Circuit Breaking Principles

il

I

t li

CHAPTDR V

Principles o f Circuit-breaking N the occurrenceof a short-circuit within a system,a particular breaka in the aftectedline will l* ttippud oB.6nautomatically fuaction of this breaker is to carrsethe fault curreirt to oeaseflowing assoonaspossible,and to dp so without harmful effeqtgeitlier to the systemor to itsdfi It was shown in Cbapter IV that contact separa- ' tion merely creates an arc discharge i'ithofit caueing appreciable altercion oJ the current initially. This is perfectly desiaSle becagse any suCden cut-off of the current at that stage might give rise to harmful induecd voltages in the circuit. The function of the breaker is tihen, more epecificalln to introducea high impedanceinto the circuit in such a rnanner as to force, or allow,tte current to reach zero without possibility of the arc being res'truckthereafter, and without causing the current to r€ach an abnorsralrate of decrase. The possibility of daoage to thb breaker hasbeen mentioned because,as well as being subjected to hating and electromagaeticforce in common with the other componentsof the alfectedline, the breaker is also the 'nain receptacleof ttre shoftcirnrit energy becauseof the resistive nature of the dischargewithin it A primary factor in high power circuit-breaker design is the Prwision of mechanical strength sufrcient ts withstand the sudden Qeleascof large quantities of eiergr of the order of, say 500 kilopatteeconds(i.e. more than 100,000 calories). Early extinction of the arc helpsto keep tlis eneg5r releaseie 3 minimum. D.G" Circuit-brea&lng. Fig 5.1c rqrresents a eimple d.c. *.u! *-prising a'geneoior, rdi"tor, teacior and circuit-breaker. t-he latter is assuued to break the lmd curreotr I - ElR. -la diagramb, the line t is the currenf-voltage characteri*ic.'bJ-the arc u'hen'its current is decreasing; line 2 is a resistanceline and represeilq.lhe voltage E-;& thai, is, tlrc: e.m.t'. remaining when the restgljV-_q drop is:heducted.. 5he initial current at contart'$:innrtiol

.

r riJ l

76

sWtTCg€BAB PRINCTPLES

ir thc frrll veluer, and it is clearfrom thc diagram'l thst thc src uostablcd qtrrents betw.ccnr an{ 11 ei"ce tni voitages ,.q;rd b ", (.4_r"). qr greater thantheamounts.iil.ui"

:: 11:ijg TTj howcver, tte arc currenttdndsto i"cre"s; 11-T,:""9 -!tb rmcethis sectionof!::Iine

! lics-berowrine2, andin ruch conditionr of thc fault currentis not possible.UnAer tJ"ryp":l circuit-hhing oonditions, howcycr,ttc arcis bciig tengtncnJti tihescparetioa of tlrc contacrs, andtlc ecot'is;d;,fr;*;;,: voltagccts*crsistic so ttst it lies abovdthc !*istanac ti;Thl

g.D.

T ctr

(-

(a). Fig.5.l. D.C. qc choactsisth-+otatt

deaca4nE

progrcssivelengthening,yh$ qly be assisrcdby etectromagnaic forcein cstain designg,is.a basic'r.equircment oi th" d";;il|r.

breater.Ncverthelds-ti i; th" t;;rA;;;iA;G* ;;;;; nalntainth9,arcwhishresuttsin the rai-sing .f til';h;;illtl

rnr{any adon which brings this about ,*} i_pro*;;;;; performance.Thus, any m-eans',r'hi"hin"""a.e thc.r.ate

of enercy try. pg unir'teFthof'gc'arc'cotumn nay-uca""l*uu=io-.id#i -. -..--.

to thc lengthpning. . The grnerarcfi.ectof thc circuit inducranccduring thc inter. nrption ie of spccialimeresr. While the *I"1, i"j"-iir";;;l: to d whcrethc zolace-.a--r,R ir ressd -q,r;ftb ,"L bahncbof voltsgeir

gdl up by c, - fup,doo..hi" iil nqativc with dc. crc$ng curEcnt an4 in a@rdance '*frti;r,; I;;A;;; qcnd to.

4fnraintlrso'rcnt Til ;Jrt"gaa.b";'"ti.fiJ

.rsis6 ia istcrrustion, but ir ad G'to increascite indud volageehthc u;bblerangce .u-*l--l dcarfron diaeranC. Morcovcr, "f *t-.t"F"io. m ec'ount d ol s.t .r"*iffiTi#

*rr*o.[J"* r6i-o highvetucr frl1#,ntnt_o Idcaun thcd.c" u'.,r",g"rrJg.p*# "*y iiln ;tri"HJ.f,Xfll#:fr F rlrilcdlelcarof thoscristrnc! *

tiilil";dt*

oo ,qioo bcrpca

I Il, I I

I lt'

PRINCIPLES OP CIRCUIT.BNEAKINO

7i

L andf" should vanish, and such that induced toltagesjust prior to *tinaio" are g Einimum. Note that alteration of the generator Ir.f. giuo a correponding elteration of the positioi of the reietancc tine2 iho"e slope, however, remainsfixed by the circuit resistance' iii* ,ft" hish;r-the circuii e.m.f. the morl eficient'must be the iJoi.f thelreaker in lengthcning and de-ionizing the arc to raiee thscharactsistic abovethc resistanceline.

af

Fig.5.2. D.C. oc cnaactl**laawt rg arralc&n

dadc&h'g 6td I^Su

[ig. 5.?.a shows the arc characteristic 6E it might bc- when

nodified by lengthening during the brarLing-opgetign- .fh9 orc voltagcie At piiou greatcr than the ahount-E+? anCthc batancc "i !r grveuVy t. il4at wnicn i" proPortignal t9 lhc gradie$ o{ the el orrenrtirie curvc. Thie curvc msy therefore be deduccd, '-fhir rhownin.fig. 5.?,h,and prondes thc totd arcing timcr Pgibd ney b€ an-appreciable-fraction of the total time of bhort-circuit crrrrentflow and must be kept as ehort ss possibla Fig. 542showB thartlu arcingtime may be reducedby incrcasiogthc retesof Gurrent frlt but this i;volve nigber ralues of tlhc indued voltagec,ca, and a orrcapondinglyhighcr-poaition of thc chrracteri*ic" - In thc-typicel casei[uatratcd-by fig. 5.2t thc e*tincion voltage of th. arc h ebcut tnicc the generaior-voltagq ehowing that thc dielectric etrcngth of thecontactq'apat ertinction has rcachedan adequatelevcl ' Summiizing, euccessfuld.c. circuit-prca&'in! must-involve s ostrin amountof-compromise.'fhe nccessriryfunction of th9 9t q.t il to raise the arc' charasteristic iufficiently to avoid rtability. - Dy increasingihe e$gctubeyond the minimtrm neccssary,a shot!3r' r' Iri.rg :ii". and,-tieiefoie less.lcverc heating and eleetronnsg3t€Bic'

('

78

swrrcucgaR pn.tNctplEs

str6scr' will bcobained fthough the rikefihoodof ovcrvortagcsua! a.isa In goeral tlre minimumir-ciog e-e p.r_ioi[i. -*ti"p*i upon the 2r''ount of inductanccin tf,c circiia : A.C Circuit-brea&ing. In atternatiai_curlent cy*erns -.1* tfu whole probtem.of circuir.f,seakiogi" difiHt-i; io sFtems. This is becar.seo.6...h"nEesdirectico t"ice l"-;"G;; G

.

andis therdoreinsangleoylf zdo twicep*.y"d l" i,nort, til ac" circuit-brealer gdoits th6 poperty bi pd;ddttre c'rrcn from

rising again afti d ?Ero&"i.. It i,l.neithcr i"""o"ry ,i, desirablcto crrt ofi the anrrcot at any otherpoint"" trr" Jrr"i",{ wavebecausesucha distoriion of the nattrrar--"o"1i*"*"a a*

'

produceorervoltages.Idealy, tn" *tre"t.sni rru u"-"ri*"a to r.u normallyto zero,at whichiiiant Oe aiU""t lirt between'the circuit-breakercontactsrnourJ rapiary"il""o*" ""gO;ilh;;; *c prever:tre-strikingof the ar',c"The aim of desigtrerj tni"*."", i, i brea&er wriichcao-invariably accomplish tni" irr" cirt *rredt zcm after contast separation, "t c,r.,ent *i$hi' t" ,d;?;I "tto"ny caplgiry'and without harur it"er. rn" pfid;;"" ;fihi problernin designis srtremelydifr"olq;A-;;" lf ;;y;[rd

would not probablybe economical Efucts gf.pouer-Futot. The ac. circuit-breaker atte'pts b preve-ntre-striking of the arc after the fault curcnt and the contacr gap must theiefore ;h;;-.h",'ctJuit it"u "'orq vorag. appearing there at the" zcro instlnl This instantaneousvortage dcpendsopg1 tl" po\rer factor of the circuii tn" i"-"fa f tn is relativeryrigt, tLetreaker -"y r,""" a "f-." aJ-""1v,lii.ome natul of voltjrgc appreciably tess ttai .h" ono. Nevert'eless,in setecing a breakerr*'" ;-J;,-t""".lrr"r,r, possibh, o"*r*iar";'* fzults and associatedpoier f"d; n r"t i" in any event,it shouldbe rimeurbered "Uor"Ji* "nd,, o", O" a"*rl-ii*l", gapms

..contend volage *.n" oni**iiiii", intemrs iT,t\.p*t d +hst tion, if not ii*""t (rssuming, of course, the 50c./r &equeocy ".tra$, *ppti).

r \

t -*y et"u.,-breakeclocationson h.v.se 1fpoira 'Erstems, the factorunderf""tt *;Jttdd-#;;iearty.zao; grvi"g virt'a,Iy pcakvorageJ*r*a zero.s;this is.beca'scof tro

*,* tr Hf.ry *. olaiin,c"Eciary-"uirl.lTgl"ilt"g

*"* ., llf*il:ffi "ffu?"ffi ffi$?r#"TiH caseof aslmnetdcal *d;-R"f;;;-d:%:"rn ,"1"

and uninorlco-ps.f q pyrg.t

i i

,,i,j--

i""r currcnt on 50 c/s sgcms lgrc gsears aaaie.ilGfriffiu[i*-fir$, aqdthr, ....Y*o*

-'-:_:

:

'

P R I N C I P I , . E SO F C I R C U I T . B R E A K I N C

lt,-

79

.cur€nt zerosdo not thereforecoincidewith p&ks of vottage. This phasedisptacementdccrcases,of cource, as ihe d.c. comp-onentof thc. asrynmetrical current decays. The apparent phase shift -to with asymmctrical current tends associated give an easier inrcmrption and breaker duiy. illuch of this benefit-man however, bclost-bgcauythe asymmetrydecaysappreciablyduring ihe opening time of the circuit-breaker. Even with ; power factor Is low as'0.1, for-examplgmost of the asymmetrydiesaway dudng the 6rst q or L ' g!!o: and this irrterval may be ehort comparadrpith that ffom fr.rlt initiation to contact separation. Much depends,howevei upon the ipfting characteristicsof the protectivl scheme, and-uiorn tfie breakermechanismand gontact d*ign, and partially asym-metriel currentsoftcn occur during arcing. A circuit-breaker has to deal with normal load currents, as well asfaults, and since the former are generally aseociated.with high powerfactorsthe interruption duty is then rit"tivety easy: The;. load-breakingsrvitch dso always-benefitsfrom tais condition, of oourBe. Current Zero Period. After contact eeparation in an ac.circuit-breaker,an ars is drawn but and its curient alternatessuch that an instantaneouszero (rccurs periodicafly according to the . supplyfrequency. Final extinctionofthe arc is onty po*ib=l" at one ot theie current zeros and will occur if the volage availatle at Buch an instant is not enoughto break dowa the de-iinizea i.esidual arc column. The possibilitiesof breakdownor final --xtinction occurring at any current zero depend upon the phyisical conditions cxisting then. 'I'his does not mean tlat the ..cirtient zero period'; on Uu rcgardetlas being independent of conditions existing prwiously. rYe.-navealready shown.tb* power loss and rate of current fall, IT:r_tg zero, havea markedefrect upon the t€mperaturc aird.degree of ioniCationof the irc,at the z;C imt nL 'The all-imfortot Plyt:g state of the contact g-apat cun€fit zero thus dependsupon :lho,l* beenhappcniagto tf,e'arcjust pleviously and thus includes

1!

.T-.e"

9f grrryit conditionsaswell asde-baizingand ionizing

**Tn#!lrffii*f;

orthe'roiaoal"or,,mn in thecontacr

Bap.erists when the current.reacheszero. 'f,he next stageis tho riseof voltagefrom zero(at the cprrent zeroinetant)t what 13irtvatug.is appropriateto the circuit'conditions.It is Ois rlSng 3;el whigh_nray or-inaynot.breakdowrrthe gap and re-strikethi :_":,"Cj arc. Just as the p.re-zeloconrlitioncart inpoiant in determining

g0

slvtTcHcEAR PRINCTPLES

the conductivity state of tle column ^t ?.arc,,the very s$ort perid of rising voltageimmediatelyaften'ards is widently equally cdtical The generalaatureof arcing conditionsin circuh-breakss hasbeen describedin ChapterIV and we uray now prcaeedto a closeretrrdy of the relationshipb'etweenthe arc aad the circuit conditionsarouni current zero. This involvesa consid€ratiorrof: (i) dre efrectsofan vob.rgeonthe currentwave-form;(2) the efieoof circuit capacitancc in parallel t'ith the breakergn'current wave-form; (3) .the nanncr in which the volage acrosst.hebreaks dfes frdm ao to tte circuit

II

iI

r La,-L-cb

Ancwltagse,er*Idil,

a

oB ack war.d "eurnen t, i6

I'ig..5.3.Distortbnof A.C. currcn,ttax,cb1'urcevltuge value appropriateto thc current zero instant; and (4) the current zero period in relation ro sone of the physical lrrosessesoccurring in the residualcolumn in the contactgap. Arc Vollage and Curreat \[ave-form. It is convenientto . .think of the a.c. arc ail a circuit dement whoseresistancegenerally variesin an inversernannerwith the current during any half-cycli Thus.the powerfactor of the circuit in which a Urdf,erib oper;ting may changesomeqriatduring any half-cycleof arcing. For instanrr; an arc of 1,000A r.m.s., iri oil" severalco in tength,iiU havea total resistanceof a fraction of an ohar.ovs moet oia naf-"yae p€rio4 but asth:eqnreot falls towardszero, tbe resistancemay'rise Lpidty to sr,veralhun&edsor thcusandsof ohms. The effectsoi this deirmi v.erX-nuch_upon circuit conditions. Coileiaer a sinple inductivc drcuit as shownin 69. 53. Thd cirsrit-bree&sr ts elced and tb

PRINCI PLES Or CTRCUTT.DREAKINC

(

8t

currqrt is sinusoi&I, loqS,nq90" electrically behind thc gcnerstor llthen the circuit-breatcr opcnsand an arc is €stabtishcd e'.f. thc voltage eo modifies.tr" pn u; rehtion .f;;;;;;i arc voltaga The efiect is comparativetysmall while tbe *-";i,;li;; . rnd arc resistane is low, but a m"rked efiect can;;; ;;JJ"?; approachezergt An irnpott ot tourtl" tan" zero is ehiftcd from its naturalposition in tirire and so coincides*ith lessthan ";;;;;;; lglta.ge + p*k This distortioo of thu *r-"r-;;;""j rhe displacementof ?Ero,:n y b" 9f iittfi-p;;;; rcurrenl systems, h.v. but on l.v. s1'stemswheri the ;i t compamblewith the supply "re"g "Jdg; the currerit _"y-b, J;;a -"?lqg", crmpletelyforcedintg nhxe wittr th-esupptyr.rrg" oi. Thiecondition considerabryeasesthe di..y ; th;;;;;;;;1't" ".-;;L, volugeavailablet is-iclatively.v"ry r'_"[ "*,ro lrc is the more readily achieved. ""d;Ae;ffi;; .Fr.orn-fig.. 53 the-true current during arcing may be . esilv

dcernined, giveathecircuite.m.f.andth;r""d;;;U

#a

ilrrfrri ea_arcvoltage characteristic. actual current l. wilt be the -The difrcrencebetrveenthe- no,rmal *o"rrt-r, which floil *t"i tfr". .,backwarJi breaker-is

closed,andthe *iL*t ,;, intt e circuitif thearcingroltage,"pro"oiJif,J *n "l-,'r"finli ,.fi".*fi;;; IT "yrt voltageand;'backJardit;;;-; rehtedasindicated. in thediagram.Summarizing, we-"y oy tlrt t" g*;a;;;;;

isgontroltedappreciabtyby.ihe;";iJ;h;""ild;fi

{r

controlledover the.remainderof the half-cycl* Th;*"*, hhs

;ill:

G; i, fic. t.tf";;tffi.-'leenexaggerared Cgac-itanJg.The*"airio* shownin fig.5.3bavc .^^_P_1ttr"J oeensontewha.ldrlized

for simplicity; in practicetno?" cenainamountof c-Facrtancein parallel *irh.;r;id.;k;?;;

fu;;

to (a)thecapacit"""Jt * iitt" mu andonnections i3.j.ej* onthegenerator sideof thebrea&awhenaneartl r""rt]ol*"- oilil ddu:"g (Dltp *pu"l.** u"irlccnli sonact l3l andthet"nk untbanoil circub-brealcr.. FigaS.&rpro*t, capacrtaae "i-prkr*ltt c, tu t'ilara " stuh *i.t thGbrca&cr queto +" theeanhfrn* on the"F;;; lmd side.itff".ry i_**;ffi 'ot parallet I*:t .o

capacianciare its innuenl on the;vorage acrocc tac gap aft r corrTt zero,with.wniin we shsll deal later on,

th. *ro"i j; b"f;r thezcro. is rcachcd. S,lrltn:. "ry -8. "r" ).) shosBttre circuit_of . S.l in eimplified for*-T;; !g. omple equations rqrresent the b€heviorr o{ t[c currcDts and vorqges"I.e*,;. .-,i. *r"o..i; t: ;6"; *rroo. (thlougb th3 G

EwrrcgcEARPRTNctPLBg lnductang); l], - cafacitanoe curr?nt; c ' E sis,sa - aryptt c.m.f., and c. - aie voltsge. Ttc arc dl instsatsrcou! velucsof the variablcs. Thc equetionof thc voluges is: c'- L "dinldt 1- c, The equationof currcntsis: i. 1- in .- l, The voltagcequafon nry slso b€ wittaz.ttri.ltlt - (c -. c.lfl. eC

(

F4.5.4. Capacitances h ?oalh, @irhc*adt-bteahaoc Let us study the eventsoccurringfrgm the time when eobeginl to riee appreciablyastlrc current approacheszso. The current and voltage curves in fig. 5.5 are shownnfor clarity, on a somewhd expandedtime-scde. As in the previoussimple exampl.e,the riein6 arc voltageforcesthe main current t,oto collapsemore quickly thm if the breaker had remained closed. This is clear also from tlc exprcs.siontot di-ldt, as e at rtris period of the cqrrcnt cycle ii ncgativeand the arc vohagee. is addedto the generalorvoltrgq d giving a iorreslonding increasein the rate of current fall. Thl rising arc vottagi also causesa divcrsionof currcnt into the capach. ance, the value.o{ this current at any rnstant bcing given byr 'This io - C . dc"lilt. current is positive(clockwisein hg.-S.S1 d4 sincc i i; it is obtainedat the expenseofthe arc currd which therefore starts to fdl more rapidly than does ,L. As th! instant a is approachedthe rce of fall of .li" diminishes eincei, it nowdecreasingwith eoapproachingite pok value. At the instanta . *rle 'Immediately :0, r, - 0, and in - i-i this condition is illustratcd in 69; 5.54. afterwards,the capaciance darts to discharsd giving a negitive (anti-clocftnuise)currcnl: The 'nain and rrc, currens Tg podtivc,s9 rtrat the ar! is eupplicil both by tho "til by the capacitahce,and ri is then gdi* than {,.. At t[c :ource Td. rnstant, the main qrrrent r€ach€ zeroand the arc is then eupplid .eolelyby die'capacitance, nrr 6g. 5.i6.' Aftci thig iirsisnt td;d"

-.,i.

FRlNgIPLES OF CIRCUIT-DREAETNG

)Current wftlr 6neeket,closcd J \

Arc caneat zeFo

eo(pek) Yoltogezero

Ln

( r

"tl d (q) $) FiS.55. Efrcat of palWcq.ibnt

oruc aninailw!@a.?o

.+r:{iz!a !

83

-::r\

84

switcllcEAl

PRTNcIPLEB

eir,encurrent rnoybc much lcssw.hentac eurrenthasbccndecggasing lpiafy ttran in the static casa Tbig is on ocoouttt of thcsmd i-".t"roi, in tle arc and is obviously a disadvantagein circuitao il;eakirg. Fhwever, the static condidon Tly--bo-a-pproeched.to the actualrate of fall of the r.'trent bcforc i"F"" Jing upon (")-and ;.*, (b) the el6ciency of dc-ionizCtio.n coti.iaite4 tf* io**. We are hereconccrnedwith'the erulent zcro insterg 88d ttre PgrS immJat"ly afterwarriswhcn $e gaP-ie subjectcd e qt cirsqit soltage. Any mean"which cal bc e-iii'loyed to. pcrsuade*".*:.' posible satic conditionsat this stsgearedeeirable" t ."ily '.c,rch-as "" capacianceis evidendyfavourablc,sincg-rtv!.{y The efect of parailel rnuch reducestheratc oi fu11of ar" c.trt"nt iW beforczero,allowing the arc temperatuie and conductivity morC time to accommodatc ttemselvcsto the current zso condition Note that it is the arc currentthet is apprecirblydtaed'in fotn, rather tfian the main cilrrent paneingthrough the seriesrgct{ce, andthe questiouof high inductivevoitageodoesnot neceaoarit-v-arise. Neverthcless,the riiversionof current from the arc increaseethe arc voltage (owing to tall of temPereture,ionization, etc.) eld its ratc of rii which in turn increasesthc cepecitivecurrent. The procm tlius terrdsto becomecumulativl, rail in oar1docircumstance'to bp sturtierllater, canteadto the production of ovcrmltrge. Thc extent to which current ie divcrted from t'he arc during this pre-zeroperiod dcpdndsupon the velue of rapritarcc ard tbe rate of rise of arc' ooit"go. Whcre thcs€two hctors arc appreciable,the effect on tbe arc tirnpcraturc and didectric *rength of tbe g{p 8t zGromay nrake current intcrruption comparativdycesy. The ratp of fdl pf surrent it zcro rvith asymmetricalcurrcnt ie generally lees thaa with the cquivalentq'rnmetrlcat current.,and therdolg in accordanccu{th tte aboveprinciple, Bomecasencntof brcakerduty msy be obtaincd. wish the fcrmcr. Re"etrikiog Vottago" Tte volagc available at thc circuit' breakeret a current zerodependsupon Powcr factor, aqd thcrdu€, has somefinite nalue, unl€s8tbc circoit is ri5tualy icaistivc. Ttc way in which the re*cri&in! volagc risec fron ztro (tnthc qrq€nt zcro instant) to the approprice tzlue it of thG gr€stcstimportane" thir voltagcriec would For instance,in o circuit sithout bc instantaneol&snd therc would bc litdc chanceof any circuit' breaker gap having aufficient'dielcc{ric ruength to witirstand it without breakingdown. Forauratclyoucbconditioneerc hypochcticat fior thse io alwayeeomecapecitand.ipiercnt in thc circuit" Thc

.t,\ -r':',

)

PRINCIPTES OF CIRCUIT-BREAKINC

8'

cffcct of this is to modify the rate of rise of rLstriking voltage (R.R.R.V.)such that a few vital microsecendselapsebetween Strictly zeroand.the cstablishment of a voltage of any significance. ",rrr.it speaking,there are two conditions to consider: (a) breaker and capacitancein parallel, and (&) breaker and capacitance in series. 11'eshall examine both, although the former is of much greater importance as it is nearly always associated with fault conditiong. Pnctical systems are usudly much more corrrplex than the basic circuits referred to. in the following, although reasonable eimplificadon of actual conditions may permit a rlseful mathematical treatment usingthe same principles. Alternatively, itudies of rate of rise of rc-striking voltage may be carried out with the aid of a network rnalyser. A simple example of parallel capacitance is seen in fig. 5.4 wherethe line fault places C1 effectively across the breaker. A pair of voltage equations is needed which, when sblved, will gi're the voltageacro$ the breaker after current zero. Thus:

e:

L d i ^ l d t+ ' l ! C x ! i " . d t .

and ll@ I i".dt : r (i* -

i")

wherea is the generatorvoltage and r the breaker resistanceat z,ero. 0n solving them we obtain three possible casesaccording to ihe valueassignedto a l. If r is made infinitg and assuming no other resistancein tlre circuig we have o (4crossbreaker) :.e (l - cos {Ii@ . tl. In this hypotheticatinstancethe voltage rises-from zeroto iwice the tnstananeorncircuit voltage and tlen returns to zero and continue .toccillate in this way about the value e. 2. "lf. r is less than infinity but greater than the quantity:- then the cosineterm in the expreseionin ca6e ! !4e (in-clr) l. ts multiplied by a factor exponential decoy. This lmply meansth* the high-frgquency gecillation is gradually damped row irnd the voltagerapidly setrl€odown to thc circuit rblue a. Thc freqr5ncyof this edlation is giien by: f/(2zr) x {ire and ie usually puch greter than the normal 50 c/s altrirnation of thc

!.ril'

86

'

' - '

SWITC'TGEAR PRINCIPLT.S

i-!H;f;!i:i,"{o"

,("

Cun"ent zeto instant

(e)

5O'c.p.s.

G:Jt!{{}-

____s_\r

,(

zero' instant

ft)

Fig.5.6. Rllrfihiryaoh4c(a)aith ldgh-ttc' @d(blarfrcarydetFl gercy &tin voltrgg Lc. r feult bndition hving zcro Pows factor. The ftit pcet of tbo trrneicnt ccillrtion doeonc qtdte rcach lpicc c on rccouatof tbc danping cfiect of bleakcr sesknca 3. Thc hst cssciBthat in which the breakcrresietanccr is lcc rtm thc quaotiry* {Ti-C,eAqilth thir cqnditionthc volagc ir not osAUstorybut instcaddsesloggithnicatly to the vdue c ar.rdSg follorrr tho normd 50 c/e veve-form. This is iltustrotcd in 69. 5.6bandthc rc-otdkingvoitqgcis saidto be logarithmi;a[y d+s,p"{ Now lct uestudy an exanrplcwherethe caEaoanaoig tii scnt with thc brcakcr. Fig. 5.7shourcasimplc circuit in whicb wp ns$Clc

PntNctPLES O? clRcOrt-Bn8AB8N6

gt

noftult' b't imaginethst thc brca&ccis to opcawhilc the camcfuive is flowing. We may faitly neglcc G,l thc rUetlvely maft ryt ridcrablccrpaciteneto carthof e longtnusnisdon litia This lcavcr ur sith e rinpte sc of inducanq-capacitenccand reristancc(of the brcalc^) in rccica Thc sppropristc voltegc cquation L: 0 - L di@ + ,i + UC x t i.dt.- As bcfolc, tbr; cerer ero obtrind &on'thc rohfiion, but with r vcry inportrnt itiffcrcaco &omthc "pniall€1"ci!€:

cr>c, Ffg.5.7. CcpacitaueoJlng narlltt&rion fuu sai6drhbcohah

e'

-a(1 -e.u@4 l. If rbserothcvoltageformie:o 2. ll t is greetcrthanzerobut lessthan the i"*tity 2 {re

thcn the .cosineterm is multiplied by an exponcntially doying fr$df. the volage ic - 3, lf tiegeder than thc quantity Z @taf,n togadthmicaltydamped. The foregoing discussionshoya that the form of volage rieo dter zeroig infuenced by the value of the breikcr rcsiltance stzeio in rdatioa to the drcuii inducance and capacianca. Fdrthcr, in ordcr to daop out the h.f. ccillation" a rninimum rralue of thc btca&crrcsistanceis reqniredwith mries capacitanceand a rnoririum vduc with parallel capqcitance.-It will be secn tater that such $.pitg ie dnmt alwaysdeshable. When this is not pcsible we mut take account:of d most important factor, namety, the ratc at thigh the voltagerises up to its fiist peak; .This is clearly'dependentupon both e:and ihe natural froguency of the circui! which 'fhe. ' ie vcry high for lorv'valuesof epacitance and inductance.

- +

88

swtrcgcgAR pRrNcrpr.Es

rate of rise of volage may be high, thercfore, et e location ctce to t t|re generator. 'transient", ..inherent The tcrm restriking often met witlr, aimply mgry that volage trensient which. would bc obtaind on .c@unt of the onln without nodification by the breker, "it*i i.e. when r is infiirite with parallel capacitance,aod ;. with scries capacitance.In practicg of coune, there is always4 srnall esurunt of resisrancein the circuit itsdf and thus the inhgrent transient iB of

thc darnpedoccillatorytpe. For simpliciqf*tl";;;;G;i factorwhichb T-: lfllg u-gor,the vatueto whichihe ""o rises{fter zso. The initial.'anpfitudeof thu t"""d;iil"lu.tio""olgge i, notin fact e the circuit volage,but a + er.yherc c*(sie ig. S.A *i* ir is the pre-zercarcvoltagepe"k. This is realaity is remembered ""a6tA a"" tiin, that the cciltationsof tjre transient rapid- interchapging of energy between thc ir-rductance "r *a- tuo

capacitance,and that-this oscillation, therefgre, u"Ein, ,"Lul the ca_pacitance "iontty is chargedto its.maximum just.UlforJ ,;;';; a in fig. 5.5. The amount of energy in,thJ op""itnn"" instant'manifets itsgl.f a1 a vottageidditioo, ,; tb .h;*"r"f "i,U, oscillationinitial amplitudea f'he student ihould alwaysappreciatethe physical aspect of eu-chphenomenaand nor be conteni with tt ,nutir"i..ti""r;;;;^.tration. In this way insight and intuition " utti_otdJ-u"q;r"a In thepresentinstance, kiepin:gin mindtr,."r" iaqoi ii.eilnliii

energyintcrchange to occurbetweenthe circuit

-

ro, iii" Aity

seenthat, rvith serieserpacitancethe breaker reistance "t"-*t", rr.iu; tf

*ry dircctlv absorb_ to prevent gncrgv ltstr similurly, rvith para'et capacitance,a veryt "-'"Jd';f-;il1;"sr. tow bieokei ioi.ao*

will prdvent sscillation by_-diversionof the €nergy. &ilical Resisttpzce.Wnh parallel capacita-rice the maximum

valueof breakerresiltance y[gb-will preveit*ell"tiool idr"g try volageis givenbyt {If1,pd inttre ecriescasethc-nioinun ralue.is ?1/L!C. Thesearetermedcriticalraistancq.A; b"* seenearlier, the series genelfly -", *n* o$ir[;-d*ryir missionline wittr to carth. rn such enditionsthe criticar"ppr"ciatrte-crpaatLl gi"tance lr ,i.tii. io*-r00 ohns or l€sc*nd g€n:rauylessthan the curitnt roi*"o". of a circuit-breaka, sottrat "iio

criticar&nnpingis rrsuarypdseiure. rulr *liiitilF"uy desirable, however, secpage:l@, f"i -o* a.,"ff"a?t.fr" :apacitive-cu*enrT=nttr*.tr,*i{uiripal*[i#*.e, " "f ti:cbrcaker

is notalwave ableto oni."aa"-fdJnii.uiJr-ra,

,a . i-t

PRTNCIPLESOF CTPCUTT-BNEAKINO

&'

I

89

cxasrdleof a fault to earth on t&e line side of the b;caks in thc snren shown in dg. 5.4, the paraltel capacitane is Cs whilh is .iudry small. The &itical resistancethen may be of thc order of 1.000ohms. This may be more or lessthan tAe current zero resist'enceof the breaker, depending upon the Erpe of breaker,t$e fault orrent magnitude and the circuit Paramet€rs. (lrrrent Zero "Pause". Thc upp€r part of fig. 5.8 showt a typicalrecord of arc current, arc voltageand cirorit voltagein which tso current zeros occur. Thc first invoilves breakdown and resuikingof the arc for a further half-qde atthe end of which the gap withstandsthe rising voltage; this is seen to be of the.ccillatory form.beforefollowing the'normal op-en-circuitvoltage wavc. With such a record the arc current aPPearsto Pass straight through ,erd it-, the lrreakdown case(a) and to cohe to an ab"ruptstoP at the ctarance (&). The expanded records below show that these curreirt is forced may'be false. Just before zeto ardnrearc impressions down abnormally due to the pardlel capacitanceeffest The arc' reshtence,although finite, is high enough irhmediately after zero for the currcnt to remain very small-a frastion of an amperefor a periodof microseconds,or even,tens of microseconds,bdore the , breakdownat.which the curreat resum€sthe normal sinusoi&l form. / This very small current existingbetween.zso and breakdown-'+ften termed"pre-breakdov'n" surrsnt-may adopt various charagteristic forms. It may exhibit a small peakand thcn decreascsliglrtly.before ggg.rapi$y on b-reakdownof the gap. qis paniorlar form rnight _be'6btainedwf,iri a powerful de-ionizing action exfutsbnt where thc t,' rateof. rise'of re-.trikit g volage is v6y high. On the other hand, . vhere little de-ionization is achieved and a very high short-circuh ctrrrentb involved,the current may passmore or lessstraightthrough zerowith the breakdownoccurring at a low point on the re-suiking. vokagewave Clearly it is not possibleto generalizeconcerningpre. breakdowncurrent form, but the idea of a errrent zgo pauseistnown to bejusafied uodr q'any conditions. - 2e. zero Ein fig. t.8 th" arc is successfutlyinternrpted aod in tho" lowerdiagramwe see'ihat the current does not ceeseabnrpdy but continuesin the r&ersed direction'after zso fot a shon pciod of eomeinicrosecondsbeftlre frnally ceasing..That "pet arc" .cuircnt Soultt erist is only to be expecte-a sincetEe ascresistanceisrstill 6nitc 8t crirr€st zeio.thougb'rapidly'increasing, and such cufrents can actuallybe of the order of an asrp&e in certain cirirrmstances. Noticc - 'thai tlt pre'zero arii vottbg" pJ"t.t zer6 g,-'.slpupper diagiaic-iir

90

SEI"CEGEAN

PRINCTPLES

Anc curnent '"

Itcvoltaqe peok -

Arc 'twltsle

rAne \ voltegg' \ ,v.ee6tt. ,

PEEK

I Cincuic I voltafe

I

Sreokdowi'

.IVormct s/t*r-tq/"to

?,"9 t)

showin$successire ca?"ent -ze leedind to- 6neak{owh " eny' 'cledronce" "os

\4rc

cunnent

oPist-erc"

(6) Expo ndednecordshowing curnent Vffij,;;r,!i:W Frg.5.8. Cstat zoaoe-forns nca sero smaller than the re-striking voltage peak This is a result of thl increasingcolumn resistancedu:ing the zero period: Further, iq ts assrrmedthat the de-ionizing effects have becone more pronounceC at zer.oD,and the arc voltagepak ju.t prior tg thig zerg is sho'i'n dis' tinctly higher than a1zer6a. Accordingln th9 pre-zerofcrcing dow

P R I N C T P L E SO F C I R C U I T . B R E A K I N C

<, .i

9I

of arc current is shorvnto be more appreciabreat D in the cxpandcd diagram. De-ioaisationat Zoo paase. Any detailed considerationof the proccsesT y"r\ Tti" the arc "residual corumn" tn" p"ur" 3rcnot onlv heyondthe scopeof this volumebut are "tout of""ro phl for the very good reasonthst much has still to be clarifiea o*iic t" th" glr€Te nmplexity of curent zero phenornena. Neverthel""* ,h; follo*ing broad outline should be useful, bearingi";;d th'i th; pfenymag vary widely, accordingto the type of 6roker, f"fi-";: ilo'de and qr€uit conditions. \4/ith any given circuit-breaker tf,c mostimportant factor at currcnt zero is proiabry tr," -o"l"o,pu*tureof the residualcolumr\ sincethe amo"nt of lonir"tion"i;fi:;" wr,v critically d-ependentupon this. Agir, th" *or,t'""ro-tlroper-sturegenerally depcnds upon the rate of current decreaseiust beforczeroan! r9eo3r the amount of cooringand de-io;i;ti;il ril;

ti: rh9f period. Thus,at currentzerowe t uu".ioiauJ;!";; still subjectedto de-ionizingeffectsand possessi"g * coretemperature-bdieved to approach r0.000.K ,ir,a.r "ppr;;;li; rom" .on-

ditions-and therefore co'tainirig- considerabledensitiesor etect orp andpositiveand ncgr,'tiveions. The rising voltageafter zeroa""a" L the energycontent.ofthe col.rmniy cfJetdng the i;;,;;, ilT.f. '*.l.oy tendrng to maintain the tempera.tureand ionizationlevels. eventually lead to breakdownancire-establishment 'I ^:1t.^p-y1may he power inpuj :t_1|t::* !o the arc is, however,"ery srnullat firet is tow, and the continued cooling, by externalmeans :31".n: :oltage sucn_rs uast, natural convection and conduction, may predominate

andfunherincrease thecorumnresistance ro it,'t it.#ilffi.g" l:::TTn""aiiqlv tessefiecronthet -pl."r". Sirin"iiy,itGc i: lt:ilg rapidly lengthened the rate'of energtr input ltlcngth 6er unit

of thccotumnduelo theris11gvgltage*ttt#;ildrd;i; reduced-suchconditionsmaywe[ ieaato'successru aea"inceiffil

Qer6'ylossescontinue to preiominata .. This is aot necessarilythecase,howwer, for we baveonlv con-

sidered the immediate posr-zerc periodr-.-t"-p'otffii#;

otetgycontentof the column It seems, very probabih th*;61"; can still ocsur, evcn wn* tUetelipo"ture hasfallen onL:fO*:* account of energy losses,provided there is still some "rqcraDry-'oD cegre of.ionization^rernaininf-when tt r;cilL-o ;;;: ,,r{n lcvet. The genernlproce$ of " "oft"gu brskdonn in this :::l) i, one of rapid ioniration by collisions betneen eleGrons ""ht il'.i:rentti' of ..,8n crr€r6Tand gas particles. . Little irj knorrn about the poeci;

92

slvITcTIcEAR ltRINcrPLEs

rnechrnism h.t it seemslikely that-it nray bc starterl by i,small num_ ber of elccr.rns-ha'ing liigh vclocitie derived from ihe n"ri ii .i. 'l'his 'f

gapmaybc akinto spar-k"type breakdown, asa;rti""iiroi the " thermal" qrpervherethecolumniesistancefallscozt;""o"rti- iri to in-creasing 9ne-qycontent. Tlesetwo generarror- or Crl"t-to-*-o distingui$gd are by the form of voltage-collapse, asseen;;;; muchexpandgd timescale(sea fig. 5.9). T'heispark" ur*t.r"*ri cxtreryelyrapidandthe voltagemaycollapsefrom severalth.usarrds

(

'TVermol"6reo/<down taermol-hreal<down

Spork"breakdown

F;g.5.9, Eutr7bs of beahdozonat cTrtrentzero

of votts to atmostzero in_ao more than a microsecond. The thermal 94rc, on the other h** p generally ^ very much slowe, oro& qlibiting a-quiteroundedyhag" &ot misocconds. A^ *y be imagi-ned"i""tu""o ""i ;;;j#d^;;, appar to exist in whicha prooess startsasa therial rF*a t" ;;d;-uyip*r formation" and vice versa Bcsidcr the simple th€rmat and spark breakdown mechaniems

outlincd it is orobalrcthat

p-"o.o

may bc important du;qg the cTrref zerg "*r"*-J,iiii"ty pcriod. For instancc,the cxistcncc

of a hightempeiaturc in thc rcsiiual;;G; riderabh photo-emissionof cl"d;]";;

;;;;ffi;;;

.h. cathodeif the radiation froin thc gasto thc erectrodc t htgh l;;b .yp" prfr"',il. cethodcitsdlf ma1'havcvarying "r b thc.state of io "ClJ ""riging

PRTNCIPLESOF CIRCUIT.BREAKTNG

93

gfaoe and the purity of rnetal. The attachmentof elcctronsto atorns .o form neg'ativeions ocgtris in orylen and is, ih cfiecq virtualtv a rbionization processsincethe mobility of such ions is relatively verv bv, Energy is given up when a gas reverts from the atomic io thl .aolecular state with falling tempemture. This ..dissociative reccabination" may be occrirring to someextent during the zeroper.iod mdthecnc.rgygiven up ma5rcontribute towards ionization. Ali these rind others, qrobably occur during the critical period, and prooerees, oal4rge extcnt the problem is to decidewhich of them predominate 'mder epecifed conditions. Much researchis being divoted to the .understandingof thesephenomena It is often stated that the outcome of any current zero period drp."d" upon thj "race" betweenthe rising voltage and the'rising didectricitrength of the residual column inthe breakergap. This is in gentral tnre, but'may be wrongly interpreted. It should not be uten as meaninq ttrat the dielectric strength and appliei voltage increase independently of each other. For insance, if the energ;,-'balanceprinciple is applicable to the 6eturnn at ?.ero,the rate of changeof tlielestric strength, or resistance,is ccrtainly very much " afrected by the rising voltageas well as by energyloss. It is useful to :notethe definition of dielectric strength of the residual column as grvenby Dr. J. Slepian: the dielectric strength is measuredby thc ' roltageneededto maintain the reistance of the column constant. . Thus,if V, W, R arcthe dielectric strength, power lossand resistancc rspectively,IPIR-W. Thet is, the power input and lossareequa! andthe resistanceis steady,or V - (Wn'f'. 'fhis definition treatst'he columnasa simple variableresiqtorwhoseresistanceis somefunction cf the enerry content. It musl be emphasized however, that thc phenomena associatedwith the current zero period now appearto bc muchtoo complex for simple mathematical.treatment Although uncertaintiesexist about the precisenature of current ztro processis,we can be sure that time is always one vital factor in thepo*-zero period- Consequently,the rate o?.ise of the voltagc lltcr z€roaiross the gap may exercisea consid€rableinfluenceupon .theoutcomeof t.hezero period, and.it is b€ause"bf thii that somc rpacewasdevotedto ratesof rise of re-striking voltageearlier. ol De-ionizatioz. l'here is an aspectof ac. circuit. Efieaioeness not.alwaysfully apprepiated. I'he importanc€ of powerful ItTUing ae'ionizationaction at, and immediately after, current zero is beyoad.

,. -.*-!.d.-+.--j

94

swtrclIGEAR :'lttNclPr.gs

Interruption of the arc current, arrdde-ioniza *'ell es unnecessar,v. tion, are not regriired at any other irntant er(cePtzero. ln mc rnoderncircuit-breakersthe de-ionizing mearurare neccrisarilyappliec noie or less continuously; hence, liberated from the arc are largr arnounts of energy which have liale efiest on the de-ionization a zero. The stateof afiairs at z"rc Ccpcndsgenerallyon conditicns ovet a very short priol period, becausean arc is generally able to adju* itself to changingconditions extremelyrapidly. A highly efrcient de ionizing force applied, say in tbe middle of the current loop, merely meansthat more voltage is required to maintain the current theq while much mostly uselessenergf is ideased which must be safel; disposedof. Thus it is that de-ionizing meanswhich arc a functioa oftlre currcnt, such aselectromagneticdevices,are not necessarilya eft'ectiveas might be supposed,for they have little br*rring on e"int et crtrrent zero, tlloi is, when most needdd. There are excepdonr however,where the energy liberated in the middle of the loup may have a residud effect at current zero, and these will be met later. "Cut:ent Chopping." The common phcnomenon knorrynas "current chopping" is an undesirableconsequenceof the genent irnperfectionmentionedabove,and is alsoa causeof overvottage.U a brc:rkercxertsthe samede-ionizing force for all currenB rvithin it short-circuit capacitn then this force must be great enough, in variably, t9 givc intermption at the highet current ratin!. Ideally, ir wo.uldbejust sufficientfor this but not high enoughto iatrse uniuc distortion of the current wave. Ilou'ever, whcn such a brea&er! ' called_upon to break,saythc normalroadcurrent, rvhich may be les than 5 pcr celq bf thc maximumbreakingcurrcnt, this smali curred maysuffersuchdistortionastobeactuallyforccd straightdorvnt. zero from a relatively high value bdore tne nat'rat zera. l'his is termed yo-! chopping.The idea of :i breaker with a constant de-ionizing force hasbden invoked to illustrate this poasibitrty,.but typea wttici grod119e_varying {egreesof this fqrce in tneir operation arilnerrertho lessllabLeto "chop" when breaking cgrrents.. 'fhe bfiect of r "-roll pracicauy instrntaneouscolrapseolthe arc cutrei,g.even ofonlyr fcw am:peres,is-potentiall-vvery serious from the p"l* overvoltageswhich may rcsult in the systern- An example"i"f.*'a will illus, trate this: a 220 kY oil circuit-breaklr.,internrpting ia tr-ansformct maSnetizingcurrent of ll A'r.m-s. chop.ithis rinrrit ut.in i*t o. . taneousvalue of ? A. The naluesof inductanc" u"a c"p""iun& io

cirolitye 35.2_H and,0.0023 pP;n4 ;"*t"c.d;U-.ilil;; enirgy is transferredto the cabacita*i

:trrrb;;;;

""roo

;il;"d.

f'

pRtNclf r,E80tr otRcutT_DRSArrNO

Curreot fiaollu suppnessed66fone ndturol zeno--.

f.i

'ilbturol' carten€ zerv -!..+r _.Gl

trnre

Anc

I I I I I

cul.ent

I I

Recovetu taanshllt

-.i I

I I I

Anc voltoge

Ancvoltogep?o* , ,

i

iiii;

iii# iii,,t L_

1 c3.

i;fbir;m$:"-rrl'* 'f"/

choPPtng

t&.sJe. Effsorexadcio4fuS

.

ffi;ffiffi;iffiffi*sffitr

tclado 3e rh/q phcaomcnon ;"4 ;.fi" .no$s

to.

of 6g. 5.fO, u,htch a sone*n"t "id wcmryrcc hor "irnpunJ_ilrii"ililenent, hgyo \Faencurr€ntchopprng

T*T:*g

sctsrn.

r"nil*"if ,:f#'#"T:lTiiffff Le*rlffi qrrreat' At a cenain rarue;il;;;" *rrent,

'!r onaccouhr inrtab'ity rcr "titi&i or-tn" aisptopoiioo't ri'i"?gl dc-ionizingforcc.and

96

swrrcrcEAR PRTNcTPLBS

there is a.l drnost instantarreous collapseto zero. This is ohownin thc diagram a.sthe lirst chop. The current in the arc at the tinre wrr florving from the soureethrough the inductanceto t[e brea].er. Inductive curent cannotceasernstantarreouslnand the actualchoppiag of the arc current is, therefore, acbompaniedhy a simultaneoui diversion of the main current from the breaker to the capacitancc. Consid€f,what must be involved when such a rapid collapseof arc curent takesplace. Ifthe current choppedis i. thin the voltage os ' the gpacitance rises initially at a rate: doldt - i/C (neglectingany small capacitancecurrent existing befogethe chop). But this voltasi is alsoacrossthe breakerand acis in the sameway asthe normal pte. zero rise of arc voltagein incresingthe rate of fall of the rzaft current corring from the sourceand through the inducance. The maxinrum possible voltage across the cpacitance and breaks is thereforo obtairredwhea the main current reachs zero; that is, when atl thc electromagneticenergv associatedwith the chbpped current in the in
Te"n?, a1a 1!!etter re-strikingvoltageurayresule This b nd eop" ;y dxrq *nowourrtit e :hoyl io !S1.5.10.Successrr." M *9p btihgorhc currentio z zixo.pr6..turay""*ith ;; i"nno sincethegapis now,inan advance stag6of de-ionizatioo I!*l*fog rEwtu

be-appr€crated that surrentchoppingis reallyan extremeforo or.thc nhenonenonilustrated,ill 5.5. The fottowiig genenl poins shouldben6ticcd: ' I "- 't.Fg. l

a

pnrNctpt.nsoF crRcutr-BREAKtNc

( t

97

l. Thc maximuininstantaoeous current whiqhcan be ctropped given a r.m.s. current by-I particular breakeris increasedby sith thi amount of paralld capacitance,sincethis cnabtesari. increasing to set in at a higher crurent a ?*lrois approached. instability 2. The rise of voltage when the currcnt c6,ttap.esmay be damggdby the ellects of eddy currents a'ri hystercsirtm in sinal transformermagnetizingcurrenL. 3. The risirig chop voltaggmay be limited by.the circuit_brcakcr in two ways:(a) the breakergapis generallyabti to allow re-striking beforethe pe.& of the prospectivevoltagi is reached:and (D) thl' rnodeof operation of the breakerlimits the instantaneousvatuc or curent that it can chop. 4. \l-'her- the cap-acitance and inductance involved are fairly snali,the rate of rise of volage is ftlst and re-strikine occun at quiie l,rn'raluesof the voltagg sin-e the time for further?e-ionizatiol oi thegapis correspondingly small. A consequence of this is tlut.thc rcsurrence flegucncl of successive chopsmay be veiy high, with thc possibility-ofa dangcrousresonanoe condition being sri up in the system..This is often the most undesirablefeatureof current chop: -tf,e ping,since although the prospectivevoltagesas.sociated witt choppr.ng are not generallyreached,overvoltagesrnay occur due to suchiesonance.yT"lt choppin-gphenomenonis an interestingaspectof ^, Thg thethcrmal hybteresisefl'ectin a.c. circuit-breakerarcs.'.Supposc, tor _irstance,tJrat therc were no such lag effec1and thit in arc couldadjustitself infinitely quicktyto Til ili; ",rrrJnt-"-n"ig* mernthat a circriit-breaker with Lfficient de-ioniziig meanswourd aluiysprodrrcechoppingn€ar current zero, irrespecti-ve of tlre r.m.s. sal_ue o[ current being_broken In actualfacq however,the nonis not obtainedwhen the r.[Ls. current is increased ihenomeubov romc criticalvaluedependingupon therbraker and circuit. This shorv! Inrt $ suchhigher currents tlie tenp€rature and conductivity of thc arcrcmaintoo high for instability to occur near?.sro. {.lapacitive.currest Breaeing. There is another @rnmon imp.rfecdo" j"E;i;: 1our5 of overvoltagewhich is dug-to "n I*|9r behaviour,and the conilitions hereare'thd oUtairwawnen capacitive.cuLeni a1 for erample,in the openingof a long ::::"*i,lq uul.'adeil transmission . 'wili fine- Sggh a lini, although lntoalea in thl

l"*t1l scnse,

aduatycairy ,;"tt- ii;;il &; i;;i " on a9c9u-nt9? t!" op-"cilnc. to card. :urr.enr

lill.llt". ur rtte lrnGitself,

even although it bb "ppieci"bte open at the far end.

98

srvrrelrcEAR PRINctPLeS

Fig. 5.lt showsthe simple equivalentcircuig and the voltageand qnrent.phenomena theoretically possible when breaking a small capacitanceqrrrent The latter is assumedto be intemlpted c instanta whenthe circuit volgge is at its peakvalue Zn,in the positive dir.edion..'This dfectivety separatesthe ge,netatorside of thC'circuit from the line sideand, sinceit is achievedwhile the commonmltage is of .magnitude{ Vrs,the unloadedline is isobted with this positivc voltage upon ia The circuitvoltage cpntinueson its nor. mal sinusoidaloourseand tbesetwo difierem voltagesexi;t at the respectivecontactsof tle breaker. Aftei in;stacta the breakergap ir -the theiefore zubjectsdto theclifiaanceof voltagesV, ard% ti*, aftcr onequaner-c,ydeperiod from a, the circuitvoltage haereac-hed ,:"o -"od the voltage acroesthe breaker is theo Zrr; thereeftcr the circuit rcltage Z, increasesin the negative direiiion and tic voit4geacrossr"hebreakergap becomeseven greatcr und! at i is rziue is tvice (u. ,L.suae aow that this abnormallyhigh irressing of the gapresultsin re-striring the arc. The two previously-ssparated parts ot the circuit wilt now be joined effectively,b-van arc of very low resistancc.The liue capacitancedischargesat onceto reducetlii vottageacrossthe breakerto its appropriarci"gligiUl" value,and this is accompanied by thc tamiliar h.f. ciscillationset up when.acapacitor disci'rarges in an inductive-capacitivecircuit. It is most iriporant to apprecia!9tha! the zero,aBit were; of the voltageswing isnot erth potentialbut the circuit voltageat the iastant 9f discharge"l.lnrs sc see,tlt.p.yill swing right down to the value -3 Z*lbelow earth potergtid),siacethe arnplitudeof the voltageswing is? i/,o, neglect:ng dampingefiects. l'he restrike crrreni quicHy ,octil it"-nst zero, sinceits pciodicity is in accordancewith tire cltcuit oatural frlquepcy, 3"q-*iU prcbably ceaseafter this one half-cy.cle. Thc voltagson tAe llnc is noy -3 Ve nL once ag?in,the tsvi halvesof the ciro-it are squ:zted and tlc"line is isolateda this poteotial.. At this stage-imarediatelyafter L-the vbltageaaossthe'brea&eris, of @urse'only-twice z* since'thecircuir votage is itsetf ai its instastaneousmaximumin tbc aegativedirecrion" Tte poiential dificrenco acrossthe.breaker.gp*otii.*to increase,d;;, * I/, b"*; lessnggativean{, aI the instant c, th:,eetressingreaches4 f *. $,th, Sapshguldbreakdown againat this point thd the events&*rrlog .one-half-qydeearlierareiepeated oo * eoeo-or" fo"-iarSi" Ji the voltage suriagnill_now & I Ve ana tn" rine may tien Uete$ 3s rsolercdat a potcntial 5 Z; abdveeai.h vlhFnthe trans-ient,e_eciilc cuTclrtertiguisheqiustafterc.. - . .- .':.-. .'."

( l

PRINC.IPLES OF CTBCUIT.ARAABTNO

99

t

76xYgp \

I

I

\

/t'-i Dze \

-

r

.

r

t

{ ! \...-./

L:J

eE. asxTgpt

Capocitivecunnent6eforcinten uption lnonsientcurneotof resCrike V. oltoge to eoethongenewto" sr.de of ilrcokcr.v_ Yoltogeto eort;hof trunsmissionlinervs ' (g) lnternuption of copocttire caF?qrt (b) firc ri*nike

i"i sririili*;r" lUo.e*-n"orunl r3. 5.ll. rrW

ohenbicahhg

6-y nhenomcnonmay procccdindef,nitctn lceving _-,^-1o Yoltage on thc $e line of nagnitud€si V*l V*nd - oo. Th-ri rnd practiceusuallydifierihoweve, aild diondb"=as hcrc. thl PTTcatfartors'whichprecludca theorcticallyixtsibic conditionrrc utessilg. Thec limiting - factrire arc asrociccd rrith catain 1 sT{o* mads namely:

**ff 'l'.l"'ffi .H;!*",:ff*Jnf'[tfru":,Ifl !'tFpp{'], is it niy fd! quirciapidtyon aioum

llll|n. ilffittona

of tcritoq thelattc bcingPard@lirlycficctivcrc vcryhigh

l0o

swlrcHGEAR PRINCIPLSS

?- The instsnt of re;striking. From the point of view of voltagc guing amplitude the worst possibteinstarts rperechosentd showthc iL of the sitgation- It shouldbe rememberedthag after the ".*ilifi of the capacitivesurrent at a, the dielectric strength of theil."k;|f i" ii"n r"ing dt i rate which-ariy be comparablewith the rate of """ of the ioltage stressingof tbe gap after a. The variation of ilE* attcr zctrrrent zero is subjec to.the influene s; di"f"cttic *t "gtn nafi''e an4 under thesecondition', a teof a nildom |i'"t"-"no &it" l" posible at rarious times after the normal interruption, and not necessarilyone half-cyde latet. Notwithstanding ttree pracical limitetions, howbver, serious catrocc,it in th" mannershownwhenbreaking.caplgtive overvoltages currents,;d the vduesobtainedmay amoustto asmuch as3 Zrpon tL" tir," or on the breaker. The solecauseof this qpe of overvoltagg G it noted,is the inability of the circuit-breakerto provide adbquate dielectricstrengthin the contastgap after intemrpiion' In the simplifiedexaurplejust given,fig. 5.Il' the voltageacross the breaker is'shown as zero u'hen the capacitive.:urrent is extincuishcdandtlrcnnsesrclatirrly slowly,i.e.attlresupplyfrequency. A irore practicalexamplewould indude ttre reactancebetweenthe generatoi qnd the breaker; the lead.ingcapacitivecurrent qaus€se ioltage rui in this reactance.At the zerosof the -capacitivecu-rr-ent the vilage onthe supplysideof the breakerthereforetendsto fall to the geneiatorvalue and,ln doing so' sets up a transient re'striking voltige oscillation- T'his trairsient appearsfor the samebasicrerlsorts asrviih the normalintemrtltion of short-circuit iaductive current but' . whcreasthe latter transientis bised on the pekof the firll recovery voltage,the capacitivecurrent re-striking transientis a funstjon of the smallir peak vollage a6o$s the serie$reactance..Fol this 1asol capacitivecurrent G rebtively.basyto internrpt Paradoxicallys'rfh goodperformanceis undesirablgfor h raeansthat dre corditions for i"-"ttifir,g (and oven'ohages)tend to arisewhile the contact gap io sti[ shoft-an4 therefore,the srore likely to allow re-striking. Siuri' larln a fast rate of rise of re-srtrikingvoltageis'actually.desirableat thiJ stage. Preferablythe breakershould clearthe capryfive current orily w[en a gap tength has be€n redhrd wirich will prevent sub re-striking. sequent The .'Id€a!" Gircuit-Breaker. It is hdpful to have in mind the conceptionof the ideal, or perfect,breakerfrom the intemrpdon ihough this nay b!' io point of viem. Impossibleof ac.biervement definition is simple: tlu ?qfcc, t*dtit-bleafu ofrctsseto in?cd&re

I'

i l.Li] fi P

P R I N C T P T . E S( , F C I R C U T T - B R E A K I i I C

(

101

&fna intantption and infnite impedancc thncaltet For an a-c. circuit-brcakerwe_ma-!relaborate upon the dcftnition somewhatby lpcifying our perfect breakeras one that offers zero impedanceuntii flcfirst curremtzero after contacrseparation,at rrlrich instant infinitc impdance !s ofrered, Coruider the implications of such nerfection. Zcro impedance before thc first current zero meerut: (a) no arc voltaggq) ry getg)' releaeedwithin the breakcr, a featurc greetly o bc dcsircd (c) no distortion of the current wavc. (d) no currerrt chopping.The instananeouschangefrom zero to infinite impedanC n the 6nt qrrrent zao would ensure,(a) invadable clearenccat tte 6nt zcro and (6) impcsibility of re-siriking, and conscqucntly no orcrvoltages. It is interestingto notethat in practicea circuit-breakerdeaigrred to hsvea powerfrrl de-ionizing force to obviate re-stri&ing is, for ttat vcryreason,mor€likdy to producecurrent chopping. This inustrates . ocll &c need for a complete-changeof condltions at cu'rrent zero. Thc neard precticat e2proach to-the ideal breaker woutd be one crpatlcof applp.g very powerful de-iouizing fore instantaneously " after contact scparition- Such co-ordiaatiot a the fitst currGntzero baween thc ncchanical and electrical aspects of circuit-breakcr opntion b undoubtedty bcyond rcolization at presenl The ac. b, in eomcforms ar lang nearerthe idesl thgn thc d"c. typc _ brca&,cr . rhich, by io very sstion in mbing the rclagc of ttrc erc to achieit dnctioq.is a looi way indeed from thc of ao arc roltrge prior to intcrruption rl.C Gtrcuit-breeecr Rstinp fucahing Capacity. \ige hrvc rccnthet thc r.m.r. vdue of Lrrrent rcsulting frorh a shbrt-cirodt nay trry epprccieblyrrith tine oa aeount of dccrcncnt and.erymmgtry, nd tturta cirerit-trealcr ie q*liad with having broken tli vt"c ii cxirting d thG cottracr spparation imr.at This is-drrya ry ldcd ssen r.m.s. vrluc eod ie cr[cd 6e brcnling crncat. If thc we;o . bailt aqynmcrricalccoatect acparrtion it ir nJorrcrnually caneatm Ft,rctricol bcahdrgqra& braHng cEtat iB tlN T\c sywical ltm-givco to ths r.o.r. veftreof tbe as (tc. symadcat) c@ (afiS:22) of thc sByGat@ntee ecparition Ttg paformane of g brcelrcr in dcating with a given bru&in6 clltrent'elsiidcpcnda ee wc hgvc soen,upon the voltage anilable tC tscrikc the arc gt caeh eutlsrt zcrb. Now thie particnlar reluc nqontyuponthcr.n-r. trluo olvoltage availablc but upol :tqrdr rr:-Polcr fs6or of rhe feultcd cirbrG,.-For the prceentpurpcc h il ebdorutybe* o rrsunc zeo powc fector rinci this gi;'c.ihc tld

- ar:

5 ro2

gwtrcrrcEAR pRrNcrpLEB

scvenebTaFng-duty. The arailable volage is subjcct gsteraly to eomevariation duririg short-circuig and the am-e. valuJof intric* regardiqgtbc U le,,ttlg gnfty is t.hatof the voltagq which "ccotlc?tt ig that acoss the circuit-breaker contasrs iineaLtay fi-l cxtiastion oftf,e arcs. The pcfoloance of a circrrh-bieafcr"ita oo circuit test is assessed "hiott by t tringaccountof varioushitori in iilitidon to the breaki4gcurr€ut aad recoverlrvolagg but theselmcr are t*gro quanthiesof imnediate coqcernin the considerdion ofe bttatd fqr a-giveofadt le'eL -The bre&a copacitl of a citodt-breaka,is, thereforg,the current that it can brea&-ata-statedruoov€rJr*rtrd aad thisagzia is-genaally expressedmore coopletely of ft -the in 6s synnctr;c( aot-d. osytmnetricalcapacities,i.e. ani {1--etfot asyryet$cal-b.reakingcurrents respectively vOliage_ "t " "t"[ed is s&€ralrt The breakingapacity rctittggivento a circuit-bftaker one of a set of standardsand is usudiy txpressedin ter-"-or ltti 1nd kV. Thuc tbe breakiry gpacity rating of a single_phasedA;. -is g!l'en, icraccordar.cevath British coo"ritior,, as iolious: Ratedbrlking cafalg @\r4) - rated.s1,mmefiicalbrealcing current ([A) x rated servicevolage (kV). F-o;a 3-piase brcakerthe raled x{vA is similarly obtainedbynultiplyrag the product of rated breaking current ,'oiug" I ""a "o"i"" €;.flb fago-rappgarshere,as.inany3-phase circuig whea$wer, or VA, is calculatedia tc.rmsof line *.rer,t and voltage 56;"" ,f"! by_meansof suitabletests it has been pro""t.fut F"3 3-phzsc.break." can iovariably interrupt "-fio"i .ym-t ic"l.SrdUnf current of 13.5kA r.m.s, with a tesr recovery " voluge of fZ tl..-ii s)ometrical breakingcapacltyis, therefore:

13-5(kA) x 12(kv) x 6

_ 280.5(Mva)

The uearesr*andard ratilg lelow this is 250 MVA at lI kV, wtri.S a syurnetilcal breakiogc.urr€rt of 13.1 irA, ;J$ii :^ITp"l * p r:ung mrgEttrerefore be assignedto the bresker. rn facg aswe shen

cTamgtrertestreqrdents nust usuallytefut:fjl $1fT_:t uued,'etor-e ary snchratiag can be giVur.

The abovestate'meat of c".pagty nting rs a British g.S. one, taken from the ,"I*TC tfO: otl.otTit-breaken, and,.it will be noted, is in terrnsof -.-:.11t. s]Enmetrical brq,king current. Noir since tbe lattcr may onlv be t$c of a short-circuitcurrentwave,..Au ?c- con.Itonent that a-breakerratedat, for instancszSOftAdfl "tut.fo*Jt_;f*. il kV;;"# ersa8irn aqfonetrical of Otd r.n.s. valuegreat€frr."n.l3.l

..

(

- a.-

1l

FRTNCTPLES oF CrtCUlT-D'RBtKtNo

103

tA. This is indeedso, and the mattsr ie taten carl of in the proying testswhich ensurethaq before beiag allotted the ratiog, the Lreakei caninternrpt currents having tAe rated symmetricalvalue together withat leastasrnqch asymm€tryas might occur in senrice. In Americait is the practiceto givecircrrit-breakersasynmetrical brcakingcapacityratingr. That is, the rating is in terms of aq";mand the servie vottage; any particular rnetricalbreaking breakerwould thus be given a higher rating in America than in Britain The rdative merits of the two methodsof rating have'been the subja t of no litdc discussionand argument, for the matter is fairly complex. An outline of thie questionis, however,included in ' ChapterX. A(ak'ng Capaqity. The possibility of a circuit-breaker completinga gull short circuit on being clgsedmust be taken accountof, andit nrustthereforebe testedandproved ig this respegt. la mahing clt acity is thereby-assessgd a1d a coirespondingrating gir.en-. Now thecapacityof a circuit-breaker te "make" cutrents dependsvery much_upon its ability to withstand, and to closeeucceisfuilyagainsi, the effectsof electromagneticforces. The maximura force in'uny plurseis a function of the square of the maximum instantaneou's currentoccurringin that phaseon dosing. .We thdrbforg-findthat nakingcapacityis statedin tcrms of apeakvalueof currcnt insteaclof en r.m.s. value. The making cunent is thus the pedk value of the yalmu1 clrrent loop, induding d.c.ncomponent,in any pha.se duringthe first cycle of current when the circuit-breaker is ciosed. Themakingcapacityis, accordingly,the making current it can make *1 -qrty instantaneouslyat the rated senrice voltage. These definitions,from B.S. 116:1952,are oonaernedonly wiih the first cycleof current on closing the breakbr. This is clear'when it ie . rcmembered that the maximum peak currem possibleoccurs in the **r" in any phaseof Ito-"yA-" only, wheir marirnum aslnmmetrJr tbebreaker. The miaking of a breaker intended for a given faultlevel must therefor"-ne at leastequalto the first peaf of aTuiy aslmmetricalcurrent wavewhoseas. comp?tcnthas.anr.nus. valui cqualto the grmrerrfraf fauh lwel To calculatcthi" pon, we there'fore multipty the symmetricalbreakingcurrent by {irceonverr thir t19mr,yt.s.to peak,and thcn by 1.8to indude tte .. doubling effgct,' or maximumasymmery, thb total multiplicrtion factor being 2.55. qrnnot be lessenedby llottcc tlut the closing duty o! the breaks $rrtm decrementand decai of the d.c. cornponentes rray occur i; tlc breakingduty.

a

tOl

swrrcncsAn PRrNctpLEs

Fault occrrrrd oftca iavolvor thc'flor fffrl/.-era Mg. of hcavy qrrrcut tlrtough a sircuit-brca&cr not rcquirpd to opco. This neans thst circuil-breakcrsgenerallyshorld bc ablc to carrv high crrrcnt eafdy frr eomespecifiedshort pcriod whilo remaiainr cJccdi that rs, thcy shorrld hsve a prov€o elu*erc airy. N ehott-time ctrlcnt, which is $ocrelly not tessthan rhe sydtctricet breaAingcurrerit, is uqgly requfrcdto be carriedfor'g f;iod of uo to 3 eernnd+acordingtg thc protecive schemcrcquir;co6. Td short-frtrrereting wS+ go F gr"* depcndsupon thc ability of thc breakerto witbstand(a) the electronegneticforce erficct, end (r) th tcmpcraturerise. Ihe greatp4dectrooagaaiclorcc billosedfirr is the makingcapacitymting, bot th" efiees of szsrairadvibrrrtions'ein be:rwealed only-in-a test for ghoit-time rating. The questioaof temperatlrrerise is alsovery important, asthis riie may bc sdde4 h service,to tbat alreadyexisdngdue to frrll load currqit Nonnal Ct reTt Rdilng. This is the curr.,ntshic& caobe.csnicd continuouslyby a circuit-breakea The only limiatioa in t}ir caocia the Gmperaturerise of the current-carryingpriits. A circuit-brenter is genbrallyrderred to tnief,y in terns of ir ![9 and $ngs, and a typical examplemight bei.i,500 A 11000 MVA, 33 kV, 3-seon4 $phase oil cirenit=bieaka',. Fromdris descriptionwe have: Ratednormal curr-errt-1,5(X)A r.m.e.

Rate'tsymmaricafbroki"g;;;--

t,m(gr

r' l-

x 6l I

Rated making currcnt: li:; ?;.'S'"

_ 44.62bt(I,cee)

Short-time rating-l?.5 kl[ r.n L for 3 eeonds. Itsted servicevottage-33 kV r.m-s. (tne voltage[

,!