Ezhilarasi P1,*, J.

Electrical Systems 18-4 (2022): 582-599

Ramesh L2,
Jens Bo Holm- Regular paper
Nielsen3
Modellingand analysis of cost
effective smart meter with
decentralized CIS framework
towards optimal network traffic

The smart meter has become necessary technology due to the high demand for energy-relevant
real-time data to monitor and control the smart grids. According to UN-SDG 7. b, smart meter
plays a vital role in infrastructure expansion and technological upgrades to ensure affordable
and sustainable energy services in developing and underdeveloped countries. Although the smart
metering approach yields many benefits, it involves complete infrastructural changes on utility
providers and consumer premises. Meanwhile, the research society widely concentrates on
replacing existing metering infrastructure with the new smart meter. But this changeover results
in uneconomical large-scale smart meter deployments, including communications infrastructures
and maintenance costs. Unfortunately, very few researches were conducted on utilizing the
existing metering infrastructure to achieve the benefits of smart metering. And also, the viability
of such an implementable model is less common. Aside from this, as advanced metering
infrastructure emerges, smart grid communication is experiencing an increase in data traffic
due to the collection of huge volumes of data. Consequently, this paper proposes a Network
Enabled Smart Energy Meter that can network existing multifunctional digital energy meters
without replacing them for smart meters, while reducing network traffic through a decentralized
framework. The proposed design is simulated and tested in Proteus software to validate basic
and decentralized data processing operations. Obtained results show that without changing
metering infrastructure at consumer premises, smart metering can be incorporated using
Network Enabled Smart Energy Meter with lower deployment cost for a group of houses
compared with existing metering technologies. Moreover, decentralized data processing
frameworks are used to reduce data traffic in smart grid communication networks, and their
results are discussed in the comparative data volume analysis section. In addition, we conducted
a residential consumer preference survey to validate the acceptance rate of the proposed
technology for low-cost smart meter.

Keywords:Smart metering; smart meter; decentralized processing; smart grid; big data; data traffic.

1. Introduction

Handling future energy demand with a minimum carbon emission approach is a critical
issue for energy sectors. Moreover, the conventional electrical grid has been degraded over
time and cannot meet the ongoing power sector transition [1],[2]. In addition, the lack of
real-time supervision and control makes the electrical grid unbalanced in terms of supply
and demand. In order to handle the dynamic grid’s nature with the minimum supply and
demand gap, the electrical grid needs to be modernized into a smart grid [3]. In that
process, mart Meter (SM) plays an important role since it exchanges real-time data between
consumer and utility providers[4]. Though several terminologies like smart energy meter,
remote meter, and intelligent meter are used among the research community, the SM is
widely accepted. In the course of history, the measurement of electricity consumption had
*
Corresponding author: Ezhilarasi P, Dr.M.G.R. Educational and Research Institute, Department of Electronics
& Communication Engineering, Chennai, India, 600095, E-mail: [email protected]
1
Dr.M.G.REducationa and Research Institute, Department of Electronics and Communication Engineering,
Chennai- 600095, India
2
Dr.M.G.R. Educational and Research Institute, Department of Electrical and Electronics Engineering, Chennai-
600095, India
3
Aalborg University, The Faculty of Engineering and Science, Aalborg, Denmark

Copyright © JES 2022 on-line : journal/esrgroups.org/jes

 J. Electrical Systems 18-4 (2022): 582-599

undergone numerous changes along with technological advances. Because of this, the
energy meter used for calibration has continuously evolved with added features and
functionalities over time. In conventional meter a reading person has to visit consumer
premises to take the reading every month once, which leads to errors in reading, bribery of
the reading persons, and delays in generating bills [5]. The above-said issues are removed
by SM where the human workforce is not needed for meter reading and bill generation.
Compared with the conventional meter, a SM can record additional fine-grained real-time
information and communicate it to utility providers through communication infrastructure.
Consequently, the SM offers a diverse range of intellectual benefits to consumers, utility
providers and society which are illustrated in the Figure 1.

Figure. 1SM benefits to consumers, utility provider and society

Furthermore, these benefits can only be achieved if smart meters can read the following
parameters in real-time current, voltage, power, reactive power, power factor, and energy
consumed [6],[7]. So, the transition of energy meters from conventional to smart will make
the electric grid reliable in terms of electricity generation, transmission and distribution.
Furthermore, other smart sensor nodes collect grid information and send it to the utility's
Head End System (HES) in order to fully automate and modernize the grid. Data collected
from various sensor nodes can be processed by smart grids after collected in HES using
different management systems such as Geographic Information Systems (GIS), Meter Data
Management Systems (MDMS), Outage Management Systems (OMS), Customer
Information Systems (CIS), and Demand Response Management Systems (DRMS) to
achieve a wide range of applications [8]. As regards the above management systems, CSI
focuses more on consumer benefits than utility benefits. With centralized data processing,
all collected data is fed into the central server to enable the different management systems,
causing data traffic in the communication network as a result of continuous data pooling.
Furthermore, the CIS is heavily reliant on consumer data that is collected by the SM at the
consumer's house. Recent days the SM has been receiving exceptional research attention
due to its role in electrical grid modernisation. The rollout of smart meters is estimated to
reach 729.1 million in 2019 and a decadal growth rate of 3013% [9].
Although, the cost for deploying smart meters at a massive level will play a critical role
in determining its benefits since, the cost is directly proportional to SM functionalities [10].
Potential investment and benefits of smart meters can be quite different in various countries
and circumstances. As a result, developing and underdeveloped countries face difficulties in
the massive rollout of smart meters due to the lack of consumer acceptance and different
economic strata within the countries. Recent studies and research on smart meters have

583
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

focused on installing a new SM in every consumer premises to enjoy the benefits of smart
metering. However, this will put financial pressure on weak economic countries as well as
low-income consumers results in dumb node (consumer without smart meter) in the smart
grid eco system. Although the design and development of smart meters have been studied
and investigated intensely, little attention is paid to utilising the existing metering
infrastructure for smart metering instead of a complete revamping metering infrastructure.
This paper discusses the importance of utilising existing metering infrastructure without
replacing it for smart metering to focus on the lower cost deployment strategy. In addition,
some of the features of CIS (such as energy consumption information and bill details) are
implemented in a decentralized framework by programming Network Enabled Smart
Energy Meter (NESEM) to report consumer information based on consumer queries. A
consumer opinion survey is also conducted among the residential consumers to validate the
NESEM design for lower deployment cost strategy.

2. Related works

Literature researches on smart meters cover a wide variety of topics like meter data
analytics, communication analysis, demand or load management, smart grid management
with SM data, load estimation or forecasting and peak load estimation, state estimation, SM
privacy and security, SM tariff, smart grid management with SM data, etc. Noelia Uribe-
Pérez et al. [11] briefly discussed the trends and status of global SM deployment. Yasin
Kabalci conducted a detailed survey on smart metering methods along with communication
standards [1]. Comparison of SM communication standards are discussed by Sabine
Erlinghagen et al. for European countries [12]. Challenges and issues of SM are detailed in
[13] by Soma ShekaraSreenadh Reddy Depuru. Lulu Wen et al. discussed the challenge and
benefits of big data collected from smart meters and smart grids [14],[15]. Domenico
Capriglione et al. discusses the need for decentralised metering when compared with the
centralised approach [16]. Ezhilarasi et al. [17] detailed the major works done by various
authors on the SM on different topics. Unfortunately, less focus is given to the cost-
effective design and architecture of smart meters. SM design normally includes one
microcontroller for sensing and calculating the energy details, which is communicated to
the desired data center through different communication modules. Details some of the
literature related to the SM in the aspects of design and architecture done by various authors
are reported in Table 1.

Table 1: SM Literatures in the aspects of design architecture

Data routing destination Software
Author Microcontroller Communication
used
Shang-Wen Luan et
al.[18] dsPIC30F ZigBee Rear-end processing system MULTISIM

Giuseppe Del Prete et

STM32F107VCT6 CAN bus Data aggregator ----
al.[19]
H. M. Zahid Iqbal et Utility Administrator and
dpPIC30F3014 GSM
al.[20] consumer (on demand)
----
Himshekhar Das et
----- GSM Utility server
al.[21]
----
L. C. Saikia et al.[22] Arduino GPRS Utility web hosting, ----

584
 J. Electrical Systems 18-4 (2022): 582-599

customers mobile phone

Nithin B et al. [23] Arduino Virtual terminal Server proteus
base station
V.Preethi et al.[24] ARM7 Zigbee ----
regularly
National Instruments
Internet or electrical
R. Morello et al [25] Single-Board RIO Control station ----
network
9626
Fadhela K. Handhal
Arduino RF Transreceiver Remote station
et al.[26]
S. Elakshumi et
Arduino PLC Main server MATLAB
al.[27]
Shashank Singh et
Arduino GSM/Transceiver server MATLAB
al[28]
Teddy Surya
Arduino Ethernet Server MATLAB
Gunawan et al. [29]
SaikatSaha et al.[30] Arduino Wi-Fi Cloud PROTEUS
Abdul Ahad, et
Arduino GSM Central server ----
al.[31]
QasimMalik et al.[32] Arduino GSM User mobile phone PROTEUS
Farah Sabir et al.[33] IC ADE-7753 Wi-Fi module Web server ----
Zohaib Sultan et
PIC18F452 GSM Connectivity server ----
al.[34]
Himanshu K. Patel et
Arduino GSM System server ----
al. [35]
NagibMahfuz et
Arduino Nano GSM Cloud server ----
al.[36]
G.I. Rashed et al. [37] PIC18F452 GSM Server PROTEUS
M.Leelavathi et PIC16F877
GSM Server ----
al.[38] microcontroller

All the research mentioned earlier and the global scenario insists on a separate SM for
every house premises to develop AMI. But these concepts are economically infeasible for
the developing and underdeveloped countries with weak economic status. Almost all
literature focuses on designing and developing a new SM for smart metering in households.
And also, the designed SM always communicates the data either directly to the utility
server or to the local DCU, which in turn route the data to the server. In the AMI network,
houses without smart meters are considered dump nodes or blind spots. From the cost
analysis report [39] of ISGF in the Indian context, all meter nodes need to be replaced with
smart meters for the full benefits of AMI. According to the current scenario, the
deployment cost for replacing all old digital meters to smart meters will increase the budget
of SM deployment. In developed countries, people can purchase smart meters regardless of
the cost. But situations in India are restricted to the SM cost. In addition, the big data (103
of tera byte) generated by advanced metering infrastructures is one of the major concerns of
smart grid technology [40]. The realization of a fully functional smart grid requires the
interaction of a large number of smart nodes that gather real-time information and send it to
a central server for further processing. The smart grid will collect a greater volume of data
as the number of smart devices such as smart meters, AMIs, RTUs (Remote Terminal
Units), Phase Measurement Units, WAMSs (Wide Area Monitoring Systems), SCADAs
(Supervisory Control And Data Acquisition), and IEDs (Intelligent Electronic Devices)
increases substantially in the near future. Furthermore, a variety of semi-structured, quasi-
structured, unstructured, and structured data are collected to enhance grid analytics. As
shown in the following Table 2, the above- mentioned data varieties are described in more
details.

585
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

Table 2: Characteristics of big data from smart grid

Variety of the data Formats
Structured Comma-Separated Values (CSV)
Relational Database Management System (RDBMS)
Online Analytical Processing (OLAP)
Relational tables containing customer information
Electrical consumption data in numbers and strings
Semi-structured XML and JSON data files
Web service data
Load monitoring
Power quality data
Quasi - structured Web clickstream (contains erratic formats) and
Data values
Web scrapping data
Search engine results
Un- structured Census and text
Social media streams
Tweets
Audio
Video
Photographs

According to [14], if the data volume increases in the network, the data congestion
encountered by the AMI network is high. This data congestion results in delay and loss of
granular information. In addition, [16] discuss the importance of the decentralized metering
approach in smart grid for smart metering since smart grid invokes a massive transition of
whole grid infrastructure at different level like design, communication and management
software with huge investment cost. In a centralized framework, processing structured data,
along with other structured data varieties, is tedious due to pre-processing (data integration
& data cleaning) [40]. This problem can be mitigated with the decentralized data processing
(structured data format only) CIS framework incorporated in NESEM. This decentralized
approach reduces the amount of storage needed on the central server to store consumer data
and the time required to respond to consumer queries in real time. LTE's widespread
presence makes it a popular choice for smart grid communication, reducing utility
providers' financial burden. Because LTE is used for other services such as
communications such as video streaming, web browser, voice, chat and e-mail services
[41], so a limited amount of bandwidth is allocated to SM communications. Consequently,
inefficient use of limited bandwidth will increase communication network data traffic. This
can be mitigated with a decentralized data processing framework enabled in NESEM. To
validate the traffic reduction in the LTE communication network due to the decentralized
framework, this paper analyses only billing information, customer energy consumption
information, and outage/blackouts information from CIS.
To deal with the above-said problems (higher deployment cost and network data traffic),
we proposed a NESEM, a common meter for a group of houses without replacing old
metering infrastructure with reduced overall meter deployment cost. Data congestion of the
existing smart metering concept is mitigated in this proposed metering infrastructure with
the help of decentralized CIS processing. This work is one of the objectives of IGEN.
PAASHAM Alliance Research Group which is affiliated under The Institution of Green
Engineers (IGEN) to mitigate the research gap and enhances further research activities. The
Objective of PAASHAM (High Performance Affordable Smart Meter) is aimed to design
and develop next generation smart meter, which will help to achieve the targets of UN

586
 J. Electrical Systems 18-4 (2022): 582-599

Sustainable Development Goal 7. In line to justify the need of the research, residential
consumer opinion survey conducted and its details discussed in the next section.

3. Consumer opinion survey

Current projects and studies on the SM are in the piloting stage in India. There is no
active implementation of metering infrastructure model to the consumers which would play
a key role in better acceptance of the smart metering concept [21]. To examine the
acceptance of this proposed metering concept among residential consumers opinion survey
is conducted on existing smart metering technology and NESEM at Tamil Nadu, a southern
state of India. The questioners were set in simple terms that explain the overall concept of
smart metering and proposed NESEM to reach commoners. A pilot survey was conducted
among 10 persons to analyse the participant's understanding level on survey questioners.
According to results from the pilot survey, questioners were modified, and the final survey
was conducted among 3000 participants through online mode (google forms). According to
Indian household electrification status, Tamil Nadu has a total of 1,02,83,678 houses with
electricity [42]. Out of this count, 75% need smart metering (consumer with digital meter-
77,12,758). So, for this survey, the population size is considered 77,12,758 with a
confidence level of 95% and margin error of 1.9% results in a sample size of n=2660. Since
the sample size is more than the received response, the survey outcomes' results are
considered appropriate to represent the total population size with a confidence level of
95%. The outcomes of important questioners are discussed and illustrated in the following
section. From the outcomes illustrated in Figure 2. residential building is equally shared
between the single and double bedrooms compared with the bungalow. This question is
included to formulate the load model for proteus simulation platform. In addition, with that
91.8% of participants are interested in reducing electricity bills by monitoring their
consumption, which shows the importance of SM deployment

Do you want to know, how many units of energy utilized

What type of house do you live? in your home regularly to reduce your monthly
electricity bill?

3.9 % 8.2 %
Bungalow

Double bedroom Yes

Single bedroom No
47.7 % 48.4 %

91.8 %

Figure. 2 Responses set I

From Figure 3., among the population size, 79.5% have the digital meter in their
households, which supports the idea of converting old meters to smart meters without
replacing old ones. The above results indicate the viability of using existing metering
infrastructure for smart metering. And also 70.3 % of participants are interested in the
national SM rollout program which shows the active participation of consumers in SM
rollout.

587
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

National Electricity Board plans to change all installed electricity

meter to smart meter with an estimated cost of 4000 to 5000 Rs.
Which type of energy meter installed in your home? for single-phase & 8000 to 9000 Rs. for three-phase. Do you
agree with it?

20.5 %
Digital Meter 29.7 % Yes

Analog Meter No

79.5 % 70.3 %

Figure3: Responses setII

From Figure 4., on explaining our proposed idea without motioning the cost compared with
existing smart metering technology 31.6% of respondents are interested in NESEM. But
with cost for proposed idea reduction, 73.4% respondents are agreed to our proposed idea
for smart metering when compared with the counterpart and is reported in right side of
Figure 4.
"We propose a new centralised meter called NESEM(Network Enable Smart
Energy Meter) which can act as common smart meter for group of houses.
With NESEM, Electricity Board can individually read every houses
electricity bill without replacing old meter ". After this explanation which
With the following two option, you can get all smart
one do you prefer for smart metering? metering functions, which one do you prefer for your
house?

By connecting an add-
on device to the old
26.6 % digital meter with an
30 % 31.6 % estimated cost range
No idea from 1000 Rs. to
1500Rs.
Proposed NESEM for
group of houses
73.4 %
without replacing old
34.3 % digital meter By replacing the old
meter to a new
Individual new
meter with an
smart meter in
each houses
estimated cost
with replacing range from4000 Rs.
old meter - 8000 Rs.

Figure 4: Responses related to the proposed NESEM design acceptance

From this residential consumer opinion survey, it is concluded that consumers are willing to
accept the concept of smart metering over conventional metering. In spite of this, the cost
of upgrading infrastructure to a SM makes them accept NESEM for smart metering because
it is low cost.

4. NESEM

4.1. Design Methodology

Global scenario and literature research heavily focus on developing various smart-meter
designs, and all are for individual households to do the advanced metering. In addition, all
works on smart meters focused on centralized processing to establish smartness in the
electrical grid. Consequently, a separate SM for each household results in huge deployment
costs. With the existing approach, all the data collected from each SM is collected by Data
Concentrator Unit (DCU) and send the utility server through wide area network
communication technology. A centralized approach to processing all collected data makes
the grid unestablished because of the volume of information processed in real-time. The
comparative architectural network layout of the proposed NESEM and Existing SM
illustrated in Figure 5. The design and development of the proposed NESEM consists of
three parts 1. Designing an add-on device to extract the real-time data from the existing

588
 J. Electrical Systems 18-4 (2022): 582-599

digital energy meter.2. designing the central NESEM to process the extracted data for the
advanced metering process. 3. Formulating decentralized smart metering and CIS
algorithms.This proposed work considers a digital meter with serial communication since,
according to the survey, it is estimated that 72.9% of households have digital energy
meters. To mitigate the problem caused by dumb nodes in AMI, all households are
equipped with one add-on device along with the digital meter. Adding a communicable
add-on device replaced the original dumb node with a communicable digital meter.

GIS
DM – Digit al Met er
AD – Add on Dev ice OMS
CLUSTER 1
Home 1 DRMS
Home 2 Ut i l i t y
MDMS
Home 2 Home 1 h es
CLUSTER n

Ut i l i t y
(HES) Smar t
CIS met er i n g f or
DCU NESEM up t o N
GIS
ad h ou ses
DM
OMS
CIS
Home 3 DRMS
CLUSTER 2
MDMS
Home 3
Smar t
Up t o
Home 4 n h ome Home N met er i n g f or
10 h ou ses

Up t o 10 home
Home 4 Home 10

Existing smart metering Proposed NESEM metering

Figure 5: Comparative architectural network layout of the proposed NESEM and Existing
smart metering

In accordance with the simulation and microcontroller specification, a NESEM can be used
to do smart metering for up to ten households. Add-on device modelling involves two
process flows, the first extracting real-time data, and the second communicating it to the
central NESEM. On the other hand, NESEM is designed to collect and process data routed
by various add-on devices to enable advanced metering. In Figure 6, the schematic block
diagrams of the add-on device and NESEM are shown.
ADD-ON DEVICE NESEM

GSM
Microcontroller Microcontroller
module
Router

DIGITAL
ZigBee
ENERGY ZigBee
Relay unit coordinator
METER
Local storage
RS485 Display
(Data logger)

Ut i l i t y
su ppl y

Figure 6: Block diagrams of add-on and NESEM

Clusters of consumers are created based on their distance from NESEM, each covering ten
houses maximum. The NESEM network is based on a star topology, in which every add-on
device from a cluster is directly connected to NESEM. The add-on device uses the RS485
serial communication port of the digital energy meter for data extraction with the help of a
microcontroller. Microcontroller is programmed to control and manage the data extraction

589
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

and communication process in real-time. Data extracted in real-time are transmitted to

NESEM via a wireless module (ZigBee transceiver). Since ZigBee's wireless
communication module is low-power, low-cost, and offers high nodal coverage, it was
selected for this study. A Zigbee communication module can be categorized into three
types.1. End devices 2. Zigbee coordinator 3. ZigBee router. Therefore, add-on devices in
each household are configured to be Zigbee end devices, while the NESEM is configured to
be a Zigbee coordinator. As NESEM and Zigbee operate in a master-slave configuration,
each add-on device (end node) is assigned a slave ID (consumer ID), which is controlled by
master NESEM. Add-on devices communicate data to the ZigBee coordinator in NESEM,
which is then transferred to the microcontroller (Arduino Mega) for further processing.
Instead of sending all data to the server, collected data from different households is stored
in the local storage of NESEM. A display unit is also included in NESEM for visual and
graphical indication of metering activity. As part of NESEM, a GSM module provides
notifications and alerts to consumers, delivering messages based on their requests. SM are
simulated in two case studies based on the SM benefits to the utility (remote metering and
controlling such as remote metering, remote supply connect/disconnect) and benefits to the
consumer (decentralized CIS such as energy consumption, billing and IHD update). A
summary of the tested results is provided in the results section for each case study. In
Figure 7, the flow chart outlines the different processing mechanisms for each function of
the proposed NESEM.

Figure. 3 Algorithms for processing mechanism of NESEM

4.2. Simulation results and discussions

Proteus 8 Professional is opted for designing and simulating the NESEM and add-on
module to validate the viability of smart metering. Proteus has the virtual emulation
capability of Arduino and other required tools for proposed work. Along with that Arduino
IDE, open-source software used for writing codes for Arduino controllers. After execution

590
 J. Electrical Systems 18-4 (2022): 582-599

of the written codes in IDE, .hex file is generated and inserted into the proteus software for
simulation. Virtual Serial Ports Emulator (VSPE) software is used to connect all ports of
communication modules (Zigbee) internally. With this COM ports pairs were generated and
assigned to respective Zigbee modules. Architecture of the proposed work consists of five
different designed models, namely 1. Load model (Resistive and inductive loads), 2. Digital
meter model, 3. Add-on device model, and 4. NESEM model apart from inbuild proteus
library models Zigbee modules. Test is conducted for two case studies such as case 1-
Remote metering and controlling, case 2-Decentralized CIS. The simulation architectural
layout of the NESEM and simulation block of add-on device is shown in Figure 8, 9
respectively.

Figure 8: Simulation design model of the add-on device

The digital meter model is responsible for measuring the amount of energy consumed in the
household with the reading from current and voltage sensors. Analog read functionality of
Arduino the values of current and voltage are measured in the digital meter block. RS485 to
TTL module used to establish the communication between the serial ports of the digital
meter and the addon device. Driver Enable (DE) and Receiver Enable (RE) pins of RS485
to TTL are connected together with pin 2 addon device Arduino. By decaling pins 7,6 as
software serial pins, the Driver Input (DI) and Receiver Output (RO) pins of RS485 to TTL
can send data by connecting 7 and 6 with DI and RO. Extracted data is communicated to
the NESEM through the Zigbee connected in TX0and TX1 of add-on device Arduino.

Figure 9: Internal simulation design of the proposed NESEM

591
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

NESEM model consists of the main controller (Arduino MEGA), ZigBee transceiver,
GSM, and esp8266 in its architecture. Upon receiving the communicated from the Zigbee
Transceiver at the add-on device Zigbee coordinator collects the information and store in
the local memory of NESEM. Arduino Mega has 4 dedicated hardware serial port and in
that TX0, RX0, TX3, R31 and TX2, RX2 are connected with GSM, esp8266 AND Zigbee
module respectively. Because of the simulation software constraints, the different case
studies mentioned earlier are simulated tested separately and are discussed in the following
section. esp8266 enables server communication to receive the control signal for billing
updates. GSM is programmed to send consumer consumption updates upon receiving the
request command from the users.

Case 1: Utility benefits – remote metering and remote supply connect/disconnect.

The utility's smart metering process entails in two key steps: remote metering and remote
supply connection and disconnect. Traditionally, household energy consumption is
measured and billed by human workers. During this process, a great deal of time was
required, transportation costs were incurred, and labour costs were increased, resulting in
larger emissions of CO2. Contactless monitoring and instant metering are possible with
remote metering, reducing both CO2 emissions and processing times. NESEM able to send
consumption detail in real time to the server using es8266 and 12 hours energy
consumption of five houses are plotted and illustrated in the following Figure 10.

Figure 10:12 Hours energy consumption plot for five houses

With remote supply connect/disconnect features, monitoring and maintaining household

payment status can be accomplished without a manual visit. NESEM is programmed to
check each consumer's payment status remotely and disconnect or reconnect the
household's supply without human intervention. This feature is initiated by a command
signal from the utility center's data base. Virtual terminal windows are used to feed the
billing status of the five houses to the NESEM in this simulation. As a result, NESEM
determines whether to connect or disconnect a consumer's supply based on the local
connection status. The NESEM sends a control signal to the programmable relay in the add-
on device based on the decision status. The simulation result of this functionality is detailed
in the following Table 3.

592
 J. Electrical Systems 18-4 (2022): 582-599

Table 3: Characteristics of the thermal units

Command initiated Consumer ID Supply connection Billing status Relay Decision status
from the HES status of ID status signal to relay
YES Home 1 Connected Paid ON ON
NO Home 2 Connected Not paid ON No action
YES Home 3 Disconnected Paid OFF ON
YES Home 4 Connected Paid ON ON
YES Home 5 Connected Not paid ON OFF

Case 2: Decentralized CIS – remote billing, energy consumption updates and IHD
updates
Efficiency in energy utilization contributes greatly to the reduction of energy generation. In
order to accomplish this, consumers need to participate in energy conservation programs,
which is achieved when consumers know what they are consuming in real time. NESEM
directly notifies consumers regarding real-time consumption details without requiring a
central server. Due to NESEM, consumers do not have to constantly log into utility servers
in order to obtain consumption details. In the NESEM, a GSM module sends the necessary
data in the form of SMS messages based on the requests raised by the consumer. It is
possible for consumers to access this information through their mobile phones. A major
feature of the CIS is the billing, consumption, and IHD updates that help consumers
conserve energy and save money. In energy consumption updates, NESEM reports the
details of energy consumption based on the checking queries raised by consumers through
SMS, and updates the billing information on a monthly basis based on utility regulations.
Simulation results of these feature are show in the Figure 11.

Consumer 1 mobile display

Consumer 2 mobile display

Consumer 4 mobile display Consumer 3 mobile display

Figure 11:Remotemetering and billing

The IHD service delivers real-time consumption and grid control data to consumers in real-
time to promote energy conservation. Since IHD covers most of the data transfer between
utility companies and consumers, the communication network has a very high data volume.
By avoiding the constant downlink of data from the utility HES, decentralized
programming of the NESEM sends the real time data by itself results in reduced data traffic
in the smart grid communication.

593
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

5. Data volume and traffic analysis of Decentralized NESEM frame work

Smart grid big data are categorised into different entities such as volume (amount of data
in bytes), velocity (speed of data transfer), variety (type of data), veracity (messiness of the
data), and value (quality of retrieved data). Additionally, bidirectional communication from
the consumer end to the utility end is divided into two parts: 1. Uplink data volume, 2.
Downlink data volume. Monitoring and controlling the grid depends mainly on uplink data,
while monitoring and controlling consumer energy consumption depends on downlink data,
as illustrated in the following Figure 12.

Figure 12:Benefits of uplink and downlink data

In a smart communication network, the frequency of uplinks and downlinks is
determined by the requirements of both entities, and this frequency determines the volume
of data transferred. Structured data plays a major role in billing, consumption, and IHD
updates, and they can be obtained easily from the consumer end, so decentralized CIS
frameworks can be adopted to solve big data issues associated with smart grids. As part of
this study, Volume and Variety characteristics are considered in analysing uplink and
downlink data [44] related to consumer monitoring and control. We formulate a
mathematical model to analyse the volume of data transferred in communication medium
for both uplink and downlink data. The mathematical model is examined for one year with
billing, energy consumption, and IHD updates information for 1 million consumers in two
scenarios 1. Existing Centralized data processing framework and 2. NESEM decentralized
data processing framework. As part of the analysis, some prerequisites are referenced and
assumed from the existing literature. Pocket sizes vary based on details contained in the
pocket, such as meter id, total energy consumption, tamper records, and other meter
information [45]. As referenced in [46] the following Table 4 summarizes the pocket size.

Table 4: Pocket size of the data from SM

Pocketinformation Sizeofthepocket(byte)
MeterID 7
Totalenergyconsumption 4
TamperStatus 2
AvailableCredit 4
Date Stored 4

For this analysis we consider only Meter ID and total energy consumption are encapsuled
in the pockets (11 bytes) which send from the smart meter. Time period and total number of
consumers for this study are pre fixed values such as 106 and 24 months respectively.

594
 J. Electrical Systems 18-4 (2022): 582-599

Centralized data processing

(1)
(2)
Decentralized data processing
(3)
(4)

= bytes from smart meter = 11,

,

m= number of NESEM=105

In a centralized CIS framework, data collected from consumers is routed first to a nearby
DCU, then to a central HES for processing. The uplink data volume in this scenario is
restricted to SM data only, not other smart devices like PMUs and RTUs. SM updates the
consumption details of 11 bytes of consumption details for every 15 minutes (b=44 bytes
for 1 hour) through nearby DCUs. Thus, the frequency of pooling energy consumption
details is 96 times a day and 35040 (f) times a year. For 1 million consumers, it is
calculated that 385.440x109 bytes of uplink data is transferred to a central server at
frequency of 35040 by using equation 1. In case of downlink, the total volume of data
transferred is the cumulative of billing data, energy consumption update, IHD updates.
Billing information is updated to the consumers for every once a month or twice a month
depending upon the utility policy. In this paper, the monthly billing method is considered
for the analysis, so the information flow occurs once every month ( =12). The size of a
standard SMS is 140 bytes ( =140). For 1 million consumers, one year will result in
1.68x109 bytes of billing data (downlink data). As for IHD downlink volume, it is
calculated assuming that from the central server to IHD, 80 bytes ( ) of information is
transferred on average. According to [42] a single consumer checks with IHD on average
five times a day ( ) for updates. A total of 146x109 bytes of downlink IHD
data is transferred each year from the utility HES to the 1 million consumers. Based on the
consumer request ( =1825, similar to the frequency of IHD checking), the downlink
data volume for one year with packet size of 140 bytes ( ) is 255.5x1099 bytes is derived
by equation 2. Over a period of 24 months, 385.440 x 109 bytes of uplink data and 403.18 x
109 bytes of downlink data are transferred between SM and HES for 1 million consumers.
Unlike centralized frame work the Decentralized CIS frame work with NESEM, data
routing mechanism involves mainly communication between add-on device and NESEM in
order to implement the CIS. In this scenario, the HES receives only the uplink data, while
all downlink data is downloaded from the NESEM with decentralized CIS programming.

595
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

Furthermore, according to the NESEM design, 1 lakh NESEMs will cover one million
consumers (simulation aspect). From the simulation it is assumed that NESEM updates 20
bytes ( ) of 10 consumer information for every 15 minutes
( ). For 1 lakh NESEM the uplink data volume is calculated
for m=105 (NESEM count) is resulting in 70.08x109 bytes for one million consumers
associated with a 1 lakh NESEM for one year using equation 3. However, the downlink
traffic originates from NESEM for billing and energy consumption updates. From NESEM
1.68x109 bytes of information are used for consumer billing every month
( ) for the period of 1 year. In contrast to centralized processing, in
this case the downlink traffic is created by NESEM, not by HES. Energy consumption
updates via SMS are also generated by NESEM with ( =1825) for one year, similar
to IHD checks) downlink data volume of 255.5x109 bytes for one year and packet size of
140 bytes ( . Unlike in centralized data processing in decentralized case grid control
data of 20 bytes is transferred from the HES and the remaining 60 bytes are generated from
NESEM to IHD. As a result, during a one-year period with display checking frequency of
five times a day ( ), the downlink volume of 3.65x109 bytes is delivered
from HES to 1 lakh NESEM and 109.5x109 bytes generated from one lakh NESEM to
consumer end. So as total of 113.15x109 bytes of IHD data transferred from HES to 1
million consumers. In conclusion, 70.08x109 bytes of uplink data were transferred from
consumer to HES, 3.65x109 bytes were transferred from HES to NESEM, and 370.18x109
bytes of downlink data were transferred from 1 lakh NESEM to consumer end (for 1
million consumers) and is calculated using equation 4. These details are illustrated in the
following Figure 13 and Table 5.
Tot al uplink
Downlink
GIS For
consumer s to HES --> 70.08x 10 9 by tes

Uplink dat a for HEAD END OMS One million consumer s

wir eless communicat ion bet ween

DCU and consumer
SYSTEM DRMS Tot al Downlink
NESEM to single consumer --> 370.18x 10 3 by tes
For
DM – Digit al Met er MDMS One million consumer s HES to 1 lak h NESEM --> 3.65x 10 9 by tes
AD – Add on Device 70.08x109 bytes of
uplink data

3.65 x109 bytes of

Home 1 downlink data
dm AD
370.18 x103 bytes
Home 2 of downlink data
dm AD CLUSTER n
3
36.5 x10 bytes of
Home 3 downlink data Smar t met er ing
NESEM
dm AD f or 10 houses up
CIS 36.5x103 bytes of
downlink data to N
Home 4
dm AD
36.5 x103 bytes of
downlink data
1 0 Up CLUSTER 2 n
o
ho t o to l l i
m
e Up mi
n e me
Home 10 Smar t met er ing =o h o
N
dm AD f or 10 houses
Smar t met er i ng
f or 10 houses

Figure 13:Decentralized frame workuplink and downlink

Table5:DatavolumeincentralizedanddecentralizedCISframework
Data
Centralized CIS Decentralized CIS
volume
9
Uplink Consumer to HES 385.440 x 10 bytes Consumer to HES 70.08x109 bytes
HES to 1 million 9 HES to 1 lakh NESEM 3.65 x109 bytes
Downlink 403.18 x 10 bytes
consumers NESEM to 1 million consumers 370.18 x109 bytes

596
 J. Electrical Systems 18-4 (2022): 582-599

Functionality of SM is not restricted to certain limits and varies from meter to meter based
upon vendor, consumer, and utility preference. All smart meter’s support remote metering,
monitoring, controlling, consumer alert notifications, and remote disconnecting. We
simulate and test all such functionalities in this paper. In comparison with existing smart
metering concepts, NESEM has a smaller uplink and downlink data volume due to its
decentralized CIS framework. Due to this decentralized metering process, a powerful HES
is not required to store large volumes of big data collected from different meters.
Consumers in developing and underdeveloped countries have difficulty switching to smart
meters because they are expensive. In contrast to existing smart metering technologies,
NESEM networked a group of homes without replacing the digital meters. As a result,
deployment costs in the smart grid are reduced on a major scale and dumb nodes are
reduced. In addition to mitigating the need to replace old meters with new ones, the
NESEM enhances consumer participation in the AMI network and reduces network traffic
due to a decentralized CIS configuration. In comparison to existing SM and smart metering
processes, NESEM has several benefits for consumers and utilities. And also, the target 7.b
of UN- SDG 7 affordable and clean energy can be realized by NESEM for developing and
under developed countries. NESEM and existing SM are compared in Table 6 for their
benefits and features.

Table 6: Comparative benefits and feature of existing SM and NESEM

Existing SM and metering process NESEM and metering process
Huge deployment cost Less deployment cost when compared with
existing smart meter
Maintenance cost is high since all houses have smart Maintenance cost is less a common NESEM is
meter present for group of houses
Replacement of old meter Replacement of old meter is not needed
Network congestion is high due to high volume data to be Network congestion is less due to minimal
processed. amount of data processed
Centralized CIS framework Decentralized CIS framework
Dump node is more because of inactive consumer Dump node is less because of networking of
participation digital energy meter

6. Conclusion
The NESEM design simulated and its data volume of decentralized CIS framework is
calculated which shows that reduced network congestion for smart grid communication
with the lower-cost deployment of SM. Through the results and outcomes of the simulation,
it is concluded that smart metering is viable and implementable through the proposed model
of NESEM without replacing existing metering infrastructure. In addition, the uplink and
downlink volume analysis of decentralized CIS indicate that traffic produced by the big
data in smart grid communications is reduced. The existing SM common functionalities like
remote metering, remote supply connect/disconnect, remote billing along with
decentralized CIS are incorporated in the NESEM. The justification of the proposed idea
among public is validated with the residential consumer opinion survey. With these
obtained results and outcomes, it is concluded that the proposed model is suitable to
mitigate the certain problems associated with the existing SM related to UN-SDG 7. b. The
future scope of this work is expanded in terms NESEM will be added with decentralized
processing MDMS to increase the benefits of it to consumers, utility, and society.

597
 Ezhilarasi P et al: Modelling and analysis of cost effective smart meter with decentralized CIS
framework towards optimal network traffic

References
[1] Y. Kabalci, “A survey on smart metering and smart grid communication,” Renew. Sustain. Energy Rev.,
vol. 57, pp. 302–318, May 2016.
[2] H. K. M. Y. Amal A.Hassan, Faten H. Fahmy, Abd El-Shafy A. Nafeh, “Control of Three-Phase Inverters
for Smart Grid Integration of Photovoltaic Systems,” J. Electr. Syst. 6(11), 951–952., vol. 1, pp. 5–48, 2019.
[3] J. Alfonsopalacios, A. Aguado José, J. Leiva, A. Palacios, and J. A. Aguado, “Smart metering trends,
implications and necessities: A policy review,” Renew. Sustain. Energy Rev., vol. 55, pp. 227–233, Mar.
2016.
[4] D. Kumar and A. H. Khan, “Roll-Out Strategy for Smooth Transition of Traditional Meters to Smart
Meters,” in Lecture Notes in Electrical Engineering, vol. 764, D. S. Pillai R.K. Dixit A., Ed. Springer
Science and Business Media Deutschland GmbH, 2022, pp. 287–294.
[5] W. Aslam, M. Soban, F. Akhtar, and N. A. Zaffar, “Smart meters for industrial energy conservation and
efficiency optimization in Pakistan: Scope, technology and applications,” Renew. Sustain. Energy Rev., vol.
44, pp. 933–943, Apr. 2015.
[6] K. Sharma and L. Mohan Saini, “Performance analysis of smart metering for smart grid: An overview,”
Renew. Sustain. Energy Rev., vol. 49, pp. 720–735, Sep. 2015.
[7] S. M. M. T. R. RoofegariNejad, S. M. Hakimi, “A Novel Demand Response Method for Smart Microgrids
Related to the Uncertainties of Renewable Energy Resources and Energy Price,” J. Electr. Syst., vol. 2, pp.
419–441, 2016.
[8] H. Daki, A. El Hannani, A. Aqqal, A. Haidine, and A. Dahbi, “Big Data management in smart grid:
concepts, requirements and implementation,” J. Big Data, vol. 4, no. 1, pp. 1–19, 2017.
[9] B. K. Sovacool, A. Hook, S. Sareen, and F. W. Geels, “Global sustainability, innovation and governance
dynamics of national smart electricity meter transitions,” Glob. Environ. Chang., vol. 68, p. 102272, 2021.
[10] M. Malloni and J. Paris, “Understanding the total cost of smart meter operations,” 2014.
[11] N. Uribe-Pérez, L. Hernández, D. de la Vega, and I. Angulo, “State of the Art and Trends Review of Smart
Metering in Electricity Grids,” Appl. Sci., vol. 6, no. 3, p. 68, Feb. 2016.
[12] S. Erlinghagen, B. Lichtensteiger, and J. Markard, “Smart meter communication standards in Europe - a
comparison,” Renew. Sustain. ENERGY Rev., vol. 43, pp. 1249–1262, Mar. 2015.
[13] S. S. S. R. Depuru, L. Wang, and V. Devabhaktuni, “Electricity theft: Overview, issues, prevention and a
smart meter based approach to control theft,” Energy Policy, vol. 39, no. 2, pp. 1007–1015, Feb. 2011.
[14] L. Wen, K. Zhou, S. Yang, and L. Li, “Compression of smart meter big data: A survey,” Renew. Sustain.
ENERGY Rev., vol. 91, pp. 59–69, Aug. 2018.
[15] A. Kumari and S. Tanwar, “A secure data analytics scheme for multimedia communication in a
decentralized smart grid,” Multimed. Tools Appl., vol. 81, no. 24, pp. 34797–34822, Oct. 2022.
[16] D. Capriglione, L. Ferrigno, V. Paciello, A. Pietrosanto, and A. Vaccaro, “Experimental characterization of
consensus protocol for decentralized smart grid metering,” Measurement, vol. 77, pp. 292–306, Jan. 2016.
[17] P. Ezhilarasi, and L. Ramesh, “Review on Smart Energy Meter for low cost design,” in 2019 5th
International Conference On Computing, Communication, Control And Automation (ICCUBEA), 2019, pp.
1–8.
[18] S.-W. Luan, J.-H. Teng, S.-Y. Chan, and L.-C. Hwang, “Development of a smart power meter for AMI
based on ZigBee communication,” in 2009 International Conference on Power Electronics and Drive
Systems (PEDS), 2009, pp. 661–665.
[19] G. Del Prete, C. Landi, S. Università, G. Del Prete, C. Landi, G. Del Prete, and C. Landi, “Real-time smart
meter with embedded web server capability,” 2012 IEEE Int. Instrum. Meas. Technol. Conf. Proc., pp. 682–
687, May 2012.
[20] Z. I. Rana, M. Waseem, T. Mahmood, H. M. Z. Iqbal, M. Waseem, and T. Mahmood, “Automatic Energy
Meter Reading using Smart Energy Meter,” in Proc. Int. Conf. Engineering & Emerging Technologies,
2014, pp. 1–5.
[21] H. Das and L. C. Saikia, “GSM enabled smart energy meter and automation of home appliances,” in 2015
International Conference on Energy, Power and Environment: Towards Sustainable Growth (ICEPE), 2015,
pp. 1–5.
[22] L. C. Saikia, H. Das, N. B. Dev Choudhury, and T. Malakar, “GPRS enabled smart energy meter with in-
home display and application of time of use pricing,” in 2016 IEEE Annual India Conference (INDICON),
2016, pp. 1–5.
[23] B. Nithin, S. Bhaskaran, and S. Ullas, “Advanced Metering Infrastructure ( AMI ) with combination of Peak
Load Management System ( PLMS ) and Theft protection,” 2016.
[24] V. Preethi and G. Harish, “Design and implementation of smart energy meter,” in 2016 International
Conference on Inventive Computation Technologies (ICICT), 2016, vol. 1, pp. 1–5.
[25] R. Morello, C. De Capua, G. Fulco, and S. C. Mukhopadhyay, “A Smart Power Meter to Monitor Energy
Flow in Smart Grids: The Role of Advanced Sensing and IoT in the Electric Grid of the Future,” IEEE Sens.
J., vol. 17, no. 23, pp. 7828–7837, Dec. 2017.

598
 J. Electrical Systems 18-4 (2022): 582-599

[26] C. Paper and F. K. Handhal, “Design and building a single-phase smart energy meter using Arduino and RF
communication system Design and building a single-phase smart energy meter using Arduino and RF
communication system,” 3rd Int. Sci. Conf., no. March, 2018.
[27] S. Elakshumi and A. Ponraj, “A server based load analysis of smart meter systems,” in 2017 International
Conference on Nextgen Electronic Technologies: Silicon to Software (ICNETS2), 2017, pp. 141–144.
[28] S. Singh and M. P. Selvan, “A Smart Energy Meter Enabling Self-Demand Response of Consumers in
Smart Cities of Tamil Nadu,” in 2019 IEEE International Conference on Smart Cities Model (ICSCM),
2019, pp. 1–6.
[29] F. Teddy, S. Gunawan, M. H. Anuar, M. Kartiwi, and Z. Janin, “Development of Power Factor Meter using
Arduino,” 2018 IEEE 5th Int. Conf. Smart Instrumentation, Meas. Appl., no. November, pp. 1–4, 2018.
[30] S. Saha, S. Mondal, A. Saha, and P. Purkait, “Design and Implementation of IoT Based Smart Energy
Meter,” in 2018 IEEE Applied Signal Processing Conference (ASPCON), 2018, pp. 19–23.
[31] A. Ahad, S. Mitra, O. Morshed, I. Khan, R. Sarker, and S. S. Chowdhury, “Implementation and Feasibility
Analysis of GSM Based Smart Energy Meter for Digitalized Power Consumption with Advanced Features,”
in 2018 International Seminar on Intelligent Technology and Its Applications (ISITIA), 2018, pp. 403–407.
[32] Q. Malik, A. Zia, R. Ahmad, M. A. Butt, and Z. Ahmad Javed, “Design and Operation of Smart Energy
Meter for Effective Energy Utilization in Smart Cities,” in 2019 IEEE Conference on Sustainable Utilization
and Development in Engineering and Technologies (CSUDET), 2019, pp. 219–223.
[33] F. Sabir, M. A. Chohan, and B. Pirzada, “Design and implementation of smart metering device using
specialized IC ADE-7753 and its integration with web server for smart grid,” in 2019 Second International
Conference on Latest trends in Electrical Engineering and Computing Technologies (INTELLECT), 2019,
pp. 1–5.
[34] Z. Sultan, Y. Jiang, A. Malik, and S. F. Ahmed, “GSM based smart wireless controlled digital energy
meter,” in 2019 IEEE 6th International Conference on Engineering Technologies and Applied Sciences
(ICETAS), 2019, pp. 1–6.
[35] H. K. Patel, T. Mody, and A. Goyal, “Arduino Based Smart Energy Meter using GSM,” in 2019 4th
International Conference on Internet of Things: Smart Innovation and Usages (IoT-SIU), 2019, pp. 1–6.
[36] N. Mahfuz, M. Nigar, and N. Ulfat, “Smart Energy Meter and Digital Billing System for Bangladesh,” in
2020 11th International Conference on Computing, Communication and Networking Technologies
(ICCCNT), 2020, pp. 1–4.
[37] G. I. Rashed, H. Haider, and M. B. Shafik, “Improving the reliability of energy supply utilizing smart energy
meter system based on GSM: case study in Pakistan,” in 2020 5th Asia Conference on Power and Electrical
Engineering (ACPEE), 2020, pp. 1080–1084.
[38] M. Leelavati and K. Aswini, “Smart Energy Meter with Reading Indication using GSM,” IRJET, vol. 2,
2015.
[39] R. K. Pillai, R. Bhatnagar, and H. Thukral, “AMI rollout strategy and cost-benefit analysis for India,” in
2016 First International Conference on Sustainable Green Buildings and Communities (SGBC), 2016, pp.
1–6.
[40] B. P. Bhattarai, S. Paudyal, Y. Luo, M. Mohanpurkar, K. Cheung, R. Tonkoski, R. Hovsapian, K. S. Myers,
R. Zhang, P. Zhao, M. Manic, S. Zhang, and X. Zhang, “Big data analytics in smart grids: State-of-theart,
challenges, opportunities, and future directions,” IET Smart Grid, vol. 2, no. 2, pp. 141–154, 2019.
[41] C. Karupongsiri, M. Farhad Hossain, K. S. Munasinghe, and A. Jamalipour, “A novel scheduling technique
for Smart Grid data on LTE networks,” 2013, 7th Int. Conf. Signal Process. Commun. Syst. ICSPCS 2013 -
Proc., no. Dp 1096276, pp. 1–5, 2013.
[42] Ministry of Power, “Household Electricity ststus.” [Online]. Available: https://saubhagya.gov.in/.
[43] D. M. Jaiswal and M. P. Thakre, “Modeling& designing of smart energy meter for smart grid applications,”
Glob. Transitions Proc., vol. 3, no. 1, pp. 311–316, Jun. 2022.
[44] Y. Zhang, T. Huang, and E. F. Bompard, “Big data analytics in smart grids: a review,” Energy Informatics,
vol. 1, no. 1, pp. 1–24, 2018.
[45] K. G. Ngandu, K. Ouahada, and S. Rimer, “Smart meter data collection using public taxis,” Sensors
(Switzerland), vol. 18, no. 7, 2018.
[46] [46] M. Aiello and G. A. Pagani, “The smart grid’s data generating potentials,” 2014 Fed. Conf. Comput.
Sci. Inf. Syst. FedCSIS 2014, vol. 2, pp. 9–16, 2014.
[47] P. Xu, J. Shen, X. Zhang, X. Zhao, and Y. Qian, “Case Study of Smart Meter and In-home Display for
Residential Behavior Change in Shanghai, China,” Energy Procedia, vol. 75, no. August, pp. 2694–2699,
2015.

599
 © 2022. This work is published under
https://creativecommons.org/licenses/by/4.0/legalcode(the“License”).
Notwithstanding the ProQuest Terms and Conditions, you may use this
content in accordance with the terms of the License.