UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

DC distribution systems for home application Iyer, Shreya 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2015_november_iyer_shreya.pdf [ 16.4MB ]
JSON: 24-1.0165791.json
JSON-LD: 24-1.0165791-ld.json
RDF/XML (Pretty): 24-1.0165791-rdf.xml
RDF/JSON: 24-1.0165791-rdf.json
Turtle: 24-1.0165791-turtle.txt
N-Triples: 24-1.0165791-rdf-ntriples.txt
Original Record: 24-1.0165791-source.json
Full Text

Full Text

DC Distribution Systems For Home ApplicationbyShreya IyerB.E, R.V College of Engineering, 2013A THESIS SUBMITTED IN PARTIAL FULFILLMENTOF THE REQUIREMENTS FOR THE DEGREE OFMaster of Applied ScienceinTHE FACULTY OF GRADUATE AND POSTDOCTORALSTUDIES(Electrical and Computer Engineering)The University of British Columbia(Vancouver)August 2015c© Shreya Iyer , 2015AbstractUnprecedented expansion of native direct current (DC) powered equipment (com-puters and consumer electronics) has increased household DC electricity consump-tion over the past decade. Since power utilities deliver alternating current (AC)rather than DC, the conversion process (rectifier) used to supply DC loads is veryinefficient. The research investigates the suitability of employing conventional ACwiring to distribute DC power to supply loads directly, in particular around out-let/switch arcing issues. The problem of arcing in DC system is very predominantand needs to be addressed to meet safety requirements while improving the effi-ciency of the system. In order to overcome the arcing issues, an alternative flatDC wiring system is proposed which offers improved transient electrical and ther-mal characteristics for household wiring. The flat wire solution employs the sameraw materials and provides improvements in parasitic values associated with arcingwhile reducing thermal resistance. The proposed flat wire geometry is expected toachieve reduction of arcing and improve the overall efficiency of the distributionsystem.Simulations of the two preliminary AC and DC systems are provided for typicaldomestic loads and switching events. These characteristics are verified by conduct-ing similar tests on house wiring system prototype created in the lab. Furthermore,the switching behaviour is observed on loading the system through the outlet.iiPrefaceThis research was done with the aim of using DC for homes while keeping thecurrent architecture and design as far as possible. The study included comparingand testing the existing system with the proposed system.The project was supported by Natural Sciences and Engineering Research Coun-cil of Canada (NSERC) under Grant CRDPJ 434659-12 & The Institute for Com-puting, Information and Cognitive Systems (ICICS) and Telus People & PlanetFriendly Home (PPFH) Initiative at The University of British Columbia (UBC)throughout the term of research.This work is based on research performed at the Electrical and Computer En-gineering Department of the University of British Columbia by Shreya Iyer, underthe joint supervision of Dr. William Dunford and Dr. Martin Ordonez.A version of Chapters 1,3 and 4 have been published and presented at the IEEEPower and Energy Society General meeting (PESGM) 2015 [1].Hardware design, assembly and testing of the DC home wiring set-up was doneby Shreya Iyer under the guidance of Dr. Dunford and Dr. Martin Ordonez.As the first author of the above-mentioned publication and work, the author ofthis thesis developed the theoretical concepts and wrote the documents, receivingadvice and technical guidance from supervisors Dr. William Dunford and Dr. Mar-tin Ordonez. The author developed simulations and experimental set-up with helpof Dr. Dunford, Dr. Ordonez and Dr. Ordonez’s research group.iiiTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiGlossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xAcknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiDedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Literature Review . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2.1 History Of AC Power . . . . . . . . . . . . . . . . . . . . 21.2.2 AC Versus DC . . . . . . . . . . . . . . . . . . . . . . . 31.3 Existing Technologies . . . . . . . . . . . . . . . . . . . . . . . . 41.4 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61.5 Organisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 DC Distribution For Homes: Theory and Application . . . . . . . . 8iv2.1 Design Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.1.1 System Topology . . . . . . . . . . . . . . . . . . . . . . HVDC Distribution Systems . . . . . . . . . . Choice of Voltage Standards . . . . . . . . . . . 102.1.2 Protection . . . . . . . . . . . . . . . . . . . . . . . . . . 102.1.3 Types Of Loads . . . . . . . . . . . . . . . . . . . . . . . Fluorescent Lighting . . . . . . . . . . . . . . . Motor Load . . . . . . . . . . . . . . . . . . . Electronic Loads . . . . . . . . . . . . . . . . . 122.1.4 Other Advantages . . . . . . . . . . . . . . . . . . . . . . 122.2 Transmission Lines . . . . . . . . . . . . . . . . . . . . . . . . . 132.2.1 Transmission Line Effects . . . . . . . . . . . . . . . . . 132.2.2 Transmission Line Parameters . . . . . . . . . . . . . . . 132.2.3 Equations . . . . . . . . . . . . . . . . . . . . . . . . . . 142.2.4 Types Of Transmission Lines . . . . . . . . . . . . . . . . 162.2.5 Termination Schemes . . . . . . . . . . . . . . . . . . . . 172.3 DC Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193 Modelling And Design . . . . . . . . . . . . . . . . . . . . . . . . . . 213.1 Electrical Model Of Transmission Line . . . . . . . . . . . . . . . 223.1.1 Equations . . . . . . . . . . . . . . . . . . . . . . . . . . Pair of Conductors with Circular Cross-section . Parallel Plate Type Conductors . . . . . . . . . 243.1.2 Existing Wire Parameters . . . . . . . . . . . . . . . . . . 253.1.3 Proposed Modification . . . . . . . . . . . . . . . . . . . 263.1.4 Experimental Verification . . . . . . . . . . . . . . . . . 263.1.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . 273.2 Thermal Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.2.1 Thermal Model Design . . . . . . . . . . . . . . . . . . . 293.2.2 Equations . . . . . . . . . . . . . . . . . . . . . . . . . . 293.2.3 Verification Of Model . . . . . . . . . . . . . . . . . . . 313.3 Hardware Set-up . . . . . . . . . . . . . . . . . . . . . . . . . . 323.3.1 Materials Required . . . . . . . . . . . . . . . . . . . . . 33v3.3.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . 343.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . 374.1 Tests On Regular Cable . . . . . . . . . . . . . . . . . . . . . . . 384.1.1 Open Circuit AC Tests . . . . . . . . . . . . . . . . . . . 384.1.2 Open Circuit DC Tests . . . . . . . . . . . . . . . . . . . 424.1.3 Waveform Properties and Measurements . . . . . . . . . . 474.2 Tests On Modified Cable . . . . . . . . . . . . . . . . . . . . . . 494.3 Load tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 544.3.1 Resistive Load . . . . . . . . . . . . . . . . . . . . . . . 554.3.2 Inductive Load . . . . . . . . . . . . . . . . . . . . . . . 594.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 635 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.1 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 645.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67A Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71A.1 Verification of the Cross-sectional Area . . . . . . . . . . . . . . 71A.2 Line parameters and PSIM Line Model . . . . . . . . . . . . . . . 72viList of TablesTable 3.1 Line parameters Of Regular Cable . . . . . . . . . . . . . . . 26Table 3.2 Line parameters Of Modified Cable . . . . . . . . . . . . . . . 26Table 3.3 Measured Parameters Of Regular Cable . . . . . . . . . . . . . 27Table 3.4 Measured Parameters Of Modified Cable . . . . . . . . . . . . 27Table 3.5 Regular Cable Summary . . . . . . . . . . . . . . . . . . . . . 28Table 3.6 Modified Cable Summary . . . . . . . . . . . . . . . . . . . . 28Table 3.7 Thermal Parameters Of Line . . . . . . . . . . . . . . . . . . 31Table 4.1 Transmission Line Parameters 1 . . . . . . . . . . . . . . . . . 48Table 4.2 Transmission Line Parameters 2 . . . . . . . . . . . . . . . . . 49Table A.1 Simulation Parameters for Regular Cable . . . . . . . . . . . . 72Table A.2 Simulation Parameters for Modified Cable . . . . . . . . . . . 72viiList of FiguresFigure 2.1 Common Types of Cables . . . . . . . . . . . . . . . . . . . 16Figure 2.2 Transmission Line Characteristics . . . . . . . . . . . . . . . 17Figure 2.3 DC Switching . . . . . . . . . . . . . . . . . . . . . . . . . . 20Figure 3.1 Transmission Line Model . . . . . . . . . . . . . . . . . . . 22Figure 3.2 Cable Cross-section . . . . . . . . . . . . . . . . . . . . . . . 23Figure 3.3 Thermal Equivalent Model . . . . . . . . . . . . . . . . . . . 30Figure 3.4 Thermal Model of Cable . . . . . . . . . . . . . . . . . . . . 32Figure 3.5 House Wiring Set-up . . . . . . . . . . . . . . . . . . . . . . 34Figure 3.6 Testing and Measurement Set-up . . . . . . . . . . . . . . . . 35Figure 4.1 Open-circuit Test Model . . . . . . . . . . . . . . . . . . . . 38Figure 4.2 Steady-state Configuration For Regular Cable . . . . . . . . . 39Figure 4.3 Steady-state AC Waveform . . . . . . . . . . . . . . . . . . . 39Figure 4.4 Open Switch Configuration . . . . . . . . . . . . . . . . . . . 40Figure 4.5 Step-up Transient 1 . . . . . . . . . . . . . . . . . . . . . . . 40Figure 4.6 Step-up Transient 2 . . . . . . . . . . . . . . . . . . . . . . . 41Figure 4.7 Closed Switch Configuration . . . . . . . . . . . . . . . . . . 42Figure 4.8 Step-down Transient 1 . . . . . . . . . . . . . . . . . . . . . 42Figure 4.9 Step-down Transient 2 . . . . . . . . . . . . . . . . . . . . . 43Figure 4.10 Open Circuit Test Model . . . . . . . . . . . . . . . . . . . . 43Figure 4.11 Open Circuit Test With Square-wave Input . . . . . . . . . . 44Figure 4.12 Regular Cable DC test 1 . . . . . . . . . . . . . . . . . . . . 44Figure 4.13 Regular Cable DC Test 2 . . . . . . . . . . . . . . . . . . . . 45viiiFigure 4.14 Simulation Verification 1 . . . . . . . . . . . . . . . . . . . . 45Figure 4.15 Regular Cable DC Test 3 . . . . . . . . . . . . . . . . . . . . 46Figure 4.16 Regular Cable DC Test 4 . . . . . . . . . . . . . . . . . . . . 46Figure 4.17 Time Delay Calculation . . . . . . . . . . . . . . . . . . . . 47Figure 4.18 Frequency Calculation . . . . . . . . . . . . . . . . . . . . . 48Figure 4.19 Open Circuit Test Model . . . . . . . . . . . . . . . . . . . . 50Figure 4.20 Open Circuit Test . . . . . . . . . . . . . . . . . . . . . . . . 50Figure 4.21 Step Response 1 . . . . . . . . . . . . . . . . . . . . . . . . 51Figure 4.22 Step Response-Simulation Result . . . . . . . . . . . . . . . 52Figure 4.23 Peak Overshoot . . . . . . . . . . . . . . . . . . . . . . . . . 53Figure 4.24 Test Circuit For Load Tests . . . . . . . . . . . . . . . . . . . 54Figure 4.25 Test Circuit For Resistive Load . . . . . . . . . . . . . . . . . 55Figure 4.26 Test Circuit For R−L Load . . . . . . . . . . . . . . . . . . 55Figure 4.27 Load characteristic 1 (Resistive Load) . . . . . . . . . . . . . 56Figure 4.28 Load Characteristic Simulation (Resistive Load) . . . . . . . 57Figure 4.29 Load Characteristic 2 (Resistive Load) . . . . . . . . . . . . . 58Figure 4.30 Load Characteristic 1 (R−L Load) . . . . . . . . . . . . . . 60Figure 4.31 Load Characteristic Simulation (R−L Load) . . . . . . . . . 61Figure 4.32 Load Characteristic 1 (R−L Load) . . . . . . . . . . . . . . 62Figure A.1 Cable Cross-section . . . . . . . . . . . . . . . . . . . . . . . 71ixGlossaryUBC University of British ColumbiaNSERC Natural Sciences and Engineering Research Council of CanadaPPFH People & Planet Friendly HomeICICS The Institute for Computing, Information and Cognitive SystemsDC Direct CurrentAC Alternating CurrentPCB Printed Circuit BoardHVDC High Voltage Direct CurrentEMF Electromotive forceMCCB Moulded case Circuit BreakersAFCI Arc Fault Circuit InterruptersAWG American Wire GaugeIEC International Electrical CodeRCD Residual Current DeviceXLPE Cross-linked polyethylenePVC Polyvinyl ChloridexLV Low VoltagePSIM PowersimNEC National Electrical CodeRMS Root Mean SquareIEC International Electrotechnical CommissionLCR Inductance(L), Capacitance(C), Resistance(R)xiAcknowledgmentsI offer my enduring gratitude to the faculty, staff and my fellow students at UBC,who have inspired me to continue my work in this field. I owe particular thanks tomy supervisor Dr. William Dunford and my Co-supervisor Dr. Martin Ordonez,whose penetrating questions taught me to think more deeply. They have helped medevelop curiosity and made me more self-motivated. They have taught me to beorganised and consistent with my work. Above all, they have kept constant faith inmy work and encouraged me throughout the process.This work was supported by PPFH initiative and I want to thank Dr. Panos Na-siopoulos and Dr. Mahsa Pourazad for accepting me as a student in this initiative.I would also like to thank Ion Isbasescu, David Campos Gaona and RafaelPena-Alzola for their help during my experiments. Special thanks to Mark Finnissfor helping me out with all the materials and the tools required in the workshop. Icouldn’t have done this without their help and support.I also take this opportunity to thank my lab mates, Matin Rahmatian, Veron-ica Galvan and all other colleagues who have encouraged me and also guided methrough some tough technical issues related to my research.My heartfelt thanks to my friends and well wishers to keep me motivatedthroughout my research term.Special thanks are owed to my parents and my sister, whose have supported methroughout my years of education, both morally and financially. They have alwaysstood by me during my success and failure.xiiFor my parentsxiiiChapter 1Introduction1.1 MotivationDirect Current (DC) distribution systems are becoming very popular nowadays andare being used for various applications. DC is largely used in telecommunicationapplications and certain renewable energy sources produce DC output. With theincreasing number of alternative energy resources such as solar power coupledwith batteries, there is abundant energy in the form of DC available.DC as power source has been used in many devices, DC is also being used asa part of transmission in High Voltage Direct Current (HVDC) systems; but littleis done to implement DC power as the primary source of power at the distributionlevel.The main challenges while using DC in low voltage application such as homes,buildings and commercial applications is the compatibility with existing systemand the costs involved in modification to suit the new system.Thus the primary aimof this work is:-1. To evaluate the DC distribution systems for homes and check for compatibil-ity in terms of power rating, load types and relevance to the existing system.2. To propose a DC system configuration having better system characteristics,(i.e. switching transients, thermal properties) while maintaining the similarvoltage and current ratings to the existing AC systems.13. To modify the existing wiring systems by constructing a modified cablestructure with line parameters such as to suit the system requirements andhaving the same current rating as that of the existing electrical wiring.When implementing DC, defining a voltage standard for DC distribution becomesvery important. For Telecom application, 48 V DC input is considered as a stan-dard and is within safety limits. The same DC input voltage can be used for lowvoltage DC application but is not suitable for high power loads that are presentin homes. The voltage standards of existing loads are 120 V or 240 V in NorthAmerica, and 230 V in Europe. Currently, a +/− 190V [3]is being proposed fordata centers, is low compared to the existing voltage standards. A slightly highervoltage would be compatible with either Alternating Current (AC) or DC input.Establishing a voltage standard is an important step while implementing DC forresidential applications. Some of the possible DC voltage standards are evaluatedin[4] and [5].1.2 Literature Review1.2.1 History Of AC PowerIt all began in the time of Edison, when he built a power station to light incan-descent bulbs using DC power. The main drawback observed was that the powerdiminished with distance and it was concluded that the power houses must be nearthe source of electricity. This is mainly due to the losses in the cable or the lossesin the resistance of the wire, popularly known as the losses due to I2R.Power delivered to the load (anything which uses power) is defined as the prod-uct of Amps and Volts, or P=IV. On the other hand, line losses (the energy lost inthe transmission wires) are determined by the product of Amps squared times ohms(the resistance of the wire), or I2R. The voltage plays no part in line losses.With AC, then, the transformer (a low cost device that are only applicable toAC systems) stepped up the voltage (or down, depending on the ratio of the numberof winding of input and output) to hundreds of thousands of Volts. Naturally, sinceoutput power must equal power input (minus losses), the AC current decreased inthe same proportion[6].2Thus, super-high voltage and super-low current meant very low line losses irre-spective of how far you needed to send it. Of course, very high voltage is dangerousfor appliances, lights, and motors. With AC, however, once you get the power tothe home, farm, or shop, a second transformer (on the utility pole) would step thevoltage back down for use. The use of AC power continued until the demand forpower became ever-increasing.1.2.2 AC Versus DCThe current distribution system has been predominantly AC for a very long time.Considering DC systems to replace at least certain parts of the existing systemwould prove to be better with respect to many aspects. The most talked about ad-vancement would be in the field of distributed generation and Micro-grids. Theelectric power systems has become increasingly distributed while trying to includemore and more renewable energy resources. Clearly, some of the renewable re-sources such as Photo-Voltaic are inherently DC supplies and cannot be incorpo-rated in the current system without power electronic interfaces.The provision of energy storage is virtually non-existent in AC distributionsystems but storage of DC energy in batteries is very practical. The DC distributionsystem will also minimize conversion losses incurred in the existing systems alongthe way [7].In addition, DC offers greater capacity per power line, reducing the weight andsize of power equipment throughout the grid and home, and can help to improvethe reliability of power supplies as it is easier to control the voltage on DC lines.Thus the main benefits of DC based grids would be:• Greater energy efficiency• Higher power quality• Smaller equipment• Less complexity and lower costsThough DC has more advantages over AC systems, some of the critical pointsthat need to be addressed while using DC systems are the Voltage transformation,3Circuit breaker protection and Voltage stability. The voltage transformation in ACis possible with ease using transformers but similar function in DC systems can bequite complex while using DC converters[7].The circuit breakers used in AC systems are more prominent and more devel-oped in comparison to the existing technologies for DC systems. Some of the TripUnit technologies used in DC circuit breakers are the Thermal-magnetic Mouldedcase Circuit Breakers (MCCB) and pneumatic-Magnetic MCCB. The usage of cir-cuit breakers with electronic trip units is not possible for DC circuits since most ofthese use current transformers to sense currents[8].The tripping time of the circuit varies in AC and DC systems. While inter-rupting a circuit, the inductance of the circuit stores energy that will expend itselfthrough an electric arc during circuit switching or interruption. This is importantmainly in DC circuits due to the absence of periodic zero voltage crossing, where acircuit breaker have higher probability of extinguishing a fault current arc. The keypoints that are necessary to extinguish an arc is to raise the voltage across the arcor by reducing the conductivity of the plasma region by decreasing temperature[8].DC distribution clearly has lot of potential in terms of reliability and securityof the system, energy efficiency and compatibility with multiple sources,load anddemand side management. The use of DC microgrids (could be in any scale), andthe practicality of such a system in today’s scenario where electronic loads areever growing, can change the centralised generation and distribution of the powersystem to one that is more accommodating of the load and more flexible in termsof power generation.[9].1.3 Existing TechnologiesThere has not been a lot of work done on implementing DC for homes. Some of thework done includes the development of a DC compatible plug by companies suchas in reference[10], which addresses the issue of arcing. A magnetic circuit is usedto spread the energy of the arc to enable extinguishing the arc easily. Addition-ally, apart from the commercial products, research has been done on reducing thetransients by using effective snubber circuits across the plug and load side[11][12].An arc fault is an unintentional arcing condition in an electrical circuit, which4may lead to fire hazards in case of sustained arc formation. The causes of arcare mainly due to damaged wires, breakdown of electrical insulation, overheatingof cables and wires.Having identified the main problem, arc fault detection is notvery simple. The key problems while detecting an arc is the current absorption inmany appliances that are used in homes, that tend to mask or attenuate the arcingcharacteristics[13]. These reasons may allow arcing to go undetected, which isvery crucial when using an Arc Fault Circuit Interrupters (AFCI).Since in AC systems the arc can reappear every half cycle, understanding thebehaviour of DC systems with respect to the time taken for arc formation and thetime to failure of the system becomes necessary.With the advent of power electronics and modern control strategies, more andmore number of loads use rectified AC mains supply. The growing number ofloads not only affect the power quality but are vulnerable to power disturbances.As mentioned in [14], power disturbance not only affects the loads but also theoperation of the power systems. For example, harmonics can create erroneousreading in meters, can affect operation of devices and also lead to heating of cablesas well as produce eddy currents in transformers. A thorough study on the usabilityof alternative power distribution in residential and commercial buildings has beendone in [15]. Various loads such as electronic appliances are on the rise and arecompatible with DC supplies. Additionally the AC line voltage creates currentdistortion and complicates the construction when using rectifiers to supply theseloads.Feasibility analysis of DC distribution systems is done in [3] , in which thelosses in cables with respect to the voltage chosen (48V , 120V or 230V DC) is cal-culated to estimate the ideal voltage standard to be chosen in order to ensure min-imum modification of existing cables and architecture. This analysis helps studythe importance of choosing the right voltage standard for a new DC distributionsystem.DC switching is another main area of concern. The ideal switching scenariowould be when there is a physical disconnect between the source and the load [11].But practically this is not the case. The current continues to flow to the load byionizing the air around the switch resulting in the formation an arc.In the real case,the wiring and the load have inductive behaviour. There is energy stored in these5elements, which refuse to change instantaneously with the opening and closing ofthe switch. The only way to dissipate this energy is through the arc formed acrossthe switch, causing power loss. In AC systems the arc would ideally quench in thenearest zero crossing of the current wave, but DC systems would need an externalaid to duplicate this behaviour.Typically, allowing an alternative path for the inductive energy would solve theproblem. Adding a snubber circuit as demonstrated in the work in reference [11]and [12] would be useful. There can be many drawbacks when including such acircuit in terms of the charged capacitor which continues to hold some voltage.This is not very desirable in terms of safety in homes, while accessing the outletsfor plugging and unplugging loads. In most cases using a bleeding resistance wouldserve the purpose, but it also depends on the discharging time of the capacitor. Thecharging constant must be comparable to the time taken to plug-in or plug-out theload [11][16].1.4 ObjectivesThe primary objective of this research is to propose a new and improved distribu-tion systems for homes with very little modification to the existing AC system.As a consequence, some the secondary objectives would be:1. Improving thermal and electrical properties of the wiring2. Improving switching characteristics3. Better integration of various energy sources4. Cost effective solution to using DC power5. Incorporating Power line communication1.5 OrganisationThis work has been organized as shown below.• Chapter2 describes the characteristics of DC distribution systems and in-cludes the application of DC distribution for Low voltage application, i.e.6DC homes. It also compares the existing system with the new system interms of the loads, protection & control methodologies. The drawbacks ofDC as a power source has been dealt with in detail in this chapter as well.• Chapter3 emphasises on the design aspects and explains the experimentalwork carried out. The theory and implementation of DC systems for homesapplication has been addressed in this chapter. The electrical and thermalmodel comparison of the line parameters when using DC or AC source isclearly indicated along with verification experimentally.• Chapter4 enumerates the results thus obtained while conducting tests on ahome wiring system while using AC supply. These results have been simul-taneously compared by substituting the source to a DC supply. Finally, thesame set of tests were carried out when a modified cable is used in the homewiring set-up when supplying DC. Some conclusions and analysis are drawnbased on these tests.• Chapter5 concludes the results and developments of the research and de-scribes the future scope and extension to this work.7Chapter 2DC Distribution For Homes:Theory and ApplicationA typical home runs its loads through AC supply. Considering a DC form of energywill bring about many changes in the existing homes. However, a DC source ismore advantageous over AC as described in the following sections.Today, energy is available from multiple sources, like utility, Photo-Voltaicpanels, wind turbines, possibly a small gas turbine. There will also be energystorage in the form of batteries and probably thermal storage. Certainly much ofthe electrical energy will be produced and used in DC form, so a DC wiring systemis a logical way of connecting the sources and the loads.Considering the various factors mentioned in Chapter 1, using DC in powerdistribution will help in moving towards a more green and efficient system.2.1 Design AspectsWhile considering the hardware required and distributing energy around the build-ing, there are several aspects that need to be evaluated. DC systems have been notvery popular because of the high cost of controlling and protecting them; the mainproblem being arcing while switching. In DC applications no natural zero crossingexits, hence quenching the arc can be challenging.In a house, portable loads are mostly used wherein electrical plugs become8inevitable. The electrical plug and outlet is similar to a switch. It has to be designedproperly for the safety of the users as well as the loads. The absence of zerocrossing in DC systems makes developing a switch a challenge[10].2.1.1 System TopologyConventional AC grids have many stages of conversion. The various componentsare ; Generating station, power transformer, transmission and distribution lines andautomation devices.The AC generated voltage is stepped-up and stepped-down in various stages.Initially the AC voltage is stepped-up before being fed to the transmission lines,to reduce losses while transmission. The voltage is again stepped-down at thereceiving end, i.e. at the transmission substation. The AC voltage is further re-duced using the distribution transformer before being fed to various low-voltageconsumers. Thus the AC systems are bulky and have high conversion losses.The topology of some of the existing DC systems can be evaluated before im-plementing a 400V DC Distribution system, e.g. A typical data center,Telecomcenter. In a data center, the high voltage AC line is stepped down and then con-verted to DC, to be able to be paralleled with Battery-Backup systems. The DC isthen converted to High-Voltage AC for distribution within the building, then con-verted yet again to Low voltage DC and then to voltages for other low voltagecircuitry using DC-DC converters. This has 4 stages of voltage conversion.In the case of telecom systems, there are two major stages of voltage conver-sion, but each of these stages are quite inefficient. Firstly the AC voltage is con-verted to 48 V DC and combined with the back-up batteries, which is also providedto the DC-DC converters that can supply the local low-voltage circuitry. HVDC Distribution SystemsLooking into an HVDC distribution system, the number of conversion stages aresimilar to that of a telecom center, but there is a large improvement in the systemreliability and conversion. The AC voltage is rectified to 380 V-DC (nominal)voltage, with a battery back-up operating at the same voltage. The DC voltageis then distributed to various low-voltage loads through DC-DC converters. This9system can draw power from AC grid, the batteries and also be integrated withother renewable sources such as wind and solar depending on the power demand.The concept of HVDC and protection have been ventured into since a very longtime. The Reference [17] talks about the principle and usability of the DC circuitbreakers. Choice of Voltage StandardsThe choice of voltage and phase configuration determines the ease with whichthe the current system can be modified to the suit the DC system. The cost ofconversion from an AC system as well as the compatibility to existing loads areimportant factors[11]. Sticking to the current voltage standards will reduce the costof wiring,cost of conversion of voltages and enable working with existing loads.A lot of research has been done on finding the optimum voltage standard forDC distribution, keeping in mind the voltage of distributed energy sources (whichis smaller than the existing AC voltage standards). Choosing a high voltage canhave issues relating to safety while choosing a low voltage can lead to major powerloss in cables. Based on the two important criteria of ’Voltage Drop” and ”PowerLoss”, the voltage standard and a suitable area of cross section of the cable hasbeen evaluated in Reference [5].2.1.2 ProtectionA conventional AC circuitry has circuit breakers and fuses provided for the purposeof protection and isolation.In a DC system,some protection is provided by the AC system circuitry thatfeeds it but additional protection circuit breakers and controls have to be providedfor each DC sub-circuit. The main disadvantage when it comes to DC system isDC switching; AC current periodically passes through a zero crossing making iteasier to extinguish arcs at current zero. Thus it becomes a costly and risky affairto install DC switches, circuit breakers and fuses which are difficult to design forvery high voltages.Apart from switching protection, protection from over current must also be ex-plicitly provided for DC circuits due to the absence of current limiting transformer10reactance as in case of AC circuits.2.1.3 Types Of LoadsMost types of loads that exist in homes operate on existing voltage standards andare easily suitable to DC systems. The common types of loads are explained below. Fluorescent LightingThe use of fluorescent lights are known to be energy efficient when compared toconventional incandescent lights. The efficiency of fluorescent lights is increasedfurther by using an electronic ballast. These ballasts basically improve the lightoutput per watt of power. A typical electronic ballast has an inbuilt rectifier thatrectifies the input AC voltage to DC and then it is converted to high frequency,which implies that usage of DC input would cut down the cost of the ballast whilecontrolling harmonic distortion that would otherwise affect the power system [11].The common phenomenon of ”flickering” of fluorescent lights is due to theoscillating nature of AC power. The high frequency ballast producing double ACline frequency components can be eliminated when DC input is used instead. Theflicker is affected by the power quality, i.e. the continuity of power, variation ofvoltage magnitude, transients and harmonic content in AC power [11]. This willno more be an issue when dealing with DC. Motor LoadCurrently, the motor driven appliances contain an inbuilt inverter. The ability tocontrol the motor based on the usage is provided by this inverter. Typically, a AC-to-DC converter is connected to a DC-to-AC converter to connect to the motor. TheDC-to-AC converter helps in controlling the characteristics of the inverter.Thus tomodify such loads to work with DC would be as easy as eliminating the AC-to-DCconverter stage [10].For instance; Various day to day household loads, like washing machines, thatrun on motors use a DC link that supply a stiff voltage to the variable frequencydrive. The front end has a rectifier supplying the DC link giving a pulsed inputcurrent to the inverter. Using a DC source would eliminate the DC link filtering11and the need for a rectifier, although pulsed current will still be drawn from thesupply by the inverter. It will be interesting to study what would happen if theinverter input capacitor were to be supplied directly from transmission line. Therewould be a potential inrush current at the beginning and perhaps an inductor wouldstill be required at the input to draw a smooth current[10] [11] [18]. Electronic LoadsThe common day-to-day electronic loads such as computers,television, mobile andlaptop chargers etc., mostly consume low DC voltage(about 3.3 V). There existsan AC-to-DC converter fed to a DC-to-DC converter to step-up or step-down to thenecessary voltage levels to match the load requirements. Hence, these electronicloads having a diode bridge supplying a high frequency inverter,can instead bepowered by DC too. But a similar issue with regard to the inverter input currentarises.2.1.4 Other AdvantagesThe presence of transients in terms of harmonics(due to 60 Hz oscillation) cancause interference with neighbouring communication lines. The power-line com-munication can be interfaced more easily when using a DC input [11].Another very common phenomenon of audible noise in fluorescent ballastsand transformers that include magnetic components. This is due to the vibration ofmetallic components at 60/120Hz due to the AC voltage oscillation. These heatingelements experience a ”humming” noise with AC excitation due to the pulsatingforces between adjacent current carrying wires. No such problem exists in DCsystems.There have been some health concerns due to the electric and magnetic fieldsassociated with AC transmission. The World Health Organisation gives statisticson the magnitude of AC and DC magnetic field effect on human health [19]. Thesestatistics imply that there exists no such adverse effects on health when exposed tostatic DC magnetic fields.122.2 Transmission LinesThe effects of transmission lines affect the transfer of signal from the sending endto the receiving end. An experimental set-up similar to the panel for homes wasfabricated to carry out tests to study switching behaviour of home distribution sys-tem. The set up conformed to all the rules as listed on the Canadian electrical code[20]. It was powered by the power supply i.e. 1φ 120V AC supply, which is thedistribution level voltage standard in North America. The testing was aimed toprovide important observations with regard to transmission line effects while usingDC power over AC, and based on various load characteristics.2.2.1 Transmission Line EffectsThe transmission cable used in the experiment has a specific material and lengththat causes it to exhibit certain non-ideal characteristics. The material of the cabledetermines the velocity of propagation, but in addition to that, the cable behavessimilar to a ”transmission line” which brings in various other factors in the pic-ture. Some of the effects that can be taken into account using the transmission linetheory include the signal reflections and the associated reflection noise, cross-talkbetween closely spaced traces, simultaneous switching noise due to inductance inthe power supply path connecting active drivers (this is of major concern especiallywhen using DC source for reasons mentioned previously). In signal and systemsterminology, these non-ideal effects are referred to as signal integrity, since thedesired signal is corrupted by these effects as explained in the reference [21].2.2.2 Transmission Line ParametersThe transmission line has various line parameters that can be determined from thephysical properties, i.e. geometry and material, of the line. The propagation veloc-ity can be calculated by considering the material properties, i.e. relative permittiv-ity and permeability with respect to those of air. The dimensions of the line playno role in the velocity of an electrical signal. The relevant formulae are presentedin Equation 2.1.The other line characteristics involved are the characteristic impedance of theline. This is the intrinsic impedance offered by the line to the flow of signal. Unlike13the propagation velocity, the impedance depends on the geometry as well as thematerial properties[21]. The calculations involved are mentioned in Equation 2.2& 2.3During these reflections, there may be instances where the voltage at the loadmay reach beyond the applied voltage due to positive reflection constant. The timerequired to propagate also determines the time taken for the signal to reach steadystate which depends on the propagation constant. The line parameters are calcu-lated by using the formulae below (Assuming zero resistance line). PropagationDelay can be calculated as shown in Equation 2.4.The Transmission line parameters appear as inductance, capacitance and re-sistance. The propagation constant is used to describe this characteristic and isshown in Equation 2.6. The resistance is offered by the geometry of the cable, asmentioned before. But the inductance of the line is due to the current flowing inthe conductor. The flow of current induces as magnetic field which is varying sinu-soidally with the voltage. Thus a varying magnetic field produces an Electromotiveforce (EMF) which opposes the current flow in the line and appears as the inductiveeffect on the line.A conductance parameter arises in the line due to the leakage current that flowsfrom the conductor to the ground. In most cases, this is neglected. But there is apotential difference between the phase conductors which gives rise to an electricfield. These lines act similar to the parallel plates of a capacitor. This arrangementgives rise to the capacitance parameter of the transmission line.2.2.3 EquationsThe main line parameters of a transmission line are the velocity of propagation,Characteristic impedance, Time Delay and the Propagation constant as explainedpreviously. The equations of the line parameters are listed below in detail.1Velocity of propagationThe velocity of propagation, also called wave propagation speed or velocity factor1where, L= Inductance per unit length & C= Capacitance per unit lengthD= Length of the cable & R= Resistance per unit lengthG= Conductance per unit length & w is the angular frequency defined as w= 2pi f , f=frequency ofsignal, i is the imaginary unit14of a transmission medium is the ratio of the speed at which a change of the electricalvoltage on a copper wire passes through the medium, to the speed of light in avacuum [22].v=1√LCm/s (2.1)Characteristic ImpedanceThe characteristic impedance or surge impedance (usually written Zo) of a uniformtransmission line is the ratio of the amplitudes of voltage and current of a singlewave propagating along the line; that is, a wave travelling in one direction in theabsence of reflections in the other direction. The general expression for the char-acteristic impedance is [23]:Zo =√R+ iωLG+ jωCΩ (2.2)The terms are defined in the footnote defined at the end of this page. For a losslessline, R and G are both zero, thus the Equation 2.2 reduces to;Zo =√LCΩ (2.3)Propagation DelayPropagation delay, as it relates to transmission lines, is the length of time it takesfor the signal to propagate through the conductor from on point to another [24].TD=Dvsec (2.4)Propagation Constantγ is the Propagation constant and is defined as the ratio of amplitude at the sourceof the wave to the amplitude at a certain distance (say x) such that [25]:AoAx= eγx (2.5)γ is a complex quantity and has a real part α and an imaginary part β . The realpart αis called ”attenuation constant” and the imaginary part β is called the ”phase15constant”2γ = α+ iβ = (R+ iωL)(G+ iωC)β = ω√LC (2.6)2.2.4 Types Of Transmission LinesLooking into the various types of transmission lines,there are the parallel platetype line, micro strips, regular two wires/ twin lead cables, coaxial cables. In theexperiments, the types of cables used are the regular two wire type cables (housewiring cables). Some of the analysis done using a modified cable is similar to theparallel plate type cable. The types are shown in the book [21]and the two maintypes that are being used in this thesis has been reproduced with guidance from thebook, in Figure 2.1.Figure 2.1: Common Types of Cables2where,γ =Propagation constant, α =attenuation constant & β = phase constant162.2.5 Termination SchemesWhen a signal propagates along the transmission line, it causes a wave to start totravel down the transmission line. For a DC signal on the line, this would be a volt-age or current wave. The amplitude of the wave is depicted by the line impedanceof the line and the impedance at ”input” or ”generator” end and the ”load” end.The difference in this impedance causes reflections of the waves back and forth.If the line is unloaded, i.e. open circuited, the voltage would produce reflectionswhile the current would be zero; while in case of a short circuit, the voltage wouldbe zero but the current across the load would be non-zero. In each case, after acouple of reflections, the voltage and current approach their steady-state values.The reflections of the voltage and it’s propagation through the line is shown infigure 2.2. This figure was taken from a supplementary material prepared for abook as cited in Reference [2]. The cause of these reflections is when a finiteFigure 2.2: Transmission lines: (a) Short-circuited; (b) Open-circuited.Adapted from [2]17length transmission line in not terminated in its characteristic impedance. Theextreme case would be when the far end is short− circuited. The explanation forthe short circuit situation is demonstrated in figure a) in 2.2. Initially, the sourcevoltage (or the driver), views the line’s characteristic impedance and send acrossa voltage wave of ’Vsrc/2’ (Refer figure 2.2). As the wave propagates across theline and reaches the end at time T , it views the short circuit. Thus to satisfy the thekirchoff’s laws, a voltage wave propagates in the opposite direction, to cancel theoriginal waveform. The far end will have the reflected original wave, and a shortcircuit is observed at time 2T .Another type of termination is the open− circuit, which is demonstrated inFigure b) in 2.2. In the beginning,this case is similar to the short − circuit caseexplained previously. But when the wave hits the far end at time T , the current hasnowhere to go and hence a voltage of the same amplitude propagates back, addingto the original voltage. Twice the voltage Vsrc/2 is observed at the driver side attime 2T . These reflections occur both at the near end as well as the far end of theline, if the source impedance (Rsrc) does not match the line impedance(Zo).The turning ”on” and ”off” of switch in a room connected to an incandescentbulb exhibits the reflection phenomenon as well. The movement of electrons alongthe wire is governed by the impedance mismatches along the cable. The resultingcurrent is perturbed due to mismatches either at source or the load end, causingreflections. In the practical sense, these reflections usually last for a very shortperiod of time (in nanoseconds) and hence can be ignored. But in certain cases,these reflections can cause significant amount of transients, causing power lossesin forms such as heat dissipation.The termination of the line is very crucial is depicting the reflections in theline. The correct termination scheme can eliminate the severe mismatches and thisprocess of eliminating the reflections is known ”Impedance matching”. There arevarious effective transmission schemes, e.g. diode termination that gives flexibilitywhile allowing to connect a load at the line end keeping the reflections minimum.Terminating the line with a non-resistive load, or reactive terminations can havea different effect on the transient. In the real scenario, most line terminations con-tain some reactance, like shunt capacitance or series inductance[21].182.3 DC SwitchingOne of the most important safety requirement while implementing DC system isthe quenching of arc while switching and prevention of the re-striking of arc. Themain reason for the formation of arc is due to the stored energy in the inductiveelement on both source and load end. There needs to be way to either store ordissipate the 12Li2 component of the inductive elements on both load and sourcesides.Direct current presents a different problem than alternating current with regardphenomena associated to the interruption of high value currents since arc extinctionresults to be particularly difficult. This problem is more prominent in DC systemssince AC systems can normally quench the arc in the subsequent zero crossingof the current. This is due to the oscillatory nature of the AC voltage. Some ofthe stored energy in the inductive element gets dissipated in the arc and the restcan return to the source and the load. The ionization of the air could continue tonext half cycle, but by this time the switch leads must be far enough to be ableto re-strike. In the case of DC systems there exists the possibility of sustained arcbecause of the absence of natural zero-crossing of the current. Figure 2.3 shows thecircuit elements in a typical DC circuit indicating the storage elements, source &load inductance. The steady-state equation of the circuit shown is described below.3Ldidt=U−Ri−Ua (2.7)To guarantee arc extinction, it is necessary that:didt< 0 (2.8)Equation 2.8 implies that for the arc to be extinguished, the Ua (arc voltage) mustbe higher than theU (source voltage) such that equation 2.7 becomes negative. Theextinction time of a direct current is proportional to the time constant of the circuit,τ = LR and also to the extinction constant. The extinction constant is a parameterthat is dependent on the arc characteristic and the circuit supply voltage [26].3where U is the rated voltage,Ls is the source inductance, L is the load inductance,R is the resis-tance of the circuit & Ua is the arc voltage19Figure 2.3: DC SwitchingThe second sub-figure under figure 2.3 is a generic DC switch circuit with adecoupling capacitor C1 which helps in closing of the switch and a flywheel diodeD1 ,across the R− L load. The capacitor C2 is pre-charged to the line voltage,thus at the instant of opening of the switch, the source voltage is impressed uponit, forcing the current across the switch to zero. The key element in the design isthat for a given load or overload current the capacitor voltage must fall at a ratecompared with the switch blade opening such that the switch voltage is always lessthen the ionization voltage of the air [11].20Chapter 3Modelling And DesignAccording to the electrical installation guide [27], Low-Voltage installations areusually governed by a number of regulatory and advisory texts. Based on thesetexts, the voltage ranges in North America have been defined based on the Na-tional Electrical Code (NEC)1voltage standards and recommendation. The commonnominal voltages (three-phase four-wire or three-wire systems) are 230/400V or380/660V in many countries like Europe[27][29]. The electrical Installation Guideis based on relevant International Electrotechnical Commission (IEC)2standards[20].The power distribution utility connects Low Voltage (LV) neutral point to itsMV/LV distribution transformer to earth. All LV installations must be protected byResidual Current Device (RCD) and all exposed conductive parts must be bondedtogether and connected to earth[27]. A residential wiring system usually includesa meter and in some cases an incoming supply differential circuit-breaker whichinclude over-current trip.1NEC, or NFPA 70, is a regionally adoptable standard for the safe installation of electrical wiringand equipment in the United States. It is part of the National Fire Codes series published by theNational Fire Protection Association (NFPA), a private trade association [28].2IEC is a non-profit, non-governmental international standards organization that prepares andpublishes International Standards for all electrical, electronic and related technologies collectivelyknown as ”electro-technology” [30]213.1 Electrical Model Of Transmission LineThe behaviour of a cable can be modelled as a transmission line. As explained inthe previous sections, we can model a line using the per unit Capacitance and perunit Inductance. The line model is shown in Figure 3.1. The model shown wasdeveloped using Powersim (PSIM) software. Additional parameters of Rser andRpar are taken into account for line losses. The line parameters are:1. The series inductance that arises due to the emf produced by a varying mag-netic field.2. The shunt capacitance arising due to the voltage difference between thephase conductors.3. Series resistance offered by the material and geometry of the line itself.4. Shunt resistance or conductance developed by the leakage current that flowsfrom the line to ground.Figure 3.1: Transmission Line ModelThe main problem that arises due to these parameters is arcing while switching(refer 2.3). To reduce these effects, we need to minimise the transients in the line.This is done by reducing the impedance offered by the line. The main objectiveof modifying the existing wire would be to reduce the inductance of the wire andincrease the capacitance, to effectively reduce the line impedance.3.1.1 EquationsThe equations used for this analysis are listed below. For the regular conductor withthe circular cross-section the calculation of the capacitance of the 14AWG wire can22be calculated based on the equations described in Section 3.2.2. The dimensionsand configuration is explained in Figure3.2.Figure 3.2: Cable Cross-section3.1.1.1 Pair of Conductors with Circular Cross-sectionThe ordinary 14/2 cable used for home wiring is the NMD90 14/2 Romex cablewith a specified insulation thickness as listed in Reference [31]. Wire types forNorth American wiring practices are defined by standards issued by UnderwritersLaboratories, the Canadian Standards Association, the American Society for Test-ing and Materials, the National Electrical Manufacturers Association and the Insu-lated Cable Engineers Association. The common insulation type used is XHHWwhere XHHW stands for ”XLPE (cross-linked polyethylene εr = 2.25 [32]) HighHeat-resistant Water-resistant”. XHHW is a designation for a specific insulationmaterial, temperature rating, and condition of use (suitable for wet locations) forelectrical wire and cable [33][34][35].The inductance of the pair of cables can be calculated as described in equation3.1. It can be seen that the calculations are done for a 12AWG cable instead ofthe 14AWG. This is because of the inability to construct a flat cable with the samecross sectional area as the 14AWG, due to the unavailability of tapes of convenientdimensions. Though there are a variety of tapes of varied thickness and width, theone that had the exact match to the commonly used twin wire cables was the 12/223cable. The calculations are provided in the appendix section to verify the aboveobservation (refer Appendix:A.1). 3.The type of cable is the two wires/ twin lead type of cable, where the conductors(live and neutral) are placed side by side with insulation around each conductor aswell as a jacket around the whole assembly. The equations used are used from areference book for transmission lines [21] [38].L= 4×10−7cosh−1(D2r)H/m (3.1)The capacitance of the pair of cables can be calculated from the equation 3.2.C =piεcosh−1( D2r )F/m (3.2)The dimensions of the cables are mentioned as used in the calculation later in thechapter. Parallel Plate Type ConductorsAlong similar lines, the line parameters of the parallel plate type or the modifiedflat type cable needs to be calculated. It is quite obvious that this type of designwould increase the capacitance between the pair of cables, since they behave likethe plates of a parallel-plate capacitor. Hence the equations would be similar to thecapacitance equations for two parallel plates.Clearly, the capacitance of the cables is much higher than the regular typeof cables but the inductance might not be significantly lower. But the effect ofincreased capacitance and almost the same inductance might have a large influenceon the transient characteristics of the cable.To verify the above assumptions, the line parameters were calculated for a cablethat is designed to have:1. Equal cross-sectional area as that of the conventional cable : to ensure thesame current rating as that of the regular cable.3where,r = radius of the conductor (12AWG wire of 0.08”mm)D= distance between the conductors = 2.38mmε = εo× εr = 8.85pF/m(Permittivity of Vacuum)×2.25 (Relative permittivity of Polyethylene[32])242. Same insulation thickness around the cable: to adhere to the insulation rulesas stated in electrical guides such as References [31],[36] and [37].3. Maximum width while maintaining the same area of cross section: to allowbetter electrical and thermal characteristics. Thus the cable can be madeas thin as possible depending on the convenience of fabrication and alsoinstallation.In our study we have considered a width suitable to be fabricated using availableflat wire strips, while ensuring improvement in electrical and/or thermal character-istics. The equations for inductance and capacitance of the cables are listed below.4.L=µo×DwH/m (3.3)C =ε×wDF/m (3.4)The dimensions of the cables are mentioned as used in the calculation later in thechapter.3.1.2 Existing Wire ParametersTo study the wire parameters of a regular cable, the parameters were measured toanalyse the magnitude of the line parameters. These parameters are quite insignif-icant over a short transmission distance, but in homes and other applications, thelength of the cable is not negligible.The common wire gauge used for homes is 14 American Wire Gauge (AWG).The diameter of such a wire is around 1.62mm in diameter with a RW90 Cross-linked polyethylene (XLPE) insulation of 0.030” (0.76mm) thickness[36][37]. Thedimensions are shown is Figure 3.2.4where,w= width of the conductor(L×w= pir2)Taken w= 0.01”; implied L= 0.5”D= distance between the conductors = 0.09”(2w+2×insulation thickness of 0.03”)µo = 0.4piµH/m (Vacuum Permeability)ε = εo× εr = 8.85pF/m(Permittivity of Vacuum)×2.25 (Relative permittivity of Polyethylene[32])253.1.3 Proposed ModificationWhen considering the flat wire configuration (laminated cable), it can be comparedto that of a parallel plate capacitor; one plate being the live wire and the other beingthe neutral wire with a dielectric medium of the insulating material between them.Clearly, this creates a uniform high capacitance along the length of the transmissionline reducing the inductive nature of the wires. Such a cable has a convenientgeometry when being used to mount on flat surfaces like walls of a home.The calculations for the electrical parameters rectangular cross sectional wirewere done based on the equations for design of a micro-strip for Printed CircuitBoard (PCB) (Printed Circuit Boards). The PCB have the similar problem ofswitching transients where the transmission line model displays wave character-istics. These boards have signal interference among neighbouring cables which areetched on the same plane. A similar characteristic can be adopted while modellinga flat cable wire, in which the conductor will be a copper sheet with the same in-sulation thickness (according to the NEC standards) for a 14AWG wire. The resultsare tabulated in Table 3.1 and Table 3.2.Table 3.1: Line parameters of Regular Cable(Refer 3.1.1)Line Parameters Inductance CapacitanceFormula 4×10−7cosh−1( D2r ) piεcosh−1( D2r )Calculated Value 0.46µH/m 53.98pF/mTable 3.2: Line parameters of Modified Cable (Refer 3.1.1)Line Parameters Inductance CapacitanceFormula µo×Dwε×wDCalculated Value 0.18µH/m 110.62pF/m3.1.4 Experimental VerificationA 50 feet long cable was used when fabricating the hardware set-up. This length ofwire that was chosen accounted for the parasitic involved in a two- storied home.26With the aim of verifying the theoretical values for the line parameters, they weremeasured using an impedance meter, by conducting the open circuit and short cir-cuit tests. Using the cable, the inductance and capacitance per meter length werederived. The results of the total inductance and capacitance are as tabulated inTable 3.3.The per unit parameters can be calculated easily by dividing the total value thusobtained by the 50 f t (15.24m) of wire which were verified to be matching to thoseobtained theoretically, as tabulated in the previous Tables 3.1 and Table 3.2. Thecalculation assumed a lossless line, but in real case there exists a series impedanceand a conductance.Table 3.3: Measured Parameters Of Regular CableShort Circuit Test Ls = 0.41µH/m Rs = 12.55mΩ/mOpen Circuit Test Cp = 47.88pF/m Rp = 86.41MΩ/mTable 3.4: Measured Parameters Of Modified CableShort Circuit Test Ls = 0.88µH/m Rs = 37.07mΩ/mOpen Circuit Test Cp = 82.78pF/m Rp = 84.23MΩ/m3.1.5 SummaryTo summarize the values obtained for the line parameters using the well-knownformulae and to verify with those obtained from the test results using the Induc-tance(L), Capacitance(C), Resistance(R) (LCR) meter for the cables, the data hasbeen tabulated in Tables 3.5 & 3.6.From the values summarized, it can be observed that there is an improvement inthe line parameters of the modified cable, and has higher capacitance between thelines. This means that this modified line model is better in transmission of voltagealong the line, resulting in improved transient characteristics (reduced reflectionand peak overshoot)in the line. In other words, the line model now offers smaller27Table 3.5: Regular Cable SummaryLine Parameters Inductance CapacitanceTheoretical 0.46µH/m 53.98pF/mMeasured 0.41µH/m 47.88pF/mTable 3.6: Modified Cable SummaryLine Parameters Inductance CapacitanceTheoretical 0.18µH/mm 110.62pF/mMeasured 0.88µH/m 82.77pF/mline impedance when compared to the one with lower capacitance per unit lengthof cable.Commenting on the error margin between the calculated and the theoreticalvalues, it can be seen from Table 3.5 that the values are very close to each other.But when looking at Table 3.6, there’s quite a bit of variation in the measured andcalculated values. The cause of error in the modified cable measurements could beeither due to the mechanical construction or due to errors during measurement.The flat cable was fabricated by placing the insulation layer by layer on theconductor strips such that the insulation layers were sandwiched between copperconductor plates.1. While placing the layers of insulation and the copper tape, there could havebeen air bubbles in between the layers which modifies the medium in be-tween the conductors.2. Due to human error while placing the insulation, there could have been cer-tain areas of the conductor that might have been exposed without any insu-lation.3. The insulation coating around the flat copper strips was done using a dou-ble sided adhesive insulation tape. This is an acrylic adhesive of unknownrelative permittivity εr and could be acting as an additional insulation jacketthus modifying the electrical properties of the cable.283.2 Thermal ModelThermal behaviour of the cables is attributed to the material of the cable compo-nents; the conductor, the dielectric and the sheath losses. The conductor is usuallymade of a single copper wire or many thin strands of copper twisted to form astrand of wire. The insulation jacket surrounds the conductor pair, generally witha ground copper wire running along the center. The insulation is made of Nylon orPolyvinyl Chloride (PVC).The cable carrying current tend to dissipate power through heat. The insula-tion is designed to withstand a minimum temperature rise caused by these currentcarrying conductors. The ampacity of a 14 AWG (American Wire Gauge) wire isabout 20 amps, in which some of the power is used in heating of the conductor.Ideally this heat, that causes temperature rise in the conductor, would be uniformthroughout the copper material whereas the insulation would have a non-uniformheat flow depending on the proximity of the heat source. To summarize, the con-ductors are single point heat sources and the dielectric medium distributes heatalong the length of the cable.3.2.1 Thermal Model DesignThe thermal model of a cable is as significant as the transmission line model. Thethermal characteristics can be modelled as shown in Figure 3.3. A current carryingconductor emits power in the form of I2R loss, which is converted to the form ofheat. The insulation is affected by the heat loss emitted by the conductors. Theinsulation as well as the geometry of the conductor can determine the efficiency ofheat transferred into the environment.The thermal model shown in the figure is for a two conductor model, withJ being the heat source, i.e. the current carrying conductors, and Rth being thethermal resistance offered by the insulation material. TA is the ambient temperatureand T1 and T2 are the temperatures at the two conductors.3.2.2 EquationsThe thermal resistance of the insulation (or conduit), made of PVC, has thermalresistance of k = 0.51W/mK. The thermal resistance is calculated as shown in29Ambient Temperature:TA = 20◦C(Refer Figure ??)Figure 3.3: Thermal Equivalent Modelequation 3.5. 5Rth =Lk×A (3.5)where k is the thermal conductivity of the insulator material, L is the length of theflow of heat (i.e Diameter of cross section in this case) and A = 2pir2l (l is thelength of the cable) is the area of the curved surface. For the proposed flat cable, Lwill be the breadth of the cross section of the conductor and A = l×b will be thearea of the lateral side.The electrical resistance of the cable is given by a similar formula shown in5considering; k = 0.51W/mK (for Polyethylene insulation)30equation 3.6. 6.Re =ρ×LA(3.6)where ρ is the Electrical resistivity of the conductor material, l is the length of thecable,A is the area of cross section depending on the geometry of the cable.The power loss in the cable can be equated to the temperature rise in the insu-lator due thermal resistance offered by it (refer equation 3.7).J =V 2Re(3.7)T = TA+(J×Rth) (3.8)Using the above equations, the temperature of the conductor can be calculated asin equation 3.8. Below are the calculated results in table 3.7. 7Table 3.7: Thermal Parameters Of LineConductor Temperature(◦C) Temperature(◦C)Regular cable Modified cableLive Conductor 21.80 21.80Neutral Conductor 21.80 22.773.2.3 Verification Of ModelThe Thermal behaviour of the two conductors was analysed using Elmer softwarepackage. Elmer is an open source modelling software and is very handy is carryingout simple thermal modelling for the purpose of research[39].The temperature distribution was modelled based on the resistivity of the cop-per in the conductor and the thermal resistance of the PVC (Poly-vinyl Chloride)insulation.Figure ?? is the cross section of the flat cable such the conductors are placedabove each other. The sub-figure (a) is the cross section of the regular cable where6considering; ρ = 1.68×10−8 (for copper conductor)7Ambient Temperature: TA = 20◦C31the conductors are placed side by side. The ground wire runs in between the liveand the neutral wire.The maximum temperature of the cables shows improvement in the model.Additionally, the heat distribution in the flat cable is much more efficient owing tothe larger surface area.Figure 3.4: Thermal Model of Cross-section of Cable3.3 Hardware Set-upPower is distributed to residences and cottages through overhead wires or under-ground cables. The service supplied is a three-wire service consisting of two liveconductors and a neutral wire. Three-wire service provides 120 V, 120/240 Voltsand 240 Volts capabilities. In North America, individual residences and small com-mercial buildings will usually have three-wire single-phase distribution, often withonly one customer per distribution transformer. Standard frequencies of single-phase power systems are either 50 or 60 Hz. Single phase is commonly divided32in half to create split-phase electric power for household appliances and lighting.Since all the voltages in a 1φ phase supply vary in unison, a revolving electric fieldcannot be produced unlike the 3φ voltage supply. Thus motor loads are not selfstarting and a single phase motor would require an additional circuitry for starting.3.3.1 Materials RequiredThe standard way of measuring the thickness of a wire is AWG. It indicates thethickness of the wire only (not including the insulation).The wiring of homes generally use Multi-conductor cables. These are basicallycables consisting of two or more conductors. The general house wiring wouldhave two insulated conductors, for the live and the neutral , and one bare copperconductor for the grounding. Such a cable is referred to as the ”14/2 with groundcable”, since the cables are 14 AWG thick conductor pair.The commonly used wire sizes for wiring in homes are 14, 12, 10, and 8 AWGconductors. Thickness of the conductors determines the current rating of the ca-bles, and is decided based on the amperage required for a particular circuit, whichis determined by the connecting load in case of home applications. For eg: A 14AWG wire can be used for 15 Amp circuits such as supply loads like receptaclesand switches.A higher amperage would mean a thicker cross sectional area and lower AWG;for example, a 12AWG wire is used for 20A circuits and so on. Thus the currentrating of a cable is known as the Ampacity and is defined by the NEC as the maxi-mum amount of electrical current a conductor or device can carry before sustainingimmediate or progressive deterioration.The switches used in the wiring uses a single pole switch that is used to controla light or an outlet from one location only. Other types of switches used are thethree-way switches and the four way switches based on the number of controlpoints.The house wiring requires the installation of receptacles (outlets) and must beof good quality in areas where there is usage of outlets constantly.333.3.2 ImplementationBased on the 2012 Residential Wiring guide and the North American Electricalcodes, the wiring of a panel of a home wiring set-up was modelled as shown in fig-ure 3.5. The set up shown was done using the regular 14 AWG wiring. A modifiedFigure 3.5: House Wiring Set-upflat type cable was designed to carry out tests for similar conditions. The flat cablewas insulated and tested using DC signal to analyse the transient characteristics ofthe cable, the results of which are shown in Chapter 4. The modofied cable and themeasurement and test set-up are shown clearly in figure 3.6.The input plug was given either an AC/ DC signal using the 120V power sup-ply/ Laboratory DC power supply. The house set-up included one mechanicalswitch and one plug outlet to observe the transient response under various con-ditions.The mechanical switch is the standard switch used for homes and buildings.The switching action while constant turn-off or turn-on action over time can causedamage to the properties of the mechanical contact. These switches seem to havean inbuilt snubber circuit to prevent switching at peak voltages (oscillating ACsignal).The outlet is a double pole plug with two loading points. Some loads are me-chanically pulled out or plugged in without the use of switches which affect thehealth of the plug. The transient while switching or plugging are the main areas of34Figure 3.6: Testing and Measurement Set-upstudy in the following chapter.The wiring used in the set-up (as shown in 3.5) clearly shows the excess lengthof cable used that is would in a coil. This extra length was used to take into accountthe length of a typical home wiring. A maximum length of about 50 feet (i.e15.23m) was used to model a electrical design close to reality.The distribution box used is typical of a home wiring, from where multipleparallel circuits are made. A main circuit breaker is installed in the module topower the whole house circuitry. In addition to which, there needs to be inserted asecondary circuit breaker for controlling each parallel circuit mainly to isolate thecircuit under emergency, to enable normal operation of other circuits.3.4 AnalysisBased on the results tabulated and displayed above, it can be clearly seen that:1. The inductance of the rectangular cross section wire is much smaller thanthat of the existing wire. The capacitance of the rectangular cross section35wire is much larger than that of the existing wire.2. Thermal characteristics of the flat cable is better for the same current carry-ing conductor with the same thickness insulation as seen in the regular cable.The configuration of the conductors in the modified cable (placed one abovethe other) has a better temperature distribution when compared to the regularcable configuration (placed side by side).3. The regular conductor model is shown to experience higher temperature gra-dient and sudden changes in temperature. Thus, to transfer the same amountof current, a thinner copper conductor would be sufficient in case of themodified cable structure. This also improves cost of the construction of thecable.4. Also mechanically, the flat cable would be more convenient to clamp downand will have more contact area with the clamping surface making it neaterwhile wiring.5. The capacitance and the thermal characteristics can be further improvedmaking thinner conductors, having wider geometry. The more the area ofthe plates, higher will be the capacitance in addition to having better heatdissipation. But making the conductors very flat, can make them more frag-ile and susceptible to damage.36Chapter 4Experimental ResultsThe switching characteristic at the load outlet and at the switch were studied toanalyse the effect of the cable parasitic at the load end. When applying an ACsupply, the instant at which the load is unplugged is very crucial. At the peak of ACvoltage, the switching transients are more pronounced due to the high voltage thatis impressed across the open switch. To be able to analyse the high instantaneousvoltage across the switch, a number of tests were carried out to study the behaviourof the transmission line (cable).The primary aim of the tests were to study the existing switching transients inthe current wiring architecture and to bring about improvement with the modifiedcable. Various tests were conducted on the cables with DC supply to study theswitching behaviour of the cables.A low-voltage DC about (40V ) was given tothe system instead of the AC supply itself. It was interesting to see the ringingand the oscillations due to the line inductance and capacitance. There could beadditional reasons to the ringing effect; like the mechanical characteristics of theswitch itself.Voltage at the load (plug outlet) is expected to be same as the inputsquare wave varying between 0 and 40V . On the other hand the voltage acrossthe switch must be zero since the circuit is switched on. Since there is no loadconnected to the outlet, there will be no current drawn at the output.374.1 Tests On Regular CableThe primary focus being on the transient characteristics of the house wiring, vari-ous tests were conducted to study the behaviour of the house wiring while switch-ing and under loading conditions. The following are the tests that were carriedout.To observe the transients that occurs while switching, the transmission linereflections in the output voltage and the input current, series of tests for both ACsupply voltage and low voltage DC supply were conducted. The results are shownin the following sections.4.1.1 Open Circuit AC TestsThe general test set-up for AC test is as shown in Figure 4.1. Depending on whetherthe load is being plugged in or out or an open circuit test is being performed,the corresponding switch would be engaged. For example; an open circuit testwill not involve the study of load switching, hence switch S2 (Refer Figure 4.1)will be permanently in a closed position. The switch S1 would be the one that’sbeing turned on and off. In case of the home set-up, the switch S1 is the circuitbreaker provided in the distribution box itself and switch S2 is the mechanicalswitch installed to control the supply to the outlet.Figure 4.1: Open-circuit Test ModelBeginning with the steady state performance of the test set-up, the set-up is pluggedinto the AC supply. The results of the tests are shown in Figure 4.3 and the switchcondition for the same is indicated in Figure 4.2.38Figure 4.2: Steady-state Configuration For Regular CableFigure 4.3: Open-circuit Current(Ch1) & Voltage(Ch2) Waveform DuringSteady-state Tested On Regular CableTo observe the switching characteristics, the switch S1 is turned on and off to studythe peak voltage and the time taken for the oscillations to die out. The first test isdone to view the switching behaviour while turning on the switch S1, the switchingcondition for which is shown in Figure 4.4, followed by similar tests to observe theturn-off characteristics. The open circuit test of a long transmission line is expectedto provide the same voltage drop at the outlet as that supplied at the source end.The absence of load would produce no current at the output. Though there’s no39Figure 4.4: Open Switch Configurationcurrent drawn, the switching action causes reflections in the form of oscillations inthe current as well; refer figure 4.5 & figure 4.6. It is observed that the Root MeanSquare (RMS) value of the peak to peak voltage i.e.2×Vac×√2 = 360V and thecurrent is 0A.Figure 4.5: Open-circuit Current(Ch1) & Voltage(Ch2) Waveform DuringSwitching (Turn-on) Tested On Regular CableThe transients observed in Figure 4.6 can be due to more than one factors. The mostobvious reason is the transmission line behaviour of the cable, but this could alsoarise because of the source impedance offered by the AC source (the single phasepower supply). The mechanical ”switch bouncing” effect can also contribute to the40Figure 4.6: Open-circuit Current(Ch1) & Voltage(Ch2) Transient WaveformDuring Switching (Turn-on) Tested On Regular Cabletransients obsereved. Thus we cannot fully isolate the cause of the transients dueto the multiple factors involved in this case.The high frequency transients also arise the possibility of an additional inbuiltcircuitry in the switch that forces turn-off at zero voltage, forcing the current to zerovery quickly. In the following tests it was also observed that the turn-off alwayshappened at zero voltage, strengthening the previous assumption of the presenceof an inbuilt snubber circuit.The same set of experiments were carried out, but this time while the switchingoff switch S1, the switch condition for which is indicated in Figure 4.7. The resultsare followed in Figures 4.8 & 4.9. As the switch is turned off quickly (mechanicalswitch installed in the home set-up), the transients were captured in the normalmode of the oscilloscope. As expected from the transmission line characteristics,there’s a sudden step-down in the voltage which creates transients in the currentflowing through the line. The voltage must rise to 2× Vp−p. But during the testsas observed in 4.8 and 4.9, the switching never happens at the peak voltage; it isassumed that the switch has an inbuilt snubber (refer [40] & [41])circuit to turn on/turn off at zero voltage only.41Figure 4.7: Closed Switch ConfigurationFigure 4.8: Open-circuit Current(Ch1) & Voltage(Ch2) Waveform DuringSwitching (Turn-off) Tested On Regular Cable4.1.2 Open Circuit DC TestsThe test set-up for the DC open circuit test is as shown in Figure 4.10. Since asquare wave input is applied as a source, the mechanical switching action is notrequired. The switching condition is shown Figure 4.11. Here the source sideswitching is done using square pulse, which eliminates the factors concerning themechanical nature of the switch. Hence the transients observed can be safely as-sumed to be mainly due to the transmission line characteristics.The results follow in Figures 4.12 & 4.13.42Figure 4.9: Open-circuit Current(Ch1) & Voltage(Ch2) Transient WaveformWhile Switching (Turn-off) Tested On Regular CableFigure 4.10: Open Circuit Test ModelThe simulation results using the same cable model created for the AC tests are usedand the result is as indicated in Figure 4.14.Similarly, the transients during turn-off (or step-down of square pulse input) is asindicated in Figure 4.15.The verification of the above results obtained during the experiments are doneusing the PSIM model and is as shown in Figure 4.16. Looking at the simulation43Figure 4.11: Open Circuit Test With Square-wave InputFigure 4.12: Open-circuit Output Voltage(Ch2) & Input Voltage(Ch3) TestedOn Regular Cableresults in both the cases i.e. 4.14 and 4.16, a good agreement between the simula-tion and the test results can be observed. The line model created on PSIM did notincorporate the mechanical nature of the switch or even the arc model, which wouldhave been necessary when using a DC source and a mechanical switch to study thetransients. For the DC test, the source is replaced by a DC supply ,using the squarewave input from a signal generator. The purpose of using a square waveform is toobserve the step response. The switching happens at the source end,i.e. S1 is usedin figure 4.10. A low voltage DC supply is used to observe the behaviour, based onwhich similar conclusions can be drawn for higher DC systems (220V )compatible44Figure 4.13: Output Voltage(Ch2) & Input Voltage(Ch3) Transient WaveformWith Step-up Input Voltage Applied On Regular CableFigure 4.14: Verification Of Transients During Step-up Voltage Using PSIMFor Regular Cable45Figure 4.15: Output Voltage(Ch2) & Input Voltage(Ch3) Transient WaveformWith Step-down Input Voltage Applied On Regular CableFigure 4.16: Verification Of Transients During Step-down Voltage UsingPSIM For Regular Cable46for house supply.The interference of source impedance can be neglected since DC supply isgiven as input from a signal generator. But in the real case scenario, the inputvoltage must be bridged using a high impedance, i.e. using a large capacitor at thesource end. As mentioned previously, we also eliminate the mechanical nature ofthe switch from contributing to the transients observed.As observed in figure 4.12, the output follows the input voltage. The voltageat the load end lags the source end by a small phase angle. The inductive nature ofthe transmission line is verified by this phenomenon.The time delay is measured in the waveform observed in the oscilloscope asdemonstrated in figure 4.17. Similarly the frequency of oscillations in the stepresponse are also verified as shown in figure 4.18.The transient response can be observed in figure 4.13 and figure 4.15. It is clearthat the output peak to peak voltage jumps to 2×Vp−p = 2× 31.8V as expectedfrom the transmission line model explained in Section Waveform Properties and MeasurementsFigure 4.17: Time Delay Calculation Using Cursors In OscilloscopeThe time delay in transmission of time period T is calculated from the wave-47form as observed in Figure 4.17. The time delay was observed to be 140ns. Thecalculation of the line parameters are listed in Table 4.1. The line parameters for thenew and modified , which is expected to have much lower characteristic impedance,is tabulated in table 4.2 is shown later in the section. 1.Table 4.1: Transmission Line Parameters Calculation For Regular Cable(Refer Section 3.2.2)Velocity of propagation Characteristic Impedance Time Delayv= 1√LCZo =√LC TD=Dvv= 2.01×108m/s Zo = 92.38Ω TD= 75.77nsFigure 4.18: Frequency Calculation Using Cursors In OscilloscopeThe time period for two cycles as observed from the waveform in Figure 4.18 wasabout 1.52µs. Thus the frequency is given by 1/2T . According to observationfrom the waveform; the frequency is given by :fw =12×T =12×1.52µs = 6.58MHz (4.1)1L= 4.60×10−7H/m & C = 5.39×10−11F/mD= 50 f t & R= 0Ω (considering lossless line)48According to theoretical calculation, the time period T is given by:fth =Dv=2.01×108m/s50 f t=2.01×108m/s15.24m= 13.19MHz (4.2)The time period T as observed from Figure 4.17 are of the same order and can beverified by the above calculations.The line parameters of a the laminated cable is tabulated in table 4.2 2. We cancompare the parameters and see an reduction in the line characteristic impedanceby more than 50%. The waveform comparing the two types of cables are shown inthe following sections.Table 4.2: Transmission Line Parameters Calculation For Modified Cable(Refer Section 3.2.2)Velocity of propagation Characteristic Impedance Time Delayv= 1√LCZo =√LC TD=Dvv= 2.24×108m/s Zo = 40.34Ω TD= 68.03ns4.2 Tests On Modified CableThe fabrication of the flat cable, demonstrated in chapter 3, has better electricalproperties when compared to the existing type of wiring as verified in the theoret-ical calculations. However it is yet to be verified experimentally, if the proposedcable is worth investing in.For this reason, the flat cable was fabricated using copper tapes of desiredthickness and width (Refer Appendix A.1) based on the area of cross section ofthe existing type of cables. The cable was then subjected to DC (one copper tape is”+ ve” electrode and the other being ”− ve” electrode), to observe the behaviour.Although, it should be noted that the cable measurement and construction wasdone using readily available parts, i.e. the copper tapes (50′× 0.5”× 0.01”) andPVC insulation with double sided adhesive(50′× 1”× 0.03”). The aim was to be2L= 1.80×10−7H/m & C = 11.06×10−11F/mD= 50 f t & R= 0Ω (considering lossless line)49able to get the best possible substitute for an existing circular cross-section cableusing the flat type cable.The step response of the modified cable and the corresponding regular, circular-cross section cable were to be compared to draw final conclusions. The test set-upfor the DC open circuit test is as shown in Figure 4.19. The waveform for the flatFigure 4.19: Open Circuit Test Modelcable is verified using the test set up illustrated. The transients while switching areobserved to compare with that obtained when using the regular cable. The switchcondition is shown in Figure 4.20 and the subsequent waveform in Figure 4.21.To verify the results obtained, a PSIM model of transmission line was modeledFigure 4.20: Open-circuit Test With Square-Wave Input For Modified Cablesimilar to that done in the section 4.1.1 and 4.1.2. The line parameters of the perunit length segments would be different in case of the modified cable as measuredand verified in 3.1.4. The simulation results are illustrated in Figure 4.22.50Figure 4.21: Step Response. Input Step(Ch1), Regular Cable(Ch2) & Modi-fied Cable (Ch3)51The peak over shoots captured when comparing the transients of a regular ca-Figure 4.22: Step Response Obtained Using PSIMble with that of the modified (flat cable) are as shown in Figure 4.23. The peakovershoot (peak to peak) value of the modified cable is about 10V lower than theregular cable. Although the experimental results did not reach the expected lev-52Figure 4.23: Peak Overshoot. Input Step(Ch1), Regular Cable(Ch2) & Mod-ified Cable (Ch3) 53els of improvement as predicted from the simulation results, there’s an observedimprovement in characteristics.Comparing the results obtained on PSIM with those captured on the oscillo-scope, the results are similar in showing the reduced transients when using themodified cable, however the improvement in performance is far more pronouncedwhile studying the simulation results shown in Figure 4.22. From the simula-tion waveforms,there is a reduction in the peak overshoot by a large margin. Thischange in the observation can be attributed to the following; difference in materialof insulation, non uniform insulation and/or trapped air bubbles, change in mate-rial properties from the extra adhesive layers, errors while construction of cable ormechanical punctures while coiling the long cable.Commenting on the ”transient time”, it is clear from the test results that thetime taken in both the cases (Regular Cable and Modified Cable) is similar whichis in contrast to the observations from the results obtained on PSIM. There is alarge improvement in transient time in the simulation results. This discrepancycan be attributed to the switch bouncing effect of the switch. The transient time isdependent on the mechanical bouncing, which is similar in both cases as we usethe same mechanical switch in both of the circuits.4.3 Load testsFigure 4.24: Test Circuit For Load TestsTo compare the results further, some load tests were performed with commonloads i.e the resistive loads commonly seen in household loads like the toasters,and the inductive load which represents most household loads. The ”switching-on”54and ”switching-off” characteristics of the various kinds of loads are demonstratedbelow. The test set up is shown in the figure Resistive LoadFigure 4.25: Test Circuit For Resistive LoadFigure 4.26: Test Circuit For R−L LoadThe transmission lines were terminated with a pure resistance of 9.1Ω. Theload was switched on and off using a regular 15A mechanical switch and the tran-sients were to be noted. The mechanical switch had its own constraints in termsof the ”switch bouncing” effects and the transition time to disconnect or connectthe load. Figure 4.27 shows the transient voltage and current at the load end whileusing the two types of cables. The characteristics are almost similar in both thesecases. The test case is illustrated in figure 4.25. The PSIM results for the same testcase is shown in figure 4.28. On the other hand, the ”switching-off” condition forthe same load shown in figure 4.29 shows some improvement. It can be observedthat the peak to peak value of the voltage is about 180V using the regular cable, but55Figure 4.27: Load Characteristics (Resistive Load) During ’Switch-on’ State.Load Current(Ch2), Load Voltage Of Regular Cable(Ch1)& Load Volt-age Of Modified Cable(Ch4)56Figure 4.28: Load Characteristics (Resistive Load) During ’Switch-on’ StateObtained Using PSIM57the voltage is about 100V when using the flat cable. There is a improvement andreduction in peak to peak voltage by about 44%. The time taken for the current torise is almost similar and is more than the time frame shown in the figure.Figure 4.29: Load Characteristics (Resistive Load) During ’Switch-off’State. Load Current(Ch2), Load Voltage Of Regular Cable(Ch1)&Load Voltage Of Modified Cable(Ch4)584.3.2 Inductive LoadAlong similar lines, an inductive load, i.e. an R−L load was used to load the lines.The inductive termination of the transmission line represents most loads since mostloads exhibit inductive characteristics.To verify the above results, PSIM results are as shown below in figure 4.31.The switching action was captured using a inductance coil of 3mH in series witha resistance of 9.1Ω. Figure 4.30 shows the ”Turn-on” behaviour while measuringthe load end voltage and current. Clearly, there is not much improvement in thepeak to peak voltage or the transient time. The test case is illustrated in figure 4.26.As in the case of the purely resistive load, there is a reduction in the peak topeak voltage in the case of ”switching-off” the R−L load. The recorded peak topeak values show reduction from 160V in the case of regular cable to about 130V inthe case of the laminated cable. There is an improvement of about 19% in terms ofthe reduction of the voltage. The switching-off characteristics are shown in figure4.32.The ”switch-off” characteristic on PSIM using either kind of loads showed notransient behaviour. Since the PSIM model did not incorporate the mechanicalcharacteristic of the switch or the arc characteristics, the switching off characteris-tic was observed to be that of an ideal switch.59Figure 4.30: Load Characteristics (Inductive Load, i.e. R − L) During’Switch-on’ State. Load Current(Ch2), Load Voltage Of RegularCable(Ch1)& Load Voltage Of Modified Cable(Ch4)60Figure 4.31: Load Characteristics (Inductive Load, i.e. R − L) During’Switch-on’ State obtained using PSIM61Figure 4.32: Load Characteristics (Inductive Load, i.e. R-L) During ’Switch-off’ State. Load Current(Ch2), Load Voltage Of Regular Cable(Ch1)&Load Voltage Of Modified Cable(Ch4)624.4 AnalysisThe difference in the observed waveform and the simulation can mainly be at-tributed to the mechanical nature of the switch that was used in the hardware set-up.Thus in some cases, there was no significant improvement seen in the test resultswhen compared to the simulation results. The mechanical switch model was notconsidered while simulating the the transmission line parameters. The primary fo-cus of this research being on reducing the line parameters, the ’switch bouncing’effects involved with the mechanical switch was not taken into account. As thecomponents of the switch settle into their new position, they mechanically bounce,causing the underlying circuit to be opened and closed several times. This non-ideal behaviour of the switch causes multiple electrical transitions each time thestate of the switch is changed.Additionally the error margin between the expected and the actual results canalso be influenced by the difference in the properties of material used while fabri-cating the laminated cable. Since the cable was constructed manually for a lengthof about 50 f t, there is scope for a lot of imperfections and non-uniformity betweenthe layers of the cable.A lot of work has been done on representation of DC arc in the form of variousmathematical equations have been done to study its characteristics. Incorporatingthe arc model to simulate the real case scenario would perhaps help in understand-ing the arcing behaviour in DC switches more accurately.The mechanical characteristics of the switch or the arc equations in the simu-lation model is complex and shifts the focus of this research. Our main objectiveis bring about practical improvement in DC switching.63Chapter 5Conclusions5.1 SummaryThis thesis investigated the properties and characteristics of the wiring in the cur-rent residential wiring. A new and modified wiring was proposed with the aimof improving both electrical and thermal characteristics, especially useful for DCpower distribution systems in residential application.The proposed flat-wire system ensured minimal modification to the existingload and power configuration,while having better switching characteristics-whichis one of the main concerns while implementing DC as power supply.Over a period of time the arcing affects the lifetime of the switch and callsfor replacement more often. Improving the cable to address this issue not onlyimproves the lifetime of the switch but also is a long-term solution in the wiring ofhomes.In most other scenarios, DC power is more user-friendly and more suitable forhomes. The compatibility of loads with DC power supply was studied before car-rying out various tests to prove improved electrical characteristics of the proposedwiring system.A DC homes wall set-up was fabricated to do these tests. Additionally, theflat wire cable of about 15 meters in length was laid with insulation, manually, tobe able to compare the proposed wiring system with the regular existing wiringsystem.64A thermal analysis was also carried out using a software called Elmer to ob-serve any improvements in this field. Subsequently, all these tests were verifiedwith simulation by modelling a transmission line model in PSIM with the mea-sured line parameters of each type of system.After conducting innumerable tests and running simulations, the modified ca-ble thus proposed has the following benefits:-1. Improved electrical properties: Line parameters of the cable are better thanthe regular cable parameters, contributing to better electrical properties.2. Better switching characteristics: Large margin of improvement in the peakovershoot as well the transient time implying lesser risk due to arcing interms of faster arc quenching properties.3. Efficient and convenient mechanical structure: Flat architecture would bemore convenient to install and can be mounted in a neat fashion.4. Possible reduction in cost of manufacturing: Same amount of copper used inthe modified cable can be rated for higher current. Thus the cost of copperrequired for the same current rating would be lesser in the modified-flat typecable than the regular type of cable.Having proven and stated the above conclusion, the main objective of thismodification was to improve the switching characteristics. The main question thatcomes to mind is if there still is requirement of using a snubber across the switch,or if substituting with the modified cable is sufficient to observe much improve-ment.The main concern would be the problems related to the practical fabrication ofthese cables and implementing them in a large scale. There needs to be thoroughanalysis in terms of the effort versus improvement in efficiency in comparison tothe exiting, conventional cables.5.2 Future WorkThere is still scope for improvement and advancement to this work. A lot of re-search is already being done in this field. In terms of carrying forward this partic-65ular work, the final objective of this research would be the fabrication of a profes-sional flat-cable set-up to replace the house wiring. Also to be considered would beto match the safety standards specified for home application- Since DC switchinghas a lot of stored energy, plugging in and out of loads can be dangerous and canpose severe electrical hazards. To prevent this each DC outlet must be coupled toa DC switch which must be turned off before being able to manipulate with theload. The design of this outlet-switch arrangement would be similar to the systemimplemented in various countries where each plug has a designated on/off switch.The usage of a plug lock system to prevent unplugging when the power is on isanother possible way of avoiding electric shocks.Designing of the DC arc characteristics in the simulation study cases wouldalso help in understanding the effect of change in line parameters on the arc be-haviour and the time taken to quench the arc. Additionally, considering the effectson the output characteristics due to the mechanical switch properties like the switchbouncing effect, turn-on and turn-off times, can be of interest. Alternatively, it willbe interesting to notice the change in characteristics if the switch were perhaps re-placed with an electronic switch with better turn-on and turn-off times as well asreduced switch bouncing effects. This will help in understanding the detrimentaleffects in switching characteristics due to the mechanical switch.Future work would also involve fabrication of wiring components like plugs tobe readily installed in home distribution system while incorporating the solutionto arcing while plugging in and out of the loads. Having done this, another majorstep would be to conduct cost analysis to check for feasibility of installing andmodifying the existing system and to ensure that this modification is worth theinvestment.If DC distribution systems for homes becomes successful,there would be con-siderable reduction in the power consumption of the loads, since the power ratingof individual loads can be lowered with DC [42].In addition to the other obviousbeneficial outcomes by using DC power, overcoming the main drawback of DCpower today (DC switching) will help in the promotion of a new distribution sys-tem altogether.66Bibliography[1] S. Iyer, W. Dunford, and M. Ordonez, “DC Distribution for Homes,” in Powerand Energy Society General Meeting, IEEE Preceedings, July 2015.→ pagesiii[2] J. F. Wakerly, “Supplementary material to accompany ’Digital Design Princi-ples and Practices’, Fourth Edition.” [Online].Available:http://www.cde.com/tech/design.pdf, 2006. → pages 17[3] A. Sannino, G. Postiglione, and M. Bollen, “Feasibility of a DC network forcommercial facilities,” Industry Applications, IEEE Transactions on, vol. 39,pp. 1499–1507, Sept 2003. → pages 2, 5[4] K. Engelen, E. Leung Shun, P. Vermeyen, I. Pardon, R. D’hulst, J. Driesen,and R. Belmans, “The Feasibility of Small-Scale Residential DC Distribu-tion Systems,” in IEEE Industrial Electronics, IECON 2006 - 32nd AnnualConference on, pp. 2618–2623, Nov 2006. → pages 2[5] W. Li, X. Mou, Y. Zhou, and C. Marnay, “On voltage standards for DC homemicrogrids energized by distributed sources,” in Power Electronics and Mo-tion Control Conference (IPEMC), 2012 7th International, vol. 3, pp. 2282–2286, June 2012. → pages 2, 10[6] E. Csanyi, “AC Vs. DC.” [Online].Available:http://electrical-engineering-portal.com/ac-vs-dc. → pages 2[7] D. Hammerstrom, “AC Versus DC Distribution SystemsDid We Get itRight?,” in Power Engineering Society General Meeting, 2007. IEEE, pp. 1–5, June 2007. → pages 3, 4[8] G. Gregory, “Applying low-voltage circuit breakers in direct current systems,”Industry Applications, IEEE Transactions on, vol. 31, pp. 650–657, Jul 1995.→ pages 467[9] P. Savage, R. Nordhaus, and S. Jamieson, “DC Microgrids: Benefits and Bar-riers,” → pages 4[10] K. Techakittiroj, S. Patumtaewapibal, V. Wongpaibool, and W. Three-vithayanon, “Roadmap for implementation of DC system in future houses,”in Harmonics and Quality of Power, 2008. ICHQP 2008. 13th InternationalConference on, pp. 1–5, Sept 2008. → pages 4, 9, 11, 12[11] W. G. Dunford, “The Implementation of DC Utilization in Distributed Gen-eration Systems,” in PEDAC 04, March 2004. → pages 4, 5, 6, 10, 11, 12,20[12] T. Iino, K. Hirose, M. Noritake, A. Nakamura, K. Kiryu, and J. Sekikawa,“Characteristics of 400 V dc plug and socket-outlet for DC distribution sys-tems,” in Renewable Energy Research and Applications (ICRERA), 2012 In-ternational Conference on, pp. 1–6, Nov 2012. → pages 4, 6[13] G. Artale, A. Cataliotti, V. Cosentino, and G. Privitera, “Experimental char-acterization of series arc faults in AC and DC electrical circuits,” in Instru-mentation and Measurement Technology Conference (I2MTC) Proceedings,2014 IEEE International, pp. 1015–1020, May 2014. → pages 5[14] P.-W. Lee, Y.-S. Lee, and B.-T. Lin, “Power distribution systems for futurehomes,” in Power Electronics and Drive Systems, 1999. PEDS ’99. Proceed-ings of the IEEE 1999 International Conference on, vol. 2, pp. 1140–1146vol.2, 1999. → pages 5[15] J. Ferreira and H. van der Broeck, “Alternative power distribution in residen-tial and commercial buildings,” in Power Electronics and Applications, 1993.,Fifth European Conference on, pp. 188–193 vol.7, Sep 1993. → pages 5[16] W. G. Dunford, V. Buchholz, J. Neilson, M. Wang, and D. Casson, “DCDistribution Systems for Commercial Buildings,” May 1995. → pages 6[17] A. Greenwood and T. Lee, “Theory and Application of the CommutationPrinciple for HVDC Circuit Breakers,” Power Apparatus and Systems, IEEETransactions on, vol. PAS-91, pp. 1570–1574, July 1972. → pages 10[18] M. Starke, L. Tolbert, and B. Ozpineci, “AC vs. DC distribution: A losscomparison,” in Transmission and Distribution Conference and Exposition,2008. T x00026;D. IEEE/PES, pp. 1–7, April 2008. → pages 12[19] W. H. Organisation, “Environmental Health Criteria 69,” 1987. → pages 1268[20] I. E. Code, “Low voltage electrical installations - Part 1: Fundamental prin-ciples, assessment of general characteristics, definitions,” vol. 2005, p. 10,2005. → pages 13, 21[21] A. F. Peterson and G. D. Durgin, Transient Signals on Transmission Lines:An Introduction to Non-Ideal Effects and Signal Integrity Issues in ElectricalSystems, vol. 3. 2008. → pages 13, 14, 16, 18, 24[22] Wikipedia, “Velocity factor.” [Online]https://en.wikipedia.org/wiki/Velocityfactor. → pages 15[23] Wikipedia, “Characteristic impedance.” [Online]https://en.wikipedia.org/wiki/Characteristic impedance. → pages 15[24] Wikipedia, “Propagation delay.” [Online]https://en.wikipedia.org/wiki/Propagation delay. → pages 15[25] Wikipedia, “Propagation constant.” [Online]https://en.wikipedia.org/wiki/Propagation constant. → pages 15[26] “ABB circuit-breakers for direct current applications.” [Online].Availablefor download:http:http://www04.abb.com/global/seitp/seitp202.nsf/0/6b16aa3f34983211c125761f004fd7f9/$file/Vol.5.pdf, 2007. → pages 19[27] L. Mischler, “Electrical installation guide.” [Online].Availablefor download:http://www2.schneider-electric.com/sites/corporate/en/products-services/product-launch/electrical-installation-guide/electrical-installation-guide.page, 2015. → pages 21[28] Wikipedia, “National Electrical Code.” [Online]https://en.wikipedia.org/wiki/National Electrical Code. → pages 21[29] H. Standard and N. Horizontale, “IEC 60038 Edition 7.0 2009-06 INTER-NATIONAL STANDARD,” 2009. → pages 21[30] Wikipedia, “International Electrotechnical Commission .” https://en.wikipedia.org/wikiInternational Electrotechnical Commission. → pages 21[31] Southwire, “ROMEX SIMpull TYPE NMD90,” tech. rep., Romex SIMpull,2002. → pages 23, 25[32] Wikipedia, “Relative permittivity.” [Online]https://en.wikipedia.org/wiki/Relative permittivity. → pages 23, 24, 2569[33] Wikipedia, “Electrical wiring in North America .” [Online]https://en.wikipedia.org/wiki/Electrical wiring in North America. → pages 23[34] N. F. P. Association, “NFPA 70 National Electrical Code (NEC) 2008 edi-tion,” vol. 552, no. c, pp. Article 310.8 (B)&(C), Table 310.13(A), 2008. →pages 23[35] USA Wire & Cable, “XHHW-2 XLP insulation, 600 Volts,” tech. rep., 1853.→ pages 23[36] I. E. Code, “International Standard: Conductors of Insulated cables,” 2004.→ pages 25[37] I. E. Code, “International Standard: Calculation of current rating,” 1999. →pages 25[38] J. Grainger and J. W. Stevenson, Power System Analysis. Jan 1994. → pages24[39] C. I. C. for Science, “Elmer Tutorials.” [Online].Available:http://www.csc.fi/elmer, 2011. → pages 31[40] K. Harada and H. Sakamoto, “Switched snubber for high frequency switch-ing,” in Power Electronics Specialists Conference, 1990. PESC ’90 Record.,21st Annual IEEE, pp. 181–188, 1990. → pages 41[41] R. Saverns, “Desgin of Snubbers for Power Circuits.” [Online].Available:http://www.cde.com/tech/design.pdf. → pages 41[42] M. Rodriguez-Otero and E. O’Neill-Carrillo, “Efficient Home Appliances fora Future DC Residence,” in Energy 2030 Conference, 2008. ENERGY 2008.IEEE, pp. 1–6, Nov 2008. → pages 6670Appendix ATheoryA.1 Verification of the Cross-sectional AreaFigure A.1: Cable Cross-sectionThe dimensions of the chosen flat cable for tests are Width (W)=0.5” and thick-ness (L)=0.01”. The area of cross section of such a cable:A= L×W = 0.5×0.01 = 0.005inch2 (A.1)71A regular cable of equal cross-section was chosen.A=piD24= 0.005inch2 (A.2)D= 0.08inches (A.3)Hence D=0.08” is calculated to be a 12 AWG wire.A.2 Line parameters and PSIM Line ModelTable A.1: Line Parameters of Regular Cable in PSIM ModelSpecification L C Rs RpPer unit length values 0.46µH/m 53.98pF/m 12.55mΩ/m 86.41MΩ/mPer Segment values 0.35µH/m 41.16pF/m 9.57mΩ/m 65.88MΩ/mTable A.2: Line Parameters of Modified Cable in PSIM ModelSpecification L C Rs RpPer unit length values 0.176µH/m 110.62pF/m 37.07mΩ/m 84.23MΩ/mPer Segment values 0.13µH/m 84.35pF/m 28.26mΩ/m 64.23MΩ/mThe PSIM model was created using the line parameters calculated and verifiedby measurements using the LCR meter.Using the distributed line model, 20 seg-ments were created to model the line. It was observed that increasing the numberof segments beyond this did not have significant change in the output waveform.The line parameters obtained during the experiments were per unit length values(i.e per metre). To obtain the values of the each of the 20 segment parameters,the per unit length metre is multiplied by the total length and is thus divided intotwenty equal parts.The per unit length values and the per segment values are as shown below in Ta-bles A.1 & A.2. The per segment value is calculated by simply multiplying the permeter values by a factor of 15.23m(50 f t)20 .72


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items