technical collection cahier technique no. 183

Cahier techniqueno. 183Active harmonic conditionersand unity power factor rectifiers

E. BettegaJ-N. Fiorina

Technical collection

"Cahiers Techniques" is a collection of documents intended for engineersand technicians, people in the industry who are looking for more in-depthinformation in order to complement that given in product catalogues.

Furthermore, these "Cahiers Techniques" are often considered as helpful"tools" for training courses.They provide knowledge on new technical and technological developmentsin the electrotechnical field and electronics. They also provide betterunderstanding of various phenomena observed in electrical installations,systems and equipments.Each "Cahier Technique" provides an in-depth study of a precise subject inthe fields of electrical networks, protection devices, monitoring and controland industrial automation systems.

The latest publications can be downloaded from the Schneider Electricinternet web site.Code: http://www.schneider-electric.comSection: Experts' place

Please contact your Schneider Electric representative if you want either a"Cahier Technique" or the list of available titles.

The "Cahiers Techniques" collection is part of the Schneider Electric’s"Collection technique".

ForewordThe author disclaims all responsibility subsequent to incorrect use ofinformation or diagrams reproduced in this document, and cannot be heldresponsible for any errors or oversights, or for the consequences of usinginformation and diagrams contained in this document.

Reproduction of all or part of a "Cahier Technique" is authorised with theprior consent of the Scientific and Technical Division. The statement"Extracted from Schneider Electric "Cahier Technique" no. ....." (pleasespecify) is compulsory.

no. 183Active harmonic conditionersand unity power factor rectifiers

ECT 183 first issued, June 1999

Eric BETTEGA

After joining Merlin Gerin in 1983 as a laboratory technician in ABT’selectronics engineering and design service, he became part of theScientific and Technical Division in 1986. In 1991 he obtained hisEngineering degree from the CNAM (Conservatoire National des Artset Métiers), and is currently a senior Corporate Research engineer.

Jean Noël FIORINA

He first joined Merlin Gerin in 1968 as a laboratory technician in theACS (Static Converter Power Supplies) department where heparticipated in the development of static converters. In 1977 heobtained his ENSERG (Ecole Nationale Supérieure d’Electronique etde Radioélectricité de Grenoble) Engineering degree and rejoinedACS. Starting as development engineer, he was soon afterwardsentrusted with projects. He became later responsible for design anddevelopment for MGE UPS Systems. He is in some ways theoriginator of medium and high power inverters.

Cahier Technique Schneider Electric no. 183 / p.2


Active harmonic conditioners and unitypower factor rectifiers

Electric loads are becoming increasingly non-linear in the industrial, tertiaryand even household sectors. These loads absorb non-sinusoidal currentswhich, under the effect of circuit impedance, distort the purely sinusoidalvoltage waveform. This is what is known as harmonic disturbance of powernetworks, currently a cause for concern as it gives rise to serious problems.

We recommend that harmonic non-specialist readers begin by reading theappendix where they will find the basic concepts required to understand thevarious standard and new solutions to limit or combat harmonics. Not onlythe characteristic quantities, but also the non linear equipment, influence ofthe sources and disturbing effects of harmonics need to be known. Last butnot least, standards lay down levels of compatibility, i.e. the maximumpermissible levels.

The purpose of this “Cahier Technique” is to describe the active harmonicconditioners. This attractive, flexible and self-adaptive solution can be usedin a wide variety of cases to complete or replace other solutions. Howeverchapter 1 of this “Cahier Technique” will review other “traditional” solutionswhich should also be taken into consideration.

Contents

1 The traditional solutions 1.1 Reducing harmonic currents of non linear loads p. 41.2 Lowering harmonic impedance of the source p. 4

1.3 Carefully choosing the installation structure p. 4

1.4 Harmonic isolation p. 5

1.5 Using detuning reactors p. 5

1.6 Passive harmonic filters p. 6

2 Unity PF rectifiers and active 2.1 Introduction p. 7

2.2 Unity PF rectifiers p. 8

2.3 The “shunt type” active harmonic conditioner p. 11

3 Hybrid active harmonic conditioners 3.1 The “parallel/series” hybrid structure p. 17

3.2 The “series/parallel” hybrid structure p. 18

3.3 “Parallel” combination of passive filters and activeharmonic conditioner p. 19

3.4 The performances of hybrid structures p. 19

4 Implementing a “shunt type” active 4.1 Objective and context p. 22

4.2 The insertion point of a “shunt-type” active harmonicconditioner p. 23

4.3 Sizing a “shunt-type” active harmonic conditioner p. 24

4.4 Application examples p. 25

5 Conclusion p. 27

Appendix: review of harmonic phenomena Definition and characteristic quantities p. 28

Origin and transmission p. 29

Deforming loads p. 30

Harmful effets of harmonics p. 31

Standard and recommendations p. 32

harmonic conditioner

harmonic conditioners


1 The traditional solutions

Electricians need to be familiar with thesesolutions in order to take the right measureswhen installing polluting equipment or to take allfactors into account when designing newinstallations.

The solutions described hereafter depend on theobjective sought and on the non linear/sensitiveequipment installed.

They use passive components: reactors,capacitors, transformers and/or carefully choosethe installation diagram.

In most cases the aim is to reduce voltage totalharmonic distortion at a load multi-connectionpoint (in a distribution switchboard).

1.1 Reducing harmonic currents of non linear loads

Besides the obvious solution which consists ofchoosing non-disturbing equipment, theharmonic currents of some converters can belimited by inserting a “smoothing” reactorbetween their connection point and their input.This solution is particularly employed withrectifiers with front end capacitors: the reactor

may even be proposed as an option bymanufacturers.

A word of warning however! Although thissolution reduces voltage total harmonic distortionupstream of the reactor, it increases it at theterminals of the non-linear load.

1.2 Lowering harmonic impedance of the source

In concrete terms this consists of connecting thedisturbing equipment directly to the mostpowerful transformer possible, or of choosing agenerator with a low harmonic impedance(see appendix and fig. 1 ).

Note that it is advantageous on the source sideto use several parallel-connected cables ofsmaller cross-section rather than a single cable.If these conductors are far enough apart,apparent source impedance is divided by thenumber of parallel-connected cables.

Fig. 1 : addition of a downstream reactor or reduction inupstream source impedance reduces voltage THD atthe point considered.

1.3 Carefully choosing the installation structure

Sensitive loads should not be parallel-connectedwith non-linear loads (see fig. 2 ).

Very powerful non-linear loads should preferablybe supplied by another MV/LV transformer.

Fig. 2 : a Y-shaped distribution enables decoupling by natural and/or additional impedances.

THD

ZS ZL

Non-linear load

E

Non-linearequipment

Sensitiveequipment

Non-linear equipment supply

“Clean” power network

a) Solution to avoid b) Solution to recommended


The aim is to limit circulation of harmoniccurrents to as small a part as possible of theinstallation using suitable couplingtransformers .

Use of Y-connected primary transformers(without neutral!) with zig-zag secondary is aninteresting solution as it ensures minimumdistortion at the secondary. In this case 3 k orderharmonic currents do not flow at the transformerprimary, and the impedance Zs depends only onthe secondary windings. The inductive part of theimpedance is very low:Uccx ≈ 1%, and resistance is practically halvedcompared with a ∆Y transformer of identicalpower.

Figure 3 and the following calculation show why3 k ω angular frequencies are not present at thetransformer primary (zero sequence current is nil).

Current circulating for example in the primarywinding 1 equals:

NN

i - i2

11 2( )

where

i = = sin 3k t1 1 (3k) Ι Ι ω( )

i = = sin 3k t - 433 3 (3k) Ι Ι ω π

i = sin 3k = i3 1 Ι ωt( )hence

NN

i - i = 02

11 3( )

As regards three-phase loads, some harmonicorders can be removed by using transformers orautotransformers with a number of displacedsecondaries, a solution particularly adopted forpowerful rectifiers. The best known of thesecircuit assemblies is the rectifier consisting oftwo serial or parallel-connected bridges, suppliedby a transformer with two secondaries, one Yand the other delta connected. This assemblyproduces a 30 degree phase displacementbetween the volta-ges of the two secondaries.

1.4 Harmonic isolation

The calculation shows that the 6 k ± 1 orderharmonics where k is odd are removed from thetransformer primary. The first harmonicsremoved, which are also the highest inamplitude, are for k = 1, harmonics 5 and 7. Thefirst harmonics present are then 11 and 13.This property can be generalised by increasingthe number of rectifiers and the number oftransformer secondaries or the number oftransformers by choosing the appropriate phasedisplacement for each secondary.

This solution is commonly employed in the caseof very high power rectifiers where currentdistribution in the various bridges presents noproblems. It is frequently used by electrolyticrectifiers (up to 72 phases!).

Parallel-connected uninterruptible powersupplies (UPS) are of special interest, as theinverters share the output currents and therectifiers supplying them absorb identicalcurrents.

1.5 Using detuning reactors

This solution consists of protecting the capacitors,designed to improve the displacement powerfactor by installing a series reactor.This reactor is calculated so that resonancefrequency matches none of the harmonicspresent.Typical tuning frequencies are for a 50 Hzfundamental: 135 Hz (order 2.7), 190 Hz(order 3.8) and 255 Hz (order 4.5).

Thus for the fundamental, the battery canperform its displacement power factorimprovement function, while the high impedanceof the reactor limits amplitude of the harmoniccurrents.The switched-steps capacitors must allow for thepriority of certain resonance frequencies.

Fig. 3 : zig-zag secondary transformer and attenuationof 3 k order harmonics.

N1

N1

N1

N2

N2

N2

N2

N1 i3 i1(i1- i3)

N2

N2

N2

i2

i3


1.6 Passive harmonic filters

This case differs from the above in that acapacitor is used in series with a reactor inorder to obtain tuning on a harmonic of a givenfrequency. This assembly placed in parallel onthe installation has a very low impedance for itstuning frequency, and acts as a short-circuit forthe harmonic in question.A number of assemblies tuned on differentfrequencies can be used simultaneously in orderto remove several harmonic orders.Passive filters contribute to reactve energycompensation of the installation.

This apparently simple principle neverthelesscalls for thorough study of the installation since,although the filter acts as a short-circuit for therequired frequency, there is a possibility ofresonance risks with other power networkreactors on other frequencies and thus ofincreased previously non-troublesomeharmonic levels prior to its installation(see “Cahier Technique” no. 152).


2 Unity PF rectifiers and active harmonic conditioners

2.1 Introduction

The previous chapter reviewed the techniquesand corresponding passive systems used toreduce harmonic disturbances.These systems all modify impedances,impedance ratios or cause the opposition ofcertain harmonic currents.Other impedance monitoring means areavailable (which we shall not dare to term“intelligent”!), which use static converters of everincreasing effectiveness due to the steadyincrease in semiconductor power componentpossibilities (see table fig. 4 ).IGBT’s made possible the industrial developmentof power converters able to guarantee non-disturbance at the point of common coupling(unity power factor rectification), and harmoniccompensation of power networks (activeharmonic compensation).

c Unity PF rectification is a technique enablingstatic converters to absorb a current very similarto a sinusoidal waveform with, in addition, adisplacement power factor close to the unit: thishighly interesting technique should be used withincreasing frequency.

c Active harmonic compensationAn active harmonic conditioner is a device usingat least one static converter to meet the“harmonic compensation” function.This generic term thus actually covers a widerange of systems, distinguished by:v the number of converters used and theirassociation mode,v their type (voltage source, current source),v the global control modes (current or voltagecompensation),v possible association with passive components(or even passive filters).The only common feature between these activesystems is that they all generate currents orvoltages which oppose the harmonics created bynon-linear loads. The most instinctiveachievement is the one shown in figure 5which is normally known as “shunt” (or “parallel”)topology.It will be studied in detail in paragraph 3.The “serial” type active harmonic conditioner(see fig. 6 ) will be mentioned merely as a

Fig. 5 : “shunt-type” active harmonic conditionergenerates an harmonic current which cancels currentharmonics on the power network side.

Technology V A F (kHz)

Transistor MOS 500 50 50

Bipolar 1,200 600 2

IGBT 1,200 600 10

Thyristor GTO 4,500 2,500 1

Fig. 4 : typical characteristics of use of powersemiconductors in static converters.

Power network

Active harmonic conditioner

Load(s)

Power network

Active harmonicconditioner

Sensitiveload(s)

Fig. 6 : “series type” active harmonic conditionergenerates an harmonic voltage which guarantees asinusoidal voltage on the load terminals.


reminder as it is seldom used. Its function is toenable connection of a sensitive load on adisturbed power network by blocking theupstream harmonic voltage sources. However, inactual practice, this “upstream” harmoniccompensation technique is of little interest since:v the “quality” of the energy at the point ofcommon coupling is satisfactory in the majorityof cases,v insertion of a component in the “serial” mode isnot easy (for example short-circuit currentwithstand),v it is more useful to examine the actual causesof voltage distortion within a power network (theharmonic current sources).

Out of the numerous “hybrid” alternatives we shallconcentrate on the “serial/parallel” type combiningactive and passive filtering (see fig. 7 ) which is avery effective solution for harmonic cancellationclose to high power converters.

However this “Cahier Technique” does not aim tobe comprehensive and deliberately chooses notto treat many topologies. This is because all theother systems are merely variations on a sametheme and because the basic solutions aredescribed in this document.

Before going on to describe unity PF rectifiersand active harmonic conditioners in detail, itshould be noted that there is a certaintechnological resemblance between these twodevices, namely:c when the control strategy of a rectifier bridge(integrating, for example, a BOOST stage)imposes circulation of a current reduced merely toits fundamental, this is called unity PFrectification and the rectifier is said to be “clean ”,c when the current reference applied to thiscontrol is (for example) equal to the harmoniccontent of the current absorbed by a third-partynon linear load, the rectifier cancels all theharmonics at the point of common coupling: this

Fig. 7 : “series/parallel” type hybrid filter.

Fig. 8 : unity PF rectifier and active harmonicconditioner.

Non-linearload(s)

Passivefilter(s)


Power network

a) Unity PF rectifier

b) Active harmonic conditioner

Power network

LoadConverter

Controlprocessing

Power network

Controlprocessing

Non-linearload

Converter

is known as active harmonic conditioner .Thus the same power topology is able to meetthe two separate needs which are non-disturbance and harmonic compensation. Onlythe control strategy differs (see fig. 8 ).

2.2 Unity PF rectifiers

Whether for rectifiers, battery chargers, variablespeed drives for DC motors or frequencyconverters, the device directly connected withthe power network is always a “rectifier”. Thiscomponent, and more generally the input stage(power and control) determines the harmonicbehaviour of the complete system.

Unity PF rectification principle (in single-phase)This consists of forcing the absorbed current to besinusoidal. Unity PF rectifiers normally use thePWM (Pulse Width Modulation) switchingtechnique. Two main categories are identifiedaccording to whether the rectifier acts as a voltagesource (most common case) or a current source.

c Voltage source converterIn this case the converter acts as a back-electromotive force (a “sinusoidal voltage

generator”) on the power network (see fig. 9 ),and the sinusoidal current is obtained byinserting a reactor between the power networkand the voltage source.Voltage is modulated by means of a control loopdesigned to maintain current as close aspossible to the required sinusoidal voltagewaveform.Even if other non-linear loads raise the powernetwork’s voltage total harmonic distortion,regulation can be used to draw a sinusoidalcurrent.The frequency of low residual harmonic currentsis the frequency of modulation and of itsmultiples. Frequency depends on thepossibilities of the semiconductors used(see fig. 4).


e1

i1

iL L D

Tv

Powernetwork

Control loop, iL, Vs

u Vs

i

iL

i1

0

v

0

0

e1

0

t

Vs

0

u

Fig. 10 : single-phase diagram equivalent to a currentPWM rectifier.

Fig. 9 : single-phase diagram equivalent to a voltagePWM converter.

iL

Power network

L

t

+ E

0

- E

iL

B.e.m.f.

B.e.m.f.

c Current source converterThis converter acts as a chopped current“generator”. A fairly large passive filter isrequired to restore a sinusoidal current on themains side (see fig. 10 ).This type of converter is used in specificapplications, for example to supply an extremelywell regulated DC current.

“Voltage converter” implementationprinciple

Its ease of implementation means that thediagram in figure 11 is the one most oftenchosen (as for certain MGE UPS systems).This diagram uses the voltage generatorprinciple.

From the source viewpoint, the convertermust act like a resistance: (sinusoidal) i1and in phase with e1 (DPF = 1). Bycontrolling transistor T, the controllerforces iL to follow a sinusoidal type currentreference with double wave rectification.The shape of i1 is thus necessarilysinusoidal and in phase with e1.Moreover, to keep voltage Vs at itsnominal value at the output, the controlleradjusts the mean value of iL.

Fig. 11 : single-phase diagram equivalent to a current PWM rectifier.

iL

Power network

L

I

t

+ I

0

- I

iL


Order Max. contrib. Typical values Measuredas a% of I1 without unity PF valueas in IEC 1000-3-2 rectifier (Uccx = 1%)

3 14.65% 81% 8.03%

5 7.26% 52% 2.94%

7 4.90% 24% 3.15%

9 2.55% 6% 1.65%

11 2.10% 7% 1.09%

13 1.34% 6% 1.07%

a) Time form

b) Spectral breakdown

Fig. 13 : current upstream of a “clean” single-phase rectifier (2.5 kVA UPS - PULSAR-PSX type).

Fig. 12 : evolution of current iL compared with thereference i.

Figure 13 shows the time curve and theharmonic spectrum of the current drawn by aunity PF rectifier of a 2.5 kVA UPS. In this casethe transistor is a MOS, and the switchingfrequency equals 20 kHz.

Transistor T (normally using MOS technology)and diode D make up the voltage modulator. Thevoltage (u) thus moves from 0 to Vs according towhether transistor T is in the on or off state.When transistor T is conductive, the current inreactor L can only increase as voltage v ispositive and u = 0.The relationship is then:didt

eL

0= ⟩ .

When transistor T is off, the current in L decreases,provided that Vs is greater than v, so that:didt

= e - Vs

L 0⟨ .

For this condition to be fulfilled, voltage Vs mustbe greater than the peak voltage of v, i.e. therms value of the ac voltage, multiplied by r.If this condition is fulfilled, the current in L can beincreased or decreased at any time. The timeevolution of current in L can be forced bymonitoring the respective on and off times oftransistor T.Figure 12 shows the evolution of current iL withrespect to a reference value.The closer are the switching moments of T(i.e. switching frequency is high), the smaller theerrors of iL compared with the reference sinewave will be. In this case the current iL is veryclose to the rectified sinusoidal current, and theline current i1 is necessarily sinusoidal.

t

u

i

t

Reference iiL


The harmonics of the current absorbed arehighly attenuated compared with a switch modepower supply which does not use the “unity PFrectification” control strategy, and their level isbelow standard requirements. Filtering ofu 20 kHz orders is easy and inexpensive.

Three-phase circuitsThe basic circuit arrangement is shown infigure 14 .

We recognise the arrangement in figure 11 withreactors placed upstream of the rectifiers. Theoperating principle is the same. The monitoringsystem controls each power arm, and forces thecurrent absorbed on each phase to follow thesinusoidal reference. There are currently nothree-phase unity PF rectifiers on the market asadditional cost is high. Changes in standardsmay however stipulate their use.



2.3 The “shunt-type” active harmonic conditioner

Operating principle

The «shunt-type» active harmonic conditionerconcept can be illustrated by means of anelectro-acoustic analogy (see fig. 15 ). Theobserver will no longer hear the noise source Sif a secondary noise source S’ generates a

counter-noise. The pressure waves generatedby the loudspeaker have the same amplitudeand are in opposition of phases with those of thesource: this is the destructive interferencephenomenon. This technique is known as ANR(Active Noise Reduction).

Powernetwork

Vs

Primary noisesource S

Secondary source S'

Control microphone

Controller

Error microphone


This analogy is a perfect illustration of the“shunt-type” active harmonic conditioner: the aimis to limit or even remove the current(or voltage) harmonics at the point of commoncoupling by injecting an appropriate current(or voltage) (see fig. 16 ).

Provided that the device is able to inject at anytime a current where each harmonic current hasthe same amplitude as that of the current in theload and is in opposition of phases, thenKirchoff’s law at point A guarantees that thecurrent supplied by the source is purelysinusoidal. The combination of“non linear loads + active harmonic conditioner”forms a linear load (in which current and voltageare linked by a factor k). This kind of device isparticularly suited for harmonic compensation ofLV networks irrespective of the chosen point ofcoupling and of the type of load (the device isself-adaptive).

The following functions are thus performedaccording to the level of insertion:c local harmonics compensation: if the activeharmonic conditioner is associated with a singlenon linear load,c global harmonics compensation: if theconnection is made (for example) in the MLVS(Main Low Voltage Switchboard) of theinstallation.

The “shunt-type” active harmonic conditionerthus forms a current source independent of

power network impedance, and with thefollowing intrinsic characteristics:c its band-width is sufficient to guaranteeremoval of most harmonic components (instatistical terms) from the load current. Wenormally consider the range H2 - H23 to besatisfactory, as the higher the order, the lowerthe harmonic level.c its response time is such that harmoniccompensation is effective not only in steadystate but also in “slow” transient state (a fewdozen ms),c its power enables the set harmoniccompensation objectives to be met. Howeverthis does not necessarily mean total, permanentcompensation of the harmonics generated bythe loads.

Provided that these three objectives aresimultaneously achieved, the “shunt-type” activeharmonic conditioner forms an excellent solutionas it is self-adaptive and there is no risk ofinteraction with power network impedance. Itshould also be noted that the primary aim of thisdevice is not to rephase the fundamental U and Icomponents: insertion of an active harmonicconditioner has no effect on the displacementpower factor.

Nevertheless, if the load treated is of the“multiphase rectifier” kind, then the global powerfactor is indeed considerably improved as thedistortion factor is closer to the unit and the

Fig. 16 : principle of compensation of harmonic components by “shunt-type” active harmonic conditioner.

Source current iF

iFA

Compensator current iH

iH

Load current iF + iH

iF + iH

Non-linearload to be

compensated

Harmonic activeharmonic conditioner

Source


displacement power factor of a rectifier (notcontrolled) is close to the unit. However this ismore a “secondary effect” than an actualobjective!Although the main pupose of the device isharmonic depollution, it can also compensatepower factor. In that case the reactive currentmay be proportionally high and should beaccounted for in the product current rating.

Structure of the “shunt-type” active harmonicconditioner

This device is broken down into the following twosubassemblies (see fig. 17 ).c Power: input filter, reversible inverter, storagecomponents,c Control: reference processing, U/I controls,converter low level control.The main difference between a converter and aunity PF rectifier, described in the previouschapter, lies in the control and monitoring (as thesetpoint is no longer a 50 Hz sine wave). If the“storage” component is a capacitor or battery,the converter has a similar structure to that of the

input stage of the converter with unity PF rectifier(see fig. 18 ).A reactor can also be used (see fig. 19 ).

MGE UPS Systems chose Voltage SourceInverter -VSI- for its SINEWAVE range because

Fig. 17 : schematic diagram showing the structure ofthe “shunt-type” active harmonic conditioner.

Fig. 18 : diagram showing the “shunt type” active harmonic conditioner with VSI (voltage source inverter).

Fig. 19 : diagram showing the “shunt type” active harmonic conditioner with CSI (current source inverter).

Source

References

Vcapa monitoring

Filter i measurement

To load

Load i measurement

Inputfilter

Reversibleinverter

Control andmonitoring

C

L


of its added value in technical and economicterms: wider pass-band, simpler input filter.Moreover the VSI structure technicallyresembles inverter structure.

Control and monitoring electronicsIts main function is to control the powersemiconductors. As such it must:c control capacitor load (c) on energising,c regulate voltage at the terminals of c,c generate “rectifier” on/off patterns when it hasan inverter function so that the active harmoniccondi-tioner permanently supplies a currentcompensa-ting the non linear harmonic currents(see fig. 16).

There are 2 signal processing methods, namely:c the real time method, which is particularlysuitable for loads with ultra-fast variations in theirharmonic spectrum. It can use the “synchronousdetection” method or use Clark transformations;

c the non real time method, used for loads wherethe harmonic content of the current absorbedvaries slightly in 0.1 s. This method uses thefrequency analysis principle and is based on theFast Fourier Transform (FFT). It enables globalor selective treatment of harmonic orders.

Examples of performances obtained usingnon-linear loads

In these examples the loads do not operate onfull load, as the THD (I) is at its lowest on fullload. In the example below, the THD (I) is 30%on full load, whereas it is 80% with a 20% load.c Case of a UPSA “shunt-type” active harmonic conditioner isparallel-connected on a three-phase uninterruptiblepower supply of a power of 120 kVA. The currenttime waveforms are shown in figure 20 . Thespectrum of the current absorbed by the load isgiven in figure 21 and corresponds to an

Fig. 21 : “shunt type” active harmonic conditioner on UPS - source currents spectrum.

Fig. 20 : “shunt-type” active harmonic conditioner associated with a UPS - time waveforms of currents (20% load).

t

I

t

I

03 5

Relativevalue (% of fundamental)

Harmonicorder

7 9 11 13 15 17

10

20

30

40

50

60

70

80

c Source I without active harmonic conditionercc Source I with active harmonic conditioner

a) Load current (THD = 80% Irms = 44 A) b) Source current (THD = 4,6% Irms = 35 A)


harmonic distortion of 80%. Use of the “shunttype” active harmonic conditioner considerablyattenuates the THD (I) which drops from 80% to4.6%. The rms current drops by nearly 20%, andthe power factor increases by 30%(see fig. 21 and 22 ).c Case of a VSD (frequency converter type)An active harmonic conditioner is parallel-

connected to a variable speed drive forasynchronous motor of a power of 37 kWoperating on half-load. The current timewaveforms are shown in figure 23 andcorrespond to an harmonic distortion of 163% forthe load current.Figure 24 shows the harmonic spectrum of thesource and load currents.

Fig. 24: “shunt-type” active harmonic conditioner associated with a variable speed drive - harmonic spectrum of thesource current.

a) Load current (THD = 163%, Irms = 25 A) b) Source current (THD = 22,4%, Irms = 15.2 A)

03 5

Relativevalue (% of fundamental)

Harmonicorder

7 9 11 13 15 17 19

10

20

30

40

c Source I without active harmonic conditionercc Source I with active harmonic conditioner

Fig. 22 : “shunt-type” active harmonic conditioner on UPS: measured values.

Current Without active With activecharacteristics harmonic conditioner harmonic conditioner

Irms (A) 44.1 35.2

Peak factor 1.96 1.52

THD (I) as a % 80.8 4.6

Power factor 0.65 0.86

DPF 0.84 0.86

Harmonic Irms (A) 27.7 1.6

Fig. 23 : “shunt-type” active harmonic conditioner on variable speed drive - waveforms of currents on half-load.

I

t

I

t


Use of the “shunt-type” active harmonicconditioner considerably attenuates the THD (I)which drops to 22.4%. The rms current drops bynearly 40% (see fig. 24 and 25 ).Performance is lower than in the first case (UPS)since line current fluctuations are much faster.In this cases addition of a 0.3 mH smoothingreactor is recommended. The table in figure 26illustrates the resulting increase in effectiveness.We can conclude that the “shunt type” activeharmonic conditioner is an excellent means forremoving harmonics on a feeder or non-linearload. However:

c removal of all disturbances, even if it ispossible, is not necessarily the aim,c it is not suited to voltage power networksexceeding 500 V,c it has no effect on disturbances upstream ofthe current sensor,c technical and economic considerations mayrequire use combined with a passive component;for example a reactor (see fig. 26 ) or a passivefilter to remove the 3rd or 5th harmonic(considerable decrease in “shunt type” activeharmonic conditioner power rating).

Fig. 25 : “shunt-type” active harmonic conditionerassociated with a variable speed drive - currentcharacteristics.

Fig. 26 : “shunt-type” active harmonic conditionerassociated with a variable speed drive with smoothingreactor - current characteristics.

Characteristics Without active With activeof current harmonic harmonicon half-load conditioner conditioner

Irms (A) 25.9 15.2

Peak factor 3.78 1.95

THD (I) as a % 163 22.4

Harmonic Irms (A) 21.7 3.3

Characteristics of With active harmoniccurrent on full load conditioner and

smoothing reactor

Irms (A) 57.6

Peak factor 1.46

THD (I) as a % 3.4

Harmonic Irms (A) 2


Harmonic compensation needs are many andvaried, since the requirement may be toguarantee:c non-disturbance of a “clean” power network bya disturbing load,c or proper operation of a sensitive load (orpower network) in a disturbed environment,c or both these objectives simultaneously!

The problem of harmonic compensation can thusbe handled at two levels (exclusive or combined):c parallel compensation by current sourcedownstream of the point in question: this is the

Fig. 27 : active/passive hybrid conditioners - example.

vL

iL(h1) iL(hn)

VA.H.C

Passive filter

LoadVs(hn)

Vs(h1)

Zs

Zf


3 Hybrid active harmonic conditioners

“shunt-type” solution described in the previouschapter,c serial compensation by implementing anupstream voltage source.

The structures that we shall refer to as “hybrid”hereafter in this document are those whichsimultaneously implement both solutions, asshown for example in figure 27 .

They use passive filters and active harmonicconditioners. We have chosen to describe threeof the many alternatives available.

3.1 The “parallel/series” hybrid structure

The diagram in figure 28 illustrates the mainsubassemblies of this structure, namely:c one (or more) bank(s) of resonant passivefilters (Fi), parallel-connected with the disturbingload(s),c an active harmonic conditioner, made up of:v a magnetic coupler (Tr), the primary of which isinserted in series with the passive filter(s),v an inverter (MUT) connected to the secondaryof the magnetic coupler. The active harmonicconditioner is controlled so that:Vfa = K SHΙwhere:Vfa: voltage at the magnetic couplerterminals,K: value in “ohm” fixed for each order,ISH: harmonic current from the source.

SourceIs

Load

Vfa

Passivefilter Fi

MUT. Tr


Fig. 28 : “parallel/series type” hybrid conditioners - one -line diagram.


In this configuration the active harmonicconditioner only acts on the harmonic currentsand increases the effectiveness of the passivefilters:v it prevents amplification of upstream harmonicvoltages at the anti-resonance frequencies of thepassive filters,v it considerably attenuates harmonic currentsbetween load and source by “lowering” globalimpedance (passive filters plus active harmonicconditioner).

Since not all the power network current flowsthrough the active harmonic conditioner, thecomponents of the latter can be downsized (andin particular the magnetic coupler).

3.2 The “series/parallel” hybrid structure

The diagram in figure 29 shows that thisstructure contains the main subassemblies of theprevious structure, the only difference being inthe connection point of the coupler primary(in series between source and load).The active harmonic conditioner control law isthe same: its aim is for the active harmonicconditioner to develop a voltage which opposescirculation of harmonic currents to the source. Ittherefore acts as an impedance (of value K fixedfor each order) for harmonic frequencies.Passive filtering is thus more efficient (as thepresence of this serial “impedance” forcescirculation of the harmonic load currents to thepassive filters). Moreover, the serial filter isolatesthe load of the harmonic components alreadyexisting on the source and prevents passive filter

This structure is thus ideal for treating highvoltage and power networks, while at the sametime ensuring rephasing of fundamentalcomponents.

Its main drawback is that the passive filtersdepend on the type of load, thus requiring apreliminary study.

Finally, virtually all the pre-existing harmonicvoltages (on the source) are present on the loadside. This configuration can therefore becompared with the “shunt type” active harmonicconditioner.

Fig. 29 : “series/parallel type” hybrid conditioners. Fig. 30 : hybrid conditioner with injection by transformer.

overload. This topology is thus most oftenreferred to as an “harmonic isolator” since, insome respects, it isolates the source of adisturbing load, and, reversely, preventsoverload of a passive filter by upstreamdisturbance. It should be noted that this topologygenerates sizing and protection problems for themagnetic coupler, since:c total load current flows through this coupler,c a very high current wave is applied in the eventof a short-circuit.A possible solution to both problems may be touse a transformer with an additional secondarywinding (see fig. 30 ).Compensation then takes place “magnetically”by directly acting on the flow.

Source IsLoad

Vc

Passivefilter Fi

Vs

Vfa

MUT.

Tr.


Source LoadIs



The principle consists of “parallel” connectionof one (or more) tuned passive filter(s) and a“shunt type” active harmonic conditioner(see fig. 31 ).In this case also, the active harmonic conditionerand the passive filter prove the idealcombination. It may prove useful to limit (by theFFT technique), the action of the activeharmonic conditioner to the orders not treated bythe passive filters.

This structure is used (as applicable) to:c improve the harmonic cancellation obtainedusing only passive filters,c limit the number of orders of passive filters,c improve the effectiveness of the activeharmonic conditioner only (for the same powereffectiveness of the active harmonic conditioner).Nevertheless, this combination does not preventpassive filter overloads or the effects of anti-resonance with power network impedance.

In shortThese hybrid structures do not possess the“universal” character of the “shunt type” activeharmonic conditioner as passive filters need tobe chosen (in terms of type, number of ordersand tuning frequencies) according to the kind ofharmonic currents generated by the load. Thepresence of the active harmonic conditionerdownsizes the passive filters and reinforces their

effect. Vice versa, the addition of an activeharmonic conditioner of reduced power to anexisting installation increases the efficiency ofexisting passive filters.

3.4 The performances of hybrid structures

Prototypes have been designed, produced andtested in partnership with Electricité de France(which is the main power utility in France). Thesemodels comprised two banks of resonantpassive filters tuned on orders 5 and 11(harmonic compensation of a UPS-type load) or5 and 7 (variable speed drive load).

The result of the tests combining two types ofhybrid filters with a frequency converter (variablespeed drive for asynchronous motor) is givenbelow.

“Parallel/series” configuration (see fig. 28)

The test circuit characteristics are defined in thetable in figure 32 .

Comments : this topology is not suitable fortreating power networks with a high upstreamvoltage THD. However, its “current”

Fig. 32 : “parallel/series” type conditioner -characteristics and results.

Circuit characteristics

Source 400 V, three-phase,600 kVA, 5%,THD (Vs) < 1.5%

Load 130 kW,70% load,0.15 mH smoothing reactor.

Measurements taken

THD (IL) 35%

THD (Is) 9%

THD (VL) 2%

3.3 “Parallel” combination of passive filters and active harmonic conditioner

Fig. 31 : “parallel” connection of active harmonicconditioner and passive filters - principle.

FP1 FP2

Load



Fig. 35 : “series/parallel type” hybrid conditioner associated with a variable speed drive - reading of TH (VL) and THD (IS).

0Without filter Passive filters only Passive filter and active


2

3

1

4

5

10

20

30

40

TDH(IS)TDH(VL)

VL (in %)

Is (in %)

performances are totally respectable (the THD (I)is reduced from over 35% to 9%) (see fig. 33 ).It is thus particularly well suited for treatingpower networks with low upstream THD, or forwhich serial insertion of a device is particularlyproblematic.

“Series/parallel type”configuration (see fig. 29)The test circuit characteristics are defined in thetable in figure 34 .Comments : the performances are in this casealso totally satisfactory even if the quality of thesource voltage (THD (u) very low) does not allowto appraise performance in terms of isolation. Thesource current THD is however reduced from morethan 35% to 11% (see fig. 35 ).

Circuit characteristics

Source 400 V, three-phase600 kVA, 5 %,THD (Vs) < 1.5 %

Load 130 kW,70% load,0.15 mH smoothing reactor.

Measurements taken

THD (IL) 35 %

THD (IS) 11 %

THD (VL) 2.1 %

Fig. 34 : “series/parallel type” hybrid conditioner:characteristics and result.

Fig. 33 : “parallel/series type” hybrid conditioner associated with a variable speed drive - evolution of THD (VL) andTHD (IS).

0Without filter Passive filters only Passive filter and active


1

2

3

10

20

30

40

TDH(IS)TDH(VL)

VL (in %)

Is (in %)


Fig. 36 : summary of the various “active solutions” to combat harmonic disturbance.

Type of filter “Series” “Shunt” hybrid “Parallel” hybrid “Parallel/series” “Series/parallel type”Criterion hybrid

Schematic diagram

Sup. LoadA.H.C

Sup. Load

A.H.CP.F.

Sup. Load

A.H.C

P.F.

Sup. LoadA.H.C

Action on Uh/source Ih/load Ih/load Ih/load Ih/load, Uh/source

Performance +++ +++ +++ ++ ++

Active harmonic fund + harm harm. harm. harm. fund + harmconditioner sizing

Short-circuit great none none none greatimpact

Insertion difficult easy easy easy difficult

Improvement no possible yes yes yesof DPF

Open-endedness no yes yes no no

Resonance risk NA (not applicable) NA (not applicable) yes no no

A.H.C.: Active Harmonic Conditioner P.F.: Passive Filter

⇒⇓

P.F.

Sup. Load

A.H.C

Passive filter current remains constant and is thusrepresentative of isolation from the source.Additional tests proved that for very high upstreamdistortion (THD (V) = 11%), voltage quality at theload terminals continued to be good(THD (VL) = 4.7%).

Characteristics of the active solutionsWe have now dealt with series and parallel typeactive harmonic conditioners and with hybridstructures.

To round off this chapter, we propose tosummarise the qualities of these various “activesolutions” used to combat harmonic disturbance.

The table in figure 36 shows that, except for afew special cases, the “shunt type” activeharmonic conditioner and the parallel-connectedstructure are the solutions to be preferred in lowvoltage.


We would first like to emphasise that our aim isnot to act as a “selection guide” between thevarious types of harmonic compensationtechniques (both active and passive), but ratherto present the criteria used to size and insert theactive harmonic conditioners. Furthermore, aselection guide would imply that the varioussolutions given are available in product form.At present, given that both the “traditional”

solutions and the hybrid solutions require in-depth study and suitable solutions, only theshunt type active harmonic conditioners areavailable on the market (they require merely asimple study).We shall thus concentrate on identifying themain parameters that “potential” active harmonicconditioner users need to know in order to makethe right choice.

4.1 Objective and context

Knowing the “mechanisms”

The main problem of harmonic phenomena isundeniably linked to their very weak visibility.Although it is usually easy to observedeterioration in wave quality (voltage and/orcurrent) at one or more points, the combinationalfunction between the various sources(self-sufficient or not), loads and topology of thepower network is no simple matter!Moreover, the association between harmonicphenomena (often overlooked) and themalfunctions observed in the power networkcomponents (often random) is not instinctive.

Knowing the power network and its topologyThe first preliminary requirement thus concernsthe power network environment: implementationof an harmonic compensation technique requiresknowledge of the entire power network (sources,loads, lines, capacitors) and not just afragmented view limited merely to the zoneconcerned. This single-line diagram is in somerespects the first component of our “tool box”.

Carrying out an “inventory”

We have first placed an harmonic distortionanalyser in this “tool box”, vital for quantifyingdisturbance at various points of an existinginstallation.

Identifying and characterising disturbingequipment

We need to identify the main disturbingequipment(s) and their respective spectra. Thelatter can be obtained either by measurementsor by consulting the technical specificationsprovided by each manufacturer.

Defining the harmonic compensation objective

The second preliminary requirement concernsthe actual objective of the action considered: themethod used differs considerably according to

whether you wish to correct malfunctioningobserved, or to ensure compliance with thespecifications of power utility or a non-linear loadmanufacturer. Short term power networkchanges must also be taken into consideration.For example this stage must enable identificationof at least:c the type of compensation (global or local),c the power rating at the node considered,c the type of correction required (on voltage and/or current distortions),c the reactive energy compensation need.Once the above analyses are complete, themost advantageous technical and economicsolution must be chosen. The same objectiveoften has several technical possibilities, and theproblem is in most cases to make a choiceaccording to the individual difficulties of eachelectrical installation. For example, isolation ordecoupling by impedance of disturbing loads iseasily carried out on new installations provided itis considered in the design phase. However itfrequently generates unacceptable difficulties onexisting power networks.It is thus obvious that no “active” solutions(regardless of the type) can be systematicallychosen, but that an analytical approach isrequired in which active harmonic conditionercost alone is not necessarily the most importantfactor.Although Active harmonic conditioners haveundeniable advantages over passive filters, theyare not necessarily preferred particularly forexisting installations already equipped withpassive filters. The insertion of a serial or paralleltype active harmonic conditioner, after study, is agood solution.We shall now use experience acquired on site todescribe the implementation of a “shunt-type”active harmonic conditioner which is the simplestsolution.

4 Implementing a “shunt type” active harmonic conditioner


4.2 The insertion point of a “shunt-type” active harmonic conditioner

The connection principle of a “shunt type” activeharmonic conditioner is shown in figure 37 . Inour example it is inserted in parallel mode in the

Fig. 37 : connection of a “shunt type” active harmonic conditioner: principle.

LV switchboard of an installation, and the onlyinteraction with the power network to be treated,is the insertion of the current sensors.

Fig. 38 : the various insertion points of a “shunt-type”active harmonic conditioner: principle.

Load to becompensated

Harmoniccompensationcurrent

Supply network

LV distribution switchboard

Harmoniccurrent to becompensated


MV

LV

A

Main LVswitchboard

Subdistributionswitchboard

Terminalswitchboard

M M


B


C


As regards insertion of the active harmonicconditioner, harmonic compensation can beconsidered at each level of the tree structureshown in figure 38 .The compensation mode may be termed global(position “A”), semi-global (position “B”) or local(position “C”) according to the point of actionchosen. Although it is hard to lay down strictrules, it is obvious that if disturbance is causedby a large number of small loads, the “mode”preferred will be global, whereas if it is causedby a single high power disturbing load, the bestresult will be achieved using the local “mode”.

Local harmonic compensation

The “shunt type” active harmonic conditioner isdirectly connected to the load terminals. Thismode is the most efficient provided that thenumber of loads is limited and that the power ofeach load is significant compared with globalpower. In other terms, the loads treated must bethe main generators of the harmonicdisturbances.Circulation of harmonic current in the powernetwork is avoided, thus reducing losses byJoule effect in upstream cables and components(no oversizing of cables and transformers) aswell as reducing disturbances of sensitive loads.It is worth pointing out, however, that the“shunt type” active harmonic conditioner lowerssource impedance at the connection point, andthus slightly increases current total harmonicdistortion between the connection point and theload.


Semi-global compensationThe active harmonic conditioner, connected tothe input of the LV subdistribution switchboard,treats several sets of loads.The harmonic currents then flow between theMLVS and the loads of each feeder.

This type of compensation is ideal for multipledisturbing loads with low unitary power, e.g. onfloors in service sector buildings (officeequipment and lighting systems).

It also makes it possible to benefit from non-algebraic summing between loads, at the cost ofa slight increase in losses by Joule effect oneach feeder treated.

NB: this type of compensation can also beapplied to a single feeder, thus limiting harmoniccompensation to a single type of load(see fig. 37).

Global compensationThis form of compensation contributes rather tocompliance with the point of common couplingaccording to “power utility” requirements, than tothe reduction of internal disturbances in thecustomer’s power network. Only the powertransformer(s) actually derive direct advantagefrom harmonic compensation. Nevertheless, thisform has a serious advantage for operation inautonomous production mode as a result of thenumerous interactions between disturbing loadsand generator sets with high harmonicimpedance.However, compared with local compensation,this compensation technique results in areduction in active harmonic conditioner powerrating which benefits from the non-algebraicsumming of the disturbing loads throughout thepower network.

The main factor to consider when sizing a“shunt type” active harmonic conditioner is itspower rating (or more precisely its rmscurrent ):the rms curren ICA RMS is the current that can bepermanently generated by the active harmonicconditioner.Other characteristic active harmonic conditionerfactors are its bandwith and its dynamiccapacity :c the active harmonic conditioner bandwith isdefined by nmin and nmax, the (minimum andmaximum) action orders of the active harmonicconditioner.

The following can be written:

Ι ΙCA RMSn = n

n = nA =

min

max( ) ( )∑ CA n

2

c the active harmonic conditioner currenttracking dynamic capacity (expressed by

didt

) is

the capacity of the active harmonic conditioner to“track” rapidly varying references.

NB: these last two factors are not considered toaffect size, since they form a characteristicinherent in the active harmonic conditioner andnot an adjustable parameter.Choosing nominal rating:Provided that the spectrum of the current to betreated ICH is known, the nominal current of theactive harmonic conditioner IN CA RMS, can bedetermined such that:

Ι ΙN CA RMSn = n

n = nA

min

max( ) ≥

( )∑ CH n

2

Provided that the above condition is met, the“new” current total harmonic distortion(upstream) can be calculated once the activeharmonic conditioner is put into operation:

THD % =

+ CH2

(n) CH2

(n)n = n + 1

n

n = 2

n = n

CH

max

min

1

ΙΙ Ι

Ι( )

→ ∞∑∑

( )This formula is used to determine whether themaximum theoretical performance of the activeharmonic conditioner is compatible with thetarget objective. It can be simplified still further,if we consider the specific case of MGE UPSSystems products for which nmin = 2 and nmax = 23:

THD % =CH2

nn = 24

CH 1

ΙΙ

Ι( )( )

→ ∞∑

( )

n

Furthermore, the above nominal rating selectionrule must be weighted by the following practicalconsiderations:c the harmonic spectrum of most loads issignificant only in the band h3 to h13,c the purpose of inserting the active harmonicconditioner is not to cancel the THD (I) but tolimit it below a predefined level (e.g. 8%),c an active harmonic conditioner can be chosenwith a rating lower than IN CA RMS, and thenoperate in permanent saturation (by permanent,automatic limitation of its rms current).

Finally, parallel-connection of a number of activeharmonic conditioners at the same insertionpoint is technically feasible, a solution which mayprove of interest for upgrading of a pre-equippednetwork.

4.3 Sizing a “shunt-type” active harmonic conditioner


4.4 Application examples

Reduction of line distortions

As regards high rise buildings or buildingsoccupying a large ground surface area, the mainproblem concerns the lengths of the linesbetween the point of common coupling(MV/LV transformer) and the loads.This is because, irrespective of voltage wavequality at the origin of the installation and of theprecautions taken for the lines (choice of cablediameter, splitting,...), voltage total harmonicdistortion increases at the same time as“altitude” or distance!

As from a specific point, therefore, voltagedistortion can be considered unacceptable inpermanent mode, and the “shunt-type” activeharmonic conditioner provides an interestingalternative to traditional solutions (e.g. isolationby suitable LV/LV coupling transformers).

Let us consider the example of a three-phaseUPS supplying a set of “computer” loads at theend of a 60 m line. We then observe a voltagedistortion of 10.44% (phase to phase) and of15.84% (phase to neutral) at load level. Thisdeterioration is the result of a combination oftwo factors, namely:c UPS sensitivity (with non-PWM control) to thenon-linear characteristic of the downstream current,c the mainly inductive characteristic of the linewhich amplifies distortions.

The proposed solution is illustrated in figure 39and is based on insertion of a “shunt-type” activeharmonic conditioner as close as possible toloads. Performances are then totally satisfactorywith respect to the objective: the THD (U) dropsto 4.9% phase to phase and to 7.2% phase toneutral.

Combination of “shunt-type” active harmonicconditioner and passive componentsEffect on tariffs

A pumping station is designed to maintainconstant water pressure on the drinking waterdistribution network (see fig. 40 ). The motor-driven pump P1 is thus controlled by a variablespeed drive with frequency converter.

In this particular instance, the main objective wascompliance of the source current spectrum withthe power utility’s requirements. With no filteringdevice, the authorised harmonic emission levelwas:c greatly exceeded on order 5,c more or less reached on orders 7 and 11.

Fig. 40 : pumping station diagram (main low voltageswitchboard).

Fig. 39 : using an active harmonic conditioner to treatvoltage total harmonic distortion at the end of a 60 mcable.

n computers& similar loads

Connecting cable: 60 m/50 mm2

Non PWM UPS sn = 200 kVA


P2

Starter

P1

L

Variablespeed drive



The choice made is a combination of smoothingreactors and a “shunt type” active harmonicconditioner: performances are shownin figure 41 .c All the harmonic orders are well belowauthorised emission limits.c The current total harmonic distortion isreduced by 89%.

An advantage particularly appreciated by thecustomer is the reduction of his contractedpower (in kVA).This example also shows that the combination ofan active harmonic conditioner + smoothingreactor is particularly suitable in view of the highdegree of disturbance.

Fig. 41 : pumping station - spectral representation of harmonic currents.

02 3 4 5 6 7 9 11 13

5

10

15

20

25

15 17c authorised I c I without compensation c I with compensation


5 Conclusion

The profusion of non-linear loads makesharmonic distortion of power networks aphenomenon of increasing amplitude, the effectsof which cannot be ignored since almost all thepower network components are in practiceaffected.

Up to now the most popular solution was passivefiltering. However, an attractive alternative to thiscomplex and non risk-free solution is nowcommercially available in the form of activeharmonic conditioners.

These devices use a structure of the static powerconverter type. Consequently, semiconductor

progress means that converters, which arenormally harmonic disturbers, now form efficient,self-adaptive harmonic compensation devices .

The easy to use, self-adaptive “shunt- type”active harmonic conditioner, which requiresvirtually no preliminary studies prior to use, is theideal solution for harmonic compensation on anon-linear load or LV distribution switchboard.

However it does not necessarily replace passivefilters with which it can be combinedadvantangeously in some cases.


Joseph FOURIER proved that all non-sinusoidalperiodic functions can be represented by a sumof sinusoidal terms, the first one of which, at therecurrence frequency of the function, is said tobe fundamental, and the others, at multiplefrequencies of the fundamental, are said to beharmonic. A DC component may complete thesepurely sinusoidal terms.

FOURIER’s formula:

y t Y sin n t - nn = 1

n = ( ) = + ( )

∞∑ Yo n2 ω ϕ

where:- Yo: DC component value, generally nil andconsidered hereafter to be nil,- Yn: rms value of the nth harmonic component,- o: angular frequency of the fundamental,- ϕn: displacement of the nth harmoniccomponent.The notion of harmonics applies to all periodicphenomena irrespective of their nature, andparticularly to AC current.

rms value of a non-sinusoidal alternatingquantityThere is similarity between the normalexpression of this rms value calculated from thetime evolution of the alternating quantity (y(t))and the expression calculated using its harmoniccontent:

Yrms = 1T

y t dt = Y2n

2

n = 1

n = ( )∫ ∑

∞

0

T

Note that when harmonics are present, themeasuring instruments must have a widebandwidth (> 1 kHz).

Total harmonic distortionTotal harmonic distortion is a parameter globallydefining distortion of the alternating quantity:

THD % =Y

n = 2

n =

1( )

∞∑

100

2Yn

There is another definition which replaces thefundamental Y1 with the total rms value Yrms.This definition is used by some measuringinstruments.

Individual harmonic ratioThis quantity represents the ratio of the valueof an harmonic over the value of the fundamental(Y1), according to the standard definition or overthe value of the alternating quantity (Yrms).

H %n ( ) = YY

n

1

(Frequency) spectrumRepresentation of harmonic amplitude as afunction of their order: harmonics value isnormally expressed as a percentage of thefundamental.

Power factor (PF) and Displacement PowerFactor (DPF)

It is important not to confuse these two termswhen harmonics are present, as they areequivalent only when currents and voltages arecompletely sinusoidal.

c The power factor (λ) is the ratio betweenactive power P and apparent power S:

λ = PS

c The displacement power factor (cos ϕ1) relatesto fundamental quantities, thus:

cos 1ϕ =PS

1

1

In pure sinusoidal waveform:

cos ϕ1 = cos ϕ = λ

Distortion factor

The IEC 146-1-1 defines this factor as the ratiobetween the power factor and the displacementpower factor cos ϕ1 :

cos : = cos 1

1ϕ λ

ϕν

It is always less than or equal to 1.

Peak factor

The ratio of peak value over rms value of aperiodic quantity.

Fc = YpeakYrms

Appendix: review of harmonic phenomena

Definition and characteristic quantities


Linear and non-linear loads

A load is said to be linear when there is a linearrelationship (linear differential equation withconstant factors) between current and voltage. Insimpler terms, a linear load absorbs a sinusoidalcurrent when it is supplied by a sinusoidalvoltage: this current may be displaced by anangle j compared with voltage.When this linear relationship is not verified, theload is termed non-linear. It absorbs a non-sinusoidal current and thus harmonic currents,even when it is supplied by a purely sinusoidalvoltage (see fig. 42 ).

Voltage and current total harmonic distortionA non-linear load generates harmonic voltagedrops in the circuits supplying it. In actual fact allupstream impedances need to be taken intoconsideration right through to the sinusoidalvoltage source.

Consequently a load absorbing harmoniccurrents always has a non-sinusoidal voltage atits terminals. This is characterized by the voltagetotal harmonic distortion:

THD %

Z

U

nn = 2

n =

1( ) =

( )∞

∑ 100

2Ιn

where Zn is the total source impedance at thefrequency of harmonic n, and In the rms value ofharmonic n.

The greater the non-linearity of the load, thelarger the voltage distortion and the higher theorder of the harmonic currents (inductive sourceimpedance 2 π f1 n L).Remember that current total harmonic distortion is:

100

2

n = 2

n =

1

Ι

Ι

n

∞∑

In order to illustrate the main types of behaviourof the main sources, figure 43 shows theevolution of their impedances as a function offrequency.

For further details, readers can consult“Cahier Technique” no. 159.Do not forget that large diameter cables aremainly inductive and that small diameter cableshave a non-negligible resistance.

Fig. 43 : output impedance of the various sources as a function of frequency.

Fig. 42 : current absorbed by a non-linear load.

1F

0

I

U

π

0

50

100

50 250 500 750

Ratio of output impedance over nominal load impedance

PWM inverter

F (Hz)

Zs

Zc%

AC generator X"d = 12 %

Transformer Uccx = 4 %

Origin and transmission


Deforming loads

Most deforming loads are static converters.They may be powerful and few in number, orlow-power and plentiful. Some examples are:c fluorescent lamps, dimmers,c computers,c electrical household appliances (televisionsets, microwaves, induction plates).

Nowadays the proliferation of low power devicesis chiefly responsible for increased voltageharmonic distortion in power networks.

Figure 44 illustrates the current absorbed by afew loads, and figure 45 the matchingharmonic spectra (typical values).

Fig. 44 : curve of the current absorbed by some non-linear loads.

0

U

i

u C

i

1: Light dimmer or heating regulator

2: Switch mode power supply rectifier,for example:c computerc electrical household appliances

3: Three-phase rectifier with front endcapacitor, for example: variable speed drivefor asynchronous motors

4: Three-phase rectifier with DC filteringreactor, for example: battery charger.

5: Three-phase rectifier with AC smoothingreactor, for example: high power UPS

0

e

α = π/2

i

e

i

R

e1

i1

C

i1e1

e2

e3

i2

i3R

e1

i1

C

Lc

i1e1

e2

e3

i2

i3R

e1

i1C

i1e1

e2

e3

i2

i3R

Type of converter Diagram Current (& voltage) waveform


Fig. 45 : example of the harmonic spectrum of currents absorbed by the loads in figure 44 .

N° H3 H5 H7 H9 H11 H13 H15 H17 H19

1 54 18 18 11 11 8 8 6 6

2 75 45 15 7 6 3 3 3 2

3 0 80 75 0 40 35 0 10 5

4 0 25 7 0 9 4 0 5 3

5 0 33 3 0 7 2 0 3 2

Harmful effects of harmonics

consideration, a standardised empirical formula(NFC 52-114) is used to calculate the deratingfactor k to be applied to a transformer.

k =

1 + 0.1 Hn2

= 2

n =

1

1 6 .

n

n

∞∑

where H = nn

1

ΙΙ

For example whereH5 = 25% ; H7 = 14% ; H11 = 9% ; H13 = 8%,the factor k is 0.91.

Effect on ac generators

In the same way as for the transformer, theharmonics create additional losses in thewindings and magnetic circuit. Furthermore theharmonics create pulsating torques whichgenerate vibrations and additional overheating inthe damping windings. Finally as thesubtransient reactance is relatively large, thevoltage total harmonic distortion quickly riseswith the increase in harmonic currents.In practice, limitation of the current totalharmonic distortion to a value less than 20% isaccepted, with a limit of 5% for each harmonicorder. Beyond these values, manufacturers mustbe consulted as to the spectrum of current reallyabsorbed by the loads.

Effect on power lines and in particular on theneutral conductor

Harmonic currents create additional losses inconductors accentuated by the skin effect.Losses are even more serious whensingle-phase loads absorb harmonic currents3 and multiples of 3. These currents are inphase and are added together in the neutralconductor.With, for example, an harmonic 3 of 75%, thecurrent flowing in the neutral is 2.25 times thefundamental. The current in each phase is only

1 0.752+ = 1.25 times the fundamental.

Special attention must thus be paid to the sizingof the neutral conductor when non linear loadsare present. The TN-C earthing system isstrongly advised against.

Effects on low current appliances andsystems

Harmonic distortion may cause:c malfunctioning of certain appliances which usevoltage as a reference to generate semi-conductor controls or as a time base tosynchronise certain systems;c Disturbances by creating electromagneticfields. Thus when “data transmission lines”circulate in the vicinity of power lines throughwhich harmonic currents flow, they may besubjected to induced currents able to causemalfunctioning of the equipment to which theyare connected;c finally circulation of harmonic currents in theneutral provokes a voltage drop in thisconductor: thus in the case of the TN-C earthingsystem, the frames of the various devices are nolonger at the same potential, which may wellinterfere with information exchange between“intelligent” loads.Moreover current circulates in the metallicstructures of the building and creates disturbingelectromagnetic fields.

Effects on capacitors

Capacitor impedance decreases as frequencyincreases. Consequently if voltage is distorted,relatively strong harmonic currents flow in thesecapacitors whose aim is to improve the DPF.Furthermore the presence of reactors in thedifferent parts of the installation reveals a risk ofresonance with the capacitors which mayconsiderably increase the amplitude of anharmonic in the capacitors. In practice,capacitors should never be connected oninstallations with a voltage total harmonicdistortion greater than 8%.

Effects on transformers

Harmonics generate additional losses in thetransformers:c losses due to Joule effect in the windings,accentuated by the skin effect,c losses by hysteresis and eddy current in themagnetic circuits. To take these losses into


Harmonic Class 1 (sensitive devices Class 2 (public and Class 3 (for connectionorder and systems) industrial supply networks) of large non-linear loads)

2 2 2 33 3 5 64 1 1 1.55 3 6 86 0.5 0.5 17 3 5 78 0.5 0.5 19 1.5 1.5 2.510 0.5 0.5 111 3 3.5 512 0.2 0.2 113 3 3 4.5

THD 5% 8% 10%

Standards and recommendations

Electricity is today regarded as a product,especially in Europe with the directive ofJuly 25th 1985.

The EN 50160 standard defines the maincharacteristics at the customer’s point ofcommon coupling for a low voltage public supplynetwork, and in particular the harmonic voltagelevels (class 2 levels in the table in figure 47 ).These are the levels of compatibility in terms ofelectromagnetic compatibility (see fig. 46 ).In addition to this European standard, themaximum levels of the various harmonic ordersare defined in IEC 61000.

c For low voltage public supply networks:IEC 61000-2-2 and CIGRE recommendations.

c For medium and high voltage public supplynetworks: IEC draft standard for mediumvoltage and CIGRE recommendations.

c For low voltage and medium voltageindustrial installations: IEC 61000-2-4.By way of illustration, the table taken from thisstandard gives the harmonic levels ofcompatibility in three standard situations(classes) (see fig. 47 ).

To ensure these levels are not reached, limitsmust be set for the disturbances emitted(emission level) by devices either consideredseparately, or for a group of devices as regardstheir point of connection to the power network.IEC 61000-3-2 deals with low voltage anddevices absorbing current of less than 16 A,and the IEC 61000-3-4 draft guide deals withdevices absorbing current greater than 16 A.Although there is no standard for industrialapplications, there is a sort of “consensus”concerning the notion of stages for authorisationof connection to the public supply network:stage 1 means automatic acceptance for lowpowers with respect to contracted power, stage 2means acceptance with reservations (a singleconsumer must not exceed a level representinghalf the level of compatibility) and stage 3 meansexceptional and provisional acceptance whenthe previous level is exceeded.Finally, to guarantee proper operation of devices,these devices must be able to withstandlevels of disturbance greater than the levelsof compatibility given in figure 47 shouldthese levels be overshot (permissibletemporarily): this is their level of immunity.

Level of disturbance

Level of susceptibility: level from which a device or system starts malfunctioning.

Level of immunity: level of a disturbance withstood by a device or system.

Level of emission: maximum level authorised for a consumer on the public supply network orfor a device.

Level of compatibility: maximum specified disturbance level that can be expected in a givenenvironment.

0

Fig. 46 : the various levels of disturbance for compatibility of non linear/sensitive equipment.

Fig. 47 : ratio (as a %) of acceptable harmonic voltages (compatibility).

Schneider Electric Direction Scientifique et Technique,Service Communication TechniqueF-38050 Grenoble cedex 9Fax : (33) 04 76 57 98 60

DTP: Sodipe - ValenceEdition: Schneider ElectricPrinting: Imprimerie du Pont de Claix - Claix - France - 1000- 100 FF - ©

199

9 S

chne

ider

Ele

ctric

06-9976729

technical collection cahier technique no. 183

Documents