHSs and the utility. When
are on the same local area network or connected through the number of SHS increases, it may be costly as many
a single access point. On the other hand, the pricing sig- switches and links are required. In LAN networks the signal is
nals and status updates are exchanged between aggrega- of good quality i.e. SNR ratio is very low. But, long physical
tor or UCD and SHS/SM which belong to different local area links have risk of breakdown, hence wired networks are less
networks. preferable considering cost for remote locations of SHS based
In a practical scenario, the network may include traffic ad-hoc microgrids.
sources other than the SHS. In order to include this in the When SHS is used in a group or small communities, which
simulation results, additional traffic sources have been added results in few clusters, WiFi communication is advantageous.
and their impact on the overall delay performance has been WiMAX is also good option for SHS clusters spread over
investigated. a large area. But for SHS clusters in a small area WiMAX
Table 8 and 9 show increase in the ETE delays and average networks have relatively higher delays compared to WiFi net-
number of packet loss in the network for different type of works. That’s because, in WiMAX networks, receiving base
messages when background traffic (of 1 Mbps and 10 Mbps) stations have more subscribers since the range of WiMAX
is introduced. This is due to the fact that, if the transmission networks is comparatively large. While in WiFi networks,
of large client/server packet (i.e. background traffic) starts, receivers are available for every cluster, hence the queuing
then the other data packets has to wait in queue until this of data in receiver’s buffer is less. Thus, queuing of data
large packet is transmitted. The packet loss for status update, in WiMAX network causes increased processing time and
pricing signal and SVs with and without background traffic delay.

VOLUME 6, 2018 16993

 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

TABLE 8. ETE delay of different messages for different SHS networks under background traffic.

TABLE 9. Packet loss of different messages for different SHS networks with and without background traffic of 10 Mbps.

VI. CONCLUSIONS These systems can remain independent as well as they

In order to ensure interoperability between devices from dif- can interact with power grid to feed or take power. This
ferent manufacturers, IEC 61850 communication standard is bi-directional flow is being controlled and monitored by SHS
utilized for common information modeling. Different com- user and distant utility. The communication model of SHS
ponents such as distributed generators, fault current limiters is proposed with its functions focused on interaction with
and electric vehicles have been modeled according to the said power grid as SH2G technology. A new LN class SHCT
standard. However, despite their recent popularity, SHS and is defined for SHS as per IEC 61850-7-420. The SM is
SM are yet to be added to this collection. In this respect, this critical part in communication process between power util-
work makes a significant contribution to the body of knowl- ity and consumer end. Detailed modeling of SM with its
edge for future interoperability and modeling of microgrids. functions is done based on IEC 61850 standard. This mod-
With focus on extraction of renewable energy resources, eling will help in deployment of SMs with dynamic power
SHS are becoming popular distributed remote generations. flow. Also, smart pricing and demand side management

16994 VOLUME 6, 2018

S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

concepts are integrated to this model to advance applicability [14] T. S. Ustun, C. Ozansoy, and A. Zayegh, ‘‘Simulation of commu-
of SMs. nication infrastructure of a centralized microgrid protection system
based on IEC 61850-7-420,’’ in Proc. IEEE 3rd Int. Conf. Smart Grid
Integration of SHS and SM into smart grid and ensuring Commun. (SmartGridComm), Tainan, Taiwan, Nov. 2012, pp. 492–497.
a standard interaction is vital for the success of power grid [15] A. Ruiz-Alvarez, A. Colet-Subirachs, F. A.-C. Figuerola,
revolution. This paper makes a significant contribution to the O. Gomis-Bellmunt, and A. Sudria-Andreu, ‘‘Operation of a utility
connected microgrid using an IEC 61850-based multi-level management
body of knowledge by developing separate models for each of system,’’ IEEE Trans. Smart Grid, vol. 3, no. 2, pp. 858–865, Jun. 2012.
these components. These models are specifically developed [16] M. Manbachi et al., ‘‘Real-time co-simulation platform for smart grid volt-
to cater for different operating modes and functionalities that VAR optimization using IEC 61850,’’ IEEE Trans. Ind. Informat., vol. 12,
no. 4, pp. 1392–1402, Aug. 2016.
may be needed in future grids. [17] G. Zhabelova, V. Vyatkin, and V. N. Dubinin, ‘‘Toward industrially usable
In order to test the performance of these models over agent technology for smart grid automation,’’ IEEE Trans. Ind. Electron.,
different communication technologies, extensive simulations vol. 62, no. 4, pp. 2629–2641, Apr. 2015.
[18] I. Ali, M. A. Aftab, and S. M. S. Hussain, ‘‘Performance comparison
have been developed and run in Riverbed Modeler. Based of IEC 61850-90-5 and IEEE C37.118.2 based Wide area PMU com-
on the results, it can be concluded that the performances of munication networks,’’ J. Mod. Power Syst. Clean Energy, vol. 4, no. 3,
wireless technologies (WiMAX & WiFi) and Ethernet satisfy pp. 487–495, Jul. 2016.
[19] S. R. Firouzi, L. Vanfretti, A. Ruiz-Alvarez, H. Hooshyar, and
IEC 61850 requirements for smart grids. The performance of F. Mahmood, ‘‘Interpreting and implementing IEC 61850-90-5 Routed-
developed communication models is satisfactory and meets Sampled Value and Routed-GOOSE protocols for IEEE C37.118.2 com-
IEC requirements, even under different factors effecting the pliant wide-area synchrophasor data transfer,’’ Electr. Power Syst. Res.,
vol. 144, pp. 255–267, Mar. 2017.
network, such as other traffic sources (background traffic) [20] S. M. S. Hussain, M. A. Aftab, and I. Ali, ‘‘IEC 61850 modeling of
and packet loss. These results are important for real-life DSTATCOM and XMPP communication for reactive power management
implementation of developed standard models and achieving in Microgrids,’’ IEEE Syst. J., to be published.
[21] T. S. Ustun, C. R. Ozansoy, and A. Zayegh, ‘‘Implementing vehicle-to-
PnP in power networks. grid (V2G) technology with IEC 61850-7-420,’’ IEEE Trans. Smart Grid,
vol. 4, no. 2, pp. 1180–1187, Jun. 2013.
REFERENCES [22] P. Nsonga, S. M. S. Hussain, I. Ali, and T. S. Ustun, ‘‘Using IEC 61850
and IEEE WAVE standards in ad-hoc networks for electric vehicle
[1] L. Zhang, N. Gari, and L. V. Hmurcik, ‘‘Energy management in a microgrid charging management,’’ in Proc. IEEE Online Conf. Green Commun.,
with distributed energy resources,’’ Energy Convers. Manage., vol. 78, Nov./Dec. 2016, pp. 39–44.
pp. 297–305, Feb. 2014. [23] S. Feuerhahn, M. Zillgith, C. Wittwer, and C. Wietfeld, ‘‘Comparison of
[2] A. Kumar, P. Mohanty, D. Palit, and A. Chaurey, ‘‘Approach for standard- the communication protocols DLMS/COSEM, SML and IEC 61850 for
ization of off-grid electrification projects,’’ Renew. Sustain. Energy Rev., smart metering applications,’’ in Proc. IEEE Int. Conf. Smart Grid
vol. 13, no. 8, pp. 1946–1956, 2009. Commun. (SmartGridComm), Brussels, Belgium, Oct. 2011,
[3] M. P. Nthontho, S. P. Chowdhury, S. Winberg, and S. Chowdhury, pp. 410–415.
‘‘Protection of domestic solar photovoltaic based microgrid,’’ in Proc. [24] N. Liu, J. Chen, H. Luo, and W. Liu, ‘‘A preliminary communication
11th IET Int. Conf. Develop. Power Syst. Protection (DPSP), Apr. 2012, model of smart meter based on IEC 61850,’’ in Proc. Power Energy Eng.
pp. 1–6. Conf. (APPEEC), Wuhan, China, Mar. 2011, pp. 1–4.
[4] W. Li, M. Ferdowsi, M. Stevic, A. Monti, and F. Ponci, ‘‘Cosimulation for [25] V. Vyatkin, G. Zhabelova, C.-W. Yang, D. McComas, and J. Chouinard,
smart grid communications,’’ IEEE Trans. Ind. Informat., vol. 10, no. 4, ‘‘Intelligent IEC 61850/61499 logical nodes for smart metering,’’ in
pp. 2374–2384, Nov. 2014. Proc. IEEE Energy Convers. Congr. Expo. (ECCE), Raleigh, NC, USA,
[5] M. Z. Kamh, R. Iravani, and T. H. M. El-Fouly, ‘‘Realizing a smart Sep. 2012, pp. 1220–1227.
microgrid—Pioneer Canadian experience,’’ in Proc. IEEE Power Energy [26] Riverbed Modeler—(Formerly OPNET). Accessed: Feb. 7, 2018. [Online].
Soc. Gen. Meeting, San Diego, CA, USA, Jul. 2012, pp. 1–8. Available: http://goo.gl/72SgAM
[6] V. C. Gungor et al., ‘‘Smart grid and smart homes: Key players and pilot [27] Off Grid Electric. Accessed: Feb. 7, 2018. [Online]. Available:
projects,’’ IEEE Ind. Electron. Mag., vol. 6, no. 4, pp. 18–34, Dec. 2012. http://offgrid-electric.com/#home
[7] V. C. Gungor et al., ‘‘A survey on smart grid potential applications and [28] Mobisol. Accessed: Feb. 7, 2018. [Online]. Available:
communication requirements,’’ IEEE Trans. Ind. Informat., vol. 9, no. 1, http://www.plugintheworld.com/mobisol/
pp. 28–42, Feb. 2013. [29] M-Kopa Solar. Accessed: Feb. 7, 2018. [Online]. Available: http://www.
[8] V. C. Gungor et al., ‘‘Smart grid technologies: Communication m-kopa.com/
technologies and standards,’’ IEEE Trans. Ind. Informat., vol. 7, no. 4,
pp. 529–539, Nov. 2011.
[9] Communication Networks and Systems for Power Utility Automation,
2nd Ed., document IEC 61850, International Electrotechnical Commis-
sion, 2013.
[10] Communication Networks and Systems for Power Utility Automation Part S. M. SUHAIL HUSSAIN (S’11) received
7–420: Basic Communication Structure Distributed Energy Resources the B.Tech. degree in electrical and electronics
Logical Nodes, Eds. 1.0, document IEC 61850-7-420, International Elec- engineering from Sri Venkateswara University,
trotechnical Commission, 2009.
Tirupati, India, in 2010, and the M.Tech. degree
[11] T. S. Ustun, C. Ozansoy, and A. Zayegh, ‘‘Modeling of a centralized
from Jawaharlal Nehru Technological University,
microgrid protection system and distributed energy resources accord-
ing to IEC 61850-7-420,’’ IEEE Trans. Power Syst., vol. 27, no. 3, Anantapur, India, in 2013. He is currently pur-
pp. 1560–1567, Aug. 2012. suing the Ph.D. degree in electrical engineering
[12] I. Ali and S. Hussain, ‘‘Communication design for energy management with Jamia Millia Islamia, New Delhi, India. His
automation in microgrid,’’ IEEE Trans. Smart Grid, to be published. research interests include microgrid, power system
[13] T. S. Ustun, C. Ozansoy, and A. Zayegh, ‘‘Extending IEC 61850-7-420 communications, and smart grid. He was a recipi-
for distributed generators with fault current limiters,’’ in Proc. IEEE PES ent of the IEEE Standards Education Grant approved by the IEEE Standards
Innov. Smart Grid Technol. Conf. Asia (ISGT Asia), Perth, WA, Australia, Education Committee for implementing project and submitting a student
Nov. 2011, pp. 1–8. application paper in 2014–2015.

VOLUME 6, 2018 16995

 S. M. S. Hussain et al.: Communication Modeling of SHS and SM in Smart Grids

ASHOK TAK received the B.Tech. degree in elec- IKBAL ALI (M’04–SM’11) received the degree
trical engineering from IIT Roorkee, Roorkee, from Aligarh Muslim University, Aligarh,
India, in 2016. He was a Summer Research Intern the M.Tech. degree from IIT Roorkee, Roorkee,
with Carnegie Mellon University in 2015. He is India, and the Ph.D. degree in electrical engi-
currently with the Load Dispatch Center, Electrical neering. He is currently an Associate Professor
T&D, Tata Steel Ltd., India. He is interested in with the Department of Electrical Engineering,
smart microgrids, renewable energy integration, Jamia Millia Islamia, New Delhi. As a Princi-
power system real time digital simulation, IEC ple Investigator, he is executing research projects
61850, and smart grid communication. on substation automation, micro-grid, and IEC
61850-based utility automation funded from DST,
AICTE, JMI, and IEEE Standards Education Society. His research interests
are in IEC 61850-based utility automation, substation communication net-
works architecture, and smart grid.

TAHA SELIM USTUN received the Ph.D. degree

in electrical engineering from Victoria Univer-
sity, Melbourne, VIC, Australia. He was an Assis-
tant Professor of electrical engineering with the
Department of Electrical and Computer Engineer-
ing, Carnegie Mellon University, Pittsburgh, PA,
USA. He is currently a Research Scientist with
the Fukushima Renewable Energy Institute, AIST,
Japan. He has over 40 publications that appeared
in international peer-reviewed journals and confer-
ences. His research interests include power systems protection, communica-
tion in power networks, distributed generation, microgrids, and smartgrids.
He is a reviewer in reputable journals and has taken active roles in organizing
international conferences and chairing sessions. He delivered talks for World
Energy Council, Waterloo Global Science Initiative, European Union Energy
Initiative and the Qatar Foundation. He has also been invited to run short
courses in Africa, India, and China.

16996 VOLUME 6, 2018