Thesis - Simulating Power Quality Problems by Atp-emtp

Simulating Power Quality Problems by ATP/EMTP

by Andrew James Senini

Department of Computer Science & Electrical Engineering University of Queensland.

Submitted for the degree of Bachelor In of the Engineering division of(Honours) Electrical Engineering October 16, 1998.

Mr. Andrew Senini, 3/34 Mitre Street, St. Lucia, QLD.

4067

Ph: (07) 3371 3585 E-mail: [email protected]

The Dean School of Engineering University of Queensland St Lucia, Qld, 4072

October 16, 1998.

Dear Professor Simmons,

In accordance with the requirements of the degree of Bachelor of Engineering (Honours) in the division of Electrical Engineering, I present the following thesis entitled “Simulating Power Quality Problems by ATP/EMTP”.

This work was

performed under the supervision of Dr. Tapan Saha.

I declare that the work submitted in this thesis is my own, except as acknowledged in the text and footnotes, and has not been previously submitted for a degree at the University of Queensland or any other institution.

Yours Sincerely,

Andrew Senini

ii

To… Mum, Dad, Rebecca, Natalie, Sharon, The Boys, The Seeneys…and last but not least, my old sparring partner, Fr Greg Jordan, S.J.

iii

Acknowledgements

The author would like to thank the following people for their contribution to this thesis.

Dr Tapan Saha. Thesis supervisor. Thanks for keeping the project going and for your

encouragement and good advice throughout the year. I hope to keep in touch in the future.

Mr. Adrian Mengede. Thank you for your willingness to give a hand, and for the time

you took to provide valuable details about the University of Queensland power system.

Mr. Cristian Pippia. Thank you for proof reading my thesis, and making the changes

that were necessary. It wasn’t that bad, was it?

Mr. Adam Carr. For your advice and sense of humour as I worked through this

project. Thank you for keeping me calm when I was ready to throw the whole lot out the window. Good luck with old Johnny down in Canberra next year.

iv

Abstract

Power quality problems are a major concern in the electricity industry today. Any slight variation in voltage amplitude or frequency can cause customer equipment to fail, at a substantial cost in time and money.

The ability to simulate power quality problems in a power system is important. If a problem can be simulated, then simulating a solution is the next step.

The Alternative Transients Program (ATP) was used to simulate power quality problems occurring at the University of Queensland.

The events simulated were

capacitor switching, system faults, induction motor starting and harmonic distortion.

It was found that the ATP, when used in conjunction with the ATPDraw, is an effective and cheap method to simulate power quality problems. The results obtained largely agreed with those recorded during a site survey. Capacitor switching, sags caused by induction motor starting and harmonic distortion were all within specified limits. The cause of the harmonic distortion was most likely parallel personal computer and fluorescent light loads.

v

Table of Contents

IV

ACKNOWLEDGEMENTS

ABSTRACT

V

LIST OF FIGURES

VIII

LIST OF TABLES

X

CHAPTER 1 - I NTRODUCTION

1

CHAPTER 2 - THEORY

3

2.1 TRANSIENTS 2.2 SHORT DURATION VARIATIONS 2.3 HARMONIC DISTORTION

3 4 6

CHAPTER 3 - REVIEW OF THE CURRENT LITERATURE

10

3.1 THE REQUIREMENTS FOR POWER QUALITY SIMULATION 3.2 THE ALTERNATIVE TRANSIENTS PROGRAM (ATP)

11 14

CHAPTER 4 - SIMULATING EXISTING POWER QUALITY PROBLEMS

19

4.1 GATHERING SYSTEM INFORMATION 4.2 CONSTRUCTING THE MODELS

21 25 25 29 31 32 33 34 36 39

4.2.1 4.2.2 4.2.3 4.2.4 4.2.5 4.2.6 4.2.7 4.2.8

TRANSFORMER, CAPACITOR, CABLE AND LOAD CALCULATIONS CONSTRUCTING THE TEMPLATE SYSTEM THE ATP FILE CAPACITOR SWITCHING VOLTAGE SAGS CAUSED BY SYSTEM FAULTS VOLTAGE SAGS CAUSED BY INDUCTION MOTOR STARTING HARMONIC DISTORTION INDUCTION MOTOR STARTING – CENTRAL CHILLER STATION

vi

CHAPTER 5 - PRESENTATION AND ANALYSIS OF RESULTS

40

5.1 CAPACITOR SWITCHING 5.2 SAGS 5.3 INDUCTION MOTOR STARTING, CHEMISTRY BUILDING 5.4 HARMONIC DISTORTION, MS LABORATORY

40 43 45 50 50 59

5.4.1 MS LABORATORY MODELLED AS A LINEAR LOAD 5.4.2 MS LABORATORY MODELLED AS A PARTLY NON-LINEAR LOAD ENTRAL

HILLER

TATION

5.5 T CHE EFFECTIVENESS C S OF THE ATP 5.6

67 68

CHAPTER 6 - CONCLUSIONS

70

6.1 RECOMMENDATIONS FOR FURTHER WORK

71

APPENDIX A - THE ATP FILES

73

A.1 ATP FILE FOR FIGURE 3.3 A.2 TEMPLATE.ATP A.3 HARM.MOD

73 74 77

APPENDIX B - GUIDE TO ATPDRAW COMPONENTS USED

80

APPENDIX C - COMPLETE FOURIER ANALYSIS OF RESULTS

83

BIBLIOGRAPHY

92

vii

List of Figures

Figure 2.1- A lightning stroke current impulsive transient _____________________________________ 3 Figure 2.2 – An oscillatory transient caused by Capacitor Switching [5] _________________________ 4 Figure 2.3 – A momentary interruption [5] ________________________________________________5 Figure 2.4 – Voltage Sag [5] ___________________________________________________________5 Figure 2.5 – The CBEMA Curve. Grey indicates areas in which equipment malfunction may/may not occur[21]. __________________________________________________________________________ 6 Figure 2.6 – Breaking down a distorted waveform into sinusoidal components [1]. Note this picture is taken from an American text and thus the fundamental is 60Hz _________________________________ 7 Figure 2.7 – Parallel Resonance [1]______________________________________________________8 Figure 2.8 – Triplen harmonics [1] ______________________________________________________ 9 Figure 2.9 – Current injected into the system by a PC load (3 equally balanced phases of PCs)_______ 9 Figure 3.1 – Short Circuit Fault in a radial system _________________________________________11 Figure 3.2 – A simple harmonic circuit that can be analysed manually [1]_______________________ 13 Figure 3.3 – Graphic version of file in Appendix A. _________________________________________ 16 Figure 4.1 – Capacitor sw itching, phase A, MS Lab ________________________________________ 20 Figure 4.2 – Summary of all sags experienced at the MS Lab during site survey[22]._______________ 20 Figure 4.3 – Simplified one line diagram of Chemistry building _______________________________ 21 Figure 4.4 – Substation STL, simple one line diagram _______________________________________ 22 Figure 4.5 – Central Chiller Station _____________________________________________________ 23 Figure 4.6– Par t of the ATPDraw file, showin g Sub Board A _________________________________ 30 Figure 4.7– Substation STL____________________________________________________________ 31 Figure 4.8– Capacitor switching circuit diagram. __________________________________________ 33 Figure 4.9– Circuit used to simulate three phase and single line to ground faults__________________34 Figure 4.10– Computer Science building chiller connection __________________________________ 36 Figure 4.11– Connection of harmonic loads, parallel to MS Lab, from sub board A. _______________ 38 Figure 4.12– Central Chiller Station ____________________________________________________39 Figure 5.1 – Capacitor Switching, Phase A, MS Laboratory.__________________________________ 40 Figure 5.2 – Capacitor Switching, Phase B, MS Laboratory __________________________________ 41 Figure 5.3 – Capacitor Switc hing, Phase C, MS Laborator y. _________________________________41 Figure 5.4 – Symmetrical fault, phase A. All phases are identical. _____________________________ 43 Figure 5.5 – SLG Fault. All phases. _____________________________________________________ 43 Figure 5.6 – Standby UPS. ____________________________________________________________ 44 Figure 5.7 – On-line UPS _____________________________________________________________ 44 Figure 5.8 – Induction Motor Starting, Computer Science chiller only __________________________ 45 Figure 5.9 – Induction Motor Starting, Mechanical Services only. _____________________________ 46 Figure 5.10 – Small sag during site survey, probably from motor starting _______________________ 46 Figure 5.11 – Current to parallel PC and fluorescent light circuits. ____________________________ 47 Figure 5.12 – Current on the 11kV feed.__________________________________________________48 Figure 5.13 – Current from T3 to Sub. Board A. ___________________________________________ 48 Figure 5.14 – The voltage wav eform on the primar y side of T3. _______________________________50 Figure 5.15– Fourier analysis, voltage waveform, primary side of T3. __________________________ 50 Figure 5.16 – Current waveform, primary side of T3. _______________________________________ 51 Figure 5.17– Fourier analysis, current waveform, primary side. _______________________________ 51 Figure 5.18 – Voltage waveform, second ary of T3. _________________________________________52 Figure 5.19– Fourier analysis, voltage waveform, secondary side of T3. ________________________ 52 Figure 5.20 – Current waveform, secondary of T3. _________________________________________ 53 Figure 5.21 – Fourier analysis, current waveform, secondary side of T3. ________________________ 53 Figure 5.22 – Summary of harmonic voltage levels, primary of T3, during site survey[22]. __________ 54 Figure 5.23 – Fourier analysis, current, going from Sub. Board A to T3. ________________________ 55 Figure 5.24 – Passive 5th harmonic filter added at Sub. Boar d A. ______________________________ 57

viii

Figure 5.25 – Fourier analysis of voltage at the MS Lab. Primary after addition of 5 th harmonic filter._57 th Figure 5.26 – Current flowing in phase A of the 5 harmonic filter_____________________________58 Figure 5.27 – Fourie r analysis of curr ent in the filter. THD = 21.9%. __________________________58 Figure 5.28– MS Laboratory, voltage waveform, phases A (curve a) &C (curve b), primary side of T3. 59 Figure 5.29– Fourier analysis of phase A voltage, primary side of T3. __________________________ 59 Figure 5.30– Fourier analysis of phase C voltage, primary side of T3. __________________________ 60 Figure 5.31– MS Laboratory, current waveform, phases A (curve b) &C (curve a), primary side of T3. 60 Figure 5.32– Fourier analysis of phase A current, primary side of T3. __________________________ 61 Figure 5.33– Fourier analysis of phase C current, primary side of T3. __________________________ 61 Figure 5.34– MS Laboratory, voltage waveform, phases A (curve b) &C(curve a), secondary of T3.___ 62 Figure 5.35– Fourier analysis of phase A voltage, secondary side of T3. ________________________ 62 Figure 5.36– Fourier analysis of phase C voltage, secondary side of T3. ________________________ 63 Figure 5.37– MS Laboratory, current waveforms, phases A(curve a) &C(curve b), secondary of T3. __ 63 Figure 5.38– Fourier analysis of phase A current, secondary side of T3. ________________________ 64 Figure 5.39– Fourier analysis of phase C current, secondary side of T3. ________________________ 64 Figure 5.40 – Output from model harm.mod. ______________________________________________ 65 Figure 5.41 – The Fourier analysis of the waveform in figure 5.40._____________________________66 Figure 5.42– Induction Motor Starting, Central Chiller. _____________________________________ 67 Figure 5.43 – Motor starting recorded by the PQ Node during survey __________________________ 67

ix

List of Tables

Table 4.1 – Plant and Cable information for modelling of the system. All currents are per phase. ____ 23 Table 4.2 – The TRADY transfor mer model and recom mended values. [18]. _____________________26 Table 4.3 – Transformer data __________________________________________________________ 27 Table 4.4 – Cable Data _______________________________________________________________ 28 Table 4.5 – Loads in terms of parallel R and L components___________________________________29 Table 4.6 – Loads used for harmonic simulation ___________________________________________ 38 Table B.1 – ATPDraw components used for simulation ______________________________________ 82 Table C.1 – Fourier anal ysis of MS Lab. Primary volt age (fig. 5.14) ___________________________83 Table C.2 – Fourier anal ysis of MS Lab. Primary cur rent (fig. 5.16) ___________________________84 Table C.3 – Fourier analysis at MS Lab. Secondary voltage (Fig. 5.18)_________________________ 85 Table C.4 – Fourier analysis at MS Lab. Secondary current (Fig. 5.20)_________________________ 85 Table C.5 – Fourier analysis. Current, Sub. Board A to T1. (Fig. 5.22) _________________________ 86 Table C.6 – Fourier analysis of phase A voltage, primary side of T3. (Fig 5.28)___________________86 Table C.7 – Fourier analysis of phase C voltage, primary side of T3. (Fig 5.29)___________________ 87 Table C.8 – Fourier analysis of phase A current, primary side of T3. (Fig 5.31)___________________88 Table C.9 – Fourier analysis of phase C current, primary side of T3. (Fig 5.32) __________________ 88 Table C.10 – Fourier analysis of phase A voltage, secondary side of T3. (Fig 5.34) ________________ 89 Table C.11 – Fourier analysis of phase C voltage, secondary side of T3. (Fig 5.35)________________ 90 Table C.12 – Fourier analysis of phase A current, secondary side of T3. (Fig 5.37) ________________ 90 Table C.13 – Fourier analysis of phase A current, secondary side of T3. (Fig 5.38) ________________ 91

x

Chapter

1 Introduction

A power quality problem is defined in the text Electrical Power Systems Quality [1] as:

“Any problem manifested in voltage, current or frequency deviations that result in failure or misoperation of customer equipment”.

The changing nature of customer loads has seen an increase in the importance of power quality problems. This change is due largely to the widespread proliferation of voltagesensitive microprocessors, which are present in equipment from VCR’s and PC’s in the home to hospital diagnostic systems and automated assembly lines in industry. In some of the industrial systems mentioned above, a power interruption or 30% voltage sag lasting hundredths of a second can reset controllers and stop an assembly line, sometimes taking hours to restart. A good example is an industrial plant in the U.S., which estimates that a five-cycle interruption in power supply can cost $200 000 [2].

Power quality is therefore a very important issue in today’s competitive electricity industry. Any utility that can provide cleaner power to crucial processes, or solutions to correct the power being received will have the competitive edge over others.

Power quality problems manifest themselves in variations in the voltage being received. This variation can be in the form of transients due to switching or lightning strikes, sags or swells in the amplitude of the voltage, a complete interruption in the supply, or harmonic distortion caused by non-linear loads in the system.

1

Chapter 1 - Introduction

The purpose of this thesis is to simulate these events using the Alternative Transients Program (ATP). This will be done in a practical manner by simulating problems that

have been monitored at the Mass Spectrometry (MS) Laboratory and the Central Chiller Station, on the St. Lucia campus of the University of Queensland. Monitoring has revealed the existence of some of these events.

The importance of being able to simulate power quality problems cannot be understated. If one has the ability to simulate any problem, then the next logical step is to simulate solutions to the problem.

By fully investigating and testing any solution before

installation, serious problems may be found, possibly saving large amounts of time and money.

This paper firstly examines the theory behind power quality problems: why they happen, and the effect they have on the power system.

The following section, Chapter 3, conducts a review of literature relevant to the project. Simple hand methods for calculating the effects of power quality problems are examined, as well as the software that is currently available to simulate them. The requirements of simulating power quality for any system are determined. Finally, the ability of the ATP to simulate the power quality problems being experienced will be discussed.

Chapter 4 describes the methods used to simulate the power quality problems. Steps in the process, from gathering the system information to building the models in ATP are described.

Chapter 5 presents results and then a discussion of their significance, first comparing them to those obtained by monitoring the site, and then suggesting any solutions to the problem.

Finally, conclusions and recommendations for further work are given in

Chapter 6.

2

Chapter

2 Theory

The following is a description of the power quality problems that will be covered in this paper. The Power Quality problems to be examined are transients, short term variations and harmonic distortion.

2.1 Transients Transients can be divided into two categories: oscillatory and impulsive [1].

An impulsive transient is a sudden, non-power frequency change in the steady-state condition of voltage, current, or both, that is unidirectional in polarity. An example of an impulsive transient is given below.

Figure 2.1- A lightning stroke current impulsive transient

Lightning is the most common cause of impulsive transients. Lightning transients in the low voltage (customer) system can occur from either direct strikes to the secondary circuit or strikes to the primary circuit where transient voltages pass through the distribution transformer [3].

3

Chapter 2 – Theory

An oscillatory transient is a sudden, non-power frequency change in the steady-state condition of voltage, current, or both, that includes both positive and negative polarity values. They are classed in terms of their oscillation: high, medium or low frequency. Figure 2.2 below illustrates an oscillatory transient.

Figure 2.2 – An oscillatory transient caused by Capacitor Switching [5]

Oscillatory transients are often a part of the system response to impulsive transients. They are caused directly by capacitor switching, ferro-resonance and transformer energisation. Capacitor switching is a common problem because it is a daily occurrence on most utility systems. Sensitive equipment such as Adjustable Speed Drives (ASD’s) and microelectronics are particularly vulnerable [3] & [4].

2.2 Short Duration Variations Short-duration variations can be divided into three categories: interruptions, sags and swells. These are possibly the most important power quality concerns [5].

An interruption occurs when the supply voltage or load current decreases to less than 0.1p.u. for a period of time not exceeding one minute [1]. Interruptions can be the

4


result of power system faults, equipment failures, and control malfunctions. Figure 2.3 below is an example of an interruption.

Figure 2.3 – A momentary interruption [5]

A voltage sag is a decrease in rms voltage or current to between 0.1 and 0.9 p.u. at the power frequency for a duration between 0.5 cycles and 1 minute [1].

Similarly, a

voltage swell is an increase to between 1.1 and 1.8 p.u. for a similar period of time. Figure 2.4 below is an illustration of a voltage sag.

Figure 2.4 – Voltage Sag [5]

5


Sags and swells are typically caused by system faults or lightning. Sags can also be caused by the energisation of loads such as large induction motors, although these are usually not as severe. Generally, the effect of sags upon equipment is dependent upon the sensitivity of the equipment and the distance of the equipment from the incident that caused the sag [6].

One guide for equipment manufacturers is the CBEMA curve (Figure 2.5). This curve illustrates the voltage variations that equipment should be designed to tolerate.

Figure 2.5 – The CBEMA Curve. Grey indicates areas in which equipment malfunction may/may not occur[21].

2.3 Harmonic Distortion Harmonic distortion, occasionally referred to as waveform distortion, is a growing concern in the electrical industry. Harmonic distortion is caused by non-linear (i.e. voltage-current curve is not linear) devices in the power system. These devices draw a non-sinusoidal current when a sinusoidal voltage is applied. This distorted current then causes distorted bus voltages to appear throughout the system [3].

The cause of these problems are the advent of power electronic converters for applications such as adjustable speed drives, single phase switched mode power

6


supplies such as those used for PC’s, and saturable devices such as transformers that have steel cores with non-linear magnetising characteristics.

Harmonics get their name from the fact that these waveforms can be broken down into a series of sinusoids, each of which has a frequency that is an integer multiple (a harmonic) of the fundamental. The fundamental in this case is the power frequency (50Hz in Australia). This process is known as Fourier Analysis [7]. Figure 2.6 below illustrates a Fourier series.

Figure 2.6 – Breaking down a distorted waveform into sinusoidal components [1]. Note this picture is taken from an American text and thus the fundamental is 60Hz

Harmonic distortion causes problems such as transformer and capacitor bank overheating, reducing the life of these expensive pieces of equipment. Most frequently, problems occur when capacitance in the system causes parallel resonance. Any harmonics at or near the resonant frequency will be amplified and distortion dramatically increased [1] & [7]. The resonant frequency is defined as:

fr

=

1 2π

LC

This is illustrated below.

7


Figure 2.7 – Parallel Resonance [1]

The resonant frequency/s are the frequency/s at which impedance of the system is at a maximum. These are the peaks on the graph above.

Harmonic spectrum diagrams assess harmonic distortion. These diagrams show the relative magnitude of each harmonic of the waveform. It is also quantified by a value, the total harmonic distortion (THD), which indicates the harmonic content of the waveform: h max

THD

=

∑ h

M

2 h

=2

M

1

IEEE Standard 519 – 1992 [8] specifies a maximum THD of 5%.

Finally, one special type of harmonics that should be mentioned are triplen harmonics. These are odd multiples of the third harmonic (i.e., h = 3, 9, 15, 21…). Figure 2.8 below illustrates triplen harmonics.

8


Figure 2.8 – Triplen harmonics [1]

Figure 8 shows that the triplen harmonic currents are in phase and flow into the neutral and add. If these currents meet a grounded wye – grounded wye transformer, they will flow through unimpeded. The neutral connections of such a transformer are susceptible to overheating when serving single phase loads with high third harmonic content. The most common cause of triplen harmonics are switched mode power supplies. The current drawn by a PC switched mode power supply is given below.

Figure 2.9 – Current injected into the system by a PC load (3 equally balanced phases of PCs)

9

Chapter

3 Review of the Current Literature

Any study of a power quality problem must include the following [9]:

• • • • • •

Modelling and Analysis of the problem Instrumentation Sources Solutions Fundamental Concepts Effects

This paper is mainly concerned with modelling and analysis of the problem. This can be accomplished by time domain methods, transformed domain methods (e.g. the frequency domain) and by simulation of the existing circuit.

The purpose of simulation of the system is twofold: 1) Simulating the power system concerned to evaluate the cause of the PQ problem.

These simulations are compared to actual measurements for

verification. 2) Simulating the solution to the PQ problem

In this section, the actual task of simulating power quality problems will be examined. Firstly, the requirements for any software analysis and some simple methods will be considered. Secondly, the Alternative Transients Program will be closely examined for its suitability for the task.

10

Chapter 3 – Review of the Current Literature

3.1 The Requirements for Power Quality Simulation The obvious requirement for any system or method being used to model a power quality problem is that it needs to be able to model or take into account all aspects of the system relative to the power quality problem at hand.

For transient analysis, any system needs to be able to accurately simulate the cause of transients on the system, as well as to be able to correctly predict the system behaviour under transient conditions. The ability to model electromagnetic and electromechanical oscillations ranging in duration from microseconds to seconds, switching and lightning transients and effects of these such as shaft torsional oscillations are all necessary [10]. Two commercially available packages commonly used to simulate transient situations are ATP and SPICE [1] & [5].

The ability to model lightning strikes are also necessary to model sags/swells, as is the ability to model fault conditions such as symmetrical and single line to ground faults. A hand method to evaluate the threat of voltage sags is given in [6]. A method to evaluate a simple case on a radial distribution system will be examined briefly.

Figure 3.1 is a simple diagram of a short circuit fault in a radial distribution system.

Figure 3.1 – Short Circuit Fault in a radial system

11


To calculate the sag magnitude at the load, the point of common coupling (PCC) must first be identified. Figure 3.1 shows the resulting voltage divider. Using simple circuit analysis, it is found that

Vsag

=

Z2 Z1 + Z 2

Assuming that there is a critical voltage below which the equipment will trip, the above can be modified as follows Z2 Z1

+ Z2

Now, let Z2 = L

< Vcrit ×

z, where z is the feeder impedance per kilometre, and L the distance

between the fault and the PCC. Assuming that the X/R ratios of Z1 and Z2 are equal, then a critical distance, Lcrit, can be defined that represents minimum distance a fault must be from the PCC in order to not trip the load.

L crit

=

Z1 z

×

V crit

1 − V crit

Strictly speaking, this method is for single line systems, making it valid only for symmetrical faults. For single-phase faults, the voltage in the faulted phase can be calculated using the sum of the three sequence impedances [11]. For phase to phase faults, the sum of the positive and negative sequence impedances gives the voltage difference between the faulted phases.

[6] goes on to examine situations of sub-transmission loops, local generation and feeding from two substations.

The software package usually used to examine sags, swell and interruptions is the ATP [5].

For anything but the simplest of circuits, sophisticated computer programs are required for harmonic analysis. An example is given in [1] of a circuit configuration common in small industrial systems that can be solved easily by hand. It is a single bus system with a capacitor.

12


Figure 3.2 – A simple harmonic circuit that can be analysed manually [1]

Figure 3.2 above shows the system and its equivalent circuit. The resonant frequency can be easily determined by using the formula presented earlier. The voltage distortion due to the current Ih is given by the following:

Vh

  R + jω L =  Ih 2  1 − ω LC + jωRC 

h = 2, 3, 4….., and

ω = 2πf1h

Note that the harmonic content of the source at each harmonic is required in order for this method to work.

The essentials of a computer program for harmonic analysis can be listed as follows:

•

The ability to display waveforms, frequency-response plots and spectral plots [12]

•

The ability to perform frequency (impedance) scans at small intervals of frequency [1].

•

It should be capable of handling large networks of at least several hundred nodes

•

It should be able to display the results in a meaningful and friendly manner to the user

•

The diversity of harmonic loads requires that computer software provide user definable methods to represent the contributing loads accurately [13].

13


Some of the specialised programs for dealing with harmonic analysis, which are available in the industry, are V-HARM [12], HI_WAVE [13] and SuperHarm [5]. All come with a number of harmonic models and meet all of the criteria above.

Another more common program that can be used is PSPICE. The advantage of using this program is that it is one which is widely used in electrical engineering core courses to study linear circuits, and thus most electrical engineers are already familiar with it [14]. Presented in [14] is an example harmonic analysis, where PSPICE is shown to produce results that agree with other circuit-oriented simulators such as V-HARM and ATP/EMTP.

3.2 The Alternative Transients Program (ATP) The ATP is the PC version of the Electromagnetic Transients Program (EMTP). The EMTP is primarily a simulation program of the electric power industry. It can predict variables of interest within electric power networks as functions of time, typically following some disturbance such as the switching of a circuit breaker, or a fault [15].

It was developed at the Bonneville Power Administration in the late 1960s as a replacement for the Transient Network Analyser (TNA), which was a large analogue simulator used for transient analysis. What began as approximately 5,000 lines of code used primarily for switching studies grew into a 70,000 line multipurpose program by the early 1980s [16].

A simplistic view of a power system is that it is comprised of three categories of components: Sources, Branches and Switches. The following is a description of these components and their use in the ATP [17].

ATP has a number of different types of sources, all of which can be either current or voltage sources. Examples are:

•

Ramp functions with linear decay or rise, which is useful for simulating lightning.

14


• • •

A surge function, also useful for simulating lightning. Sinusoidal functions f(t) = Amplitude * cos(2 πft + φ) Three phase dynamic synchronous machine

Some of the branches available are:

• Series R-L-C • π-equivalent • Distributed parameter transmission lines • Surge arrestors. • Transformers More complicated networks require the impedance matrix. There are two supporting programs to obtain this data.

These programs are “Cable Constants” and “Line

Constants”. Surge arrestors are represented by non-linear characteristics built up from small linear segments. The Voltage/Discharge current characteristic is usually obtained from the manufacturer.

Transformers are modelled either as a series R-L branch, or if a more detailed study is required, support programs are available to convert nameplate and test data into a coupled R-L matrix.

Various types of switches exist. These include:

•

Ordinary Switches. Voltage drop is zero when closed, current is zero when open.

• •

Voltage Controlled Switch. Useful for simulating flashover. Systematic Switch. This is a switch that turns on and off at regular intervals. May be useful for simulating re-closing of circuit breakers.

TACS is an add-on to the ATP that was developed to simulate the dynamic interactions between control systems and electric network components in the EMTP. One of its uses is for the simulation of Silicon Controlled Rectifiers (SCRs), used in the converters for adjustable speed drives, which were discussed earlier as a source of harmonic distortion.

15


Simulation of rotating machinery is also possible in ATP. The Universal Machine model can represent single, two or three phase synchronous or induction machines, series or parallel DC machines, and separately excited DC machines. This model can be used to show the voltage sags caused by motor starting. The effects of system transients upon these machines can also be simulated.

One feature of particular interest in harmonic analysis is the ability of the program to be able to perform a frequency scan of the system. This enables resonant frequencies of the system to be found.

A relatively new addition to the ATP is MODELS. MODELS is a general purpose description language supported by a set of simulation tools for the representation and study of time variant systems [20]. This feature is important as it gives the user the capability described in the previous section, specifically the ability to model harmonic sources. In fact, [18] contains various harmonic models developed by the author of that paper, including six and twelve-pulse adjustable speed drives, PC loads and fluorescent lights. These will be examined further later.

ATP does suffer from a marked lack of usability. The program was conceived at a time when batch mode computing was the standard, i.e., the user prepared a number of punch cards, (the equivalent to one line of data) in a fixed format, and put them into the computer. In its current incarnation, ATP requires inputting information into a text file in a fixed format, with each “card” represented by one line. This makes the system difficult to become acquainted with, but once the user becomes, it becomes a lot less difficult to use. As an example, see Appendix A for the input data file of the circuit below.

Figure 3.3 – Graphic version of file in Appendix A.

16


Fortunately, a graphical pre-processor, ATPDraw has recently been made available. This program allows the user to draw the circuit in a CAD-like environment [19]. All of the sources, branches and switches, as well as the ability to use the universal machine model, TACS and MODELS have been incorporated into this program. On command, ATPDraw outputs an ATP ready text file perfectly formatted and ready for simulation.

The output of any ATP simulation consists of two files, filename.lis and filename.pl4. The first file contains a summary of the program execution and will detail any errors that the ATP found with the input file. The second file is far more useful in that it can be used with the graphical post-processor, TPPLOT [15]. It is possible to display any number of branch or node voltages, or node currents to examine transients, sags and swells. Viewing these plots can clearly show the effects of the disturbances, and this can be output to a printer. For harmonic distortion, TPPLOT can display magnitude vs. frequency plots for frequency scans, as well as perform Fourier analyses on waveforms. TPPLOT also calculates quantities such as the Total Harmonic Distortion (THD).

Hence, to summarise the characteristics of the ATP that makes it excellent for simulating power quality problems:

•

Transients can be examined through the availability of sources that can simulate a lightning strike, as well as having voltage controlled switches to simulate flashover. Capacitor switching can also be easily simulated, given the availability of capacitors as branches.

•

Symmetrical voltage sags may be simulated, with switches being used to simulate faults. Voltage controlled switches can also be set to trip out in a high voltage situation.

•

Voltage sags caused by motor starting are also

examinable through the use of the universal machine or MODELS. Harmonic studies are made possible by the existence of TACS and MODELS to simulate non-linear loads such as ASDs and the switched mode power supplies of PCs.

Frequency scans are possible to find resonant

frequencies of the system.

•

The new program ATPDraw is a graphical interface to the ATP that is simple to use and allows the use of virtually all of the ATP features.

17


•

And finally, a graphical post-processor, TPPLOT, allows viewing of time and frequency plots, as well as being able to give a spectral analysis of any waveform.

18

Chapter

4 Simulating Existing Power Quality Problems

Power quality problems have been experienced at the University of Queensland, and it was decided early on that these problems were an ideal focus for this project. Two sites in particular were examined – firstly, the Mass Spectrometry (MS) laboratory in the Chemistry building and secondly, the Central Chiller Station, where large chillers (induction motors) had recently been installed.

These loads had constantly been

tripping out, causing major disruptions, especially for the work being carried out in the MS laboratory.

Site surveys were carried out as a part of another thesis project, “Monitoring of Distribution System Power Quality”, by Andrew Meiklejohn [22]. The monitoring was carried out using a BMI/Electrotek PQ Node. A full presentation and analysis of the events recorded can be found there, but a brief summary will now be presented.

The transients recorded were confirmed as capacitor switching at the Energex substation STL, which services the university and the surrounding suburb. These transients were recorded in the morning, as the capacitors came online to provide power factor correction. A good example of the transient is illustrated below. This was one of the most severe observed.

19

Chapter 4 – Simulating Existing Power Quality Problems

Figure 4.1 – Capacitor switching, phase A, MS Lab

Short-term variations, mainly sags, were also experienced in the MS laboratory. While most of these were relatively small, one large event was recorded – a fault to ground in the St. Lucia suburb caused a large sag over the entire campus. Other causes for the smaller sags, such as starting of remote chillers in the Computer Science (CS) building will be investigated as a part of the modelling process. presented on the CBEMA curve, is given below [22].

A summary of the sags,

Figure 4.2 – Summary of all sags experienced at the MS Lab during site survey[22].

Finally, some harmonic distortion of the voltage was also experienced. The main cause of harmonic distortion was found to be the hot water switching signal, used to switch

20


hot water systems. The frequency of this signal was 1050Hz, the 21st harmonic. Other harmonics, notably the 3 rd, 5th, 7th 9th and 11 th, were also present.

The power quality problems being experienced at the Central Chiller Station were capacitor switching (see above) and voltage sags, due to motor starting. More of these events will be shown later to compare them to the results obtained by simulation.

4.1 Gathering System Information The first part of the process was to gather information on the system. This involved finding circuit diagrams, information on transformers, capacitors etc and finally the nature of the load – the size (kVA) and type (linear or non-linear). The first model built was that of the Chemistry MS laboratory. The approach taken was to start at the load and work backwards. Information was gathered from the Campus Electrical Engineer, the manufacturers of the equipment in the MS laboratory, and Energex. A one-line diagram of the Chemistry building is given below.

Figure 4.3 – Simplified one line diagram of Chemistry building

It is important to capture all of the other loads, in order to study, for example, the effects of non-linear PCs or chiller starting in the Computer Science Building. Figures shown

21


represent the loads in terms of the currents they draw. The PQ Node was positioned on the primary side of transformer T3.

It was then necessary to gather information about Energex substation STL in order to include effects such as capacitor switching, sags caused by faults in the St. Lucia suburb and the 21st harmonic hot water switching signal. Figure 4.4 below illustrates this.

Figure 4.4 – Substation STL, simple one line diagram

Further information is summarised in Table 4.1. The final information required is the circuit diagram for the Central Chiller Station.

22


Figure 4.5 – Central Chiller Station

Table 4.1 below summarises all plant/cable information required to model the system.

SUBSTATION 20 – CHEMISTRY

750 kVA, 11kV ∆ / 430V Υ,

T1

4% impedance. 1000kVA, 11kV ∆ / 433V Υ,

T2

5% impedance 30kVA, 415V ∆ / 208V Υ,

T3

4.43% impedance Cable – Sub Board A to MS Lab

27m, one conductor per phase Area of core = 6mm 2

Cable – T1 to Sub Board A

8.5m, two conductors per phase

(only significant impedance)

Area of core = 240mm 2

Chillers – CS building

2

×

105A,

Υ-∆

reciprocating starters to

reduce startup currents 1 × 360A screw chiller

23


Mechanical Services

400 A, Direct on Line (DOL) induction motor

ComputerScience

180APCs 80A Fluorescent lights

Sub. Board A (from which MS Laboratory 400A total, mixture of PCs, fluorescent is supplied) MSLaboratory

lights and normal, linear loads 22.7kVA(3Phase,worstcase) Mostly linear, a switched pump and some PCs.

ENERGEX SUBSTATION STL

33kV ∆ / 11kV Υ

T4

Normal Operation: 15.7MVA Emergency (6 Months) 17.2MVA Summer 19.5MVA Winter 2 hour: 18.75 MVA Summer 21.0 MVA Winter Impedance: 15% 33kV ∆ / 11kV Υ

T5

Normal Operation: 15.7MVA Emergency (6 Months) 17.2MVA Summer 19.5MVA Winter

2 hour: 18.75 MVA Summer 21.0 MVA Winter Impedance: 10% on 10MVA Source Equivalent Impedance

At nominal voltage (33kV), fault level is 467MVA. Z+ = 0.02 + j0.214 p.u. on a

24


10MVA base. CapacitorBank

5Mvar,1.667Mvarperphase,ungrounded wye.

Cable – Sub 7 to Sub 10

230m, one conductor per phase Area of core = 95mm 2

SUBSTATION 20 – CENTRAL CHILLER STATION

1000kVA, 11kV ∆ / 433V Υ,

T6

5% impedance 1000kVA, 11kV ∆ / 433V Υ,

T7

5% impedance Chiller 1

700A,

Υ-∆ reciprocating starters to reduce

startup currents Chiller 2

400A,


startup currents Chiller 3

700A,


startup currents Table 4.1 – Plant and Cable information for modelling of the system. All currents are per phase.

4.2 Constructing the Models This section will describe the calculations carried out and methods used to build the models to simulate capacitor switching, voltage sags through faults and motor starting, and harmonic distortion.

4.2.1 Transformer, Capacitor, Cable and Load Calculations Example calculations will now be given describing the process by which the cables, transformers, loads and capacitor bank have been modelled.

ATPDraw [19] has several transformer models. For the purpose of power quality studies, ATPCON [18] gives recommendations on the use of these models. The only

25


transformer model being used is the

∆-Y

transformer, which simplifies the process of

putting the system together. Table 4.2 describes the parameters and recommended values for use of the TRADY transformer model. PARAMETER Io

VALUE AN D CO MMENT Alwaysuse0.01

Fo Rmag Rp Lp Vrp Rs Ls Vrs Lag OUT RMS

Alwaysuse0.001 Alwaysuse999999.0 Alwaysuse0.001onhigh-side Always0.001onhigh-side Peak rated kV of the primary winding. Resistance of the secondary winding. Leakage inductance of the secondary winding. Peak rated kV of the secondary winding. -30 (degrees on secondary side with respect to primary side) Alwaysuse0 Alwaysuse1

Table 4.2 – The TRADY transformer model and recommended values. [18].

The values required by this model are consistent with the simple series inductance – series resistance model of a transformer. Vrp, Rs, Ls, and Vrs are the only variables. The variable Lag remains at –30 degrees, because all transformers to be used are ∆-Y. Series R and L values must be taken to the secondary of the transformer.

Calculations for transformer T1 will be used as an example. As listed in Table 1, this transformer is 750 kVA, 11kV

∆ / 430V Υ,

with an impedance of 4%. All R and L

values will be referred to the secondary side of the transformer.

The secondary side voltage is 430V l-l, therefore, on a 750kVA base: Z base

=

V 2 P

=

430 2 750000

=

0 . 2465

Ω

Four percent of 0.2465Ω is 9.86×10-3Ω. This represents the magnitude of Z, or jX.

26

R +


A common X/R ratio for transformers is 10 [11]. Using this value, we find that R= 9.81×10-4Ω, and XL = j9.81×10-3Ω. Using a power system frequency of 50Hz, we have L= 0.03123mH. Vrp, the primary voltage, is 11275 ∆. Therefore, the peak primary rated voltage, Vrp = √2×11275 = 15.945kV. Vrs, the secondary voltage is 430V Y, so we have Vrs = √(2/3) × 430V = 351.09V

Similar calculations were carried out for all of the transformers in the study. Values are listed in Table 4.3.

TRANSFORMER

VALUES

T1

Vrp = 15.945kV, Vrs = 351.09, Rs = 0.001 Ω, Ls = 0.0312mH

T2


T3

Vrp = 586.9V, Vrs = 169.7V, Rs = 0.0064 Ω, Ls = 0.202mH

T4

Vrp = 46.67kV, Vrs = 8.981kV, Rs = 0.127 Ω, Ls = 4.03mH

T5

Vrp = 46.67kV, Vrs = 8.981kV, Rs = 0.127 Ω, Ls = 4.03mH

T6


T7


Table 4.3 – Transformer data

The next calculations to examine are for the 5Mvar capacitor bank in the STL substation. Working on a per phase basis, there is 1.667Mvar per phase, with a line to neutral rms voltage of 11275/√3 = 6509.6V rms. Therefore, X

C

=

V

2

Q

=

6509 . 6 2 1666667

=

25 . 42

Ω

Therefore, for a power system frequency of 50Hz, C = 125.2µF per phase.

Calculation of cable impedance is very simple. Because the cables are of a relatively short length, they can be treated as pure resistances. All that is required to calculate the

27


resistance, R, of a length l of cable is the diameter of the cable core and the resistivity,

ρ, of the core.

For these cables, with a copper core, ρ = 1.724 × 10-8 Ωm.

To use the 27m cable joining Sub Board A to the MS Lab as an example, R=

ρl

A

=

1.724 ×10−8 × 27 6 ×10−6

= 0.0776Ω

All other cable resistances were calculated in a similar manner, and these figures are presented in Table 4 below. CABLE

RESISTANCE ( )

T1toSubBoardA

0.00305

Sub Board A to MS Lab

0.0776

Substation 7 to Substation 10

Sub 7 – Sub 23: 0.0112 Sub 23 – Sub 10: 0.0075

Table 4.4 – Cable Data

Various aspects of the simulation process required that loads be expressed in terms of linear components. ATPCON comes complete with models to express loads in terms of parallel resistor and inductor components. These are called LOADY and LOADD, for loads connected in wye and delta (or ungrounded wye) configuration, respectively. For example, the MS Laboratory represents a load of 22.7kVA (7.567 kVA per phase), at a power factor of 0.85.

The line to neutral voltage is 120V (rms).

Resistance and

inductance values per phase are required. The load is connected in wye configuration, and assuming a balanced load we have that P = 0.85

×

7.567 KVA = 6.432 kW per

phase, Q = √(AP – P ) = 3988.9 vars. Now, the resistor and inductor are connected in 2

2

parallel, so they each have a voltage of 120V across them.

Using Ohm’s law:

2

XL

=V = Q

1202 = 3.61Ω∴ L = 11.5mH , 3988.9

and

R

=

V 2 P

=

120 2 6432

=

2 . 24

Ω

All other Chemistry building loads were developed into linear loads in a similar manner. Figure 4.3 gives the other loads in the system as currents drawn (per phase).

28


The calculation of R and L for these loads begins by dividing the current in terms of estimated components. For example, the CS Building draws 260A (in each phase), and, due to a large PC load, it is estimated that 180A of this represents the PC load and the other 80A the fluorescent lighting. For the PC load, we have P = VI = 43.2kVA = 129.6kVA (3 phase). The reason for dividing the load like this will become more obvious when harmonics in the system are discussed. These figures are then used to determine parallel R and L components, through simple use of Ohm’s Law.

Table 4.5 below shows all Chemistry building loads, divided into R and L components per phase.

LOCATION

CSChiller SubBoardAPCs Sub Board A Fluorescent Lights

R( ), L, PER PHASE

R=0.495,L=2.544mH R=2.544,L=13.12mH R= 1.2772, L=6.56mH

Sub Board A Linear

R=7.67, L=39.3mH

MSLab

R=2.24,L=11.5mH

CS Building Fluorescent Lights

R= 3.529, L=6.56mH

CSBuildingPCs

R=1.569,L=8.057mH

Mech. Services Linear

R= 0.7059, L=3.63mH

Table 4.5 – Loads in terms of parallel R and L components

All information required to build these models in ATP is now available.

4.2.2 Constructing the template system The first step taken was to construct a template of the system. This involved drawing the circuit in ATPDraw and putting in all transformers, cable information and linear loads. The purpose for this was twofold. Firstly, a template made it convenient to make relatively small changes to switch between studies. For example, to model non-linear loads from the template, replacement of the linear loads with the appropriate model is all that is required. Secondly, the template also provided the opportunity to check that

29


the system was working correctly, and that correct voltage and currents were present throughout.

The first step in building the template was to construct the Chemistry building system. Part of the ATPDraw diagram is given below.

Figure 4.6– Part of the ATPDraw file, showing Sub Board A

This picture shows most of the components required for the simulation. Transformers and loads are obvious. The RLC components marked “AMMETER” are three phase resistors of extremely small magnitude put in place to measure the current flowing to the loads.

A full explanation of all components used for simulation is given in

Appendix B. The parameters are entered by right-clicking the component and filling in the form that appears.

The next step is to add the Energex Substation STL to the model. All that is required are two transformers and the source, as well as the source equivalent impedance. This is given in figure 4.7.

30


Figure 4.7– Substation STL

Thus, the template has been constructed, ready for use with other studies.

4.2.3 The ATP File The ATP file, template.atp, generated from template.cir, is given in Appendix A. There are a few things to note about the ATP file itself.

Firstly, it is fixed format, so it is necessary that all information be placed in correct columns. This is aided by the numerical data listed across the page at regular intervals. Occasionally (very rarely) the ATPDraw will generate node names that are too long. This must be remedied by hand, a very tedious process when not sure where to start looking.

The parameters for the simulation are set near the top of the file. These are as follows:

• ∆T: The time step for the simulation • Tmax: End time for the simulation • Freq: System Frequency. This is only used if Xopt or Copt are non zero, but it is always good practice to set this to the system frequency, in this case, 50Hz

•

Xopt: When set to zero, all inductances are in millihenries. Otherwise, Xopt is set to the system frequency, and all inductances are given in terms of their reactance.

•

Copt: similar to above, capacitances are in microfarads if Copt is 0.

31


•

IOUT: Frequency at which points are output to the screen during simulation. For example, IOUT = 500 means that every 500 th time step will be printed to the screen. It is better that this be a high number, say 500, or the simulation will be lengthy.

•

IPLOT: Frequency at which points are output to the .PL4 file during simulation. This has been set to 1, so that each point may be used for plotting.

These are the most important parameters affecting the simulation. For all simulations,

∆T has been set to 20 µs, and Tmax has been set to 0.2s.

This gives a simulation time of

ten cycles. Obviously, some system events are longer than this. However, 2-3 cycles is adequate simulation time to obtain an estimate of any variations in voltage and current. Making the simulation any longer will merely cause the entire process to become extremely tedious, especially when simulating with MODELS.

4.2.4 Capacitor Switching Capacitor switching is very simple to simulate in the ATP – it was, of course, one of the reasons the program was written in the first place.

A capacitor bank is easily modelled by adding the three phase, ungrounded wye capacitor bank, CAPD, in series with a 3-phase time controlled switch, which is programmed to switch on and off at the desired time intervals. It would be expected that a transient would occur when switching on, and none when switching off, as transients only normally occur during energisation of the capacitor bank [18]. The transients recorded by the PQ node also suggest that this will happen; transients were only present in the morning when the capacitors were brought on line.

Care should be taken to ensure that a resistance is placed in series with the capacitors. This will ensure that the RC time constant will be greater than the time-step ∆T used in the simulation, preventing possible numerical instability and erroneous results. During all simulations, a time-step of 20 microseconds was used. This means that a series R

32


greater than 0.2Ω is necessary. Resistance can be varied further to provide damping of the transient.

As can be seen from figure 4.8, the 5Mvar capacitor bank is added to the 11kV bus at Substation STL. All loads were simulated as linear loads.

Figure 4.8– Capacitor switching circuit diagram.

The voltage from the capacitor switching will be displayed at the load (MS Laboratory), on the primary side of T3, which was where the PQ Node was positioned (see figure 4.1).

4.2.5 Voltage Sags Caused by System Faults

A major voltage sag caused by a single line to ground fault in the St. Lucia suburb was recorded by the PQ Node during the monitoring of the MS Laboratory site. Unfortunately, it has proved very difficult to obtain much information about the system in the suburb, so the following approach to voltage sags has been taken.

•

The first sag to be simulated will be a symmetrical (three phase) fault to ground.

While this will result in similar sags on all phases, it will be

33


possible to calibrate the magnitude of the sag with the maximum magnitude of the sag experienced on site by varying the resistance between the fault and ground. This can be justified by pointing out that, at this stage, we are looking at simulating the events taking place at the MS Laboratory. If these events are at a proper magnitude, then a solution for the problem can be simulated.

•

The second sag to be simulated will be a single line to ground (SLG) fault, to see if similar results can be found to those experienced at the PQ Node.

These sags were simulated by placing time-controlled switches in series with a resistance to ground off the 11kV bus of Substation STL. Both the three phase and SLG cases are illustrated in figure 4.9 below.

Figure 4.9– Circuit used to simulate three phase and single line to ground faults

4.2.6 Voltage Sags Caused by Induction Motor Starting This section will examine the construction of the circuit for simulation of voltage sags in the Chemistry building, due to starting of the Computer Science chillers.

34


The starting current for a large induction motor is several times its rated load current. This high current may cause voltages on the distribution feeder to temporarily drop to unacceptably low values. The main concern is that sags from this equipment may be causing the sensitive MS Laboratory equipment to fail.

The chiller motors to be

simulated come with Y-∆ starters to reduce the amount of current drawn at startup.

ATPCON comes complete with a three phase induction motor model.

The file

IM3P.MOD was developed in [18] as a simple alternative to the complicated machine models available in ATP. All that is required for the model are a few simple pieces of data, most of which can be readily approximated. These are:

• • •

HP3P: The three phase Horse Power of the machine RDPF: Fundamental power factor at full load (have used 0.85) START: The current multiplier, times full load current. It is recommended in [18] that this is usually about five for normal motors, but a value of 3 will be used due to the ∆-Y starters on the machines being studied.

•

SDPF: Fundamental displacement power factor during startup(this has been set to 0.2 for all simulations)

•

VLNRMS: The line to neutral voltage of the bus to which the motor is attached (240-250V)

•

F: The power system frequency (50Hz)

The following is taken from [18]. It describes the function of the induction motor model, IM3P.

“For the first few cycles, the motor is treated as a fixed resistor in each phase, sized for one-tenth of its rated kVA input.

Next, it enters the starting phase with current multiplier “START” and displacement power factor “SDPF.” START is in the 5.0 range, and SDPF is in the 0.2 range.

35


For the last few cycles, it becomes a fixed sinusoidal current injector with the power and displacement power factor “RDPF” that describe the rated power condition. RDPF is in the 0.80 range.”

In order to save simulation time, the CS chillers have been combined into one large motor (worst case), with a full load rating of 350 kW (three phase), or 411.8kVA at a power factor of 0.85. This gives a full load horsepower rating of 411.8/0.746 = 552HP. The connection of this motor is shown in figure 4.10 below.

Figure 4.10– Computer Science building chiller connection

The other loads in the CHEMIND model are represented as linear loads, so that the full effect of the motor starting can be studied.

4.2.7 Harmonic Distortion

The non-linear loads in the Chemistry building system were found to be mainly personal computers (PCs) and fluorescent lights. PCs in particular inject a particularly distorted current into the system due to their switched mode power supplies. The sizes of these loads were estimated from the amount of current being drawn into each part of the system. The MS Laboratory itself has a small Adjustable Speed Drive (ASD) motor driving a small pump, as well as a SUN workstation.

36


A 21st harmonic source of suitable amplitude was also placed at Substation STL to simulate the effect of the 1050Hz harmonic throughout the system.

As discussed earlier, these non-linear loads inject non-linear currents into the system, which in turn can cause distorted voltages to appear on the bus. ATPCON has models to simulate the non-linear current injection by three phase and single phase PC loads, three phase fluorescent lights, and six and 12 pulse ASDs. It was decided firstly to simulate the MS Laboratory as a linear load, and then to add in a small ASD and PC load for the second simulation.

[18] gives an explanation of how these harmonic models work.

“The non-linear load is treated as shunt resistance for the first 3 (i.e., IA + 1) cycles (of 60 Hz), then as a sinusoidal current injector with user-input displacement power factor for the next 3 (i.e., IB + 1) cycles, and finally as a harmonic current injector for the remainder of the study (usually 4 more cycles, for a total of 10 cycles).” The information required for the non-linear load models is similar for all models. Given below is the information required by the three phase PC (PC3P) model.

• •

KVA3P: the three phase kVA of the PC load DPF: Fundamental frequency displacement power factor (set to 1.0 for all simulations [18])

• • •

PSHIFT: Connecting transformer phase shift, degrees

•

F: Power system frequency (50Hz)

VLNRMS: Line to neutral bus voltage

TRIPLE: 0 to include triplen harmonics, 1 to exclude them

For the first simulation, non-linear loads parallel to the MS Laboratory were modelled, as well as the Computer Science load. This gave the following values, estimated from the current drawn by each load:

37


LOCATION

LOAD(KVA)

Sub Board A P Cs

79.59

Fluorescent Lights

158.4

Computer Science P Cs Fluorescent Lights

129.6 57.6

Table 4.6 – Loads used for harmonic simulation

In all of these cases, triplen harmonics were included.

Figure 4.11 illustrates the

connection of the PC and fluorescent light loads parallel to the MS Lab, from Sub Board A.

Figure 4.11– Connection of harmonic loads, parallel to MS Lab, from sub board A.

For the next simulation, a small ASD and PC load were added to the MS Laboratory. These were 2 kVA (3 phase) and 1 kVA (phase C) respectively. The values for the linear part of the load were adjusted accordingly.

38


4.2.8 Induction Motor Starting – Central Chiller Station This is very similar to the induction motor section above. The same models were used to examine the starting characteristic of these motors.

The Central Chiller Plant system was added to the template, in the configuration given in Figure 4.12. The ATPDraw circuit diagram is given below.

Figure 4.12– Central Chiller Station

Again, as with the induction motors in the previous section a startup current three times the rated current was used.

39

Chapter

5 Presentation and Analysis of Results

5.1 Capacitor Switching For all figures 5.1,5.2 and 5.3 opening of the switch occurred at t = 0.1172.

Figure 5.1 – Capacitor Switching, Phase A, MS Laboratory.

The transient has a magnitude of approximately 75V, or 22%.

40

Chapter 5 – Presentation and Analysis of Results

Figure 5.2 – Capacitor Switching, Phase B, MS Laboratory

This transient has a magnitude of approximately 30V, or 9%.

Figure 5.3 – Capacitor Switching, Phase C, MS Laboratory.

There is only a small transient on phase C. As can be seen from the waveform in Figure 5.1, capacitor switching was calibrated to a very similar result to what was being seen at the MS Laboratory site. Compare this with figure 4.1. It was necessary to vary the size of the damping resistors in order to get this

41


result, but again, it must be pointed out that the aim is to simulate the power quality problems being recorded by the PQ Node at the Mass Spectrometry Laboratory.

It should be pointed out that the magnitude of the capacitor switching at the MS Laboratory is not very severe. The figure mentioned above was a worst case situation, with switching occurring at the peak voltage. This transient represented a variation of 22% in voltage for a fraction of a cycle.

Now that a good result has been obtained from the simulation, what techniques can be used to limit the effect of capacitor switching?

[1] states that the best means of reducing the effect of capacitor switching is to take measures at the capacitor bank itself.

Some solutions are pre-insertion (damping)

resistors, synchronous closing, where transients are prevented by timing closure of the switches so that the system voltage matches the capacitor voltage at the instant the contacts mate, and, finally, zero crossing switching, where each phase is switched separately at respective zero crossing points. Figure 5.3 shows a greatly reduced transient on phase C, because switching occurs near the zero crossing point.

These simulations also show that the opening of the switch at the capacitor bank produces no obvious distortions in the voltage.

42


5.2 Sags

Figure 5.4 – Symmetrical fault, phase A. All phases are identical.

The voltage has sagged down to 35%.

Figure 5.5 – SLG Fault. All phases.

For the SLG fault, phase B is down to approximately 50%, phase A is down to 70%, and phase C has remained unchanged.

43


Sags were calibrated in simulation to be on a similar level to the severest sags experienced during the site survey. The symmetrical fault produced expected results.

To realistically simulate sags, a survey of all other feeders parallel to the university would be required. At this stage, the author remains unsure as to the effectiveness of using the ATP to simulate unsymmetrical faults in a full and realistic manner.

However, as was pointed out earlier, if the simulation can be made to produce results that were similar to those being experienced at the site of the survey, then it is possible to simulate a solution.

The most common demand-side mitigation technique for severe sags is the UPS, or Uninterruptible Power Supply. Figures 5.6 and 5.7 below illustrate two different design philosophies, Standby and On-line UPS.

Figure 5.6 – Standby UPS.

Figure 5.7 – On-line UPS

44


The Standby UPS has a controller that detects the disturbance, and then switches the load to the battery backed inverter. The most important parameter here is the time it takes to switch the battery backup on. A value of 4ms is an acceptable time for switching to the batteries [1].

On-line UPS systems are always a part of the power system. The incoming AC power is rectified into DC power, which then charges the batteries and in turn is inverted back into AC for the load. When incoming supply fails, the inverter is fed from the battery backup. This provides a complete ride-through capability. On-line UPS systems can, however, be quite lossy.

These systems could possibly be simulated with the ATP using a combination of MODELS logic and TACS. The first thing to do would be to investigate the control of these processes, using MODELS to simulate this part, and then TACS to simulate the switching of any silicon controlled rectifiers in the inverter.

5.3 Induction Motor Starting, Chemistry Building

Figure 5.8 – Induction Motor Starting, Computer Science chiller only

45


Here the voltage has sagged by approximately 6.3%. This simulated the starting of the chillers in the Computer Science building only.

Figure 5.9 – Induction Motor Starting, Mechanical Services only.

This represents a sag of only 1%.

The induction motor model, IM3P, performed quite well, giving small sags of between 5% and 7%, similar to many of the smaller sags being recorded at the MS Laboratory site. Figure 5.10 below gives an example of one of the smaller sags experienced during the site survey. These can be attributed to induction motor starting by the characteristic “ramping up” of the rms voltage after the initial sag.

Figure 5.10 – Small sag during site survey, probably from motor starting

46


Perhaps the biggest surprise during simulation was the extremely small sag given as a result of starting the motor on the Mechanical Services part of the circuit. Whilst this motor is smaller than the combination of the chiller motors in the Computer Science building, it is a direct on-line motor, meaning that no means are being employed to “soften” the start of the motor. The motors in the CS building employ Y-∆ starters to reduce start-up current.

To investigate this further, figures 5.11, 5.12 and 5.13 below represent the currents in several parts of the circuit during the motor start-up. Figure 5.11 shows the current in the PC and fluorescent light circuits (parallel to the Mechanical Services motor), 5.12 shows the current on the 11kV feed to both T1 and T2, and 5.13 shows the current from T1 to Sub. Board A.

Figure 5.11 – Current to parallel PC and fluorescent light circuits.

47


Figure 5.12 – Current on the 11kV feed.

Figure 5.13 – Current from T3 to Sub. Board A.

These figures show that the start-up current being drawn by the motor is coming from two places: the parallel PC and fluorescent light circuits, and the source. As can be seen, there is little current drawn from the circuit behind the transformer T3. This can be explained by the fact that the source represents very low impedance compared to that of the Chemistry building.

48


The impedance seen from the high side of transformer T3, an 11kV

∆

/430 kV Y

transformer, can be calculated as follows:

Z 11kVside

=

V LL Vll

2

Z 415Vside

≈ 702.57 Z 415Vside

Hence, this will represent an impedance much higher than the source, and is relatively protected to the effects of the starting of the mechanical services motor. Mitigation techniques for the effects of sags caused by induction motor starting can take two approaches. The first is a similar approach to that of sags; a standby or on-line UPS, the second is to use techniques such as the Y- ∆ starters described earlier.

Fortunately, the IM3P model provided by ATPCON is fairly flexible in this regard, with the ability to simply specify how much more current the motor will draw at start-up than during normal operation. One thing that could make the model more realistic would be the ability to accurately simulate the ramp-up function during startup.

Again, it should be pointed out that the magnitude of the sags shown by the modelling is not very severe. Many of the induction motor starting events recorded by the PQ Node during the survey of the MS Laboratory were within tolerance levels, as shown by the CBEMA curve [21].

49


5.4 Harmonic Distortion, MS Laboratory 5.4.1 MS Laboratory Modelled as a Linear Load

Figure 5.14 – The voltage waveform on the primary side of T3.

Figure 5.15– Fourier analysis, voltage waveform, primary side of T3.

Figure 5.15 shows the harmonic breakdown of the primary side voltage. This wave has a THD of 4.9%.

50


Figure 5.16 – Current waveform, primary side of T3.

Figure 5.17– Fourier analysis, current waveform, primary side.

Figure 5.17 shows the harmonic breakdown of the primary side current. This wave has a THD of 3.44%.

51


Figure 5.18 – Voltage waveform, secondary of T3.

Figure 5.19– Fourier analysis, voltage waveform, secondary side of T3.

Figure 5.19 shows the harmonic breakdown of the secondary side voltage. This wave has a THD of 4.04%.

52


Figure 5.20 – Current waveform, secondary of T3.

Figure 5.21 – Fourier analysis, current waveform, secondary side of T3.

Figure 5.21 shows the harmonic breakdown of the secondary side current. This wave has a THD of 3.45%.

53


Figures 5.14 – 5.21 show the results given for the harmonic studies.

Note the

st

dominance of the 21 harmonic, Hot Water Switching signal (HWSS), which is used to switch hot water systems on and off. As was discussed in Chapter 4, this signal was added at the 33kV bus to simulate the HWSS, at a proportion calculated from results of the site survey.

Figure 5.22 below gives a summary of the harmonic spectra of the voltage waveforms on all phases, for the duration of the site survey [22]. These can be compared with the results given by the ATP. When comparing these figures, it is important to keep in mind that the simulations were made on the basis of educated guesses about the type and sizes of the loads in the building.

Figure 5.22 – Summary of harmonic voltage levels, primary of T3, during site survey[22].

Examining, for example, figure 5.15, and comparing it with figure 5.22, shows that the simulation results are actually quite acceptable – the lower order harmonics are present in similar levels, and the 21st is dominant, giving similar THDs for both. Looking at figure 5.19, we can see that the 3rd and 9th (triplen) harmonics have been suppressed at the secondary side. This makes sense, given that triplen (zero sequence)

54


components should be suppressed by a ∆-Y transformer, although the 21st harmonic has passed through unsuppressed.

The interesting thing to note with these results is the current waves, on both sides of the transformer T3, given in figures 5.16 and 5.20. The harmonic spectra of these (figures 5.17 and 5.21) are quite good – again dominated by the 21 st harmonic, but in both cases with THDs of around 4%. This shows that only a very small portion of the harmonic currents injected by the parallel PC and fluorescent light loads are going to the MS Laboratory.

The reason for this becomes obvious by using a similar argument to the one used when explaining the induction motor results previously.

Table 4.4 shows that the cable

running from Sub. Board A to the MS Laboratory represents an impedance of approximately 0.0776Ω, and the impedance back to the transformer is 0.00305Ω. In addition, the MS Laboratory load, when brought to the high side of the 415V ∆/208V Y transformer, represents an impedance of approximately 17 Ω (see section 5.3 for the example equation). Thus it is obvious that only a small portion of the harmonic current will flow to the MS Laboratory. To illustrate this point further, figure 5.23 gives the breakdown of the current flowing from Sub. Board A to the transformer, T3.

Figure 5.23 – Fourier analysis, current, going from Sub. Board A to T3.

55


This is a badly distorted wave with a THD of 42.4%!

It is obvious, then, that the distortion that is occurring to the voltage is a result of the harmonic currents from the sources parallel to the site in question.

These results all give THDs less than the 5% limit set by IEEE-519 [8]. However the voltage at the primary, which has a THD of 4.9%, is very close to unacceptable.

In order to suppress the harmonic currents injected by the PC and fluorescent loads, and thus suppress the distortion of the voltage, harmonic filters can be used. Active filters [23] inject their own currents which cancel out the currents injected by the non-linear load. They are called “active” because they adjust the output current according to the levels being injected by the non-linear load. Passive filters work by providing a low impedance path for the harmonic currents to ground. th

A passive filter, to suppress 5 harmonic currents present in the PC and fluorescent light loads, will now be presented. [18] gives a guide on the design of the filter. The filter consists of a series R-L-C circuit tuned to the fifth harmonic, where R is the parasitic resistance of the inductors used. Noting from figure 5.23 that the 5th harmonic current is quite high, a large capacity filter is required. Following [18], the filter will be rated at 150 kVar per phase. Two basic rules given are:

• •

The filter must be rated to a line to neutral voltage of 1.2

× Vlnrms.

As maximum capacitor voltage occurs at a frequency just below resonant frequency, the filter shall be tuned to 4.7 rather than 5 times the fundamental [18]

Values of C, L and R can then be calculated thus: Q 150000 C= = = 5.76mF 1.2 × Vln,rms × ω 288 2 × 314.16

L=

1 2

ω res C

=

1 (314.16 × 4.7) 2 × 5.76 × 10 −3

56

= 79.6 µH


Therefore, if parasitic resistance is approximately XL/50 at 50Hz, R = 0.5mΩ. Figure 5.24 gives the circuit added to the ATPDraw harmonics circuit at Sub. Board A.

Figure 5.24 – Passive 5 th harmonic filter added at Sub. Board A.

Figures 5.25, 5.26 and 5.27 below give the result after adding the filter.

th

Figure 5.25 – Fourier analysis of voltage at the MS Lab. Primary after addition of 5 harmonic filter.

57


Figure 5.26 – Current flowing in phase A of the 5th harmonic filter

Figure 5.27 – Fourier analysis of current in the filter. THD = 21.9%.

Figure 5.25 shows that the 5th harmonic content of the waveform at the MS Laboratory primary has been much reduced.

This has also brought the THD down to 3.8%.

Figures 5.26 and 5.27 show the current in the filter, which has a strong 5 th harmonic component. This filter is impractical and very lossy – figure 5.26 shows a peak current in the filter of 800A! However, these results do reinforce the theory that harmonics from the parallel PC and fluorescent loads are causing the distorted voltage at the primary of T3.

58


5.4.2 MS Laboratory Modelled as a Partly Non-Linear Load A small 3 phase six pulse ASD was added, and a small PC load on phase C

Figure 5.28– MS Laboratory, voltage waveform, phases A (curve a) &C (curve b), primary side of T3.

Figure 5.29– Fourier analysis of phase A voltage, primary side of T3.

59


Figure 5.30– Fourier analysis of phase C voltage, primary side of T3.

Figures 5.29 and 5.30 above show the Fourier analyses of the phase A and C voltage waveforms. THDs are 4.88% and 4.9% respectively.

Figure 5.31– MS Laboratory, current waveform, phases A (curve b) &C (curve a), primary side of T3.

60


Figure 5.32– Fourier analysis of phase A current, primary side of T3.

Figure 5.33– Fourier analysis of phase C current, primary side of T3.

Figures 5.32 and 5.33 above show the Fourier analyses of the phase A and C current waveforms. THDs are 11.4% and 12.8% respectively.

61


Figure 5.34– MS Laboratory, voltage waveform, phases A (curve b) &C(curve a), secondary side of T3.

Figure 5.35– Fourier analysis of phase A voltage, secondary side of T3.

62


Figure 5.36– Fourier analysis of phase C voltage, secondary side of T3.

Figures 5.35 and 5.36 above show the Fourier analyses of the phase A and C voltage waveforms (208V side). THDs are 4.07% and 4.11% respectively.

Figure 5.37– MS Laboratory, current waveforms, phases A(curve a) &C(curve b), secondary side of T3.

63


Figure 5.38– Fourier analysis of phase A current, secondary side of T3.

Figure 5.39– Fourier analysis of phase C current, secondary side of T3.

Figures 5.38 and 5.39 above show the Fourier analyses of the phase A and C current waveforms, on the secondary of T3. THDs are 4.02% and 19.5% respectively.

Figure 5.28 shows the phase A and C voltage waves at the primary of T3. These are quite similar, as is also shown by the spectra in figures 5.29 and 5.30. Again, these compare favourably with the results of the site survey given previously. Predictably,

64


because of the PC on phase C, the primary current is more distorted on phase C than phase A.

It is interesting to note that, comparing the current breakdowns on phase C for the primary and secondary (figures 5.33 and 5.39), the triplen harmonics present in the phase C current on the primary side have only been slightly suppressed. It can be seen that the THD has gone down by approximately 7%, but the relative levels have remained similar.

One way to accurately model any solutions to the voltage distortion problem at the MS Laboratory is to design a voltage source to simulate the exact voltages present. Using the MODELS feature of the ATP, this can be done quite easily.

Using the worst case levels for each phase from figure 5.22 (which was phase C, except for the 21st harmonic) a voltage source was designed to give a similar Fourier breakdown to that recorded by the PQ Node. This is given as HARM.MOD in Appendix A. Given below are the waveform and spectrum produced. This source could then be used to test the effectiveness of any techniques employed to filter the incoming voltage.

Figure 5.40 – Output from model harm.mod.

65


Figure 5.41 – The Fourier analysis of the waveform in figure 5.40.

HARM.MOD was adapted from the method used in [8] to construct the harmonic models in ATPCON.

66


5.5 Central Chiller Station

Figure 5.42– Induction Motor Starting, Central Chiller.

This represents a sag of approximately 10%.

The results of this set of simulations are similar to those produced during the site survey.

Figure 5.43 gives a motor starting example from the Central Chiller site

recorded during the survey.

Figure 5.43 – Motor starting recorded by the PQ Node during survey

67


As can be seen by comparing figures 5.42 and 5.43, the results have been reasonably similar. For a further discussion of induction motor starting, please see section 5.3 above.

5.6 The Effectiveness of the ATP The Alternative Transients Program is a program that was derived from the EMTP, first written in the 1960s. Today, the ATP/EMTP is still used extensively throughout the power industry, for purposes like those in this project.

On its own, the ATP is a difficult program to use. It would have been an extremely time consuming and near impossible task for the author to have written the files used for this project by hand. The fixed FORTRAN compatible formatting of the files limits the size of hand written ATP simulations to small systems only. Even installing the ATP for use with the Windows 95 operating system was a task that was at times very frustrating.

However, because the ATP has been a standard worldwide, there is much collective experience in the use of the program. It was this experience which was drawn upon through the use of the models provided in ATPCON.

ATPDraw has been an essential step in the evolution of the ATP. Now it is possible to simulate large systems and generate the files in an instant.

A newer version of

ATPDraw, designed specifically for Windows 95, is available. ATPDraw for DOS was used, simply because the ATPCON models were written for the DOS version. It would be a relatively simple but time consuming task to convert these files to ATPDraw for Windows. This would perhaps be a good idea for any future work, because frequent crashes and memory problems caused by ATPDraw/ATP made the system frustrating to use.

Another problem with the ATP is simulation time. Once MODELS are introduced, the simulation process can become quite lengthy. For example, simulating a number of harmonic sources or induction motors at once resulted in simulations taking as long as

68


five minutes. Occasionally, the ATPDraw generates incorrect files. These errors can be very difficult to find.

The TPPLOT viewer for the output files is extremely versatile, but has quite a steep learning curve. A menu driven, Windows based viewer for the .pl4 files output by the ATP would be an excellent addition to the software suite.

From a technical point of view, the ATP is an excellent performer.

Most of the

simulation results were similar to those experienced during the site survey, especially when the approximations made to the system are considered.

The author is still

doubtful, however, about the use of the ATP for simulation of unsymmetrical voltage sags. More work needs to be carried out in this area.

Specialised software is available for the simulation of each of the events simulated in this project. However, the ATP is a relatively inexpensive, proven all-round performer that, with the addition of MODELS and ATPDraw, becomes more flexible and much easier to use.

69

Chapter

6 Conclusions

•

Simulation of Power Quality problems was attempted using the Alternative Transients Program

•

Capacitor switching at the distribution substation STL was successfully simulated. It was found that, although some calibration to the results of the site survey was required, results of the simulation were quite accurate, with similar magnitude and rise times to the transients recorded during the site survey. These simulations also agreed with theory that states that switching off capacitor banks causes little to no disturbance in the system.

•

The capacitor switching simulation also showed that the transient is much reduced when switched near the zero crossing of the voltage waveform.

• •

Sags caused by faults in the distribution system were also simulated. Sags caused by induction motor starting were simulated. The first induction motors to be simulated were those present in the Chemistry building system. These showed that it was possible for many of the smaller sags recorded during the site survey to be attributed to induction motors in the system. These simulations also showed that a system could be partly protected from the effects of nearby induction motor starting by a step down transformer, especially one with a high turns ratio.

•

Simulation of induction motor starting at the Central Chiller Station also gave similar results to those seen during a site survey.

•

The induction motor model provided by ATPCON[18] is extremely flexible, with solutions to induction motor starting problems, such as “soft-starting” very simple to simulate, due to the fact that magnitude of startup current can be specified.

70

Chapter 6 – Conclusions

•

The harmonic models provided by ATPCON [18] were used successfully to obtain remarkably accurate results, given the estimates of the system that were made.

These showed that the most likely cause of the voltage

distortion being experienced were harmonic currents being injected by parallel PC and fluorescent light loads. These currents then caused distorted voltages to appear.

Again, good protection against distorted currents is

provided by a step down transformer.

•

As with the induction motor model, all harmonic models used in the simulation were extremely easy to use, and gave good results

•

A harmonic voltage source was developed that can be used to accurately simulate voltages being recorded by power quality monitors.

•

One drawback of the use of MODELS in simulation is the increase in simulation time. Simulations where a number of these loads were placed in the system took as long as five minutes.

•

The ATPDraw is an ideal way to interface with the ATP. The time saved by

•

instant generation of the file for ATP processing was invaluable. The addition of MODELS to the ATP made simulating harmonic and induction motor loads much easier.

•

Some work needs to be done to improve the ATP for use with Windows. System crashes and memory problems were a regular occurrence.

•

TPPLOT, the viewer for the output of the ATP, is extremely versatile, but needs to be updated into a menu driven system, which would be much simpler and quicker to use.

•

The ATP is a good, all-round performer for the simulation of power quality problems

6.1 Recommendations for Further Work

•

The first recommendation I would make would be that the ability of the ATP to simulate system faults be more extensively researched

•

Convert the ATPCON models for use with ATPDraw for Windows 95. This newer version would possibly have many of the bugs removed, as well as

71

Chapter 6 – Conclusions

being easier to use and the advantage of (simple) portability of circuit diagrams to other programs.

•

Conduct more work on the solutions to power quality problems. The author is convinced that MODELS and TACS could be used to create solutions such as UPS systems and active filters. An ideal project for this type of work would be to conduct the simulations

•

before the building is built and the equipment installed. This would help to

predict any power quality problems before they happen. An accurate survey of the load would also be possible using electrical plans.

•

Finally, the possibility of an equivalent software package to the ATP should also be examined. Although the ATP proved to be useful for the project, simulation time is just too long for large and/or complicated systems. The program itself is prone to quirks, and at times can produce some very bizarre results.

72

Appendix

A The ATP Files

A.1 ATP File for Figure 3.3 BEGIN NEW DATA CASE $MONITOR $CLOSE, UNIT=4 STATUS=DELETE {Destroy empty date/time plot file of "SYSDEP" $OPEN, UNIT=4 FILE=FIG33.PL4 FORM=FORMATTED STATUS=UNKNOWN RECL=8000 C C ******************************************************************************* C C EMTP DATA FILE FIG33.DAT C C C 10 20 30 40 50 60 70 80 C 345678901234567890123456789012345678901234567890123456789012345678901234567890 C ******************************************************************************* C C MISCELLANEOUS DATA CARDS C C DELTA. TMAX. XOPT. COPT. EPSILN. TOLMAT. TSTART. 0.00005 0.1 50.0 50.0 0.0 0.0 0.0 C C 1 2 3 4 5 6 7 8 C 345678901234567890123456789012345678901234567890123456789012345678901234567890 C IOUT. IPLOT. IDOUBL. KSSOUT. MAXOUT. IPUN. MEMSAV. ICAT. NENERG. IPRSUP. 40 1 0 1 1 0 0 0 0 0 C C SERIES RLC BRANCH C C BUS1. BUS2. BUS3. BUS4. R. L. C. . SRC-A NODE-A 0.2 0.1 0.0 0 SRC-B NODE-B 0.2 0.1 0.0 0 SRC-C NODE-C 0.2 0.1 0.0 0 NODE-ADELT-A 5.0 0.5 0.0 0 NODE-BDELT-B 4.0 0.1 0.0 0 NODE-CDELT-C 1.0 2.0 0.0 0 DELT-ADELT-B 2.0 1.0 0.0 1 DELT-BDELT-C 5.0 0.0 0.0 1 DELT-CDELT-A 1.0 1.0 50.0 1 C C BLANK CARD TERMINATING BRANCH CARDS BLANK CARD TERMINATING SWITCH CARDS C C VOLTAGE SOURCE - SINUSOIDAL (CHANGE V. TO 1 FOR CURRENT SOURCE) C C 1 2 3 4 5 6 7 8 C 345678901234567890123456789012345678901234567890123456789012345678901234567890 C BUS1.V.AMPLITUDE.FREQUENCY. PHASE. A1. TSTART. TSTOP. 14SRC-A 0 240.0 50.0 0.0 0.0 -1.0 0.0 14SRC-B 0 240.0 50.0 -120.0 0.0 -1.0 0.0 14SRC-C 0 240.0 50.0 120.0 0.0 -1.0 0.0 C BLANK CARD TERMINATING SOURCE CARDS C C NODE OUTPUT SPECIFICATION CARD (can have more than one line of nodes to output)

73

Appendix A – The ATP Files

C

.... .... .... .... .... .... .... .... NODE-ANODE-BNODE-CSRC-A BLANK CARD ENDING NODE NAMES FOR VOLTAGE OUTPUT BLANK CARD TERMINATING PLOT SPECIFICATION CARDS C BEGIN NEW DATA CASE BLANK TERMINATION-OF-RUN CARD

....

....

....

....

A.2 Template.atp BEGIN NEW DATA CASE C -----------------------------------------------C Generated by ATPDRAW C a Bonneville Power Administration program C Programmed by H.K.H›idalen, EFI - NORWAY 1995 C -----------------------------------------------$PREFIX,C:\ATPDRAW\LIB\ $SUFFIX, .LIB $DUMMY, XYZ000 C Miscellaneous Data Card .... POWER FREQUENCY 5.0E+01 C C C

∆T

Tmax Xopt Copt | | | | 2.0E-05 2.0E-01 0.0E+00 0.0E+00

C C IOUT IPLOT C | | 500 1 0 3 0 0 0 1 0 C 1 2 3 4 5 6 7 8 C 345678901234567890123456789012345678901234567890123456789012345678901234567890 /BRANCH C < n 1>< n 2>< R >< L >< C C < n 1>< n 2>< R >< A >< B TRANSFORMER .01 .001TX00011.00E6 1.414213562E-01 4.501581581E+02 9999 1X0001AX0001B .0001 .000115.945 2X0002A .001 .0312 .3511 TRANSFORMER TX0001 TX0002 1X0001BX0001C 2X0002B TRANSFORMER TX0001 TX0003 1X0001CX0001A 2X0002C X0002AX0038A .0003 X0002BX0038B .0003 X0002CX0038C .0003 C Cable - Sub 23 to Sub 10 X0006AX0001A .0075 X0006BX0001B .0075 X0006CX0001C .0075 X0038AX0009A .0776 X0038BX0009B .0776

> ><><>0 0

3 3 3 1 1 1 3 3

X0038CX0009C X0012AX0010A X0012BX0010B X0012CX0010C

.0776 .001 .001 .001 C Cable - Sub 7 to Sub 23 X0018AX0006A .0112 X0018BX0006B .0112 X0018CX0006C .0112 TRANSFORMER .01 .001TX00041.00E6 1.414213562E-01 4.501581581E+02 9999 1X0062AX0062B .001 .001 46.67 2X0023A .1265 4.03 8.981 TRANSFORMER TX0004 TX0005 1X0062BX0062C 2X0023B

3 3 3 3 0 0 0 0

74


TRANSFORMER TX0004 TX0006 1X0062CX0062A 2X0023C TRANSFORMER .01 .001TX00071.00E6 1.414213562E-01 4.501581581E+02 9999 1X0009AX0009B .0001 .0001 .5869 2X0012A .0064 .202 .1697 TRANSFORMER TX0007 TX0008 1X0009BX0009C 2X0012B TRANSFORMER TX0007 TX0009 1X0009CX0009A 2X0012C X0038AX0031A 1.0E-5 X0038BX0031B 1.0E-5 X0038CX0031C 1.0E-5 X0038AX0019A 1.0E-5 X0038BX0019B 1.0E-5 X0038CX0019C 1.0E-5 C PCs - off Sub Board A X0038AX0030A 1.0E-5 X0038BX0030B 1.0E-5 X0038CX0030C 1.0E-5 X0043AX0011A 1.0E-5 X0043BX0011B 1.0E-5 X0043CX0011C 1.0E-5 X0002AX0043A .0075 X0002BX0043B .0075 X0002CX0043C .0075 X0048AX0032A 1.0E-5 X0048BX0032B 1.0E-5 X0048CX0032C 1.0E-5 X0046AX0048A 1.0E-5 X0046BX0048B 1.0E-5 X0046CX0048C 1.0E-5 X0048AX0049A 1.0E-5 X0048BX0049B 1.0E-5 X0048CX0049C 1.0E-5 TRANSFORMER .01 .001TX00101.00E6 1.414213562E-01 4.501581581E+02 9999 1X0062AX0062B .001 .001 46.67 2X0023A .1265 4.03 8.981 TRANSFORMER TX0010 TX0011 1X0062BX0062C 2X0023B TRANSFORMER TX0010 TX0012 1X0062CX0062A 2X0023C TRANSFORMER .01 .001TX00131.00E6 1.414213562E-01 4.501581581E+02 9999 1X0001AX0001B .0001 .0001 15.56 2X0048A .001 .0297 .3535 TRANSFORMER TX0013 TX0014 1X0001BX0001C 2X0048B TRANSFORMER TX0013 TX0015 1X0001CX0001A 2X0048C 1 X0062AX0055A .626 11.49 2 X0062BX0055B .294 7.03 3 X0062CX0055C .294 7.03 /SWITCH C < n 1>< n 2>< Tclose >< Ie X0023AX0018A -1. 1. X0023BX0018B -1. 1. X0023CX0018C -1. 1. X0048AX0002A 1. -1. X0048BX0002B 1. -1. X0048CX0002C 1. -1. /SOURCE C < n 1><>< Ampl. >< Freq. ><

0

1 1 1 1 1 1 1 1 1 3 3 3 0 0 0 1 1 1 1 1 1 1 1 1 0

0

.626 11.49 .294 7.03 ><

A1

75

><

T1

.626 11.49 type

>

>< TSTART >< TSTOP

>


14X0055A 0 26944. 50. 14X0055B 0 26944. 50. -120. 14X0055C 0 26944. 50. 120. C MS Lab, Linear Portion C User specified object: LOADY C RA = 2.2E+00 C LA = 1.1E+01 C RB = 2.2E+00 C LB = 1.1E+01 C RC = 2.2E+00 C LC = 1.1E+01 C VOLTS = X0010A $INCLUDE, LOADY, X0010, 2.24, 11.5, 2.24, 11.5, C CS Chiller C User specified object: LOADY C RA = 4.9E-01 C LA = 2.5E+00 C RB = 4.9E-01 C LB = 2.5E+00 C RC = 4.9E-01 C LC = 2.5E+00 C VOLTS = X0011A $INCLUDE, LOADY, X0011, .495, 2.544, .495, 2.544, C Linear - Sub Board A C User specified object: LOADY C RA = 7.7E+00 C LA = 3.9E+01 C RB = 7.7E+00 C LB = 3.9E+01 C RC = 7.7E+00 C LC = 3.9E+01 C VOLTS = X0019A $INCLUDE, LOADY, X0019, 7.67, 39.3, 7.67, 39.3, C Sub Board A PCs C User specified object: LOADY

-1. -1. -1.

2.24,

11.5

.495, 2.544

7.67,

39.3

C RA = 2.6E+00 C LA = 1.3E+01 C RB = 2.6E+00 C LB = 1.3E+01 C RC = 2.6E+00 C LC = 1.3E+01 C VOLTS = X0030A $INCLUDE, LOADY, X0030, 2.554, 13.12, 2.554, 13.12, 2.554, 13.12 C Fluoro's - Off Sub Board A C User specified object: LOADY C RA = 1.3E+00 C LA = 6.6E+00 C RB = 1.3E+00 C LB = 6.6E+00 C RC = 1.3E+00 C LC = 6.6E+00 C VOLTS = X0031A $INCLUDE, LOADY, X0031,1.2772, 6.56,1.2772, 6.56,1.2772, 6.56 C CS Building - Fluoro's C User specified object: LOADY C RA = 3.5E+00 C LA = 1.8E+01 C RB = 3.5E+00 C LB = 1.8E+01 C RC = 3.5E+00 C LC = 1.8E+01 C VOLTS = X0032A $INCLUDE, LOADY, X0032, 3.529,18.128, 3.529,18.128, 3.529,18.128 C CS Building - PCs C User specified object: LOADY C RA = 1.6E+00 C LA = 8.1E+00 C RB = 1.6E+00 C LB = 8.1E+00 C RC = 1.6E+00 C LC = 8.1E+00 C VOLTS = X0046A $INCLUDE, LOADY, X0046, 1.569, 8.057, 1.569, 8.057, 1.569, 8.057 C Mech Services - Linear

76

1. 1. 1.


C User specified object: LOADY C RA = 7.1E-01 C LA = 3.6E+00 C RB = 7.1E-01 C LB = 3.6E+00 C RC = 7.1E-01 C LC = 3.6E+00 C VOLTS = X0049A $INCLUDE, LOADY, X0049, .7059, 3.63, .7059, 3.63, .7059, 3.63 BLANK BRANCH BLANK SWITCH BLANK SOURCE X0012AX0009AX0009BX0009CX0001AX0011AX0062AX0062BX0062CX0038A X0038BX0038CX0002AX0002BX0002C BLANK OUTPUT BLANK PLOT BEGIN NEW DATA CASE BLANK

A.3 HARM.MOD MODEL HARM COMMENT HARM.MOD IS A HARMONIC VOLTAGE SOURCE DEVELOPED FROM THE ATPCON [18] MODEL ASD6P. INSTEAD OF INJECTING CURRENTS, THIS MODEL OUTPUTS A HARMONIC VOLTAGE, USING TACS SOURCE 60 AT THE OUTPUT. ANDREW SENINI, 2/9/98. PHASE B AND C LAG AND LEAD PHASE A BY 120 DEGREES, RESPECTIVELY. SINE SERIES HARMONIC VOLTAGE. THE ONLY INPUTS REQUIRED ARE THE LINE TO NEUTRAL VOLTAGE REQUIRED, AND THE FREQUENCY ENDCOMMENT OUTPUT VOLTSA,VOLTSB,VOLTSC DATA VLNRMS,F VAR VOLTSA,VOLTSB,VOLTSC,WT,ANG,ASCALE VAR PI180,A120,FT,ASTEP VAR A1,A2,A3,A4,A5,A7,A9,A11,A13,A15,A16,A17,A19 VAR A20,A21,A22,A23,A25,A27,A29 VAR AMAG1,AMAG2,AMAG3,AMAG4,AMAG5,AMAG7,AMAG9,AMAG11,AMAG13 VAR AMAG15,AMAG16,AMAG17,AMAG19,AMAG20,AMAG21,AMAG22,AMAG23,AMAG25 VAR AMAG27,AMAG29 VAR ANG1,ANG2,ANG3,ANG4,ANG5,ANG7,ANG9,ANG11,ANG13,ANG15,ANG16 VAR ANG17,ANG19,ANG20,ANG21,ANG22,ANG23,ANG25,ANG27,ANG29 CONST IA {VAL : 2}, IB {VAL : 2} COMMENT SET THE INITIAL VALUES ENDCOMMENT INIT VOLTSA:= 0.0 VOLTSB:= 0.0 VOLTSC:= 0.0 PI180:= PI / 180 A120:= 120 * PI180 ASCALE:= VLNRMS * SQRT(2) ASTEP:= ASCALE / 100.0 * timestep / 1.667e-5 COMMENT SET THE MAGNITUDES AND ANGLES OF EACH HARMONIC THESE CAN BE ADJUSTED EASILY TO FIT ANY REQUIRED VOLTAGE SOURCE ENDCOMMENT AMAG1:= 1.000000 AMAG2:= 0.002365 / AMAG1 AMAG3:= 0.025068 / AMAG1 AMAG4:= 0.001488 / AMAG1 AMAG5:= 0.025409 / AMAG1

77


AMAG7:= 0.011441 / AMAG9:= 0.016214 / AMAG11:= 0.008959 / AMAG13:= 0.010242 / AMAG15:= 0.010540 / AMAG16:= 0.001576 / AMAG17:= 0.009682 / AMAG19:= 0.009333 / AMAG20:= 0.002581 / AMAG21:= 0.037794 / AMAG22:= 0.001706 / AMAG23:= 0.005931 / AMAG25:= 0.005627 / AMAG27:= 0.002139 / AMAG29:= 0.002077 / AMAG1:= 1.0000 A1:= 0 * PI180 A2:= 0 * PI180 A3:= 0 * PI180 A4:= 0 * PI180 A5:= 180 * PI180 A7:= 180 * PI180 A9:= 0 * PI180 A11:= 0 * PI180 A13:= 0 * PI180 A15:= 0 * PI180 A16:= 0 * PI180 A17:= 180 * PI180 A19:= 180 * PI180 A20:= 0 * PI180 A21:= 0 * PI180 A22:= 0 * PI180 A23:= 0 * PI180 A25:= 0 * PI180 A27:= 0 * PI180

AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1 AMAG1

A29:= 180 * PI180 ENDINIT COMMENT FIRST SET THE VALUES FOR THE ANGLES, THEN OUPUT THE HARMONIC WAVEFORM ENDCOMMENT EXEC FT:= F * T WT:= 2 * PI * FT ANG:= 0 ANG1:= A1 + ANG ANG2:= A2 + 2 * ANG ANG3:= A3 + 3 * ANG ANG4:= A4 + 4 * ANG ANG5:= A5 + 5 * ANG ANG7:= A7 + 7 * ANG ANG9:= A9 + 9 * ANG ANG11:= A11 + 11 * ANG ANG13:= A13 + 13 * ANG ANG15:= A15 + 15 * ANG ANG16:= A16 + 16 * ANG ANG17:= A17 + 17 * ANG ANG19:= A19 + 19 * ANG ANG20:= A20 + 20 * ANG ANG21:= A21 + 21 * ANG ANG22:= A22 + 22 * ANG ANG23:= A23 + 23 * ANG ANG25:= A25 + 25 * ANG ANG27:= A27 + 27 * ANG ANG29:= A29 + 29 * ANG VOLTSA:= AMAG1 * SIN( WT + ANG1) + AMAG2 * SIN( 2 * WT + ANG2) + AMAG3 * SIN( 3 * WT + ANG3) + AMAG4 * SIN( 4 * WT + ANG4) + AMAG5 * SIN( 5 * WT + ANG5) + AMAG7 * SIN( 7 * WT + ANG7) + AMAG9 * SIN( 9 * WT + ANG9) + AMAG11 * SIN(11 * WT + ANG11) + AMAG13 * SIN(13 * WT + ANG13)

78


+ + + + + + + + + + +

AMAG15 AMAG16 AMAG17 AMAG19 AMAG20 AMAG21 AMAG22 AMAG23 AMAG25 AMAG27 AMAG29

* * * * * * * * * * *

SIN(15 SIN(16 SIN(17 SIN(19 SIN(20 SIN(21 SIN(22 SIN(23 SIN(25 SIN(27 SIN(29

COMMENT PHASE B LAGS BY 120 DEGREES ENDCOMMENT WT:= WT - A120 VOLTSB:= AMAG1 * SIN( + AMAG2 * SIN( 2 + AMAG3 * SIN( 3 + AMAG4 * SIN( 4 + AMAG5 * SIN( 5 + AMAG7 * SIN( 7 + AMAG9 * SIN( 9 + AMAG11 * SIN(11 + AMAG13 * SIN(13 + AMAG15 * SIN(15 + AMAG16 * SIN(16 + AMAG17 * SIN(17 + AMAG19 * SIN(19 + AMAG20 * SIN(20 + AMAG21 * SIN(21 + AMAG22 * SIN(22 + AMAG23 * SIN(23 + AMAG25 * SIN(25 + AMAG27 * SIN(27 + AMAG29 * SIN(29

* * * * * * * * * * *

WT WT WT WT WT WT WT WT WT WT WT

WT + * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT * WT

COMMENT PHASE C LEADS BY 120 DEGREES ENDCOMMENT WT:= WT - A120 VOLTSC:= AMAG1 * SIN( WT + + AMAG2 * SIN( 2 * WT + AMAG3 * SIN( 3 * WT + AMAG4 * SIN( 4 * WT + AMAG5 * SIN( 5 * WT + AMAG7 * SIN( 7 * WT + AMAG9 * SIN( 9 * WT + AMAG11 * SIN(11 * WT + AMAG13 * SIN(13 * WT + AMAG15 * SIN(15 * WT + AMAG16 * SIN(16 * WT + AMAG17 * SIN(17 * WT + AMAG19 * SIN(19 * WT + AMAG20 * SIN(20 * WT + AMAG21 * SIN(21 * WT + AMAG22 * SIN(22 * WT + AMAG23 * SIN(23 * WT + AMAG25 * SIN(25 * WT + AMAG27 * SIN(27 * WT + AMAG29 * SIN(29 * WT COMMENT OUTPUT TO THE ATP ENDCOMMENT VOLTSA:= VOLTSA * ASCALE VOLTSB:= VOLTSB * ASCALE VOLTSC:= VOLTSC * ASCALE ENDEXEC ENDMODEL

+ + + + + + + + + + +

ANG15) ANG16) ANG17) ANG19) ANG20) ANG21) ANG22) ANG23) ANG25) ANG27) ANG29)

ANG1) + ANG2) + ANG3) + ANG4) + ANG5) + ANG7) + ANG9) + ANG11) + ANG13) + ANG15) + ANG16) + ANG17) + ANG19) + ANG20) + ANG21) + ANG22) + ANG23) + ANG25) + ANG27) + ANG29)

ANG1) + ANG2) + ANG3) + ANG4) + ANG5) + ANG7) + ANG9) + ANG11) + ANG13) + ANG15) + ANG16) + ANG17) + ANG19) + ANG20) + ANG21) + ANG22) + ANG23) + ANG25) + ANG27) + ANG29)

79

Appendix

B Guide to ATPDraw Components Used

COMPONENT

USE & PARAMETERS

Resistor Inductor Capacitor 3 phase series RLC. Enter R, L and C in per phase values Time controlled single phase switch. Specify the opening and closing times Time controlled three phase switch. Specify the opening and closing times (all phases open and close simultaneously) 3 phase AC source. Specify: - Starting and stopping times - Peak magnitude - Frequency - Voltage or Current source - Phase shift, in degrees or seconds TACS Source Type 60 Specify start and stop times, voltage or current source 3 phase pi-equivalent model (used for source impedance) Specify mutual and self resistance, capacitance and inductance for each phase Voltage probe Specify 1,2, or 3 phases Splitter. Separate 3 phase line into individual phases.

80

Appendix B – Guide to Atpdraw Components Used

Loady [18] – Grounded Y 3-phase load Specify parallel R and L for each phase Capd [18] – 3 phase ∆ /ungrounded Y capacitor bank Specify capacitance in each phase TRADY - ∆/Y Transformer (see Table 2.2) PC1P [18] – Single phase PC load. Specify: - Single-phase kVA -Fundamental frequency displacement power factor. - line-to-neutral voltage - Power system frequency IM3P [18] – Three phase induction motor load. (see section 4.2.6)

ASD6P [18] – Three phase 6 pulse ASD. Specify: -Three-phase kVA - Fundamental frequency displacement power factor.(1.0 for voltage-source drive, 0.80 for current source drive) - Line to neutral rms voltage - Connecting transformer phase shift (degrees) - Shape, 0 for current-source drive, 1 for low-ripple voltage-source drive - Power system frequency PC3P [18] – Three phase PC load. (see section 4.2.7)

81

Appendix B – Guide to Atpdraw Components Used

FL3P [18] – Three phase fluorescent light load. (identical to the PC3P model)

HARM – Three phase harmonic voltage source. Specify: - line to neutral voltage - frequency

Table B.1 – ATPDraw components used for simulation

82

Appendix

C Complete Fourier Analyses of Results

HarmonicC osine Sine Complex number coefficient coefficient amplitude

Percent of fundamental

1

-1.74E+02

2.79E+02

3.29E+02

99.99999

2

-6.00E-02

-2.21E-01

2.29E-01

0.06973

3

-1.25E+00

4.80E+00

4.96E+00

1.50881

4

-3.84E-02

-1.22E-01

1.28E-01

0.03885

5

6.16E+00

2.52E+00

6.65E+00

2.0253

6

-8.63E-02

-1.19E-01

1.47E-01

0.04462

7

-8.83E-01

-6.42E+00

6.48E+00

1.97316

8

-7.79E-02

7.54E-03

7.83E-02

0.02383

9

-4.89E+00

2.00E+00

5.29E+00

1.60874

10

1.83E-02

-1.68E-02

2.48E-02

0.00755

11

1.68E+00

3.15E+00

3.57E+00

1.08633

12

-4.43E-02

-7.83E-02

8.99E-02

0.02737

13

6.18E-01

-1.58E+00

1.70E+00

0.51594

14

-6.15E-02

-9.89E-03

6.23E-02

0.01896

15

-2.12E-01

-1.33E-02

2.12E-01

0.06461

16

-5.71E-02

-8.46E-03

5.78E-02

0.01758

17

-6.42E-02

-1.13E-03

6.42E-02

0.01955

18

-7.60E-02

1.11E-02

7.68E-02

0.02337

19

-1.02E-01

3.39E-02

1.07E-01

0.03268

20

-1.81E-01

9.65E-02

2.05E-01

0.06251

21

-8.22E+00

6.08E+00

1.02E+01

3.11256

22

1.50E-01

-1.45E-01

2.09E-01

0.06355

23

6.34E-02

-7.94E-02

1.02E-01

0.03093

24

3.38E-02

-5.72E-02

6.65E-02

0.02024

Derived from table: 1) RMS value = 2.32622162E+02

2) THD = 4.90101385E+00 %

Table C.1 – Fourier analysis of MS Lab. Primary voltage (fig. 5.14)

HarmonicC osine number

Sine

Complex

coefficient coefficient amplitude


1

-3.85E+01

1.78E+01

4.24E+01

100

2

-8.34E-03

-1.37E-02

1.60E-02

0.0378

3

-7.61E-03

-7.78E-03

1.09E-02

0.02564

4

-5.36E-03

-3.90E-03

6.63E-03

0.01562

5

5.77E-01

4.49E-01

7.31E-01

1.72245

83

Appendix C – Complete Fourier Analyses of Results

6

-7.66E-03

-6.15E-03

9.82E-03

7

9.87E-02

-6.85E-01

6.92E-01

0.02315 1.63069

8

-7.29E-03

6.72E-03

9.91E-03

0.02336 0.02109

9

-6.76E-03

5.88E-03

8.95E-03

10

-6.69E-03

7.22E-03

9.85E-03

0.0232

11

4.45E-02

3.79E-01

3.81E-01

0.89868

12

-8.58E-03

-2.77E-03

9.01E-03

0.02124

13

1.14E-01

-1.21E-01

1.66E-01

0.39152

14 15

-1.22E-02 -1.12E-02

1.29E-03 3.30E-04

1.22E-02 1.12E-02

0.02882 0.0264

16

-1.08E-02

7.10E-05

1.08E-02

0.02537

17

-1.09E-02

2.63E-04

1.09E-02

0.02569

18

-1.18E-02

8.49E-04

1.18E-02

0.02788

19

-1.47E-02

1.82E-03

1.48E-02

0.03481

20

-2.41E-02

3.39E-03

2.43E-02

0.05733

21

-9.71E-01

8.20E-02

9.75E-01

2.29637

22

1.38E-02

1.37E-03

1.39E-02

0.03277

23

2.78E-03

2.33E-03

3.63E-03

0.00855

24

-1.55E-03

2.32E-03

2.79E-03

0.00656


2) THD = 3.44592214E+00 %

Table C.2 – Fourier analysis of MS Lab. Primary current (fig. 5.16)

HarmonicC osine Sine Complex number coefficient coefficient amplitude


1

-9.50E+00

1.61E+02

1.61E+02

100

2

-1.30E-02

-1.33E-01

1.33E-01

0.08283

3

-1.09E-02

-8.73E-02

8.80E-02

0.05463

4

-8.00E-03

-6.05E-02

6.10E-02

0.03787

5

1.59E+00

2.80E+00

3.22E+00

2.00104

6

-2.70E-02

-7.44E-02

7.92E-02

0.04916

7

-1.38E+00

-2.82E+00

3.14E+00

1.94838

8

-6.03E-03

-1.82E-02

1.92E-02

0.0119

9

-1.07E-02

-1.97E-02

2.24E-02

0.01391

10

-1.48E-02

-1.10E-02

1.84E-02

0.01142

11

-5.59E-01

1.58E+00

1.67E+00

1.03966

12

-3.89E-03

-4.57E-02

4.59E-02

0.0285

13

1.68E-01

-7.80E-01

7.98E-01

0.49529

14

-1.45E-02

-1.12E-02

1.83E-02

0.01138

15

-1.55E-02

-1.20E-02

1.96E-02

0.01216

16

-1.77E-02

-9.89E-03

2.02E-02

0.01256

17

-2.11E-02

-6.50E-03

2.21E-02

0.01373

18

-2.69E-02

-1.36E-03

2.69E-02

0.01672

19

-3.84E-02

8.00E-03

3.92E-02

0.02435

20

-7.26E-02

3.40E-02

8.01E-02

0.04975

21

-3.49E+00

2.56E+00

4.33E+00

2.6862

84


22

6.86E-02

-6.86E-02

9.70E-02

0.06023

23

3.22E-02

-4.12E-02

5.23E-02

0.03247

24

2.03E-02

-3.18E-02

3.77E-02

0.02343


2) THD = 4.04554605E+00 %

Table C.3 – Fourier analysis at MS Lab. Secondary voltage (Fig. 5.18) Harmonic number

Cosine coefficient

Sine coefficient

Complex amplitude


1

-4.88E+01

6.92E+01

8.47E+01

100

2

-1.99E-02

-5.82E-02

6.15E-02

0.07263

3

-1.93E-02

-3.74E-02

4.21E-02

0.04973

4

-1.76E-02

-2.56E-02

3.11E-02

0.03668

5

5.38E-01

1.34E+00

1.44E+00

1.70677

6

-2.03E-02

-3.29E-02

3.87E-02

0.04571

7

-5.17E-01

-1.31E+00

1.41E+00

1.66395

8

-1.04E-02

-7.26E-03

1.27E-02

0.01499

9

-1.08E-02

-8.10E-03

1.35E-02

0.01599

10

-1.18E-02

-4.42E-03

1.26E-02

0.01483

11

-2.94E-01

6.91E-01

7.51E-01

0.88748

12

-5.74E-03

-1.97E-02

2.05E-02

0.02426

13

8.67E-02

-3.44E-01

3.54E-01

0.41873

14

-1.12E-02

-4.56E-03

1.21E-02

0.01424

15

-1.13E-02

-4.96E-03

1.24E-02

0.01459

16

-1.20E-02

-4.08E-03

1.27E-02

0.01495

17

-1.35E-02

-2.60E-03

1.37E-02

0.01619

18

-1.59E-02

-3.91E-04

1.59E-02

0.01877

19

-2.10E-02

3.61E-03

2.13E-02

0.02519

20

-3.64E-02

1.47E-02

3.93E-02

0.04638

21

-1.60E+00

1.09E+00

1.94E+00

2.28567

22

2.83E-02

-2.92E-02

4.07E-02

0.04805

23

1.19E-02

-1.75E-02

2.11E-02

0.02494

24

6.55E-03

-1.34E-02

1.49E-02

0.01766


2) THD = 3.44800854E+00 %

Table C.4 – Fourier analysis at MS Lab. Secondary current (Fig. 5.20)


1

Sine

Complex

Percent of

coefficient coefficient amplitude fundamental -3.37E+02 4.34E+02 5.50E+02 100

2

4.30E-01

-3.73E-01

5.69E-01

0.10355

3

1.74E+02

2.32E+01

1.76E+02

31.95304

4

-3.28E-01

-5.67E-01

6.55E-01

0.11917

5

3.25E+01

-1.09E+02

1.14E+02

20.71207

6

-6.03E-01

5.75E-01

8.33E-01

0.15158

7

-7.35E+01

2.27E+01

7.70E+01

13.99754

8

6.62E-01

2.79E-01

7.18E-01

0.13058

9

3.38E+01

5.12E+01

6.13E+01

11.15331

10

-3.44E-02

-6.50E-01

6.51E-01

0.11837

85


11

2.11E+01

-1.87E+01

2.82E+01

5.12679

12

-4.22E-01

1.81E-02

4.23E-01

0.07685

13

-1.13E+01

-1.29E+00

1.14E+01

2.06872

14

2.20E-02

-1.06E-02

2.44E-02

0.00444

15

2.83E-01

1.03E+00

1.07E+00

0.19405

16

-4.18E-02

-5.07E-02

6.57E-02

0.01196

17

-4.26E-02

-4.11E-02

5.92E-02

0.01076

18

-4.40E-02

-3.53E-02

5.64E-02

0.01025

19 20

-4.92E-02 -6.80E-02

-2.88E-02 -1.72E-02

5.70E-02 7.01E-02

0.01037 0.01275

21

-2.09E+00

8.13E-01

2.25E+00

0.40876

22

1.69E-02

-4.61E-02

4.91E-02

0.00893

23

-4.88E-03

-3.55E-02

3.58E-02

0.00651

24

-1.23E-02

-3.15E-02

3.38E-02

0.00615


2) THD = 4.24401741E+01 %

Table C.5 – Fourier analysis. Current, Sub. Board A to T1. (Fig. 5.22)

Harmonic number

Cosine coefficient

Sine coefficient

Complex amplitude


1

-1.74E+02

2.79E+02

3.28E+02

100

2

-6.46E-02

-2.23E-01

2.32E-01

0.07072

3

-1.22E+00

4.54E+00

4.71E+00

1.43262

4 5

-4.06E-02 6.02E+00

-1.22E-01 2.27E+00

1.28E-01 6.43E+00

0.03906 1.95895

6

-8.90E-02

-1.16E-01

1.46E-01

0.04459

7

-1.04E+00

-6.42E+00

6.51E+00

1.98167

8

-7.89E-02

9.73E-03

7.95E-02

0.02422

9

-4.99E+00

2.12E+00

5.42E+00

1.65104

10

1.86E-02

-1.62E-02

2.47E-02

0.00751 1.12199

11

1.75E+00

3.24E+00

3.68E+00

12

-4.59E-02

-7.85E-02

9.10E-02

0.0277

13

6.71E-01

-1.52E+00

1.66E+00

0.50685

14

-6.43E-02

-1.11E-02

6.52E-02

0.01986

15

-2.13E-01

-2.01E-02

2.14E-01

0.06524

16

-5.95E-02

-8.83E-03

6.02E-02

0.01833

17

-5.65E-02

3.77E-02

6.79E-02

0.02069

18

-7.89E-02

9.55E-03

7.95E-02

0.02419

19 20

-1.34E-01 -1.83E-01

5.23E-02 9.47E-02

1.44E-01 2.06E-01

0.04385 0.06278

21

-8.26E+00

6.09E+00

1.03E+01

3.12329

22

1.49E-01

-1.47E-01

2.09E-01

0.06369

23

3.02E-02

-6.84E-02

7.47E-02

0.02276

24

3.36E-02

-5.95E-02

6.84E-02

0.02082


2) THD = 4.88265467E+00 %

Table C.6 – Fourier analysis of phase A voltage, primary side of T3. (Fig 5.28)

86


HarmonicC osine

Sine

Complex

Percent of

number

coefficient

amplitude

fundamental

coefficient 1

3.29E+02

1.10E+01

3.29E+02

100

2

-8.48E-02

-1.68E-02

8.64E-02

0.02626

3

-1.31E+00

5.17E+00

5.34E+00

1.6224

4

-1.11E-01

-4.88E-02

1.22E-01

0.03696

5

-5.27E+00

4.24E+00

6.76E+00

2.05459

6

-1.16E-01

-1.03E-01

1.55E-01

0.04715

7 8

-5.04E+00 -8.05E-02

3.70E+00 -1.68E-01

6.25E+00 1.86E-01

1.90089 0.05657

9

-4.81E+00

1.71E+00

5.11E+00

1.5525

10

-9.98E-03

-1.97E-01

1.97E-01

0.05997

11

-3.63E+00

-3.49E-01

3.65E+00

1.10853

12

7.46E-02

-1.58E-01

1.75E-01

0.05314

13

-1.59E+00

-1.33E-03

1.59E+00

0.48431

14

1.12E-01

-1.16E-01

1.61E-01

0.04895

15

-5.66E-02

-8.74E-02

1.04E-01

0.03167

16

8.85E-02

-7.41E-02

1.15E-01

0.03508

17

3.67E-02

-7.46E-02

8.31E-02

0.02526

18

7.24E-02

-5.83E-02

9.29E-02

0.02825

19

1.17E-01

-4.30E-02

1.25E-01

0.03791

20

1.67E-01

-3.92E-02

1.71E-01

0.052

21

9.40E+00

3.92E+00

1.02E+01

3.09501

22

-2.03E-01

-2.55E-01

3.26E-01

0.09923

23

-7.69E-02

-2.70E-01

2.81E-01

0.0854

24

-2.02E-02

-2.34E-01

2.35E-01

0.07135


2) THD = 4.89550686E+00 %

Table C.7 – Fourier analysis of phase C voltage, primary side of T3. (Fig 5.29) HarmonicC osine number coefficient

Sine coefficient

Complex amplitude


1

-3.73E+01

2.12E+01

4.29E+01

100

2

-1.15E-02

-1.26E-02

1.70E-02

0.03965

3

-1.36E+00

2.61E+00

2.94E+00

6.86308

4

-1.14E-03

-5.38E-03

5.50E-03

0.01281

5

4.53E-01

3.30E+00

3.34E+00

7.77459

6

3.27E-03

-2.48E-02

2.51E-02

0.0584

7

1.16E+00

1.68E-01

1.17E+00

2.73123

8

-1.99E-03

-1.55E-02

1.56E-02

0.03642

9

1.14E+00

-1.80E-01

1.15E+00

2.6903

10

-1.39E-02

-1.11E-02

1.78E-02

0.04144

11

1.61E-01

-3.91E-01

4.23E-01

0.98526

12

-1.00E-02

-6.16E-03

1.18E-02

0.02739

13

1.40E-01

-5.69E-01

5.86E-01

1.36569

14

-1.21E-02

6.11E-03

1.36E-02

0.03161

15

-3.70E-02

1.85E-02

4.13E-02

0.09634

16

-8.46E-03

-1.97E-04

8.46E-03

0.01972

17

9.94E-02

-1.60E-01

1.89E-01

0.43968

87


18

-1.21E-02

6.21E-03

1.36E-02

19

1.37E-01

4.50E-02

1.45E-01

0.33707

20

-2.79E-02

5.94E-03

2.85E-02

0.06642

21

-8.65E-01

1.61E-01

8.80E-01

2.05102

22

9.16E-03

1.06E-03

9.23E-03

0.0215

23

1.18E-01

5.94E-02

1.32E-01

0.30821

24

-9.62E-03

1.45E-03

9.73E-03

0.02267


0.03161

2) THD = 1.13892031E+01 %

Table C.8 – Fourier analysis of phase A current, primary side of T3. (Fig 5.31)

HarmonicC osine

Sine

Complex

Percent of

number

coefficient

amplitude

fundamental

coefficient 1

3.55E+01

2.09E+01

4.12E+01

100

2

-1.82E-04

-2.45E-02

2.45E-02

0.05951

3

1.35E+00

-2.64E+00

2.96E+00

7.19686

4

-1.31E-02

-1.62E-02

2.09E-02

0.05068

5

-1.36E+00

-2.93E+00

3.23E+00

7.8549

6

-1.10E-02

2.55E-03

1.13E-02

0.02749

7

-2.03E+00

-3.62E-01

2.06E+00

5.01244

8

5.56E-03

-1.72E-03

5.82E-03

0.01414

9

-1.14E+00

1.64E-01

1.15E+00

2.79322

10

1.52E-02

-4.89E-03

1.60E-02

0.03887

11 12

-3.54E-01 1.51E-02

2.75E-01 -6.15E-03

4.48E-01 1.63E-02

1.08887 0.03962

13

-3.18E-01

2.22E-01

3.87E-01

0.94078

14

2.02E-02

-6.32E-03

2.11E-02

0.05132

15

4.33E-02

-1.90E-02

4.73E-02

0.11489

16

1.47E-02

5.75E-04

1.47E-02

0.03569

17

9.82E-02

1.82E-01

2.06E-01

0.50123

18

9.57E-03

-5.50E-03

1.10E-02

0.02681

19

-3.19E-02

-1.48E-01

1.51E-01

0.36723

20

1.75E-02

8.82E-03

1.96E-02

0.04758

21

5.07E-01

6.69E-01

8.40E-01

2.0392

22

-2.65E-03

-2.09E-02

2.11E-02

0.05126

23

-1.10E-01

6.01E-02

1.25E-01

0.30342

24

1.45E-02

-1.57E-02

2.14E-02

0.05192


2) THD = 1.23759346E+01 %

Table C.9 – Fourier analysis of phase C current, primary side of T3. (Fig 5.32)

HarmonicC osine

Sine

Complex

Percent of

number

coefficient

amplitude

fundamental

coefficient 1

-9.74E+00

1.61E+02

1.61E+02

100

2

-9.07E-03

-1.34E-01

1.34E-01

0.08328

3

4.67E-03

-1.60E-01

1.60E-01

0.09952

4

-5.29E-03

-6.39E-02

6.41E-02

0.03974

5

1.53E+00

2.18E+00

2.66E+00

1.65126

88


6

-2.21E-02

-7.26E-02

7.59E-02

0.04705

7

-1.47E+00

-3.21E+00

3.53E+00

2.18706

8

6.98E-04

-1.08E-02

1.09E-02

0.00673

9

-3.68E-02

1.42E-02

3.94E-02

0.02446

10

-5.21E-03

-3.34E-03

6.19E-03

0.00384

11

-5.82E-02

1.52E+00

1.52E+00

0.94449

12

-5.17E-03

-3.51E-02

3.55E-02

0.02199

13

5.51E-01

-8.07E-01

9.77E-01

0.60609

14 15

-2.58E-02 -2.48E-02

6.13E-04 -1.50E-03

2.58E-02 2.49E-02

0.016 0.01541

16

-2.72E-02

4.06E-03

2.75E-02

0.01708

17

1.31E-01

4.09E-01

4.29E-01

0.26606

18

-4.23E-02

-5.48E-04

4.23E-02

0.02624

19

2.78E-02

3.80E-01

3.81E-01

0.23642

20

-8.90E-02

2.31E-02

9.20E-02

0.05702

21

-3.50E+00

2.74E+00

4.44E+00

2.75447

22

5.12E-02

-8.33E-02

9.78E-02

0.06064

23

-3.26E-01

1.44E-01

3.57E-01

0.2212

24

1.96E-02

-5.21E-02

5.57E-02

0.03452


2) THD = 4.06981230E+00 %

Table C.10 – Fourier analysis of phase A voltage, secondary side of T3. (Fig 5.34)

HarmonicC osine

Sine

Complex

Percent of

number

coefficient

amplitude

fundamental

coefficient 1

1.45E+02

-7.20E+01

1.62E+02

100

2

2.87E-02

5.51E-02

6.22E-02

0.0384

3

1.71E+00

1.12E+00

2.05E+00

1.26494

4

1.70E-02

5.47E-03

1.79E-02

0.01104

5

-6.58E-02

-2.53E-01

2.61E-01

0.16117

6

1.03E-02

-1.54E-02

1.85E-02

0.01144

7

-2.96E-01

5.47E-01

6.22E-01

0.38419

8

1.59E-02

-5.46E-02

5.68E-02

0.03511

9

-2.40E-01

-2.39E+00

2.40E+00

1.48198

10

1.19E-02

-2.93E-02

3.16E-02

0.01953

11

-2.47E+00

-1.75E+00

3.03E+00

1.86836

12

6.54E-02

2.13E-03

6.54E-02

0.04039

13

-1.60E+00

4.95E-02

1.60E+00

0.99018

14

9.60E-02

5.11E-04

9.60E-02

0.05926

15

1.34E-01

9.54E-02

1.65E-01

0.10159

16

7.01E-02

3.73E-03

7.02E-02

0.04332

17

-3.75E-01

-3.47E-02

3.76E-01

0.23249

18

8.64E-02

2.05E-03

8.64E-02

0.05338

19

3.67E-01

-2.78E-01

4.60E-01

0.2841

20

1.25E-01

1.80E-02

1.26E-01

0.07781

21

4.42E+00

1.40E+00

4.63E+00

2.862

89


22

-4.06E-02

-5.93E-02

7.19E-02

0.04438

23

-1.18E-02

-4.57E-01

4.57E-01

0.28236

24

4.37E-02

-1.90E-02

4.76E-02

0.02942


2) THD = 4.10841751E+00 %

Table C.11 – Fourier analysis of phase C voltage, secondary side of T3. (Fig 5.35)


Sine

Complex

Percent of

coefficient coefficient amplitude fundamental 1 -4.52E+01 7.09E+01 8.41E+01 100 2

-3.25E-02

-6.36E-02

7.14E-02

0.08494

3

-1.97E-02

-7.33E-02

7.59E-02

0.09022

4

-3.09E-02

-3.45E-02

4.63E-02

0.05502

5

-1.20E+00

1.51E+00

1.93E+00

2.29032

6

-1.77E-02

-4.47E-02

4.81E-02

0.05718

7

-1.36E+00

-1.20E+00

1.82E+00

2.16

8

2.81E-03

-1.96E-02

1.98E-02

0.02356

9

-1.34E-02

-9.93E-03

1.66E-02

0.01979

10

-3.77E-04

-1.79E-02

1.79E-02

0.02126

11

-4.40E-01

1.37E-02

4.40E-01

0.52344

12

8.65E-03

-1.78E-02

1.98E-02

0.0235

13

3.06E-02

-7.74E-01

7.75E-01

0.92105

14

2.83E-03

7.75E-03

8.25E-03

0.0098

15 16

2.00E-03 2.01E-03

4.90E-03 5.11E-03

5.29E-03 5.49E-03

0.00629 0.00653

17

3.53E-01

-1.51E-01

3.84E-01

0.45638

18

-1.44E-02

1.31E-02

1.95E-02

0.02314

19

2.75E-01

-5.96E-02

2.81E-01

0.33442

20

-4.40E-02

2.79E-02

5.21E-02

0.06191

21

-1.46E+00

1.09E+00

1.82E+00

2.16865

22

1.87E-02

-1.64E-02

2.49E-02

0.02959

23

1.43E-01

2.16E-01

2.59E-01

0.30791

24

-6.10E-03

-1.22E-02

1.37E-02

0.01626


2) THD = 4.02252865E+00 %

Table C.12 – Fourier analysis of phase A current, secondary side of T3. (Fig 5.37) HarmonicC osine number coefficient 1 2

Sine coefficient

Complex amplitude


8.38E+01 9.80E-03

-2.42E+00 -2.10E-02

8.38E+01 2.32E-02

99.99999 0.02766

3

4.68E+00

-9.11E+00

1.02E+01

12.21947

4

-2.75E-02

-1.57E-02

3.17E-02

0.03779

5

-2.76E+00

-9.92E+00

1.03E+01

12.28767

6

-2.88E-02

4.12E-02

5.03E-02

0.05997

7

-5.37E+00

-1.79E+00

5.66E+00

6.75385

8

9.67E-03

3.40E-02

3.53E-02

0.04216

9

-3.96E+00

6.14E-01

4.00E+00

4.77741

10

4.78E-02

2.04E-02

5.19E-02

0.06198

11

-9.95E-01

1.37E+00

1.69E+00

2.01689

90


12

4.33E-02

3.54E-03

4.35E-02

0.05185

13

-4.54E-01

1.19E+00

1.28E+00

1.52383

14

4.47E-02

-1.34E-02

4.67E-02

0.05567

15

1.30E-01

-5.90E-02

1.43E-01

0.17024

16

3.13E-02

5.76E-03

3.18E-02

0.03794

17

9.02E-03

4.03E-01

4.03E-01

0.48132

18

2.73E-02

-8.36E-03

2.85E-02

0.03405

19

-2.00E-01

-2.15E-01

2.94E-01

0.35096

20 21

5.24E-02 1.53E+00

7.31E-03 5.32E-01

5.29E-02 1.62E+00

0.06311 1.93228

22

-1.30E-02

-2.01E-02

2.39E-02

0.0285

23

-2.66E-01

1.07E-02

2.66E-01

0.31771

24

2.72E-02

-1.73E-02

3.22E-02

0.03839


2) THD = 1.94775467E+01 %

Table C.13 – Fourier analysis of phase A current, secondary side of T3. (Fig 5.38)

91

Bibliography

[1]

Dugan, R.C., McGranaghan, M.C., and Beaty, H.W., Electrical Power Systems Quality, New York: McGraw-Hill, 1996.

[2]

Douglas, J., “Solving Problems of Power Quality”, EPRI Journal, pp. 6-15, Dec. 1993.

[3]

Burke, J.J., Griffith, D.C., and Ward, D.J., “Power Quality – Two Different Perspectives”, IEEE Transactions on Power Delivery, vol. 5, no. 3, pp. 15011513, July 1990.

[4]

McGranaghan, M.F., Grebe, T.E., Hensley, G., Singh, T., and Samotyj, M., “Impact of Utility Switched Capacitors on Customer Systems, Part II – Adjustable Speed Drive Concerns”, IEEE Transactions on Power Delivery, vol. 6, no. 4, pp. 1623-1628, Oct. 1991.

[5]

Electrotek Concepts, Electrotek Studies in Utility Power Quality [Online]. Available from http://www.electrotek.com/ps_study/utility/utility.htm [Accessed 30 March 1998]

[6]

Bollen, M.H., “Fast Assessment Methods for Voltage Sags in Distribution Systems”, IEEE Transactions on Industry Applications, vol. 32, no. 6, pp. 14141423, Nov./Dec. 1996.

[7]

Phipps, J.K., Nelson, J.P., and Pankaj, K.S., “Power Quality and Harmonic Distortion on Distribution Systems”, IEEE Transactions on Industrial Applications, vol. 30, no. 2, pp. 476-483, Mar./Apr. 1994.

[8]

IEEE Std. 519-1992, IEEE Recommended Practices and Requirements for Harmonic Control in Electrical Power Systems, IEEE Industry Applications Society and Power Engineering Society, New York, 1993.

[9]

Domijan, A., Heydt, G.T., Meliopoulos, A.P.S., Vankata, S.S., and West, S., “Directions of Research on Electrical Power Quality”, IEEE Transactions on Power Delivery, vol. 8, no. 1, pp. 429-435, Jan. 1993.

[10]

Long, W., Cotcher, D., Ruiu, D., Adam, P., Lee, S., and Rambabu, A., “EMTP – A Powerful Tool for Analysing Power System Transients”, IEEE Computer Applications in Power, pp. 36-41, July 1990.

[11]

Grainger, J.J, and Stevenson, W.D., Power System Analysis, New York: McGraw Hill, 1994, pp. 470-527.

92

Bibliography

[12]

Gilker, C., Dwyer, R.V., and Dugan, R.C., “A Program for the Analysis of Power System Harmonics”, IEEE Computer Applications in Power, pp 36-41, Oct. 1989.

[13]

Lindell, L.C., “Software Predicts Harmonic Problems and Simulates Alternative Solutions”, IEEE Computer Applications in Power, pp. 53-57, Oct. 1993.

[14]

Lotfalian, M., “Modelling and Simulation of Power Systems Using PSPICE for Power Quality Analysis”, Proceedings of the IASTED International Conference, Modelling and Simulation – MS ’94, Anaheim, Ca., pp. 164-166, May 1994.

[15]

Canadian/American EMTP User Group, Alternative Transients Program (ATP) Rule Book, Co-Chairmen W. S. Meyer and T. Liu, Portland, Oregon, 1995 ([email protected]).

[16]

Gunther, E., Grebe, T., Rambabu, A., and Mader, D., “Running EMTP on PCs”, IEEE Computer Applications in Power, pp. 33-38, Jan. 1993.

[17]

Hibbert, M., EMTP Application, South East Queensland Electricity Board.

[18]

Grady, W.M., A Guide for Performing Selected Power Quality Studies in Distribution Systems Using ATPDraw, ATP, TPPLOT, and ATPCON in Conjunction with ABB Feederdesign Cadpad Data [Online]. Available from http://www.ece.utexas.edu/~grady/ [Accessed 23 April 1998].

[19]

Høidalen, H. K., ATPDRAW Version 3.0 User Manual, Norwegian Electric Power Research Institute, Trondheim, Norway, January 25, 1996 (/www.ee.mtu.edu/atp/).

[20]

MODELS in ATP Rule Book, August 1995.

[21]

IEEE Standard 446-1987, IEEE Recommended Practice for Emergency and Standby Power Systems for Industrial and Commercial Applications (IEEE Orange Book).

[22]

Meiklejohn, A.G., Monitoring of Distribution System Power Quality, Undergraduate Thesis, Department of Computer Science and Electrical Enginering, University of Queensland, October, 1998.

[23]

Bernard, S., Papoz, S., McGranaghan, M. and Tang, L., Active Filter Design and Specification For Control of Harmonics in Industrial and Commercial Facilities [Online]. Available from http://www.powerquality.com/ [Accessed September 3, 1998].

93

Thesis - Simulating Power Quality Problems by Atp-emtp

Recommend Documents