DEVELOPMENT OF DEMAND SIDE MANAGEMENT SYSTEM:

DIAGNOSTIC TOOL AND LOAD RECOGNITION

KONG KIE MING

A project report submitted in partial fulfilment of the

requirements for the award of Bachelor of Engineering
(Hons.) Electrical & Electronic Engineering

Faculty of Engineering and Science

Universiti Tunku Abdul Rahman

April 2013
 ii

DECLARATION

I hereby declare that this project report is based on my original work except for
citations and quotations which have been duly acknowledged. I also declare that it
has not been previously and concurrently submitted for any other degree or award at
UTAR or other institutions.

Signature : _________________________

Name : KONG KIE MING

ID No. : 09UEB07437

Date : _________________________
 iii

APPROVAL FOR SUBMISSION

I certify that this project report entitled DEVELOPMENT OF DEMAND SIDE

MANAGEMENT SYSTEM: DIAGNOSTIC TOOL AND LOAD
RECOGNITION was prepared by KONG KIE MING has met the required
standard for submission in partial fulfilment of the requirements for the award of
Bachelor of Engineering (Hons.) Electrical & Electronic Engineering at Universiti
Tunku Abdul Rahman.

Approved by,

Signature : _________________________

Supervisor : Mr Chua Kein Huat

Date : _________________________
 iv

The copyright of this report belongs to the author under the terms of the
copyright Act 1987 as qualified by Intellectual Property Policy of University Tunku
Abdul Rahman. Due acknowledgement shall always be made of the use of any
material contained in, or derived from, this report.

2012, Kong Kie Ming. All right reserved.

 v

Specially dedicated to
my beloved family and friends
supervisor Mr Chua Kein Huat
and project partners Mr Pang Yap Seng and Ms Lim Khim Yan
 vi

ACKNOWLEDGEMENTS

I would like to thank everyone who had contributed to the successful completion of
this project. I would like to express my gratitude to my research supervisor, Mr
Chua Kein Huat for his invaluable advice, guidance and his enormous patience
throughout the development of the research.

In addition, I would also like to express my gratitude to my loving parent and

friends who had helped and given me encouragement throughout the research.
Without their encouragement and support, this project would not have materialized. I
am grateful for their unwavering support and guidance.

Finally, I take immense pleasure in thanking my project partners, Mr Pang

Yap Seng and Ms Lim Khim Yan who gave me the opportunity to work and learn
together as a team in the development of research. Their effort and cooperation was
the vital for the success of the project.
 vii

DEVELOPMENT OF DEMAND SIDE MANAGEMENT SYSTEM:

DIAGNOSTIC TOOL AND LOAD RECOGNITION

ABSTRACT

Electrical supply and demand must remain in balance at all time to ensure the
stability on the electrical grid. Demand side management is the ability to maximize
the energy usage at the side of consumption from limited electrical supply. It is the
modification of customer demand for energy through various methods including
energy efficiency, energy controllers, demand response and spinning reserve. Instead
of adding more generation to the system, demand side management monitor, reduce
and regulate the energy consumption of the users. Smart meter is one of the
incentives to improve energy monitoring. However, the smart meter is limited to the
energy consumption of a house level. The main challenge is to acquire appliances-
level information. The load data obtained is extremely valuable to energy auditors,
utilities, public policy makers, and appliance manufacturers, for a broad range of
purpose. For example, they can determine the load characteristic of a particular
electrical appliance and control it in a way which can reduce maximum energy usage
when more loads are added. In this project, a diagnostic tool which is capable of
measuring and analyzing the basic parameters of a load is developed. The basic
parameters include voltage, current, power factor and so on. The programming is
done using LabVIEW 2011 by National Instrument. An algorithm is also developed
for load recognition. Lastly, the implementation of energy storage system mitigated
the problem of voltage unbalance across the three phase network.
 viii

TABLE OF CONTENTS

DECLARATION ii
APPROVAL FOR SUBMISSION iii
ACKNOWLEDGEMENTS vi
ABSTRACT vii
TABLE OF CONTENTS viii
LIST OF TABLES xi
LIST OF FIGURES xii
LIST OF SYMBOLS / ABBREVIATIONS xv
LIST OF APPENDICES xvi

CHAPTER

1 INTRODUCTION 1
1.1 Background 1
1.2 Problem Statement 2
1.3 Aims and Objectives 2
1.4 Project Scope 3

2 LITERATURE REVIEW 4
2.1 LabVIEW 4
2.1.1 LabVIEW FPGA 5
2.2 Introduction to Smart Grid 6
2.2.1 Smart Grid Evolution 7
2.2.2 Smart Microgrid 9
2.2.3 Microgrid Topology 9
 ix

2.3 Demand Side Management 10

2.3.1 Energy Efficiency 11
2.3.2 Energy Controllers 12
2.3.3 Demand Response 14
2.3.4 Spinning Reserve 15
2.4 Load Recognition 16
2.4.1 Non-Intrusive Appliance Load Monitoring 16
2.4.2 Electromagnetic Transient Program 18
2.5 Voltage Unbalance 20
2.5.1 Voltage Unbalance Factor 21
2.5.2 Effects of Voltage Unbalance 22

3 METHODOLOGY 23
3.1 Overview 23
3.2 Concept of Diagnostic Tool 25
3.2.1 Prototype 1 27
3.2.2 Arrangement 27
3.2.3 Components 29
3.2.4 Prototype 2 32
3.2.5 Load Recognition 33
3.3 Voltage Unbalance Mitigation Experimental Setup 36
3.3.1 Measurement and Automation System 37
3.3.2 Transmission Line Model 38
3.3.3 Controllable Load Bank 39
3.3.4 Sunny Island 5048 40
3.4 Graphical User Interface 41

4 RESULTS AND DISCUSSIONS 42

4.1 Comparison of Prototype 1 and 2 42
4.2 Functions of Diagnostic Tool 44
4.2.1 Load Recognition at Idle State 47
4.2.2 Hair Blower 48
4.2.3 Water Boiler 52
 x

4.2.4 Load Characteristic 56

4.2.5 Voltage Unbalance Mitigation 58

5 CONCLUSION AND RECOMMENDATIONS 61

5.1 Conclusion 61
5.2 Future Improvement 62

REFERENCES 63

APPENDICES 65
 xi

LIST OF TABLES

TABLE TITLE PAGE

3.1 Specification of Prototype 1 26

3.2 Power Supply NES-100-24 29

3.3 Contactor GC2530M5 30

3.4 Current Transformer 100/5A 31

4.1 Comparison between Prototype 1 and 2 43

 xii

LIST OF FIGURES

FIGURE TITLE PAGE

2.1 LabVIEW 2012 (National Instrument 2012) 5

2.2 Categories of Demand Side Management 10

2.3 Impact of improved energy efficiency versus

demand response 11

2.4 An energy controller switches devices 12

2.5 Selection of priorities in a maximum demand

monitor 13

2.6 Existing Demand Response scheduling process 14

2.7 Allocation of capacity of generation unit 15

2.8 NALM devices 17

2.9 Load Characteristic Graph 17

2.10 Adaptive algorithm for load recognition 19

2.11 A balanced system 20

2.12 An unbalanced system 20

2.13 Symmetrical components for unbalanced system

of voltages 21

3.1 Experimental setup of Smart Energy Conditioning

System 24

3.2 Process Flow of Load Diagnosis 25

3.3 Schematic Diagram of Diagnostic Tool 26

3.4 1mm2 Plywood and 1mm2 Aluminium 27

 xiii

3.5 Front Arrangement 28

3.6 Inside Arrangement 28

3.7 Power Supply 29

3.8 Contactor 30

3.9 Current Transformer 31

3.10 Schematic Diagram of Prototype 2 32

3.11 Flowchart of Load Recognition 33

3.12 Electrical Power Tools (Convert to Waveform.vi) 34

3.13 Block Diagram for Load Recognition 35

3.14 Experimental setup of Energy Conditioning

System 36

3.15 sbRIO-9632XT with C Series Module 37

3.16 Transmission line model 38

3.17 Controllable Load Bank 39

3.18 Sunny Island 5048 40

3.19 Load Recognition UI 41

3.20 Energy Conditioning System UI 41

4.1 Prototype 1 42

4.2 Prototype 2 43

4.3 Diagnostic Tool Connected with Hair Blower 44

4.4 Voltage Waveform 45

4.5 Current Waveform 45

4.6 Power Waveform 46

4.7 Energy Waveform 46

4.8 Idle State 1 47

4.9 Idle State 2 47

 xiv

4.10 Hair Blower (Inductive) 48

4.11 Hair Blower (Resistive) 49

4.12 Power Waveform of Hair Blower 49

4.13 Write to Spreadsheet File.vi 50

4.14 Current vs Time (Hair Blower) 50

4.15 Power vs Time (Hair Blower) 51

4.16 Database 51

4.17 Voltage Waveform of Water Boiler 52

4.18 Current Waveform of Water Boiler 52

4.19 Power Waveform of Water Boiler 53

4.20 Current vs Time (Water Boiler) 54

4.21 Power vs Time (Water Boiler) 54

4.22 Creating New Data 55

4.23 Load Recognition User Interface 55

4.24 Load Characteristic 56

4.25 Combination of Load Characteristic 57

4.26 VUF with balanced load 58

4.27 VUF with unbalanced load at Phase A 59

4.28 VUF with Load and Sunny Island at Phase A 60

 xv

LIST OF SYMBOLS / ABBREVIATIONS

h height, m
I Current, A
Resistance, Ohm
P Power, W
s Time, s
T temperature, C
V Voltage, V

AMI automated meter reading

AMR advance metering infrastructure
FPGA field-programmable gate arrays
GUI graphical user interface
LV low voltage
NC normally close
NALM non-intrusive appliance load monitoring
NO normally open
VI virtual instrument
VUF voltage unbalance factor
UI user interface
UPS uninterruptible power supply
 xvi

LIST OF APPENDICES

APPENDIX TITLE PAGE

A Graphical Programming (LabVIEW) 65

 1

CHAPTER 1

1 INTRODUCTION

1.1 Background

The utility industries are facing numerous challenges which include fuel
diversification, optimizing the deployment of expensive assets, increasing power
demand, improving energy conservation and reduction of greenhouse gases (K.
Moslehi 2008). Engineers are required to focus on sustainable development which
can meet present needs without compromising future generation needs. In power
transmission and distribution, the electrical network has become complicated and
difficult to manage. In case of any breakdown, this catastrophic effect will cause
tremendous losses to the nation. Hence, the electrical energy system must be able to
provide a reliable and stable system which can tackle various faults.

The existing electric energy system is unidirectional and up-down oriented.

Studies are focused on the Low Voltage (LV) distribution due to the complexity in
managing different kind of load. The loads are the culprit which invokes the
necessary to perform diagnosis upon them. Different parameters obtained from the
load are analysed to sort out the potential predicament that may occur in the future.
In some situation, the problem cause by the load does not take into consideration
such as voltage unbalance, voltage rise or drop and network power loss.
 2

1.2 Problem Statement

The diagnostic system is found at the different stages from generation, transmission
to distribution. Power meter reading device is installed by utilities to monitor the
monthly usage of the household and implied charges base on the total usage. This
simple meter can only shows the total usage but not the individual devices. It is also
attached to distribution board and difficult to remove. Whenever the users want to
know about the energy consumption of certain devices, they can only based on the
specifications data without knowing the true condition of the device.

Furthermore, the voltage unbalance is a problem usually overlooked in the

LV distribution network. In factories which operate numbers of machine, the voltage
unbalance has a great effect towards induction motor such as overheating, reduce
lifespan and power loss in long run. According the National Electrical Manufacturers
Association (NEMA), 1% increase of voltage unbalance factor can cause 6% to 10%
temperature rise. This indirectly creates economic loss as the machinery with
induction motor has to be replaced more often.

1.3 Aims and Objectives

This project aims to develop a portable diagnostic tool and solve voltage unbalance
in the LV distribution network. A graphical interface controlling and monitoring
system is developed using a graphical programming language LabVIEW. In interest
of this objective, some tasks are set to be fulfilled as following:

To use the graphical programming language LabVIEW and NI FPGA

To design and construct the portable diagnosis tool
To apply present method for demand side management
To develop an algorithm to perform load recognition
To implement the Sunny Island 5048 which is a bidirectional inverter into the
system for voltage mitigation
 3

1.4 Project Scope

In this project, the aim is to build a diagnostic tool which contains the measurement
and automation system. The diagnostic tool is the integration of both hardware and
software to measure, monitor and control the system. This device must be portable
and easy to carry around. Besides, load recognition and voltage unbalance mitigation
are the additional functions of the device besides doing measurement on the LV
distribution network. The first part of the project is to construct the diagnostic tool.
The design of the structure must be able to cover all the necessary electrical and
electronic components. The protection design specification is followed to ensure the
safety of the device. In the second part of the project, it focuses on the programming
using LabVIEW. A Graphical User Interface (GUI) is constructed which allows the
user to monitor the load connected to it. The devices must also be able to perform
load recognition with the help of software implementation. When Sunny Island 5048
is added in to mitigate the voltage unbalance, the program must be able to perform
the task flawlessly.
 4

CHAPTER 2

2 LITERATURE REVIEW

2.1 LabVIEW

Over the years, numerous kinds of programming language have been developed to
ease the programmers in developing software applications. LabVIEW (Laboratory
Virtual Instrumentation Engineering Workbench) is a powerful programming tool
developed by National Instrument which helps to increase the efficiency and
productivity in solving todays problem. This software is unique compare to other
programming tools as it is a graphically-based programming language. For engineers,
it is used for testing, debugging, optimizing and so on from small to large systems.
(Product Information: What is NI LabVIEW 2011)

The simplicity and convenience provided by LabVIEW when constructing

and programming is by the Virtual Instrument (VI). A VI consists of the front panel,
block diagram, and icons that represent the program. The purpose of the front panel
is to show the indicator and allow users to control. Users can either input data or
collect data from a virtual instrument which is running. The block diagram is the
complete code for the whole VI, which serve as the central core of the programme.
Lastly, the icon which is the visual representation of the VI, has connectors for
program inputs and outputs.
 5

Figure 2.1: LabVIEW 2012 (National Instrument 2012)

2.1.1 National Instrument FPGA

Field-programmable gate arrays (FPGAs) are made up of reprogrammable silicon

chips. By programming an FPGA, it actually rewires the chip itself to implement the
user functionality rather than running a software application. It is targeted on NI
reconfigurable I/O hardware. LabVIEW is well suited for FPGA programming
because it clearly represents parallelism and data flow.

The top 5 benefits of using FPGAs are as below:

Faster I/O response times and specialized functionality.

Exceeding the computer power of digital signal processors.
Rapid prototyping and verification without the fabrication process of custom
application-specific integrated circuit (ASIC) design.
Implementing custom functionality with the reliability of dedicated
deterministic hardware.
Field-upgradable eliminating the expenses of custom ASIC re-design and
maintenance.
 6

2.2 Introduction to Smart Grid

The new generation electricity grid, also known as Smart grid is no longer a new
term in power industries. As the electrical system and its components are getting
much more complicated, there is a need to provide the utility companies with full
visibility and pervasive control over their system. Utilities need to introduce
distributed intelligence into their existing infrastructure to make them more reliable,
efficient and capable of exploiting and integrating diverse sources of energy (S. J.
Anders 2007). The three significant issues which the electric power industries
throughout the world are facing including an aging transmission system, the need for
a regulatory framework and an aging workforce (P. Mazza 2005).

The aging transmission system can increase the risk of failure during
operation. All the components of the electrical system included in generation,
transmission and distribution will degrade over a period of time. While all the
utilities have their own maintenance plans for this system, ever-increasing congestion
levels in many areas are making it increasingly difficult to schedule circuit outages
for routine upgrades. If this is not done well, the impact on utilities in terms of
reliability and asset replacement will be significant.

Investment on the electricity infrastructure has been declining over the years.
There is a need for a regulatory framework that will encourage more independent
investment, ownership, and management on the electrical grid. For the areas which
already have restructured framework shows significant increase in funding to
develop a new and advance infrastructure to meet the present demand. The aging
workforce required a succession plan to ensure the present of next generation of
sophisticated workers and engineers in the industry.

Advancement in technology will develop new approaches to improve the

functionality and flexibility of the existing electrical system. In other word, smart
grid can overcome the shortcomings of the existing grid. Smart grid is more likely to
follow an evolutionary trajectory than to involve a drastic overhaul. They will be
functioning side-by-sides with existing electricity grid. However, the maturity of the
 7

smart grid over time will cause the shifting from old grid to the new grid to improve
and enhance their critical services.

One of the common problems faced by most utility companies across the
globe is how they can meet the present demand as soon as possible, with minimum
cost, and without jeopardizing the critical services they are currently providing. They
should come out with strategies which they can achieve the highest return on
investment for such undertaking. For any new technology breakthrough, there is a
clear advantage for the utility in developing countries to implement it without the
need for backward compatibility with their existing network.

The existing electricity grid hierarchy system is arranged from top-bottom in

such a way, starting from central generation, transmission system, network of
substations, distribution network and finally customer loads. Above the customer
loads, the system is centralized and can control with basic data network. When it
reaches the customer loads itself, being passive operation, it has no data network.
The system is one-way pipeline with no real-time information about the service
parameter at the customers side. Hence, the grid is often over-engineered so that it
can withstand the maximum anticipated peak demand which does not occur
frequently, causing the system to be inefficient. The system stability is also
questioned with any unforeseen surge in demand or anomalies across the distribution
network. In the case of any component failures, it can trigger catastrophic blackouts
(Amin and Wollenberg 2005).

2.2.1 Smart Grid Evolution

In order to facilitate troubleshooting and protect the expensive upstream assets,

various levels of command-and-control functions are introduced. The most widely
used system is the Supervisory Control and Data Acquisition (SCADA). It gives
limited control for upstream assets while the distribution network remains outside
their real-time control.

From statistical report, nearly 90% of all power outages and breakdown have
their roots in the distribution network (Moore and McDonnell 2007). The distribution
 8

network is the main focus to move toward the smart grid. Due to the increasing cost
of fossil fuel, the utility companies choose not to expand their generation capacity
with rising demand for electricity. So, there is a need to modernize the distribution
network that can help with demand-side management and revenue protection.

In fact, the metering side of the distribution system has been the focus of
most recent infrastructure investments. The earliest type of meter is the
electromechanical meter which has extremely limited functionality. It has a lot of
mechanical moving parts which easily degraded over time. After that, the Automated
Meter Reading (AMR) is introduced. It allows the utilities to read the consumption
records, alarms, and status from customers premises remotely. As it is one-way
communication system, the AMR does not address the key issue they need to solve:
demand-side management. The system can only restricted to reading meter data. The
utilities companies realise that the AMR system do not allow the transition to the
smart grid where fully control at all levels is a basic premise. The utilities across the
globe moved towards Advanced Metering Infrastructure (AMI) instead of investing
on AMR. AMI is two-way communication system which allows the utilities to
monitor and well as control the service-level parameter. AMI is clearly the next step
for transition to the smart grid. Smart grid needs to leverage the AMI infrastructure
and implement its distributed command-and-control strategies over the AMI
backbone (A. Vojdani 2008).

The pervasive control and intelligence that embodies the smart grid has to
reside across all geographies, components, and functions of the system. The smart
grid is a grid that consists a wide variety of generation options. It allows consumers
to have interaction with the energy management system by adjusting their energy use
and reducing their energy cost. The evolution of technologies also leads to the
realization of self-healing system in smart grid. A self-healing grid is expected to
respond to threats, mal-function of equipment and other destabilizing influences by
preventing or containing the spread of disturbances. Since the smart grid requires
continuous improvement, Information & Technology is used to optimize the grid.
Thus, it is not a replacement for existing electricity grid but a complement to it
(Moore and McDonnell, 2007).
 9

2.2.2 Smart Microgrid

In smart micro grid, it is expected that some of the distributed generators and loads
can be arbitrarily plugged-in or plugged-out from the microgrid. It is known as the
plug and play operation. It consists of interconnected networks of distributed energy
system such as loads and resources that can carry out their normal operation whether
they are connected to or separated from the electricity grid. In such a plug and play
approach, different generators, loads and energy-storage elements are plugged-in or
plugged-out arbitrarily. The common AC-bus voltage magnitude and frequency have
to be stabilized in the absence of any sort of communication among the elements of
the microgrid.

2.2.3 Microgrid Topology

A microgrid can operate both in on-grid as well as islanded modes. During the
islanding mode, the distribution generator should be disconnected from the network
when the main electric grid continues to energize the distribution line. It incorporates
power plant capable of meeting local demand as well as feeding the unused energy
back to the electricity grid. It services a variety of loads, including residential, office
and industrial loads. It makes use of local and distributed power storage capability to
smooth out the intermittent performance of renewable energy sources. It incorporates
smart meter and sensors. The communication infrastructures enable system
components to exchange information and commands securely and reliably.
 10

2.3 Demand Side Management

Demand side management is a portfolio of measures to improve the energy system at

the side of consumption. It can be done in various ways such as improving energy
efficiency with the use of better materials, smart energy tariffs with incentives for
certain consumption patterns and sophisticated real-time control of distributed energy
resources.

The ability of the grid to sustain the increasing demand will soon face its
limit. The demand side management is one way to stretch these limits a bit further.
The main advantage is that it is require lower capital cost to influence the load than
to build a new power plant or install some electric storage device. The demand side
management is moving toward the customer driven rather than utility driven.
Depending on the timing and the impact of the applied measures on the customer
process, it can be categorized into four main points which is Energy Efficiency, Time
of Use, Demand Response and Spinning Reserve as shown in Figure 2.2 (Palensky
and Dietrich 2011).

Figure 2.2: Categories of Demand Side Management

The dynamic demand side management does not only reduce energy
consumption, the consumption patterns are also influenced. In an ideal case, where
energy efficiency takes place, the consumption pattern is lower than the original
consumption. When Demand Response is taken into consideration, the normal
 11

operation without rebound smoothen the peak of the graph. However, Demand
Response will sometimes cause the rebound effect to take place. The energy is
typically not saved and maybe a new peak is generated. This effect can be avoided,
but it might result in a reduced process quality. For example, when such an ideal
peak shaving applied to air-conditioning system, if it normally run at 50% and
were shed for half an hour, it is restricted to compensate the downtime with half an
hour of 100%. In Figure 2.3, the original consumption has a high peak in the middle.

Figure 2.3: Impact of improved energy efficiency versus demand

response

2.3.1 Energy Efficiency

Among the various ways to implement demand side management, it is clear that the
energy efficiency is most wanted. The goal must be to improve efficiency as it saves
energy and emissions. Improving energy efficiency of buildings or industrial sites
starts with information and energy auditing into the process involved. In every
workplace or site, there are problems that are not noticeable which cause energy
wastage. This includes compressed air leakages, misconfigured control, dirty filters,
broken equipment, etc. These problems are normally overlooked unless there are
tools for analyzing energy efficiency (Palensky and Dietrich 2011).
 12

2.3.2 Energy Controllers

An energy controller is used in the operation of equipments which need

consumption-driven adjustment. The device is usually placed at the energy meter to
monitor the consumption trend. The controller will automatically switch off the
equipments when the consumption trend reaches unwanted levels based on the
priorities and other rules set by the user. The energy controller switched device is
shown in Figure 2.4. (Palensky and Dietrich 2011)

Figure 2.4: An energy controller switches devices

It is extremely complex and complicating to configure such an energy

controller. During the adding or removing of consumers, the stability depends on a
wise choice of rules. A simple way to determine the priority level is shown in Figure
2.5.

Let assumes three device classes, starting from most prominent c1 until c3
which is least important. The dashed line in the middle is the ideal case which leads
to the desire goal. The graph increases its gradient when all the devices are allowed
to turn on initially. It reaches the c3 off line where c3 is required to turn off as it is
least important. Although the graph is flattened by a little, it is not sufficient to bring
back the graph to the ideal case. So, it will cross the c2 off line and the c2 is turned
 13

off. The resulting graph is too flat until it crosses c1 on. Since c1 is still on, so
there is no effect. Crossing c2 on will allow the device to turn on again and so
forth for the rest of the graph until it reached its goal. Even though the allowed
device to on at t1 and t2 are the same, the power consumption is different. This allows
the devices in one category to switch on at that time. Sometimes, it is not necessary
since the devices have their own independent controls and schedules. Figure 2.5
illustrated the selection of priorities in a maximum demand monitor.

Figure 2.5: Selection of priorities in a maximum demand monitor

 14

2.3.3 Demand Response

Smart grid technologies will further increase the use of demand response in daily
operation. It has become an integral part of the power system and market operational
practice (NIST 2010). Demand response is the process of controlling end-user (load)
from their normal consumption pattern with the objective of decreasing energy
demand during peak hours when energy generation is expensive, shifting demand to
off-peak times, or balancing demand through load shedding at the time of generation
deficit.

The demand response management involves four main participants who are
energy balancing authority, demand response provider, utility company and customer.
Before any demand response scheduling process starts, it requires the balancing
authority to give orders by determining its desired volume, date/time, and duration.
The order is submitted to the provider to compare with their demand response
availability and select the demand response participating customers. Taking into
account each customers distinct compliance factor, the provider determines
expected demand response total and reported back to the balancing authority
(Medina, Muller and Roytelman 2010). Figure 2.6 shows the existing demand
response scheduling process.

Figure 2.6: Existing Demand Response scheduling process

 15

2.3.4 Spinning Reserve

The spinning reserve is defined as the unused capacity which can be activated on
decision of the system operator. It is provided by devices that are synchronized to the
network and able to affect its active power. The energy management strategy should
accommodate both short-term power balancing and long-term power balancing. An
appropriate level of spinning reserve should be maintained to safeguard the power
systems without involuntary load shedding. Thus, increasing the spinning reserve can
reduce the probability and severity of loss of load (Wang and Gooi 2011).

When the generation cannot meet the demand, the reserved capacity will be
used to maintain the balance of the network. The primary control reserve is not part
of the spinning reserve because it is not controlled by the transmission system
operator. The secondary control reserve is a centralized automatic control that
delivers reserve power on short notice in order to bring the frequency back to the
target value. The tertiary control reserve provides manual change in the dispatching
in order to restore the secondary reserve. It also provides a more permanent solution
if the imbalance between consumed power and scheduled power persists. It will react
if the secondary reserve does not solve the problem. In Figure 2.7, the normal energy
usage is below the scheduled power.

Figure 2.7: Allocation of capacity of generation unit

 16

2.4 Load Recognition

Load recognition is the ability of a system to identify the type of load connected.
There are different methods proposed to achieve this specific outcome. In demand
side management, it becomes increasingly challenging due to growing energy
demands within offices and premises. It is essential to perform automated appliances
recognition and monitoring to fully utilize the available resources. Smart meter is one
of the incentives to improve energy monitoring. However, it can only measure
energy consumption on a house level, providing little information on the breakdown
of the energy spent. The main challenge is to acquire appliances-level information. If
each load is attached with a meter, the increase in the number of meter will reduce
the reliability and cause the system to be much more complicated. Through the load
profile, it provides details which appliances have been used, how much they have
consumed as well as when and why they are operated. (Zoha et al., 2012)

2.4.1 Non-Intrusive Appliance Load Monitoring

Non-intrusive appliance load monitoring (NALM) is a way to minimize the number

of instruments used for monitoring process. It is designed to monitor an electrical
circuit that contains a number of devices which switch on and off independently. By
a sophisticated analysis of the voltage and current waveform of the total load, the
NALM estimates the number and nature of the individual loads, their individual
energy consumption and other relevant statistics such as time-of-day variation.
(Hsueh-Hsien, Ching-Lung and Hung-Tzer 2008)

This method has no access to the individual components during the

installation of sensors or making measurement. It is much more convenient and
effective method compare to previous method which by means of placing sensors on
each individual components of the load. The load data is extremely valuable to
energy auditors, utilities, public policy makers, and appliance manufacturers, for a
broad range of purpose. This plug-and-play device is very easy for installation,
 17

removal, and maintenance. Figure 2.8 shows the NALM connects with the total load
using the standard revenue meter socket interface.

Figure 2.8: NALM devices

The NALM monitors the total load, checking for certain characteristics which
provide information about the activity of the appliances which constitute the load.
The refrigerator which estimated to consume 250 W and 200 VAR will have a step
increase of the characteristic size to indicate that it is turned on, and a decrease of
size to indicate that it is turned off. When heater with its own load characteristic is
added in, the graph become complicated as there is overlapping between the two
loads. However, the loads can be distinguished by knowing the time of each on and
off event and the total energy consumption. Furthermore, considering measurement
of the reactive power or harmonic current, along with the real power shown, it would
reveal even more information about the particular appliances. In Figure 2.9, the
power against time graph shows the load characteristic of refrigerator and heater.

Figure 2.9: Load Characteristic Graph

 18

2.4.2 Electromagnetic Transient Program

Another method of load recognition is through the electromagnetic transient program

(EMTP). The idea of EMTP is by analyzing and sampling the transient energy
signatures during the start up. Different class of loads has their own characteristic of
turn-on transient energy signatures. This method can increase the efficiency and
accuracy of load recognition, especially for loads which have same real and reactive
power (Hsueh-Hsien, Ching-Lung and Hung-Tzer 2008).

In general, electrical appliances may have more than one load representation
due to many physical components are involved. A hair dryer has two loads, a heating
coil and a motor while a water boiler has only one load which is the heating coil.
These appliances have their own unique power signatures that enable them to be
distinguished from one another. It can be observed by voltage and current waveforms
of the particular load. Besides, it can also be identified from processed reproductions
of these signals including harmonics, real power and reactive power (Laughman et
al., 2003).

The transient properties of electrical load are determined by the physical task
that the load performs. Transient energy may assume different forms in consumer
appliances base on the generating mechanism. Load classes performing physically
different tasks are therefore distinguishable by their transient behaviour. Hence, two
different appliances which consume similar real and reactive power may have
remarkably different turn-on transient currents. For example, a pump has long
switching-on transient while electronically fed appliances have short but high
amplitude switching-on transient (Leeb et al., 1995).
 19

A sample of time length, t is examined to separate transient representatives

of a class of loads. Each input waveforms has a set of transient energy values
representing a particular transient shape delineated by the separation process. The
maximum recognition accuracy, is maximized during the test phase as t is
adaptively changed based on the factor . The flowchart of adaptive algorithm for
load recognition is shown in Figure 2.10.

Figure 2.10: Adaptive algorithm for load recognition

 20

2.5 Voltage Unbalance

Among different type of power quality problem, voltage unbalance has significant
concern at the LV distribution level. The voltage unbalance at the fundamental
frequency is normally present in the supply voltage (Duarte and Kagan 2010). The
voltages at the generation and transmission levels are well balanced. When the
voltages reach the utilisation and distribution level, they become unbalanced due to
various issues such as:
1. Uneven distribution single phase load distribution across the three phase
network
2. Continuous changing of instantaneous demand
3. Unbalance or unstable incoming supply

In a balanced network, the three phases voltage magnitude is equal and the
phase angle displaced 120 degree from each other. Any difference that exists in the
three voltage magnitudes or shift in phase is known as unbalanced supply. When a
balanced three-phase load is connected to an unbalanced supply system, the current
drawn by the load also become unbalanced. A balanced system and an unbalanced
system are presented in Figure 2.11 and Figure 2.12 respectively.

Figure 2.11: A balanced system

Figure 2.12: An unbalanced system

 21

2.5.1 Voltage Unbalance Factor

Voltage Unbalance Factor (VUF) is resulted by the unequal magnitudes or phase

angle which causes unbalanced supply. Based to the European Standards, it is
defined as the ratio of negative sequence voltage (V-) to positive sequence voltage
(V+), as expressed below:

(2.1)

For a perfectly balanced system, both negative and zero sequence system will
not be present. As there is no phase displacement between the three voltages in zero
sequence system, when it is applied to three-phase induction motor, it will not rotate
as there is no rotating magnetic field. The following matrix expression is the negative
and positive sequence of the system voltage.

V 0 1 1 1 Va
1
V 31 a a 2 Vb (2.2)
V 2 1 a 2 a Vc

where V0, V1 and V2 are the positive, negative and zero sequence voltages and Va, Vb
and Vc are the three phase line voltages components. Figure 2.13 shows different
symmetrical components for unbalanced system of voltages.

Positive Sequence Negative Sequence Zero Sequence

Figure 2.13: Symmetrical components for unbalanced system of voltages

 22

Voltage rise is an issue to be concerned in the three phase power distribution

system. The nominal voltage rating delivered by the utilities need to be regulated
within the operating tolerance. Malaysia has the voltage unbalance factor has a
statutory limit of 1%. For primary distribution with voltage level of 240V, the range
is set at +5% and -10%. This ensures that the electrical equipments will perform
satisfactorily within the voltage range provided by the utilities.

2.5.2 Effects of Voltage Unbalance

Voltage unbalance will cause several negative effects on induction motors.

Overheating, line-current unbalance, derating, torque pulsation and inefficiency are
the adverse effects of voltage unbalance. If an excessive level of voltage unbalance
occurred, it has serious impacts on the load connected especially induction motors.
Although induction motors are designed to tolerate a small level of unbalance, they
have to be derated if the unbalance is excessive. This can highly reduce the life span
of the induction motor due to overheating in the rotor and stator. Besides, it can
cause significant additional power losses resulting in the steady-state temperature
rises of the windings. (Siddique, Yadava and Singh 2004).
 23

CHAPTER 3

3 METHODOLOGY

3.1 Overview

A microgrid network is developed in a portable heavy duty cabin behind UTAR SE

Block. The microgrid consists of supply from utility, switches and protections,
bidirectional inverter, batteries and controllable load bank. The National Instrument
sbRIO-9632XT is used as a main control unit for controlling, monitoring, and
acquiring electrical data of the microgrid through computer. A large system is
difficult to manage as things might go haywire during troubleshooting. Thus, it is
highly preferred to group the part which can be combined to form an individual
device. Figure 3.1 shows the schematic diagram of the microgird.
 24

Figure 3.1: Experimental setup of Smart Energy Conditioning System

The first design of the microgrid system is found to be not portable. It is

troublesome and difficult to remove when the system needs to move from one place
to another. Hence, the improved version of microgrid is developed.

The electrical components within the dotted line in Figure 3.1 can be group
together to form a diagnostic tool. During the design of the diagnostic tool, three
main parts that bring into consideration are safety, flexibility and portability. Safety
which is the paramount of every device is obeyed by selecting the suitable material
together with proper protection devices. It must be flexible which act as a plug-and-
play device that can connect either single phase or 3-phase load to monitor the
necessary data such as power, power factor, voltage and current. Besides, portability
is also a consideration so that it can be brought to anywhere which requires energy
diagnosis.
 25

Figure 3.2 shows the process flow of load diagnosis. After the separation of
the diagnosis part from the existing circuit, there is a need to identify the position of
the stand-alone device. The actual position is between the source and the load.

Figure 3.2: Process Flow of Load Diagnosis

3.2 Concept of Diagnostic Tool

The diagnositic tool is actually a detour from the power source to the load. The
function is similar to NALM which also place between the power source and the load.
The users can monitor and check either the single phase or three phase load.
Schneider GC2530M5 contactor is used to separate the two different power sources
and to prevent line to line fault. The current is measured by using TABB50C100
100/5A current transformer and collected using the NI9227 current measurement
card. The voltage is measured by NI9225 voltage measurement card for each phase
by connecting parallel to the output. All the data acquired is passed to the sbRIO-
9632XT and will be displayed on the computer. The schematic diagram of the
diagnostic tool is shown in Figure 3.3.
 26

Figure 3.3: Schematic Diagram of Diagnostic Tool

The specification of the prototype 1 is tabulated in Table 3.1.

Table 3.1: Specification of Prototype 1

Specification of Prototype 1
Type Single Phase & 3-Phase Diagnostic Tool
Material 1mm2 Aluminium, 1mm2 Plywood
Input / Output Voltage Phase voltage of 220 ~ 240 V
Max Power Rating 3KW
Max Current Rating 14A
Cable Rating 32A
Frequency 50Hz1%
 27

3.2.1 Prototype 1

The design specifications are listed out and strictly followed to ensure that the
diagnostic tool is operating in a safe region. 1mm2 of aluminium is chosen due to its
light-weight and durable. Another piece of 1mm2 plywood is attached at the back to
provide insulation. The aluminium can also be used to discharge excessive voltage
during fault without damaging the components inside the casing. Figure 3.4 shows
the material used to build the casing of the device.

Figure 3.4: 1mm2 Plywood and 1mm2 Aluminium

3.2.2 Arrangement

The arrangement is important to facilitate the users during the access of plug points
and switches. Proper wiring can be done when each item is placed accordingly. The
cable used is Copper, PVC insulated, 1.5mm2 and voltage rating of 450/750V. The
colour code of Red (Phase A), Yellow (Phase B), Blue (Phase C), Black (Neutral)
and Earth (Green) are followed to ease the users during troubleshooting. The
connection between the point and wire is soldered with lead and insulated using heat
shrink tube. It is crucial to prevent any leakage of current during the operation.
Figure 3.5 shows the arrangement of the components at the front.
 28

ON/OFF
3-Phase
Button
Power Source
(Output)
3-Phase
Power Source
DPDT
(Input)
Switch

Single Phase
Plug Single Phase
(Input) Plug
(Output)

Figure 3.5: Front Arrangement

The arrangement inside the device is shown in Figure 3.6.

Contactor

4 Slots for
Current Transformer

24V DC
Power Supply
Figure 3.6: Inside Arrangement
 29

3.2.3 Components

In order to match the requirement of the device, studies are carried out to determine
the specification of the components. The components include power supply,
Contactor and Current Transformer. The specification data sheets are listed as below.

3.2.3.1 Power Supply

The device requires a 100W Single Output Switching power supply NES-100-24 to
power up the board and also the contactor. This power supply also has the protection
from short circuit, overload and over voltage. The honeycomb design provides an
efficient cooling by free air convection. The power supply is shown in Figure 3.7.
The specification of the power supply is tabulated in Table 3.2.

Figure 3.7: Power Supply

Table 3.2: Power Supply NES-100-24

Specification of Power Supply NES-100-24

Input voltage range 176 ~ 264 VAC
Frequency Range 47 ~ 63 Hz
Output DC Voltage 24 V
Rated Current 4.5 A
Rated Power 108 W
 30

3.2.3.2 Contactor

Contactor can be sued to function as a switch in this device. This contactor has 1
auxiliary contact. When 240V AC is connected, it will trigger the Normally Open
(NO) to Normally Close (NC). Two contactors are needed to switch between single
phase and 3-phase source. It also provides isolation for input and output when it is
not triggered. From specification, it can withstand up to 25A and operate normally
under 50C. Contactor used in this device is the Schneider Contactor GC2530M5 as
shown in Figure 3.8. Table 3.3 shows the specification of the contactor.

Figure 3.8: Contactor

Table 3.3: Contactor GC2530M5

Specification of Contactor GC2530M5
Coil Voltage 240V ac
Contact Current Rating 25 A
Contact Voltage Rating 415 V ac
Maximum Operating Temperature +50 C
Number of Auxiliary Contact 1
Number of Contacts 3
 31

3.2.3.3 Current Transformer

Current Transformer is used to obtain the current when there is load connected to the
output of plug point. A low-voltage measuring current transformer is used to measure
the current flowing through it. The current induced will be stepped-down by the
current transformer. The advantage of using open core current transformer is that the
current is obtained by induction rather than direct current flow in series across the
NI9227 Current Measurement Card. When the current is large, it is highly possible to
damage any device which is connected in series. The current transformer is shown in
Figure 3.9.The specification of current transformer is shown in Table 3.4.

Figure 3.9: Current Transformer

Table 3.4: Current Transformer 100/5A

Specification of Current Transformer TAIBB 100/5A
Primary Current 40 ~ 300A
Secondary Current 1 ~ 5A
Operating Frequency 47 ~ 63 Hz
Insulation Class Class B ( CEI EN 60044-1)
Terminal Protection IP200 with sealable terminal cover
Weight 180g
 32

3.2.4 Prototype 2

Some modifications are made to prototype 1 to improve the overall functionality by

removing the unnecessary part. Few drawbacks of prototype 1 are listed as below:
1. The size is too big and bulky.
2. The 3-Phase Power Source is hardly used as this diagnostic tool is for low-
voltage measurement.
3. The casing is difficult to dissemble and reassemble.

Previously, the 3-phase power input and output require a large space within
the device. It also requires an additional contactor for selection between single phase
and 3-phase. The removal of 3-phase related components has highly decreases the
size and weight of the device. The casing is also redesign to two combination of U-
shape which can be easily access during troubleshooting. The new schematic
diagram for prototype 2 is shown in Figure 3.18.

Figure 3.10: Schematic Diagram of Prototype 2

 33

3.2.5 Load Recognition

Load recognition can be done by comparing the amplitude of the waveform and the
maximum and minimum current. The NI9227 Current Measurement card has the
ability to measure up to 50kS/s per channel sampling rate with 24 bits resolution.
This allows the measurement to be more accurate and differentiate the distinction
between each kind of load.

For higher-speed control, sbRIO-9632XT incorporate a FPGA that can be

programmed with LabVIEW software allows a silicon-speed processing on I/O data
from National Instrument C Series Module. With the integration of both hardware
and software, the flowchart to perform load recognition is shown in Figure 3.11.

Figure 3.11: Flowchart of Load Recognition

 34

3.2.5.1 Programming on LabVIEW

The graphical programming in Labview is user-friendly and easy to construct the

program. The information contained in the program can be extracted out or displayed
by using proper indicators. In Figure, sbRIO-9632XT acquired the data from the
NI9227 Current Measurement card and passed through the electrical power
measurement, the data is converted into AC current waveform. Then, the waveform
is multiplied with 20 to get the RMS current. Finally, it can be tapped out to perform
necessary analysis. The Convert to Waveform.vi is shown in Figure 3.12.

Figure 3.12: Electrical Power Tools (Convert to Waveform.vi)

 35

Further analysis can be done by breaking down the complex data source into
individual simple indicator. A current waveform consists of combination of current
Phase A, Phase B, Phase C and Neutral. The experiment is done using only Phase A
which is expected to have the same result for Phase B and Phase C. The amplitude of
the waveform is the first condition to determine the type of load. After that, the
highest and lowest current values are compared to make sure they fall within the
range. For example, the water boiler maximum state and minimum state in the
current waveform is 4.6 and - 4.5 respectively. The graphical programming is shown
in Figure 3.13.

Only Phase A is
extracted

Figure 3.13: Block Diagram for Load Recognition

 36

3.3 Voltage Unbalance Mitigation Experimental Setup

In this project, NI Single-Board RIO 9632XT is chosen because it is an embedded

control and acquisition device with the integration of FPGA, real-time processor, and
I/O port on a single printed circuit board. All these features allow it to be used as the
main control system. The I/O port can supply up to 3.3V which is sufficient to trigger
the solid state relay connected to the load bank. The ability to connect with C Series
Module, both NI 9225 Voltage Measurement and NI 9227 Current Measurement,
enhance the functionality of the board.

The network resistors with different set of values are used to emulate the
transmission line. Besides, the power resistor is constructed into a controllable load
bank which emulates the power consumer. The solid state relay is controlled by the
sbRIO-9632XT to on or off the load bank.

The Sunny Island 5048, which is a bidirectional inverter, is connected to

Phase A and Phase B respectively in the system. It can either absorb or deliver power
to the electrical network. Meanwhile, this bidirectional inverter also acts as an
uninterruptible power supply (UPS) which enable the gird to continue running even
when there is no supply from the utilities. The energy conditioning system is set up
as in Figure 3.14.

Figure 3.14: Experimental setup of Energy Conditioning System

 37

3.3.1 Measurement and Automation System

sbRIO-9632XT is the central processing unit in this device. The users are able to test,
control and monitor the whole system through interaction between computer and the
board. It also has I/O port with can produce an output voltage of 3.3 V for different
purposes. The C series modules which are used in this project are the NI-9225
voltage measurement card and NI 9227 current measurement card. Figure 3.15 shows
the sbRIO-9632XT along with the NI9225 and NI9227.

Figure 3.15: sbRIO-9632XT with C Series Module

 38

3.3.2 Transmission Line Model

There are altogether 4 set of resistors varies from 0.11 to 0.33 . The three phases
plus the neutral line is connected to 0.33 in order to create a more significant value
for voltage unbalance. From calculation, 0.33 representing 600 m of transmission
cable in length. Figure 3.16 shows the transmission line model which emulates the
resistance presence in the line.

Figure 3.16: Transmission line model

 39

3.3.3 Controllable Load Bank

The controllable load bank is built from a series of power resistor and solid state
relay. Each phase has five 200 W power resistor connected in series. The total load
can go up to 1000 W in each phase. The solid state relay is triggered by the signal
output from the sbRIO-632XT to turn on the power resistor. Figure 3.17 shows both
the resistor bank together with the solid state relay.

Figure 3.17: Controllable Load Bank

 40

3.3.4 Sunny Island 5048

Four batteries are required to power up Sunny Island 5048 and the batteries provide
energy storage up to 5760 Wh. This device is able to charge the excessive power
from utilities into the battery bank or supply to load when the supply from the grid is
low. The network condition programmed in LabVIEW will determine the
bidirectional inverter either inject or absorb power. This can maintain the VUF
within the statutory limit. Besides, the Sunny Island can also be used as the
Uninterruptible Power Supply (UPS) which can supply the load during power outage.
The Sunny Island 5048 which is a bidirectional inverter is shown in Figure 3.18.

Figure 3.18: Sunny Island 5048

 41

3.4 Graphical User Interface

A graphical user interface (GUI) is developed in LabVIEW. The front panel of the
user interface (UI) consists of several tabs which allow the users to access base on
their preference. There are also indicators to display the value of the voltage, current,
power and energy. The users can glance through all the information at the monitor
tab. The load recognition UI is shown in Figure 3.19.

Figure 3.19: Load Recognition UI

The UI for mitigation of voltage unbalance is shown in Figure 3.20.

Figure 3.20: Energy Conditioning System UI

 42

CHAPTER 4

4 RESULTS AND DISCUSSIONS

4.1 Comparison of Prototype 1 and 2

The major difference between prototype 1 and prototype 2 is the reduction in size.
The portability is highly increased when the device become smaller. Less material is
required to build which can highly save the cost during mass production. The power
rating, current rating and cable rating remain unchanged to make sure the device is
well protected. The prototype 1 and prototype 2 are shown in Figure 4.1 and Figure
4.2 respectively.

Figure 4.1: Prototype 1

 43

Figure 4.2: Prototype 2

The changes that have been made between the Prototype 1 and 2 are listed in Table
4.1.

Table 4.1: Comparison between Prototype 1 and 2

Model Prototype 1 Prototype 2
Type Single Phase & 3-Phase Single Phase
Diagnostic Tool Diagnostic Tool
Material 1mm2 Aluminium, 1mm2 Aluminium,
1mm2 Plywood 1mm2 Plywood
Input/ Output Single phase/ Three phase, Phase Voltage,
Voltage 220 ~ 240 V 220 ~ 240 V
Output Socket 3 single phase outlet and 1 3- 3 single phase outlet
phase socket

Max Power Rating 3KW 3KW

Max Current Rating 14A 14A
Cable Rating 32A 32A
Frequency 50Hz1% 50Hz1%
Dimension (WLH) 35cm 40cm 38cm 35cm 38cm 20cm
 44

The actual device in action is shown in Figure 4.3 where a blower is

connected for load analysis. The detail is displayed through the Graphical User
Interface (GUI) which consists of several tabs for users to access.

Figure 4.3: Diagnostic Tool Connected with Hair Blower

4.2 Functions of Diagnostic Tool

The functions of the diagnostic tool are related to load analysis and load recognition.
The ability of measuring different parameters of the load including, power, voltage,
current, power factor and frequency has invoke the idea of load recognition.

Notice that the Voltage Waveform has only 1 red-coloured waveform. Phase
B and Phase C remain approximately at zero because there is only Phase A
connected with load. Further detail of the load is displayed by the indicators under
the waveform graph. Figure 4.4 shows the User Interface (UI) of the voltage
waveform.
 45

Figure 4.4: Voltage Waveform

Current Waveform graph has two waveforms which is the Phase A and
Neutral. The amount of current flow in and out must be equal. If the electrical current
is not balanced between the incoming and returning neutral conductor, it will cause
the Residue-Current Device (RCD) to trip. This allow the users to monitor the load
connected through it is functioning normally. Figure 4.5 shows the interface of the
current waveform.

Phase A
Current

Neutral
Current
Figure 4.5: Current Waveform
 46

The power waveform is the outcome of multiplication between voltage and

current. The active power is displayed in unit W. The load measured using this
device has power less than 2KW. The power waveform is displayed to users as in
Figure 4.6.

Figure 4.6: Power Waveform

The energy waveform is the accumulation of active power per hour. It can be
used to calculate the cost of electricity bill for particular load. It is presented by an
increasing line as in Figure 4.7.

Figure 4.7: Energy Waveform

 47

4.2.1 Load Recognition at Idle State

When no load is connected, the device should be able to identify it is in the idle state.
The small changes in the current values vary the range from 0.042 to 0.047. The
ideal state when no load is connected should be 0 A current. The small offset value is
cause by the internal resistor. A negative value of amplitude is not possible to occur.
Hence, the range where no load is plugged in is set from 0 to 0.05. The range of the
amplitude without any load is shown in Figure 4.8 and Figure 4.9.

Figure 4.8: Idle State 1

Figure 4.9: Idle State 2

 48

Once the idle state has been determined, the range beyond can be categorize
into different load by proper measurement. In order to prove that this concept can
function, hair blower and water boiler are chosen to use as load.

4.2.2 Hair Blower

A hair blower has both the characteristic of inductive load and resistive load. This
can be proven by observing the current values and the amplitude of the waveform.
When the hair blower is turn on without heating up the coil, it is pure inductive load.
The amount of current draw is used to turn the motor as shown in Figure 4.10.

Figure 4.10: Hair Blower (Inductive)

When the motor and heating coil are turned on, it consumes more power. The
highest and lowest current values will change depending on the power consumption.
The inductive and resistive load of the hair blower is shown in Figure 4.11.
 49

Figure 4.11: Hair Blower (Resistive)

The load characteristic of the hair blower is recorded and discussed. The
power graph shows the process when operating the hair blower. Initially, the hair
blower is not turned on. When it is switched to cool fan mode without heating coil,
the power consumed is around 560 W. Once it is switched to drying mode, additional
power around 460 W is needed due to heat up the coil. The amplitude of current
waveform varies from 2.5 to 7.5 during the operation. The maximum current and
minimum current are 2.8 and - 4.6. With all the detail available, the process to
identify and recognize the load can be carried out. Figure 4.12 shows the load
characteristic of a blower.

Figure 4.12: Power Waveform of Hair Blower

 50

4.2.2.1 Data Recording and Storing

The data can also be recorded and extracted to Microsoft Excel for verification. The
way to record the data is by using Write to Spreadsheet File.vi. The data is bundled
and converted from cluster to array and stored in the C:\<File path>. The block
diagram for data recording and storing is constructed as in Figure 4.13.

Input
Figure 4.13: Write to Spreadsheet File.vi
data

4.2.2.2 Export to Microsoft Excel

Both current and power have approximately the same characteristic graph. The small
different are caused by the inconsistence value of voltage. This information is useful
for demand side management for different purposes. The current against the time is
shown in Figure 4.14.

Figure 4.14: Current vs Time (Hair Blower)

 51

The power against time graph is shwon in Figure 4.15

Figure 4.15: Power vs Time (Hair Blower)

4.2.2.3 Creating Database

When the detail of the type of load is determined, it is required to store the data in the
database. This can create a list of data which can recognize different type of
electrical equipment. The value of current obtained at real-time from the current
waveform is compared with the database. When the amplitude is in range and within
the current value, the type of electrical equipment will be displayed. A simple way to
build the database is shown in Figure 4.16.

Real-time value
compare with
preset value

Range of
amplitude

Figure 4.16: Database

 52

4.2.3 Water Boiler

Water boiler has only the characteristic of resistive load. The power consumption of
water boiler is higher compare to hair blower. When a water boiler operates, it is
either at Keep Warm mode or Boiling mode. The voltage waveform of water
boiler is shown in Figure 4.17.

Figure 4.17: Voltage Waveform of Water Boiler

The current waveform of the water boiler has a sinusoidal waveform as it is

purely resistive as indicated in Figure 4.18.

Figure 4.18: Current Waveform of Water Boiler

 53

The load characteristic of a water boiler is either low or high as there are only
two states. This characteristic is commonly found on those electrical appliances
which repeat the same process such as refrigerator, rice cooker and so on. In the case
of water boiler, Keep Warm mode only consumes around 33.9W while Boiling
mode consumes around 774W. The power waveform is obtained as in Figure 4.19.

Figure 4.19: Power Waveform of Water Boiler

 54

The current against time graph for Keep Warm mode to Boiling mode
and finally OFF is shown in Figure 4.20.

Figure 4.20: Current vs Time (Water Boiler)

The power consumption of the water boiler can also be determined through
the data saved in the sbRIO-9632XT. The approximation of power consumption
against time is shown in Figure 4.21.

Figure 4.21: Power vs Time (Water Boiler)

 55

4.2.3.1 Database Expansion

The database can be expanded by acquiring more load characteristic of different

electrical appliances. The newly determined data of water boiler is created and
included in the previous database. The dotted line in Figure 4.22 shows the newly
created data in the database to determine a new load.

Figure 4.22: Creating New Data

The water boiler has now become one of the electrical appliances that can be
recognized by the diagnostic tool. The interface is shown in Figure 4.23.

Figure 4.23: Load Recognition User Interface

 56

4.2.4 Load Characteristic

The load characteristic shows the consumption trend of a load when it is turned on.
When both of the loads operate together, each of them will have different load
characteristic. The duration is also different depends on how long the users operate
them. As water boiler continues to turn on at Keep Warm mode, its power
consumption does not remain at zero. During the Boiling mode, the power
consumption will increase instantaneously to a high level. Meanwhile, the hair
blower only turns on during usage. The power consumption is zero when it is not
operating. The two types of load present in the hair blower have different level of
power consumption. Figure 4.24 illustrate the load characteristic of water boiler and
hair blower.

Figure 4.24: Load Characteristic

 57

When two load characteristics are combined, the power consumption trend
will increase. At 169 s, the water boiler is turned to Boiling mode with hair blower
running only the induction motor without heating coil. The power consumption is
around 1 KW. After 194 s, the hair blower turned on both induction motor and
heating coil. The power consumption by adding both of the load yields around 1350
W. Hence, when two or more loads operate simultaneously, the power consumption
trend will have high peak. In demand side management, such peak is unfavorable and
most likely will be shaved to conserve energy. If different loads run at the same
time, this will increase the burden of the electrical network. Additional power plant
has to operate to meet the power demand. The combination of hair blower and water
boiler power characteristic produce a high peak as shown in Figure 4.25.

Figure 4.25: Combination of Load Characteristic

 58

4.2.5 Voltage Unbalance Mitigation

This diagnostic tool is able to perform a wide range of function when Sunny Island
5048 is integrated into the system. The functionalities included are peak clipping,
mitigation of voltage unbalance, power factor correction, provide uninterruptible
power supply and energy measurement.

Voltage unbalance can be mitigated when all the three phases have well
balanced voltage. The diagnostic tool has the ability to integrate with Sunny Island
5048 to mitigate the voltage unbalance. In this project, Sunny Island 5048 acts as an
energy storage device. It is able to channel excessive power to charge batteries and
supply power to the load to lower the burden of utilities. When the system has
balance load among the three phases, the Voltage Unbalance Factor (VUF) does not
rise beyond the statutory limit of 1%. When a balanced load is connected to the three
phases, the increase or decrease in load still maintain the phase voltage close to each
other. The balanced load varies the VUF in the range of 0.4% to 0.6% as in Figure
4.26.

Figure 4.26: VUF with balanced load

 59

In order to generate a significant voltage unbalance, the load at phase A is

loaded in a step of 200W for 5 steps. The loading condition can be expressed as
[LOAD 0 0]. This means that phase B and phase C are at no-load condition. When
the load at phase A keeps on increasing, the voltage at phase A will drop. This causes
the phase A to have lowest voltage among the three phases. Hence, the VUF rises
and exceed the statutory limit of 1%. The highest VUF recorded is 1.45% as
indicated in Figure 4.27.

Figure 4.27: VUF with unbalanced load at Phase A

This voltage unbalance problem is always overlooked and it will result in

overheating, reduce in lifespan, power loss and decrease in efficiency of induction
motor.
 60

When the Sunny Island 5048 is added into the system, the voltage unbalance
can be significantly mitigated. The energy conditioning system is able to detect and
take action when the VUF exceed 1%. After the condition is fulfilled, it takes action
by either supplying power to the phase with lower voltage or channelling the power
of highest phase to charge battery. In previous scenario where phase A is highly
loaded, the Sunny Island will supply the load to lower down the burden of phase A.
This solution successfully mitigated the VUF to 0.3%. The VUF waveform with
loading condition of [LOAD 0 0] and Sunny Island 5048 connected to phase A is
shown in Figure 4.28.

Figure 4.28: VUF with Load and Sunny Island at Phase A

 61

CHAPTER 5

5 CONCLUSION AND RECOMMENDATIONS

5.1 Conclusion

In conclusion, the objectives of the project have been met. The diagnostic tool has
been built and improved from prototype 1 to prototype 2. The device is programmed
with LabVIEW to conduct several experiments and the result proved that the device
is functioning. The development of the diagnostic tool allows the users to obtain the
load characteristic and further enhance the system to perform load recognition. Load
recognition can be used to provide information about particular load which helps in
developing the demand side management strategy.

The energy conditioning system developed has also successfully mitigated

the voltage unbalance in the network with the integration of Sunny Island 5048.
Unbalance load in a network may cause the violation of voltage unbalance statutory
limit. With the energy conditioning system, the excess power can be channelled to
the energy storage system for future use at the same time the voltage unbalance can
be kept at minimal value.
 62

5.2 Future Improvement

There are few limitations found on the diagnostic tool which can be improved. The
size of the device can be decreased to improve portability by rearranging the
components. The location of input plug can be moved to the rear part of the device so
that it is easier for the users to access. The device can further improve by having
external voltage measurement clip and clamp-type current transformer. This allows
the users to access different kind of load even in a compact area.

Moreover, the programming part can also be improved. Whenever a new load
is detected, the programme can create and store the new data in database. The load
recognition can increase its reliability and accuracy of identifying different load by
comparing their turn-on transient signatures.

On the other hand, the mitigation of voltage unbalance is now limited to

Phase A and Phase B as there are only two Sunny Island 5048 available. The
additional Sunny Island 5048 installed to Phase C enable the energy conditioning
system to operate effectively. The Uninterruptible Power Supply (UPS) is more
reliable with all the three phases connected to the bidirectional inverter.
 63

REFERENCES

A. Vojdani, Integration challenges of the smart grid Enterprise integration of

DR and meter data, in Proc. IEEE Electrical Power System Conf. (EPEC08),
Vancouver, Canada, Oct. 2008, p. 21.

D. Moore and D. McDonnell, "Smart grid vision meets distribution utility reality,"
Elect. Light Power, pp. 1-6, Mar. 2007

Duarte, S.X.; Kagan, N.; , "A Power-Quality Index to Assess the Impact of Voltage
Harmonic Distortions and Unbalance to Three-Phase Induction Motors," Power
Delivery, IEEE Transactions on , vol.25, no.3, pp.1846-1854, July 2010

Hsueh-Hsien Chang; Ching-Lung Lin; Hong-Tzer Yang, "Load recognition for

different loads with the same real power and reactive power in a non-intrusive
load-monitoring system," Computer Supported Cooperative Work in Design, 2008.
CSCWD 2008. 12th International Conference on , vol., no., pp.1122,1127, 16-18
April 2008

K. Moslehi, Intelligent infrastructure for coordinated control of a self-healing power

grid, in Proc. IEEE Electrical Power System Conf. (EPEC08), Vancouver,
Canada, Oct. 2008, pp. 37

Laughman, C.; Kwangduk Lee; Cox, R.; Shaw, S.; Leeb, S.; Norford, L.; Armstrong,
P., "Power signature analysis," Power and Energy Magazine, IEEE , vol.1, no.2,
pp.56,63, Mar-Apr 2003

Leeb, S.B.; Shaw, S.R.; Kirtley, J.L., Jr., "Transient event detection in spectral
envelope estimates for nonintrusive load monitoring," Power Delivery, IEEE
Transactions on , vol.10, no.3, pp.1200,1210, Jul 1995

M. Amin and S. Wollenberg, Toward a smart grid: Power delivery for the 21st
century, IEEE Power Energy Mag., vol. 3, no. 5, pp. 3441, Sept.Oct. 2005.

Medina, J.; Muller, N.; Roytelman, I.; , "Demand Response and Distribution Grid
Operations: Opportunities and Challenges," Smart Grid, IEEE Transactions on ,
vol.1, no.2, pp.193-198, Sept. 2010 doi: 10.1109/TSG.2010.2050156
 64

National Institute of Standards and Technology, NIST Framework and Roadmap for
Smart Grid Interoperability Standards, Release 1.0, Special Publ. 1108, Jan. 2010,
145 pp.

Palensky, P.; Dietrich, D.; , "Demand Side Management: Demand Response,

Intelligent Energy Systems, and Smart Loads," Industrial Informatics, IEEE
Transactions on , vol.7, no.3, pp.381-388, Aug. 2011(REAL)

P. Mazza, Powering up the Smart Grid, Climate Solutions, July 2005, pp. 2526.

Product Information: What is NI LabVIEW, 2011. [Online]. Available at:

http://www.ni.com/labview/whatis/ [Accessed 16 March 2013]

Siddique, A.; Yadava, G.S.; Singh, B.; , "Effects of voltage unbalance on induction
motors," Electrical Insulation, 2004. Conference Record of the 2004 IEEE
International Symposium on , vol., no., pp. 26- 29, 19-22 Sept. 2004

S. J. Anders, The emerging Smart Grid, Energy Policy Initiative Center, University
of San Diego School of Law, Octo 2007, pp.48.

Wang, M. Q.; Gooi, H.B., "Spinning Reserve Estimation in Microgrids," Power

Systems, IEEE Transactions on , vol.26, no.3, pp.1164,1174, Aug. 2011

Zoha, A.; Gluhak, A.; Nati, M.; Imran, M.A.; Rajasegarar, S., "Acoustic and device
feature fusion for load recognition," Intelligent Systems (IS), 2012 6th IEEE
International Conference , vol., no., pp.386,392, 6-8 Sept. 2012
 65

APPENDICES

APPENDIX A: Graphical Programming (LabVIEW)

66
67
68
69