reducing harmonic voltage at industrial area distribution network using network configuration...

8/21/2019 Reducing Harmonic Voltage at Industrial Area Distribution Network Using Network Configuration Management

1/105

REDUCING HARMONIC VOLTAGE ATINDUSTRIAL AREA DISTRIBUTION NETWORK

USING NETWORK CONFIGURATIONMANAGEMENT

MOHD SHAHED BIN LATIF

UNIVERSITI SAINS MALAYSIA

2008


2/105

REDUCING HARMONIC VOLTAGE AT INDUSTRIAL AREADISTRIBUTION NETWORK USING NETWORK CONFIGURATION

MANAGEMENT

by

MOHD SHAHED BIN LATIF

Thesis sb!i""e# i$ %&%i&&!e$" '% "he (e)i(e!e$"s%'( "he #e*(ee '%MS+, -E&e+"(i+.& / E&e+"('$i+ E$*i$ee(i$*

M.(+h 2008


3/105

ACKNOWLEDGEMENTS

This research could not been completed and this thesis cannot be written

without the scholarship and resources provided by Tenaga Nasional Berhad.

Thanks to my supervisor, Dr. Ir. Syafruddin asri, for the guidance and

encouragement during my study process. !lso thanks to my colleagues at

"elugor #ower Station, #enang who always support and encourage me and,

the staff at $egional %ontrol %entre, Bayan &epas who provided me all the

information re'uired for my research. !nd finally, thanks to my family, especially

my departed wife who offered moral support and endured this long process with

me.

ii


4/105

TABLE OF CONTENTS

1AGE

ACKNOWLEDGEMENTS iiTABLE OF CONTENTS iiiLIST OF TABLES viLIST OF FIGURES viiiLIST OF ABBREVIATION (ABSTRAK (iABSTRACT (ii

CHA1TER ONE INTRODUCTION

).) *verview on +armonic )). Standards on +armonic -).- +armonic itigation ). Time/0arying +armonic 1).1 Industrial !rea 2).2 3actors %ontributing to +armonic 3luctuation 4).4 5valuating +armonic %haracteristic 6).6 *b7ective and Scope of $esearch 6).8 ethodology 8).)9 %ontribution of This Study )9).)) *verview of Thesis ))

CHA1TER TWO LITERATURE SURVEY

.) Background )

. Basic on +armonics )

.- +armonic %haracteristic of Industrial !rea )2

. +armonic Standards )8.1 Time 0arying +armonic

.2 +armonic itigation and 5conomic %onsideration .4 Identifying +armonic Source 2

CHA1TER THREE SIMULATION AND ANALYSIS

-.) 5ffect of %onsumer &oad 3luctuation Si:e -9-. 5ffect of %onsumer &ocation -)-.- 5ffect of Different Network %onfiguration ---. 5ffect of Network Total &oad ---.1 0oltage Total +armonic Distortion %alculation --.2 Baseline for %omparison -2

-.4 5valuating #robabilistic !spect of +armonic 0oltage -6-.6 Simulation on 5ffect of %onsumer &oad 3luctuation Si:e 9

iii


5/105

-.8 Simulation on 5ffect of %onsumer &ocation in NetworkBranch

)

-.)9 Simulation on 5ffect of Different Network %onfiguration -.)) Simulation on 5ffect of !dding New &oad

CHA1TER FOUR TEST NETWORK3 MODELING AND

1ARAMETERS.) Industrial !rea Distribution Network -. %omponent $ated 0alues and Impedance odeling 1 ..) Transmission System 1 .. Transformer 4 ..- %ables 6

.. %onsumer &oads 19 ..1 +armonic Source 1).- #robability of Network &oading 1. Simulation Software 1-

CHA1TER FIVE SIMULATION RESULTS AND DISCUSSION

1.) $ated 0oltage Total +armonic Distortion 161. Simulation I $esults !nd !nalysis 181.- Simulation II $esults !nd !nalysis 21. !nalysis of Distance of Disturbance on T+Dv0ariation 2-1.1 $esults and !nalysis for %onfiguration B and % 21

1.2 !nalysis for Different Branch &oading 281.4 $esult of !dding New &inear &oad 491.6 Discussions 4)

CHA1TER SI4 CONCLUSIONS AND RECOMMENDATION

2.) %onclusions 412. $ecommendation for 3uture Study 44

REFERENCES 46

A11ENDICES!ppendi( ! / Table of $andom &oad &evel!ppendi( B / $esults for 5ffect of &oad 0ariability in %onfiguration !!ppendi( % / $esults for 5ffect of &oad 0ariability in %onfiguration !

at ;- %urrent +armonic

!ppendi( D / $esults for 5ffect of &oad 0ariability in %onfiguration !at );- %urrent +armonic

!ppendi( 5 / &oad 0ariability $esults for %onfigurations !, B and %

!ppendi( 3 / Difference in Network Branch &oad and Difference InT+DvBetween %onfiguration B and %

iv


6/105

v


7/105

LIST OF TABLES

1AGE

.) +armonic #hase Se'uence )1

. Basis for harmonic current limits based on I555 1)8/)88

9

.- %urrent distortion limit for general distribution systems !verage T+Dvfor $ange of Network&oad Demand

29

1. %onfiguration ! / #robability and %umulative #robability of$anged T+Dv

29

1.- 0ariation of T+Dv$esult for Total Tripping *f 5ach%onsumer &oad

2

1. T+Dv0ariability $esult for Total Tripping of 5ach

%onsumer Based on %onsumer Distance to #%%

2

1.1 %onfiguration B / !verage T+Dvfor $ange of Network&oad Demand

22

1.2 %onfiguration B / #robability and %umulative #robabilityof $anged T+Dv

24

1.4 %onfiguration % / !verage T+Dvfor $ange of Network&oad Demand

24

vi


8/105

1.6 %onfiguration % / #robability and %umulative #robabilityof $anged T+Dv

24

1.8 T+Dvat #%% as a $esult of !dding New &oad 49

vii


9/105

LIST OF FIGURES

1AGE

).)ethodology flow chart

)9

.) +armonic %urrent and 0oltage Distortion )-

. ! --?0 Industrial !rea Distribution Network )4

.- Balanced harmonic characteristic at industrial areanetwork

)6

. inimal levels of triplen and even current harmonic )6

.1 Typical distribution network of an industrial area )8

.2 +armonic voltage fluctuation at an industrial areaincoming feeder

-.) 3actors affecting harmonic voltage fluctuation and factorswithin utility@s control

8

-. 5ffect of consumer distance from #%% -

-.- #rocess flowcharts for calculating total harmonic voltage

distortion


10/105

1.- %onfiguration ! T+Dvpdf and cpf 2)

1. Scatter plot for different level of current harmonic 2

1.1 %orrelation between load fluctuation si:e and T+Dv

variability

2-

1.2 %orrelation between consumer load distance to #%% andT+Dvvariability range at #%% due to total tripping of eachload

2

1.4 +armonic voltage level at each harmonic for configurationB and % using the same random load level data,simulation and calculation

21

1.6 Scatter plot of T+Dvfor the three different configuration at

random load level

22

1.8 %onfiguration B T+Dvpdf and cpf 26

1.)9 %onfiguration % T+Dvpdf and cpf 26

1.)) %orrelation between difference in branches total load anddifference in configuration B and % T+Dv

28

ix


11/105

LIST OF ABBREVIATION

!SD !d7ustable speed drives

B? Breaker

%pf %umulative probability function

%I"$5 International %ongress of &arge #ower Systems

I5% International 5lectrotechnical %ommission

I555 Institute of 5lectrical and 5lectronics 5ngineers

I555 #5S I555 #ower 5ngineering Society

IS% Short %ircuit %urrent

I& &oad %urrent

% &arge #ower %onsumer

S icrosoft

0! ega 0olt !mpere

N*# Normally open position

#df #robability density function

#%% #oint of %ommon %oupling

S%% Short %ircuit %urrent

S%$ Short %ircuit $atio

S+I Shunt +armonic Impedance

T+D Total +armonic Distortion

T+Dv 0oltage Total +armonic Distortion

x


12/105

MENGURANGKAN VOLTAN HARMONIK DI RANGKAIAN 1EMBAHAGIANKAWASAN INDUSTRI MENGGUNAKAN 1ENGURUSAN KONFIGURASI

RANGKAIAN

ABSTRAK

Syarikat pembekal elektrik diperlukan untuk mengekalkan tahap voltan

harmonik di dalam sistem di bawah batas piawaian. Namun, voltan harmonik

berubah mengikut masa dan disebabkan oleh naik turun tahap arus harmonik

dan perubahan impedans rangkaian. engurangkan harmonik menggunakan

kaedah sedia ada adalah mahal untuk pembekal tenaga dan memerlukan

pertimbangan ekonomi. #emerhatian dan analisa ke atas rangkaian

pembahagian kawasan industri menun7ukkan perubahan pada impedans

rangkaian disebabkan oleh perubahan beban pelanggan dan perubahan

konfigurasi rangkaian boleh menyebabkan perubahan ketara terhadap kadar

voltan Atotal harmonic distortion@


13/105

REDUCING HARMONIC VOLTAGE AT INDUSTRIAL AREA DISTRIBUTIONNETWORK USING NETWORK CONFIGURATION MANAGEMENT

ABSTRACT

5lectric utility company is re'uired to maintain harmonic voltage level in the

system below the standard@s limit. +owever, harmonic voltage is time variant

and is caused by fluctuation of current harmonic level and changes in network

impedance. itigating harmonic using e(isting methods is costly for utility and

re'uires economic consideration. *bservation and analysis on an industrial

area distribution network shows that network impedance fluctuation caused by

consumer loads variability and changing network configuration can significantly

change voltage total harmonic distortion


14/105

CHA1TER ONE

INTRODUCTION

Demand for 'uality power supply is becoming a ma7or issue for

consumer, especially large power consumer


15/105

5ffect of high level of voltage or current harmonics can cause transformer

heating, nuisance tripping of fuse, circuit breaker and protective devices, high

current in neutral conductor and distorted voltage waveform. %apacitors are

sensitive to harmonic voltage while transformers are sensitive to current

harmonics. There are many researches which study the effect of harmonics

which affects both utility and consumers. "reater concerns have been

e(pressed by industries which have e'uipment or processes that are sensitive

to distortion on the supply voltage which affect their plant operation and

productivity.

$esonance is another problem related to harmonics. It occurs when

harmonic current produced by non/linear load interacts with system impedance

to produce high harmonic voltage. Two types of resonance can occur in the

system, either series resonance or parallel resonance, depending on the

structure of the network. This problem is most common in industrial plant due to

the interaction of series of power factor correction capacitors and transformer@s

inductance.

!ll triplen harmonics


16/105

5,2 S".$#.(#s '$ H.(!'$i+

Institute of 5lectrical and 5lectronics 5ngineers


17/105

be e(ceeded for short duration. I5% has provided a set of time/varying limits

based on percentile over a period of time i.e. 81thand 88thfor very short time


18/105

method is considered to be less e(pensive compared to active filter. The

method uses power electronic to emulate resistive behavior for harmonic.

+owever, the method is still under further study. %urrently, all harmonic

mitigation techni'ues involve e'uipment re'uired to be installed on the system.

There is yet a study on using other factors which can affects harmonic voltage

distortion such as network impedance. *ptimi:ing network impedance to

mitigate harmonic can be cost effective for utility to apply. Because of mitigating

harmonic is e(pensive, many utility company have resorted in imposing penalty

to consumer for in7ecting current harmonic above the standard steady state limit

into the system. This process re'uires method on determining harmonic

contribution by the consumers


19/105

within distribution system, especially factor that within its influence where they

can be controlled or managed. The factors which can contribute to harmonic

voltage fluctuation will be discussed in detail in section ).2.

5,; I$#s"(i.& A(e.

Setting up of an industrial area or industrial :one has become a common

practice in many countries where all industrial plant is located within a certain

geographical area. There are many reasons for the set up such as economic

consideration, safety issues and environmental concern. The development of

industrial area has also caused a uni'ue electrical distribution system with

uni'ue electrical characteristic, power 'uality and system stability re'uirements.

Due to the strict re'uirements from consumer to utility, consumers are provided

with redundant incoming feeders and the distribution network is supplied by

several sources from transmission system. The network is also operated by

e(tensive network control system to provide stable and reliable supply to

consumers.

tility monitors power supply 'uality of an industrial area at the

incoming feeder after the step down transformer from transmission system. 3or

harmonic monitoring, this point is the point of common coupling. The reason for

choosing the point is to ensure harmonic pollution from the industrial area is not

being transmitted into transmission system and vice versa, and to ensure

harmonic pollution from one branch does not affect another branches

connected on the feeder. +armonic level on the feeder is the best indication of

harmonic 'uality in the network.

6


20/105

5,< F.+"'(s C'$"(ib"i$* "' H.(!'$i+ F&+"."i'$

!nalysis into factors contributing to harmonic voltage fluctuation at

industrial area shows that changes in non/linear loads, network configuration

and number of linear loads within the network are the main factors. +owever,

utility has no control over the number and operational of non/linear load within

industrial plant which caused changes in production of current harmonic. The

only factors within utility@s control are configuration of the network and number

of consumer plants in the network. These two factors affect the network

impedance. &ooking in detail into network components, network total

impedance comprises of transmission system impedance, step down

transformer impedance, cable impedance and consumer@s plant network

impedance.

Transmission system network impedance looking from the low voltage

side of a step down transformer varies slightly over time because of the

impedance of a step down transformer dominates and does not vary much.

%able@s impedance is also constant and can be assume steady. +owever,

number of consumer plant in the network and their load demand changes over

time depending on plant operation and unforeseen tripping. *verall network

configuration can also change due to switching process. These two factors,

consumer load variability and network configuration changes, are the main

factors which utility can use to mitigate harmonic voltage.

7


21/105

5,= E6.&."i$* H.(!'$i+ Ch.(.+"e(is"i+

In order to determine the effect of the above factors on harmonic voltage,

network harmonic characteristic is important as a baseline for comparison. The

characteristic must be able to indicate the effect of time varying nature of

harmonic. Since ma7or contribution of harmonic voltage is the fluctuation of load

impedance under normal operation, development of harmonic characteristic of a

network due to load variability is crucial. There is currently no specific method

been developed to determine or predicting harmonic characteristic of a certain

network, other than fre'uency scan for resonance analysis which only

applicable for steady state analysis. 3or this study, since utility is able to

determine the statistical loading pattern of a network, the probability of loading

can be used to develop and estimate the probabilistic aspect of harmonic.

5,8 Ob>e+"i6es .$# S+'?e '% Rese.(+h

The ob7ectives of this study were to determine methods for utility to

reduce harmonic voltage in meeting standard@s steady state limit of 1 voltage

T+D and time varying limit of 81 thpercentile voltage T+D within steady state

limit at #%%. The second ob7ective is to determine methods of reducing

harmonic voltage with little or no cost. The study focused on distribution network

for industrial area which has the capability of switching into other configuration

since the network normally has different possible sources, backup and

redundant feeders to ensure reliability of the supply system. !ction plan for this

study were as followsG

). To determine whether varying consumer load increases harmonic

voltage.

8


22/105

. To determine amount of changes in harmonic voltage due to si:e

of varying consumer load.

-. To determine amount of change in harmonic voltage due to

location of varying consumer load.

. To determine changes in harmonic voltage due to switching

network configuration.

1. To determine changes in harmonic voltage due to adding

consumer load into e(isting network.

5,@ Me"h'#'&'*y

In order to achieve the ob7ectives, the following protocol had been set up.

Select and gather data on industrial area distribution

network configuration and components

Decide method on modeling of e'uipment for harmonic

analysis and method of simulation

odel the selected industrial area distribution network

Simulate identified factors affecting harmonic voltage

!naly:e data using statistical techni'ue and compare with

calculation based on design values

%onclude the research, suggest and recommend mitigating

action

Base on protocol and action plan a flow diagram of research

methodology was drawn and shown in 3igure ).).

9


23/105

Fi*(e 5,5 Me"h'#'&'*y %&'7 +h.("

5,50 C'$"(ib"i'$ '% This S"#y

The outcome of this study is important to utility in controlling harmonic

voltage and improving power 'uality without huge investment in mitigating

e'uipment. %omponents which are affected by harmonic voltage will have

longer life and cost of maintenance is reduced. %onsumers will also benefit from

the method since utility is able to provide better power 'uality. System design

engineers can use the method in planning of electrical system and control

engineers will be able to use the method in controlling harmonic voltage.

10


24/105

5,55 O6e(6ie7 '% Thesis

This thesis discusses and analy:es harmonic voltage distortion at a utility

distribution network supplying to industries due to changes in consumer load

and network configuration. The analysis determines the condition which can

reduce total harmonic voltage distortion T+Dv at point of common coupling.

$ecommendation to reduce harmonic voltage distortion was proposed which

can be integrated into the network control system.

The content in %hapter provides reader with the applicable standards

for harmonic, harmonic mitigation, probabilistic aspects of harmonic, economic

consideration and effect of network impedance on harmonic. $eviews from past

studies by researchers related to those issues were presented.

%hapter - discusses the method of simulation and the process flow of the

simulation. 5ach factors contributing to the changes to harmonic voltage at #%%

were taken into consideration for simulation. ethod of calculations and

analysis were also presented in this chapter.

%hapter contains information on test distribution network system

together with component data and test values that were used for analysis.

ethods for modeling and calculation of each component in the network were

described in details.

%hapter 1 e(hibits the simulation results and analysis together with

discussion of the overall situation. ! conclusion of the thesis was presented in

%hapter 2 which includes recommendation for future studies.

11


25/105

CHA1TER TWO

LITERATURE SURVEY

2,5B.+*('$#

The studies re'uired broad knowledge of the issues regarding harmonic

in power system, the standard limit and re'uirements, modeling and simulation,

issues related to utility and consumers especially at an industrial area, and

result from studies by other researchers. !ll this information is necessary to

address the changes and dynamic of harmonic voltage at an industrial area.

The following sections include brief knowledge of harmonics and reviews

on papers related to relevant harmonic standards and re'uirements, mitigation,

probabilistic aspects, cost of mitigation and effect of harmonic impedance

variability. The review focus on studies related to harmonic in power system with

regards to relation between utility and consumers. The reviews also pointed out

the differences and similarities between previous studies and this research.

2,2 B.si+ '$ H.(!'$i+s

I555 #5S Hinter eeting )886 provides basic harmonic theory which

according to 3ourier theorem, periodic non/sinusoidal or comple( voltage


26/105

++= )]cos([)( 0 nn qtnaatf


27/105

1

2

2

V

V

THD n

n

V

==


28/105

%urrent magnitude of all phases for all harmonic fre'uencies is e'ual for

a balanced system. &ooking at e'uations


29/105

and balanced it can be assumed that the characteristic of industrial area

distribution network is also balanced. This includes the non/linear loads. 3igure

. shows an actual e(ample of an industrial network at #enang Island.

The proportion of three phase loads to single phase loads is large.

Therefore, current harmonic produced within industrial plant and subse'uently

penetrated into utility distribution network is considered balanced. Investigation

on harmonic characteristic on a real industrial area shows that harmonic voltage

is practically balanced as shown in 3igure .-. Since step down delta/wye

grounded transformer at the entrance of industrial plant can block triplen

harmonic, triplen current harmonic is almost non/e(istence in utility distribution

network. 3urther observation shows that, even harmonic order is minimal as

shown in 3igure .. It is can be assumed that harmonic characteristic of

industrial area distribution network is similar to the characteristic of industrial

plant.

16


30/105

Fi*(e 2,2 A KV I$#s"(i.& A(e. Dis"(ib"i'$ Ne"7'(,

17


31/105

Fi*(e 2, B.&.$+e# h.(!'$i+ +h.(.+"e(is"i+ ." i$#s"(i.& .(e. $e"7'(

Fi*(e 2,9 Mi$i!.& &e6e&s '% "(i?&e$ .$# e6e$ +((e$" h.(!'$i+

3igure .1 shows a simple industrial area distribution network with

current harmonic flows from all branches and transmission system into #%%. !s

an e(ample, current harmonic from transmission system IhT flowing into

distribution network incoming feeder and combined with network impedance

creates harmonic voltage distortion at #%%. If the harmonic voltage at #%% is

18


32/105

lower than limit, harmonic voltage down the line at consumer@s feeder should be

lower. Similar situation happen with current harmonic from other branches.

2,9 H.(!'$i+ S".$#.(#s

I555 has come out with a guidelines and standard regarding harmonics

in the I555 standard 1)8/)88 KI555 $ecommended #ractices and

$e'uirements for +armonic %ontrol in 5lectrical #ower SystemsL. The Standard

is a guide in designing of power systems with nonlinear loads. The limits set are

for steady/state operation and recommended for worst case scenario. The

'uality of power is observed at point of common coupling


33/105

The standard generally provide information and guidelines on sources of

harmonics, resonant condition due to harmonics, fre'uency response and

modeling for transmission and distribution system, effect of harmonic, balanced

and unbalanced system, measurements and steady state limits. The voltage

distortion limits are used as a system design values set for worst case scenario

in a normal condition. +owever, the worst case scenario is normally referred to

ma(imum current harmonic penetration. 3luctuation of harmonic impedance in

the system can also cause an increase in harmonic voltage. This study looks at

varying factors of harmonic impedance within a distribution network and

compare with harmonic voltage distortion limit at point of common coupling

using design components values and ma(imum current harmonic penetration

from a single source. Table ., .- and . are the harmonic current and

voltage limits from I555 1)8/)88 standard.

T.b&e 2,2B.sis %'( h.(!'$i+ +((e$" &i!i"s b.se# '$ IEEE ;5@:5@@2

S%$ at #%% a(imum individual 3re'uency+armonic voltage -.9

9 .9 > .1

19 ).9 > ).1

)99 9.1 > ).9

)999 9.91 > 9.)9

T.b&e 2,C((e$" #is"'("i'$ &i!i" %'( *e$e(.& #is"(ib"i'$ sys"e!s -520V "h('*h


34/105

T.b&e 2,9V'&".*e Dis"'("i'$ Li!i"s

Bus 0oltage at #%% Individual 0oltage Distortion


35/105

2,; Ti!e V.(yi$* H.(!'$i+

Since harmonic is time variant, it is important to understand the factors

affecting the change. itigating harmonic to meet with standard steady state

limit is essential to ensure system stability. +owever, since harmonic is time

variant, it is more practical to use time varying limit as an inde( to evaluate the

state of harmonics in a system. 3igure .2 shows an actual voltage total

harmonic distortion


36/105

statistical characteristics of harmonics resulting from multiple harmonic current

in7ection in a network. This research focuses on the effect of changes on load

condition and system configuration on network impedance which caused

changes in harmonic voltage at #%%. !s for fluctuation of current harmonic, a

ma(imum value was used to determine worst case scenario. To reduce

comple(ity of simulation and calculation, a single harmonic source is used since

the purpose of this research is to determine the characteristic of load and

system configuration which can reduce harmonic voltage at #%%.

!naly:ing time/varying harmonic re'uires accurate modeling of power

system. The papers by %arbone et. al.


37/105

%omparing this research and the paper above, both agrees on

accounting system impedance and current variability and their correlation to

obtain accurate harmonic voltage distortion. This research is similar in the

sense of taking into account impedance variability but concentrate on different

type network. The paper analy:e on industrial plant distribution network which

help customer in analy:ing harmonic distortion, while this research analy:e on

industrial area distribution network which assist utility to mitigate harmonic

distortion. 3actors needed for consideration in modeling distribution network

compare to industrial plant is the line and cable capacitance due to longer

distance. !nother one is the load modeling. &oad in industrial plant can be

easily identified as resistive, capacitive, inductive or combinations of those.

odeling is easy following I555 recommendation. tility is unable to determine

the configuration within an industrial plant, hence, aggregate load model

recommended by I555 is re'uired in analysis.

2,< H.(!'$i+ Mi"i*."i'$ .$# E+'$'!i+ C'$si#e(."i'$

+armonic mitigation has been a sub7ect of many researches. The most

common mitigation techni'ue is using filters, either passive or active. The paper

by I:har et. al.


38/105

generated from customer side. tility re'uires an incentive to balance the cost

of mitigating harmonic.

There is little study to assist utility in mitigating harmonic which takes into

account the cost benefits analysis. There is a paper which studied a less

e(pensive method by $yckaert et. al


39/105

2,= I#e$"i%yi$* H.(!'$i+ S'(+e

The paper K! %ritical Impedance / Based ethod for Identifying

+armonic SourcesL proposed a method to determine contribution of utility and

consumer on harmonic voltage and current level at point of common coupling


40/105

)cos(si"uc

u

u EE

X

EIEQ ==


41/105

The general case of harmonic source detection is as followsG

%alculate utility voltage 5upcc/IpccRuand %I


42/105

CHA1TER THREE

SIMULATION AND ANALYSIS

Based on the factors within utility@s control which can affect harmonic

voltage


43/105

,5 E%%e+" '% C'$s!e( L'.# F&+"."i'$ Sie

tility company has certain control over network configuration of a

distribution system and information on loading pattern of consumer loads over

certain time. This gives utility certain ability to predict and manage harmonic

voltage at predetermine #%%. +owever, utility has no knowledge on the e(act

electrical configuration of consumer plant. 3or the purpose of harmonic analysis

and modeling, an aggregate load model was used to determine harmonic

impedance of a consumer load which has been recommended by I555 #ower

5ngineering Society


44/105

harmonic impedance. %ontribution of each consumer to total network

impedance with reference to #%% depends on the si:e of the consumer rated

impedance. +ence, changes in consumer impedance affect the total network

impedance variability and range of consumer impedance variability depends on

the si:e of consumer rated impedance. The greater the range of consumer

impedance, the more influence it impose on network impedance. Since

consumer harmonic impedance depends on load demand and load demand can

fluctuates from :ero


45/105

Lc

L

PCC

ZZ

ZZ

+

=


46/105

load further from #%% caused lesser impact on total network impedance

fluctuation range due to the load@s variability.

, E%%e+" '% Di%%e(e$" Ne"7'( C'$%i*(."i'$

Network impedance looking from #%% depends on impedance of cables

and consumer loads. %onfiguration of cables and consumers determine the

total impedance. Different configuration of cables and load produces different

impedance. The ob7ective is to minimi:e the impedance to reduce harmonic

voltage. Since from section -. shows that consumer further from #%% has less

impact, theoretically, configuration with longer branch network has less impact

on harmonic voltage variability at #%% compared to network with shorter

branches for the same amount of consumer load. This is as a result of in a

longer network branch many consumers are located far from #%%.

,9 E%%e+" '% Ne"7'( T'".& L'.#

Distribution network harmonic impedance is comprises of harmonic

impedance of transmission system, step down transformer, cables and

consumer load. %hanges in the number or configuration of these components

cause changes in the network impedance. 3rom e'uation


47/105

of an aggregate load is inversely proportion to the total load demand of the

consumer plant.

,; V'&".*e T'".& H.(!'$i+ Dis"'("i'$ C.&+&."i'$

0oltage at each harmonic level was calculated using total network

impedance and harmonic current at each harmonic level. #rocess flowchart for

the calculation is shown in 3igure -.-. The voltage T+D was then calculated

using e'uation


48/105

Fi*(e , 1('+ess %&'7+h.("s %'( +.&+&."i$* "'".&h.(!'$i+ 6'&".*e #is"'("i'$ -THD6 ." 1CC

The total network impedance for network in 3igure -. was calculated using

e'uation


49/105

2branchZ is branch number total impedance

cnZ is cable impedance at load n

LnZ is load n impedance

;; is the symbol for parallel circuit calculation where

BA

BA11

1$$

+=

,< B.se&i$e %'( C'!?.(is'$

The normal operating network in 3igure -. is referred to as configuration

!, was the referenced original network. *ther possible configurations for the

network are shown in 3igure -.1, referred to as configuration B, and 3igure -.2,

referred to as configuration %. ! calculation of T+Dvat #%% for rated or design

values was done first to determine steady state distortion reference level for

comparison for configuration !. %alculations were done using ma(imum current

harmonic spectrum


50/105

Fi*(e ,9 A KV Tes" #is"(ib"i'$ $e"7'( -C'$%i*(."i'$ A

Fi*(e ,; Ne"7'( C'$%i*(."i'$ B, B(e.e( BK2 is i$'?e$ ?'si"i'$ 7hi&e BK5 .$# BK .(e+&'se#

37

Q;$6.9S%%21!


51/105

Fi*(e ,< Ne"7'( C'$%i*(."i'$ C, B(e.e( BK5 is i$'?e$ ?'si"i'$ 7hi&e BK2 .$# BK .(e +&'se#

,= E6.&."i$* 1('b.bi&is"i+ As?e+" '% H.(!'$i+ 6'&".*e

The purpose of this simulation was to establish harmonic voltage

characteristic as a result of combination of several varying loads in the network.

This simulation was done on all configurations !, B and %. 3irstly, each

consumer load was set to change randomly in seven load levels between )91

and 9 of the rated value as shown in Table -.). The 9 load level simulates

full outage of the plant while )91 load level demonstrates additional 1 load

increase from design value. sing random function


52/105

T.b&e ,5L'.# V.(i.bi&i"y Le6e&

0a

riability

0!) )

91 6

1- 2

2 1

11 -

-

2 )1

4 9

To determine time varying aspect of harmonic, a probability density

function


53/105

range depends on probability of average T+Dvwhich falls within the load range

as followsG

= )()( a#era$e#ba# THDPTHDP


54/105

obtain a sample large enough for analysis, the simulation was performed on all

configurations !, B and %. ! correlation between both was re'uired to verify the

effect. %orrelation analysis was performed using analysis tool available in the

spreadsheet software


55/105

network with several short branches, configuration !, produce different T+D v

compared to a network with longer branch


56/105

simulation were tabulated and analy:e to determine the changes in T+Dv

comparing with design values without the additional load.

43


57/105

CHA1TER FOUR

TEST NETWORK3 MODELING AND 1ARAMETERS

This chapter describes the test distribution network used in the

simulation and e'uations for modeling of all components involved in the

network. !lso included in this chapter are component@s parameters and

assumptions used in analysis. Simulation software was described together with

conversion of component@s model e'uations into the software code.

9,5 I$#s"(i.& A(e. Dis"(ib"i'$ Ne"7'(

tility distribution network for industrial area is complicated and has

many types of configurations. In order to study the behavior of harmonic voltage

due to changes in load and network configuration, a simplify network was used

based on actual configuration including component design values. +armonic

source in the system is simplified to a single source to eliminate calculation

comple(ity. &oads in the network fluctuate in time with constant power factor.

The assumption was based on each consumer plant has series of power factor

correction capacitors that maintain the plant@s power factor. *ther components

harmonic impedances in the network such as transmission system, step down

transformer and cables vary slightly in time and are considered constant

throughout the simulation. The test network is shown in 3igure .).

44


58/105

The network is supplied from a )-?0 transmission system through a

step down )-?0;--?0 transformer. It has eleven consumers where one

consumer is connected directly at the incoming feeder and other ten consumers

are connected on two branches. There is one breaker at each branch, B?) for

branch ) and B? for branch . ! connecting line and breaker, B?-, connecting

the end of the two branches functioned as a backup line in the case of

malfunction or tripping of the other two breakers. The breaker is normally open

position


59/105

9,2 C'!?'$e$" R."e# V.&es .$# I!?e#.$+e M'#e&i$*

The component rated values are from actual system of an industrial area.

Impedance modeling was based on papers produced by I555 #ower

5ngineering Society


60/105

Q;$ ratio is the ratio of reactance to resistance in the system. The summary of

transmission system data based on average value given by utility is in Table

..

T.b&e 9,2T(.$s!issi'$ Sys"e! 1.(.!e"e(

3undamental 0oltage )-?0

Q;$ ratio 6.9

Three phase short circuit current 21 !

The impedance of the transmission system was modeled using series

resistance and reactance and calculated in e'uations


61/105

Therefore, the harmonic per unit impedance of transmission system at each

harmonic is given byG

'uhjhZ! )1626#0(0203#0)( += Vh 2n ), n),,-,EW


62/105

!ll lines connected to consumers are underground cables and lines.

#arameters for the cables are taken from manufacturer datasheet. #arameters

re'uired were length, resistance, reactance and capacitance of the cables.

Details parameters for the cables are listed in Table .-.

T.b&e 9,C.b&es D.".

%ables &ength


63/105

.

&jhXRhZ ooa&no )()(mi" +=


64/105

!ll consumers load in the network branch were assumed to be linear and

have rated power


65/105

P is the active power

Therefore, as shown in e'uation


66/105

+armonic *rder #er unit $elative angle


67/105

T.b&e 9,

reducing harmonic voltage at industrial area distribution network using network configuration...

Documents