A Low Cost Electricity Monitoring Wireless Sensor Network

Prihadi Murdiyat

, Abdul Hamid Kurniawan

, Marson Ady Putra

, Khairuddin Karim

and Supian Noor

Electrical Department, Politeknik Negeri Samarinda, Jl. Dr. Ciptomangunkusumo, Samarinda, Indonesia

spn.evolution@gmail.com

Keywords: Electricity Monitoring, Wireless Sensor Network, Arduino, LoRa, WiFi.

Abstract: As electricity price increases, the electricity consumption should be taken into account and managed wisely.

For schools, universities, or industries having a number of buildings, each building should be equipped with

a kWh meter. Installing a branded kWh meter in every building could be too expensive. This research aims

to develop a low cost electricity monitoring system that consists a number of sensor nodes (SNs), a

sink/gateway, and a server. The SNs and gateway are built based on Arduino boards. LoRa radio is used to

establish communication between SNs and the gateway, while WiFi radio is utilized to perform

communication between the gateway and campus internet network where the server is connected. Research

activities includes SN and gateway prototype building, sensors calibration, data communication testing, link

measurement, installation, and system running test. The system with five SNs has been built and installed at

five buildings in the campus of Politeknik Negeri Samarinda. According the measurement during running

test, the RSSI and SNR are mostly better than the value obtained in preliminary link test. Furthermore, the

monitoring system does not give only electricity consumption pattern, but figures of installation problems,

which is important for electrical installation improvement.

1 INTRODUCTION

In Indonesia, electricity tariff always increases almost

annually. The tariff is increased to reduce the

electricity subsidy for the middle to high classes, keep

the service reliable, and to enhance the electrification

(PT. PLN, 2021). Therefore, the electricity

consumption must be managed properly to save the

cost that may appear due to inadvisable use of

electricity.

Electrical consumption is normally measured by a

kilowatt hour meter (kWh meter) either analogue or

digital. Customers rely on kWh meter to measure the

electricity they have used during the past period.

When they found that the amount of electricity in the

previous period is high, they might consider to control

the electricity consumption in the next period. For

schools, universities, or industries having several

buildings in their area, a kWh meter should be

https://orcid.org/0000-0002-7962-1306

https://orcid.org/0000-0002-5772-3889

https://orcid.org/0000-0002-4736-881X

https://orcid.org/0000-0002-9030-3026

installed in every building. It is particularly important

as each building may have different electricity

consumption characteristic. Lack of kWh meter in

every building could make electricity utilization

policy becoming unsuitable.

Furthermore, rather than utilizing conventional

kWh meter where people must visit periodically to

read the electricity consumption such as in (Pasurono,

2013), people nowadays tend to use an integrated

monitoring system where kWh meter installed in

every building sends data report to a central station

periodically. As such, no reader is required to observe

the kWh meter. This trend rises due to the emerge of

internet of things (IoT). Moreover, the integrated

monitoring system is commonly able to measure

other electrical parameters such as voltage, current,

frequency, and power factor. Therefore, there will be

more data to analyse the consumption characteristic.

In this system, the data could be reported 24/7 with a

certain period such every hour.

278

Murdiyat, P., Kurniawan, A., Putra, M., Karim, K. and Noor, S.

A Low Cost Electricity Monitoring Wireless Sensor Network.

DOI: 10.5220/0010944100003260

In Proceedings of the 4th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2021), pages 278-283

ISBN: 978-989-758-615-6; ISSN: 2975-8246

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

The price of integrated system such as in (Eyedro,

2021) is USD 299 or over IDR 4 million per unit. It is

just the price of sensor node and does not include the

server, shipping, and installation. For industries,

installing every building with such a unit may not be

a problem as measuring the electrical load

characteristic is required. The cost of metering

procurement might not as high as the outcome of

controlling the electricity consumption, where they

could save the cost significantly. However, for many

schools and universities, it could be very costly.

Hence, a low cost electricity monitoring wireless

sensor network developed in this research could be a

solution.

The system developed in this research comprises

a number of sensing units (ie. sensor node) that is

installed next to main panel available in every

building. The sensor node (SN) measures almost all

electrical parameters and displays them on a liquid

crystal display (LCD). The data of several parameters

are also sent periodically to a gateway through

wireless network.

Once accepting the data, the gateway forwards

data to the server that is prepared for campus

resources monitoring. Since the server connects to the

campus internet network, the gateway should have an

access to the campus internet as well. The server

saves the data and could display it as a simple statistic

form. Every person who has authority would be given

an access to the data, and hence could monitor it from

everywhere every time.

To compare with similar research, below are

several research on Arduino based energy meter. In

(Rajput, 2018) an energy meter is developed to

provide prepaid system that utilizes GSM technology

for data communication. Through this system,

consumers are allowed to manage their electricity

consumption by paying up front. Another research

reported in (Mithya, 2019) builds energy meter that is

able to send message to both customer and the

electricity company through GSM. When electricity

consumption is beyond a specific threshold, an alarm

is sent. While both projects use GSM communication

technology, a smart energy meter developed in

(Kanakaraja, 2021) utilizes LoRa WAN as the

communication system. The energy meter allows the

user to monitor energy consumption every time from

everywhere. All systems above generally focus on

home energy meter and concern about single sensor

node. In contrast, the system introduced in this paper

focus on energy measuring network in institution

local area with a number of sensor nodes.

2 SYSTEM DESCRIPTION

The architecture of the monitoring system network is

shown in figure 1. That is similar with the common

wireless sensor network (WSN) architecture

described in (Akyildiz, 2002). A unit namely sensor

node (SN) is built in a panel box and installed next to

the main panel. Three current transformers and a

number of cables from the SN’s voltage sensors are

installed in the main panel to measure voltage,

current, power factor, frequency, active and apparent

power, and electrical energy. All data are processed

and displayed on SN’s liquid crystal display (LCD).

Since the main panel consists of three-phase

installation, the SN measures parameters of every

phase. The only three-phase parameter measured by

SN is the energy in kWh unit. The SN does not only

display its measurement results on LCD, but also send

several data to the gateway as well. Such data are the

active power of phase R, S, T, and the energy of three-

phase. The purpose of sending the active power to the

server is to observe the balance of the three-phase of

every building, while the purpose of sending the

three-phase energy is to measure the building

electricity consumption. To avoid cabling

complexity, the communication between SN and the

gateway is established by wireless link.

Figure 1: The architecture of monitoring system network

Long range (LoRa) radio technology (LoRa

Alliance, 2015) is considered to establish the link

between SN and the gateway as among other

technologies such as WiFi (Digi Int., 2008) and

Bluetooth (Wooley, 2020), LoRa is able to reach the

longest distance, which is up to kilometres away.

According its specification (LoRa Alliance, 2015), in

the urban area LoRa could reach up to 5 km, while in

rural area it is able to reach up to 15 km. In a specific

condition (The Things Network, 2020), LoRa could

even reach 832 km. In this case, a LoRa transmitter is

brought by a balloon flying 83 km above the land, and

as a consequence a line of sight (LoS) conditions is

achieved. To work in LoRa protocols, both SN and

the gateway are equipped with a LoRa transceiver.

A Low Cost Electricity Monitoring Wireless Sensor Network

279

The working frequency is 915 MHz to deal with the

Indonesian government regulation of low power wide

area network (LPWAN) where LoRa is included.

When the gateway has accepted data from SNs, it

forwards the data to server. The server should have a

connection to the internet to make it accessible to be

monitored from everywhere at any time. In this

research, the server connects to Politeknik Negeri

Samarinda internet network. As a consequence, the

gateway is equipped with a WiFi transceiver to

communicate with the server. Currently, the gateway

is located at the Electrical Department Laboratory

where several access points (APs) are available to

connect to.

All data coming from the gateway are collected

by the server. A simple web based application

controls, saves, and displays the incoming data. The

application could be accessed from everywhere at any

time by a person who has been given authority.

3 RESEARCH METHODS

Research activities is started by reviewing literatures

pertaining electrical parameter sensors, Arduino

based wireless sensor networks, communication

protocols, LoRa and WiFi technologies, and server

development. When significant information obtained

from the literature is sufficient, the SN and gateway

prototype are built. Along with the prototype

building, the sensors are calibrated by comparing the

measured parameter values with the value measured

by correspond laboratory electrical metering.

The process is followed by data communication

evaluation involving communication protocol test.

Prior installation, link measurement is undertaken to

observe the capability of communication system to

provide communication link and network between

SNs and gateway. Once the installation of five SNs,

the gateway, and the server is accomplished, the

system running test is performed.

4 RESULTS AND DISCUSSION

Sensor node (SN) consists of three PZEM-400T

sensors including its current transformer, an Arduino

Nano, a 915 MHz LoRa transceiver, 20 x 4 LCD

display, and 3DR 5dBi antenna. The unit is powered

by a 220 V AC to 12V DC mini power supply. All

parts are installed into a panel box with the size of 35

x 25 x 12 cm

as shown in figure 2. The circuit

diagram and other details are described in (Alfan,

2021). The accuracy of the SN measurement has been

compared with HIOKI PQ3100 power quality

analyser (Alfan, 2021), and shows that the error is less

than 3 % for voltage, current, apparent power, active

power, and frequency, while for power factor is

6.505 %.

Figure 2: Sensor node.

The cost of a sensor node unit in this system is less

than IDR 1 million including installation. It is about a

quarter of the price of a sensor node unit in (Eyedro,

2021). With a large number of production, the cost

could be cheaper.

Due to the requirement and convenient operation,

several parts of the program in (Alfan, 2021) has been

changed. For example, the parameters of every phase

are displayed consecutively and automatically

without pushing the push button as the previous

design. Also, the length of data values is increased 1

byte into 2 bytes. Hence the data format becomes 1

byte of the destination address, 1 byte of the source

address, 1 byte of message counter, 1 byte of packet

length, 1 byte of data1 type, 2 bytes of data1 value, 1

byte of data2 type, 2 bytes of data2 value, 1 byte of

data3 type, 2 bytes of data3 value, 1 byte of data4

type, 2 bytes of data4 value. The contents of every

data are phase R, S, T active powers, and three-phase

kWh respectively. This packet format is inserted into

the LoRa data packet. The packet is sent every 30

minutes.

During installation, several cables from three-

phase voltage sensor of SN are connected to phase R,

phase S, phase T, Neutral bus bar respectively.

Meanwhile, three current transformers (CTs) are

clamped on their respective phase load cable/busbar

with the CT cables are connected to the current sensor

unit.

The gateway comprises of a Wemos board and a

915 MHz LoRa transceiver. Both are powered by a

DC 12 V 5 A power supply and installed in a 35 x 25

x 12 cm

panel box as shown in figure 3. The LoRa

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

280

transceiver is utilized to receive data transmitted by

SNs. The data then are processed by the Wemos into

the packet that is suitable with the WiFi packet

format. Afterwards, the packet is sent to the server

throughout the campus internet.

Figure 3: Gateway.

The server of the monitoring network takes a

space of the Department of Electrical Engineering

server. It is developed based on Laravel web

framework. Currently, the web appearance is simple

but quite effective. The server is also prepared for

other resources monitoring such as water

consumption and solar energy monitoring, where the

monitoring system would be developed. The web

display is shown in figure 4. This display could be

accessed from everywhere as long as the internet is

available.

Figure 4: The display on web based server.

Prior the system installed in February 2021, a

preliminary measurement of the link between several

future SN locations and the gateway located at the

laboratory of The Department of Electrical

Engineering is undertaken. The results could be

found in (Batong, 2021).

During the measurement, two LoRa transceivers

are installed on a pole with the height of 6 meters.

Each had their own Slim Jim antenna. A transceiver

worked on the frequency of 433 MHz, while another

work on 915 MHz. Both transmitted packet

periodically. Two LoRa receivers where one worked

on 433 MHz and another worked on 915 MHz are

brought to five locations sequentially to measure

received signal strength indicator (RSSI), signal to

noise ratio (SNR), and packet loss. The result showed

that for both frequencies and all SN locations the

RSSI is above -120 dB. It means that all links are

acceptable to establish LoRa data communication.

In February 2021, five units of SN are installed in

two office buildings, two laboratories, and a

workshop. These buildings are selected due to various

distances and building functions. The different

function is believed would give specific electricity

consumption character that may be different from the

character of other buildings.

Three locations where the SN installed has been

observed in the preliminary measurement mentioned

above. Hence, the result of the preliminary

measurement could be compared with the current

measurement provided by the installed SN. It is with

the note that in the preliminary measurement (Batong,

2021), the gateway transmits packets to the SN, while

in the current situation, the SN transmits packets to

the gateway. The location on both situations is the

same, which is at the laboratory of Electrical

Department.

Furthermore, not all results in (Batong, 2021) will

be used. For example, as the antenna used in SN unit

is omnidirectional antenna, the comparison is only for

this type of antenna. Moreover, since the operation

frequency used in the current system is 915 MHz, to

deal with the Indonesia government regulation, the

comparison is only undertaken for this frequency.

The averaged RSSI and SNR of preliminary result

and the current measurement are shown in Table 1.

Table 1: The RSSI and SNR comparison of preliminary and

current measurement.

Location

RSSI (dB) SNR (dB)

Prelim. Current Prelim. Current

Electrical

Department

Office

-113 -100 7.41 9

UPT

Bahasa

floor

-118 -116 4.08 0.25

Chemical

Engineering

Laboratory

-99 -89 8.41 8.75

A Low Cost Electricity Monitoring Wireless Sensor Network

281

It can be seen that the RSSI measured by the

installed SN is even higher than the RSSI measured

by the preliminary observation. It is possibly due to

the use of metal panel box as the SN cover and the

antenna grounding. Such metal box might reduce

interference from outside the SN unit and improve the

antenna radiation pattern.

Meanwhile, the SNR measured by the installed

SN is commonly higher than the SNR measured in

preliminary observation. The SNR value in UPT

Bahasa that is only 0.25 dB is still reliable as the

minimum limit of SNR according LoRa standard

(LoRa Alliance, 2015) is -20 dB. However, during

heavy rain, packet transmitted from SN in UPT

Bahasa sometimes could not be accepted by the

gateway. Thus, the LoRa transmitter power will be

increased from the default (17 dB) into to the

maximum (20 dB) through the program. Also, the

antenna will be replaced by a higher gain antenna.

Several months after being installed, the

electricity monitoring system could show the

electricity consumption of the buildings. How much

electricity every building uses is clear as shown in

website menu (figure 5), the unit is kWh. By

comparing such electricity consumption with the

electrical equipment specification used in the

building and people activities, it could be calculated

whether the electricity utilization is normal or

wasteful. If it is wasteful, a significant action might

be required to reduce the consumption.

Figure 5: The display of buildings electricity consumption.

Since the active power of every phase is also sent

to the server, the phase equilibrium can also be read

to analyse whether the installation is proper or not.

The display is shown in figure 6. The unit of active

power is in watt, while the three-phase consumption

is in kWh. In Politeknik, improper installation may

occur as additional installation is undertaken without

considering the phase load balance. As such,

installation revision should be provided.

Figure 6: The display of individual phase active power.

iCAST-ES 2021 - International Conference on Applied Science and Technology on Engineering Science

282

4 CONCLUSIONS

It has been built and installed a low cost electricity

monitoring wireless sensor network as a solution to

measure the electricity consumption for every

building in the campus of Politeknik Negeri

Samarinda. Along with the electricity consumption

report required by the Politeknik directorate, the

system also gives the report of phase equilibrium that

is required by electrical experts in Politeknik to revise

and improve the installation. It is necessary to prevent

electrical hazard or installation damage caused by the

imbalance phase load. In brief, the integrated

electricity monitoring system could make electrical

resources utilization and maintenance more

accountable, which is important to achieve the

predicate of green technology campus

ACKNOWLEDGEMENTS

Authors would like to express gratitude and

appreciation to The Director of Politeknik Negeri

Samarinda and staffs for many supports including

moral and financial to make the project realized. The

author would also thank to students who involved in

this research. It could be worthy experience that

improve their self-confidence to create other applied

systems, which is help people to live better.

REFERENCES

Akyildiz I. F., Su W., Sankarasubramaniam Y., Caryici E.

(2002) A survey on sensor networks, In IEEE

Communication Magazine August p 102.

Alfan, Murdiyat P., Gunanto L. H. (2021) Rancang bangun

sensor node untuk sistem monitoring energi listrik

nirkabel pada gedung dalam kampus Politeknik Negeri

Samarinda, In Poligrid vol 2 no 1 p 1.

Batong A. R., Murdiyat P., Kurniawan A. H. (2021)

Analisis kelayakan LoRa untuk jaringan komunikasi

sistem monitoring listrik di Politeknik Negeri

Samarinda, In Poligrid vol 1 no 2 p 55.

Digi Int., (2008) An introduction to WiFi, Digi International

Inc. p 12.

Eyedro (2021) Eyedro business 3-phase energy monitor

(polyphase), online https://www.amazon.com/,

Eyedro-Business-EBEM1-LV-Electricity-Monitor/dp

/B07RPYZCH8?th=1

Kanakaraja P., Sourabh U., Kotamraju S. K., Suneela G. V.

S., Neelesh R. (2021) Design and implementation of

smart energy meter using LoRaWAN and IoT

applications, In Journal of Physics:Conference Series

1804(2021).

LoRa Alliance (2015) LoraWAN, what is it? A technical

overview of LoRa® and LoRaWAN™", Whitepaper,

LoRa Alliance, San Ramon California

Mithya V, Kosalya M., Madhumathi P. M. Manimegalai G.,

Ramya P. (2019) Intelligent energy meter using GSM

modem with Arduino, In International Journal of

Innovative Technology and Exploring Engineering vol.

8 issued 6S.

Pasurono, Handoko S., Setyawan I. (2013) Perancangan

kWh meter digital menggunakan kWh meter

konvensional, In Transient vol 2 no 1 p 113.

PT. PLN (2021). Q&A tarif listrik, online https://web.pln.

co.id/pelanggan/qa-tarif-listrik.

Rajput U. A., Rafique K., Saand A. S., Shaikh M., Tarique

M (2018) Modelling of Arduin-based prepaid energy

meter using GSM Technology, In International Journal

of Advanced Computer Science and Application vol. 9

no. 5.

The Things Network (2020) LoRa world record broken:

832 km/517 mi using 25 mW, https://www.

thethingsnetwork.org/article/lorawan-world-record-

broken-twice-in-single-experiment-1

Wooley M. (2020) Bluetooth 5 go faster go further

(Bluetooth.com) p 11.

A Low Cost Electricity Monitoring Wireless Sensor Network

283