Proc.

XXVI International Scientific Conference Electronics - ET2017, September 13 - 15, 2017, Sozopol, Bulgaria

Energy Saving IoT-Based Advanced Load Limiter

Basri Kul and Mehmet Şen
Department of Electronics,
Vocational School of Technical Sciences, Uludağ University
Bursa, Turkey
{basrikul, mehmetsen}@uludag.edu.tr

Abstract ‒ A load limiter is particularly needed in places like was developed via cloud technology and in this way, energy
microgrid business centers because the load limit that the management access was provided over the Internet.
microgrid structure can handle must not be exceeded. This study A number of relevant studies carried out on the subject are
aimed at keeping load variation under control by planning briefly summarized here.
daytime load distribution using IoT-based energy measurement
and central management software. By using IP-based smart
[1] Researchers demonstrated a grid structure that can be
meters and control methods to perform energy measurements, all remotely controlled with real-time communication speed in a
loads on the microgrid can be loaded in a balanced manner from smart grid. This system is based on extensive optimization
the centralized software. Moreover, load balance across the grid methods of the smart grid over existing network structures.
can be protected by shutting off the relevant overcurrent devices. Experimental results on measurement control via the smart grid
By setting the load limits for certain zones at specific times of the with different communication protocols were emphasized.
day, both the electrical devices and the microgrid are safeguarded. [2] This study dealt with the detection on the grid of voltage,
In heating and cooling applications, when all components on the phase, and frequency faults beyond the specified tolerances, and
grid are overloading the microgrid, other electrical appliances are
also examined sector-based control methods.
prevented from exceeding the permitted current limits. Thus,
installation of the system results in energy saving and reduced [3] This work explored the use of technologies such as IoT,
costs. Details are shown in this study. ZigBee and networking smart things in intelligent sensors and
equipment and integration with their web services and cloud
Keywords ‒ IP-based smart sensors, smart energy, load limiter, services, with emphasis on feasibility and energy savings.
microgrid, smart grid, [4] This research was concerned with algorithms and
methods for energy measurement via CNP protocol, control of
load distributions and automatic implementation of microgrid
I. INTRODUCTION structures such as those of solar energy and wind energy.
[5] This work involved ZigBee protocol allowing remote
With the progress of technology a rapid increase in energy energy measurement and control of home appliances. Access
demands is emerging. When examining recent statistics, it can was provided via mobile phone using GSM.
be seen that renewable energy sources will not be able to meet [6] Researchers developed mathematical models by Linear
energy requirements. This has led us to seek new solutions to Programming and indicated that in this way they had reduced
save our energy and energy resources from unnecessary energy consumption by 20% in the simulation model.
consumption. A priority of this concerned research was to [7] This study reported the effects of the total daily
prevent unnecessary power consumption of home and office consumption of electrical appliances, and emphasized that
equipment. This led to the idea of a controllable socket and rapid energy savings could be achieved by raising the awareness of
development of these smart sockets has since begun and is the users.
ongoing. Furthermore, in order to reduce the microgrid [8] Efforts were made to further reduce stand-by energy
installation costs in microgrid applications, it is necessary to consumption by combining ZigBee-based sensors with
keep the devices on or off at a certain time, so as to monitor the environmental monitoring, energy measurements and home
load consumption of electrical appliances online, and to load the automation systems.
system in a balanced manner to prevent overload. It is also [9] This study reported energy savings of ~ 7% as a result
necessary to determine total energy policies. of planning the working times and priorities of electrical
In this study, total energy consumption control was achieved appliances.
by preventing devices from overloading in the workplace [10] This study attempted to determine energy consumption
application. The devices were allowed to operate at different profiles without a central server by developing intelligent
times by dividing them into time slots. The sleep mode of the sockets that communicated with each other in order to measure
devices was then detected and energy consumption completely and control energy.
eliminated. In addition, integrated energy management software [11] Using ZigBee protocol, researchers developed smart
sockets that could measure and control the energy of household
 Proc. XXVI International Scientific Conference Electronics - ET2017, September 13 - 15, 2017, Sozopol, Bulgaria

electrical appliances by gaining access to smart home In this study, overcurrent devices are controlled by central
automation over the cloud. This work is similar to the present software and are prevented from consuming more power than
work, with a few important differences. The aim of the current specified for the time slots. When the overcurrent devices begin
study was to reduce energy consumption, especially in factories, to draw more current than the specified limit, they are shut down
business centers or places where local energy can be produced. by the system and in this way both the devices and the network
For this reason, the energy management planning was based on are protected from overloading. In addition, in order to
offices or machines. In addition, the high-current and stand-by determine and eliminate the stand-by energy, the device is
management, the conditions for the load to be active, and the disconnected from the central software via the cloud. If the user
timings were completely user determined. wants to use a device in stand-by mode again, by using a mobile
[12] This study involved socket-based energy measurement telephone or the internet to log into the cloud, it is possible to
and control and overload warning in smart home systems using restart any device. Specifically, the use of these applications
Power Line Communication (PLC). began in textile factories and microgrid business centers. The
[13] This is another work on IoT-based energy measurement textile mills that use this system can calculate unit costs per
and management. meter of fabric produced by the factory.
[14] Researchers presented information on increased In this study, the ADE7754 integration was used for energy
energy requirements and traditional measurement methods. In measurements and the current transformer was used to
addition, they demonstrated the overall advantages and accurately measure low currents, while the Hall-effect sensor
disadvantages of increasing measurement and control strategies. was used for measuring high currents. The current transformer
[15] This last work is about the provision of a wide range of used for low currents was connected to one of the ADE7754 di
management support, including centralized management / dt inputs. The Hall-effect sensor was connected to the di / dt
support equipped with an IP-based power distribution unit input of the other phase input of the ADE7754. Although the
having energy measurement and control capabilities and ADE7754 energy recovery integration is a 3-phase energy
configured to eliminate stand-by energy consumption. It points meter, in this study a single phase was applied to two phase
out that savings can be achieved by eliminating stand-by energy inputs. In this way, one made energy measurements of low
consumption, especially in banks and public institutions. currents, while the other phase input provided energy
measurements of high currents.
In addition to high-current control and IoT-based energy In this study, the NXP1769 cortex-m3 32-bit processor
measurement, the present study aimed to achieve greater energy having a built-in Ethernet interface was used. The overall
saving by monitoring the very low stand-by currents. A total structure of the system is shown in Figure 1.
energy management plan with cloud integration can provide
energy measurement and control and at the same time, achieve
savings in business centers having a microgrid or a large amount
of electrical equipment.

II. MATERIAL AND METHODS

Electrical appliances exhibit different energy consumption

rates according to their different operating modes. The new
generation devices using switch-mode power supplies display a
sudden increase in high energy consumption when switched on.
When a large number of devices with switch-mode power
supplies on one network enter the circuit at the same time, this
causes a very high current to be drawn on the mains. For this
reason, it is necessary for these types of devices, which are
located on a microgrid or on a local network, to switch in
sequence. Devices with increased energy consumption in
different modes will also cause high-current draw on the
network when switching to high-current draw modes at the same
time. When possible, these should be prevented from
overloading the network by not working in these modes at the Fig. 1. Structure of IoT-based system
same time. Thus, in this work, each electric device was provided
with an IoT-based energy measurement and control system and Figure 2 shows the hardware architecture of the designed IoT-
is controlled by centralized software via the cloud. Moreover, in MC unit.
this study, power limits and operating times and modes of each
designed IoT device controlled by the cloud software are
determined, measured and controlled. The IoT- energy meter
and control (IoT-MC) measures the active, reactive, Cos-Fi and
total energy consumption of each device, and the total energy
consumption and energy profiles of the electrical devices on the
system are measured and updated. This method could be used in
the same way to determine user profiles.
 Proc. XXVI International Scientific Conference Electronics - ET2017, September 13 - 15, 2017, Sozopol, Bulgaria

ACS711 and the current transformer during calibration. During

calibration, one of the current inputs of the ADE7754 was
calibrated by short-circuiting. The current sensors used are
shown in Figures 3a and 3b.

Fig. 2. Hardware design structure

Fig. 3a. ACS711 current connection diagram

III. WORKING ALGORITHM OF THE SYSTEM

The NXP LPC1769 with an Ethernet layer was used as the main
processor. The developed embedded software included Web
server, SNMP, Email, Telnet, SMTP, 320 × 240 16 million color
graphics LCD, and communication software with ADE7754.
The central server has a dual Ethernet output. One of them
continuously scans the IoT-MC network to gather energy
metrics with the SNMP protocol and the other opens the TCP / Fig. 3b. SCT-013-005 split core current transformer
IP socket on the other Ethernet output and sends it to the cloud.
In this way, all energy measurements are updated on the cloud.
The software on the cloud sends the necessary commands to the The IoT-MC analyzes a sudden current surge as a transient
central server according to the current energy data. On / off current or continuous current of the other devices in the circuit.
operations are controlled by sending the commands from the For this purpose, as seen in Figure 4, it waits for the surplus
cloud to the IoT-MC units on the local IP network via the central increase of the current to reach stability. If the rising current
server. The software on the cloud manages the energy value is continuous, unlike peak current or an increase in current
management policy required for the entire system architecture. due to the input capacitance of the switching device, and the
The management console and all other user actions are previously defined current limit has been exceeded, the energy
implemented via the cloud. The software on the cloud governs consuming device is then turned off and the user is informed by
the hierarchical organization of energy distribution, stand-by the overcurrent warning via the web interface.
management, settings of current limits, and operations to be
performed when the current limit is exceeded. In this way,
energy management of the entire structure is ensured. When a
computer is connected to the IoT-MC unit, a WOL protocol is
added in order to send a power-on command over the network
when the power is re-applied.
Energy calculations made with the ADE7754 integration are
provided by a single processor. The IoT-MC consumes up to
0.8W / h of power when the computation and Ethernet protocols
are active, and when inactive, they consume a maximum of
0.6W / h of energy.
In order to accurately measure low currents at the hardware Fig. 4. Current on capacitive loading
level, the SCT-013-005 current transformer with a maximum
measurement capacity of 8 A was used. This current transformer
measures currents from 3 mA (~ 0.7 Watt) to -5 A with 0.2% Different current limits were set from the management
measurement accuracy. This is enough to detect the minimum console via the cloud for different times of the day. Accordingly,
current required for all devices. Current measurements of up to the software running on the cloud provided total energy
30 A were carried out by the ACS711. The ADE7754 integrates management of load limits and energy consumption in other
a three-phase energy measurement with a reference transformer devices,
to reduce mains voltage and provide isolation at the same time. A part of the schematic belonging to the design is shown in
The output of the reference transformer was applied to the two Figure 5 and the PCB artwork is presented in Figure 6.
voltage inputs of the ADE7754. In this case, instead of three
phases, one phase was measured. Since the ADE7754 requires
calibration, a separate calibration sequence is required for the
 Proc. XXVI International Scientific Conference Electronics - ET2017, September 13 - 15, 2017, Sozopol, Bulgaria

Fig. 5. Part of design schematic Fig. 9. Management console interface on cloud

As the flow diagram of the script running on the cloud is very

extensive, only the screen output is shown in Figures 8 and 9.

Fig. 6. PCB artwork

Figure 7 shows the application of the IoT-MC.

Fig. 10. Hall-effect current sensor implementation on bus bar

IV. EXPERIMENTAL STUDY

In this study, the Hall-effect and current transformer were
Fig. 7. Finalized IoT-MC combined to provide low-power stand-by detection, load
limiting and total energy management at higher powers.
The central server scanned the IoT meter and control units
on SNMP at approximately 100-ms intervals. It took about 100
ms for the central server to update the data in the cloud. Energy
management was achieved with an average interval of 100 ms,
and this speed was found to be sufficient to control the load
balance.
In a construction with bus bar architecture, even when it is
not possible to connect the Hall-effect sensor in series, as shown
in Figure 10, a structure that can be placed on the bus bar is
needed. The MLX91208 can be used for this purpose.

Fig. 8. Responsive web interface

 Proc. XXVI International Scientific Conference Electronics - ET2017, September 13 - 15, 2017, Sozopol, Bulgaria

V. CONCLUSION [14] E. O'Driscoll, G. E. O'Donnell. "Industrial power and energy

metering–a state-of-the-art review." Journal of Cleaner Production 41
(2013): 53-64.
In this study, by measuring the energy consumption of 120 [15] B. Kul, M. Şen, K. Kaş. "IP based smart energy metering with
textile factory looms in inactive and activated states, the energy saving." Scientific Conference Electronics (ET),
developed system was used to calculate the energy costs of the International. IEEE, 2016.
enterprise according to the product type. In addition, the study
was able to determine the operations during which the reactive
components of the machines were in effect.
A similar architecture previously developed by the authors
[15], has been applied to a building within Uludag University in
Bursa, Turkey. The end result of the application was a monthly
energy saving of ~ 22%.

REFERENCES

[1] Z. Fan et al. "Smart grid communications: Overview of research

challenges, solutions, and standardization activities." IEEE
Communications Surveys & Tutorials 15.1 (2013): 21-38.
[2] Y. Seyedi, H. Karimi. "Coordinated Protection and Control Based
on Synchrophasor Data Processing in Smart Distribution
Networks." IEEE Transactions on Power Systems (2017).
[3] I.M. Moreno-Garcia et al. "Development and Application of a
Smart Grid Test Bench." Journal of Cleaner Production (2017).
[4] H.-M. Kim, Y. Lim, T. Kinoshita. "An intelligent multiagent system
for autonomous microgrid operation." Energies 5.9 (2012): 3347-
3362.
[5] N. Javaid et al. "Monitoring and controlling power using zigbee
communications." Broadband, Wireless Computing,
Communication and Applications (BWCCA), 2012 Seventh
International Conference on. IEEE, 2012.
[6] M.C. Bozchalui et al. "Optimal operation of residential energy hubs
in smart grids." IEEE Transactions on Smart Grid 3.4 (2012):
1755-1766.
[7] S. Barker et al. "Smartcap: Flattening peak electricity demand in
smart homes." Pervasive Computing and Communications
(PerCom), 2012 IEEE International Conference on. IEEE, 2012.
[8] J. Byun et al. "An intelligent self-adjusting sensor for smart home
services based on ZigBee communications." IEEE Transactions on
Consumer Electronics 58.3 (2012).
[9] J. Byun, I. Hong, S. Park. "Intelligent cloud home energy
management system using household appliance priority based
scheduling based on prediction of renewable energy
capability." IEEE Transactions on Consumer Electronics 58.4
(2012).
[10] A. L. Lentine, J. R. Ford, J. R. Finn et al. “An intelligent
electrical outlet for autonomous load control for electric power grids
with a large percentage of renewable resources,” in Proceedings of
the IEEE Power and Energy Society General Meeting (PES ’12), pp.
1–8, IEEE Power & Energy Society, San Diego, CA, USA, July
2012.
[11] L.-J. Kau, B.-L. Dai, C.-S. Chen, S.-H. Chen. “A cloud network-
based power management technology for smart home systems,” in
Proceedings of the IEEE International Conference on Systems, Man,
and Cybernetics (SMC ’12), pp. 2527–2532, IEEE, Seoul, Republic
of Korea, October 2012.
[12] G. Horvat, D. Vinko, D. Zagar. “Household power outlet
overload protection and monitoring using cost effective embedded
solution,” in Proceedings of the IEEE Mediterranean Conference
on Embedded Computing (MECO ’13), Budva, Montenegro, 2013.
[13] C.-F. Lai et al. "Integration of IoT energy management system
with appliance and activity recognition." Green Computing and
Communications (GreenCom), 2012 IEEE International Conference
on. IEEE, 2012.