Oper­a­tion, re­li­a­bil­ity and main­te­nance of LCI VSD syn­chro­nous elec­tric mo­tors

DEMM Engineering & Manufacturing - - MOTORS & DRIVES - By AMIN ALMASI


LCI (Load Com­mu­tated In­verter) type VSD (vari­able speed drive) con­verter sys­tems are used for syn­chro­nous elec­tric mo­tors.LCI type thyrist-or-fre­quency con­vert­ers are among the most sim­ple, cost ef­fec­tive and re­li­able power elec­tronic sys­tems on VSD mar­ket for vari­able-speed elec­tric mo­tor driv­ers par­tic­u­larly for medium and large size elec­tric mo­tors. These are among the best vari­able speed driv­ers for in­dus­trial and man­u­fac­tur­ing ma­chiner­ies. Oper­a­tion, re­li­a­bil­ity, and main­te­nance of LCI VSD syn­chro­nous elec­tric mo­tors are dis­cussed.


The LCI VSD con­verter sys­tem is a proven tech­nol­ogy with suc­cess­ful ref­er­ences. In iden­ti­cal ba­sic con­fig­u­ra­tion, and with sim­i­lar power elec­tronic de­vices, LCI con­vert­ers are em­ployed by power gen­er­a­tion in­dus­try in wide range of power rat­ings in high-volt­age DC (HVDC) trans­mis­sion sys­tems. Although, there are some dif­fer­ences be­tween VSD con­vert­ers and HVDC trans­mis­sion sys­tems; for in­stance, the DC link re­ac­tors. To avoid costly and space con­sum­ing ex­ter­nally-in­stalled air­cooled so­lu­tions, the di­rectly wa­ter­cooled and iron-cored re­ac­tor of smaller con­vert­ers (sim­i­lar tech­nol­ogy) should be up­rated to the DC link volt­age of high volt­age de­signs. Com­pre­hen­sive type tests con­firmed the suitabil­ity of such de­sign for a wide range of ap­pli­ca­tions in­clud­ing rel­a­tively high rated ap­pli­ca­tion.

How­ever, there are some is­sues and chal­lenges with LCI VSD sys­tems. Load com­mu­tated in­verter (LCI) drives are the source of volt­age and cur­rent dis­tor­tion (known as har­mon­ics); this is­sue re­quires a spe­cial at­ten­tion. There­fore, in most ap­pli­ca­tions of LCI VSD sys­tems, har­monic fil­ters are needed.

As in­di­ca­tions, the power fac­tor of a plant load should usu­ally be main­tained be­tween 0.85 (lag­ging) and 1.0 dur­ing steady-state oper­a­tion sce­nar­ios and un­der all load and op­er­a­tional con­di­tions. How­ever, there are some chal­lenges. Some­times, there are some changes in elec­tri­cal power sup­ply sys­tem, such as trans­mis­sion sys­tems, after the place­ment of pur­chase order of main ro­tat­ing ma­chine trains in­clud­ing main VSD sys­tems. Too of­ten, the ac­tual elec­tri­cal grid of a plant is rel­a­tively weaker com­pared to the­o­ret­i­cal grid in the de­sign stage. This could be due to many rea­sons such as re­vised power sup­ply ar­range­ment, some mod­i­fi­ca­tions, etc. This fea­ture can have some im­pacts on the size of the har­monic fil­ter pack­age. For ex­am­ple, in an in­dus­trial plant, an ini­tial above­ground 8km power trans­mis­sion line was changed to a 14km un­der­ground ca­ble. This new power trans­mis­sion ar­range­ment caused many changes in VSD LCI sys­tems and their har­monic fil­ter con­fig­u­ra­tions.


A thyris­tor is a solid-state semi­con­duc­tor de­vice with four lay­ers of al­ter­nat­ing N-type and P-type ma­te­rial (or ‘sil­i­con con­trolled rec­ti­fier’). A thyris­tor acts as a bi-sta­ble switch, con­duct­ing when its gate re­ceives a cur­rent trig­ger, and con­tin­ues to con­duct while it is for­ward bi­ased, while the volt­age across the de­vice is not re­versed. In other words, the thyris­tor is a four-layer, three ter­mi­nal semi­con­duct­ing de­vices, with each layer con­sist­ing of al­ter­nately N-type or P-type ma­te­rial (for ex­am­ple, P-N-P-N). The main ter­mi­nals, la­belled an­ode and cath­ode, are across the full four lay­ers. The con­trol ter­mi­nal (called the gate) is at­tached to P-type ma­te­rial near to the cath­ode; this is a vari­ant called an SCS (Sil­i­con Con­trolled Switch) which brings all four lay­ers out to ter­mi­nals. Since mod­ern thyris­tors can switch power on a wide scale power rat­ings (say be­low kW to multi MW), thyris­tor valves have be­come the heart of VSD con­verter sys­tems or high-volt­age di­rect cur­rent (HVDC) con­ver­sion sys­tems ei­ther to or from al­ter­nat­ing cur­rent.


Re­li­a­bil­ity is a ma­jor con­sid­er­a­tion for VSD sys­tems. In ad­di­tion to the usual fail­ure modes be­cause of ex­ceed­ing volt­age, cur­rent or power rat­ings, thyris­tors have their own par­tic­u­lar modes of fail­ure, in­clud­ing: 1. Turn on “di/dt”: In this fail­ure mode the rate of rise of on-state cur­rent after trig­ger­ing is higher than that can be sup­ported by the spread­ing speed of the ac­tive con­duc­tion area. 2. Forced com­mu­ta­tion: In this mode, the tran­sient peak re­verse re­cov­ery cur­rent causes such a high volt­age drop in the sub-cath­ode re­gion that it ex­ceeds the re­verse break­down volt­age of the gate cath­ode diode junc­tion. 3. Switch on “dv/dt”: In this mode the thyris­tor can be spu­ri­ously fired with­out trig­ger from the gate if the rate of rise of volt­age an­ode to cath­ode is too high. Some­times, plant spec­i­fi­ca­tions spec­ify “n+1” re­dun­dancy for a thyris­tor. How­ever, VSD sys­tem man­u­fac­tur­ers usu­ally ar­gue that mod­ern power thyris­tors are no more likely to fail than any other com­po­nent in the power cir­cuit which of­ten used with­out spare. In other words, man­u­fac­tur­ers usu­ally ad­vise a mea­sur­able in­crease in re­li­a­bil­ity can­not be ob­tained by this ex­pen­sive scheme of “n+1” re­dun­dancy. Op­er­at­ing ex­pe­ri­ences over 2–3 decades have showed that the sin­gle-thyris­tor fail­ure is a rare ex­cep­tion. Op­ti­cal fir­ing and check­back sig­nals to and from the thyris­tor is han­dled by fast mi­cro­pro­ces­sor equip­ment that al­lows both a se­rial data com­mu­ni­ca­tion with the dis­tant con­trol and mon­i­tor­ing cu­bi­cles, and the ex­act iden­ti­fi­ca­tion of a faulty thyris­tor. Of­ten “n+1” re­dun­dancy for thyris­tor is not im­ple­mented. How­ever, this “n+1” scheme may be used for spe­cial VSD units with very high ex­pected re­li­a­bil­ity.

To pre­vent im­per­mis­si­ble “di/dt” val­ues from en­dan­ger­ing a thyris­tor, ca­ble ca­pac­i­ties to the elec­tric mo­tor and to con­verter trans­former are usu­ally com­pen­sated by toroidal cores slipped onto in­ter­nal water-cooled bus­bars in the con­verter whereby pre­vi­ously used fer­rite “donuts” are re­placed by more ef­fi­cient wound amor­phous metal coils.

Com­bined heat losses of a vari­able-speed drive ( VSD) sys­tem, in­clud­ing the elec­tric mo­tor and fre­quency con­verter, are usu­ally trans­ferred by two sep­a­rate closed-loop water-cooled sys­tems with cool­ing banks. Ex­cept for the fre­quency con­vert­ers with their in­te­grated DC link re­ac­tors; the equip­ment is di­rectly cooled by water. The high volt­age po­ten­tial in­side the fre­quency con­verter re­quires a con­trolled clean­li­ness (con­duc­tiv­ity) of the cool­ing medium; this man­dates an­other closed-loop deion­ized water sys­tem which is linked to the pri­mary cool­ing sys­tem via heat ex­chang­ers. Of­ten, a so­phis­ti­cated, mod­ern com­pact plate-type heat-ex­changer is em­ployed to save space, ob­tain op­ti­mum costs and of­fer a good over­all re­li­a­bil­ity. To main­tain the avail­abil­ity of these cool­ing sys­tems, their cir­cu­lat­ing pumps are re­dun­dant with au­to­matic change-over.

The LCI sys­tem usu­ally em­ploys read­ily avail­able disk-type thyris­tors (or sil­i­con con­trolled rec­ti­fiers, SCRs) as solid state power switch­ing el­e­ments. They are usu­ally mounted in stan­dard­ised equip­ment cu­bi­cles for in­stal­la­tion at “In­door”, or in pur­pose-built power cen­tre mod­ules at “Out­door”.


For many VSD sys­tems, to match the drive’s in­put volt­age to the plant’s power line volt­age, a trans­former is usu­ally re­quired for the con­verter sys­tem. Great care should be taken for the siz­ing, se­lec­tion, re­li­a­bil­ity,

acous­tic in­su­la­tion, and fab­ri­ca­tion de­tails of such transformers.

Gen­er­ally, more care should be taken for transformers used in VSD con­verter sys­tems rather than other transformers. Par­tic­u­larly, siz­ing, in­su­la­tion, ther­mal de­sign, and bi-con­cen­tric winding re­quire spe­cial at­ten­tion. A less known, but crit­i­cal is­sue could be noise gen­er­a­tion. Spe­cial mea­sures should be taken to acous­ti­cally iso­late the coil and core assem­bly from the oil tank to re­duce struc­ture-born noises, and to in­crease the im­ped­ance of the sec­ondary wind­ings as re­quired lim­it­ing the pos­si­ble short cir­cuit cur­rent in the fuse-less thyris­tor fre­quency con­verter.

The volt­age drop at plant’s high volt­age dis­tri­bu­tion sys­tem dur­ing switch­ing is usu­ally of con­cern and such a volt­age drop should be lim­ited to cer­tain val­ues (say 2–3 per­cent). Se­ries-con­nected in­rush lim­it­ing re­sis­tors should thus be in­stalled on the high volt­age side. These are mo­men­tar­ily switched in by a sep­a­rate cir­cuit breaker for a cer­tain time pe­riod (say one sec­ond) dur­ing trans­former mag­neti­sa­tion.

Ther­mal con­sid­er­a­tions need spe­cial at­ten­tion as many transformers and gen­er­ally VSD sys­tems were failed or ex­pe­ri­enced op­er­a­tional and re­li­a­bil­ity is­sues be­cause of poor ther­mal man­age­ment. In a case study for a VSD elec­tric mo­tor, a high volt­age pri­mary trans­former was ex­pe­ri­enced sev­eral break­downs. The anal­y­sis of oil from the trans­former has demon­strated a pre­ma­ture oil age­ing be­cause of ther­mal is­sues due to im­proper ini­tial siz­ing. The heat siz­ing of the equip­ment did not in­clude an ad­di­tional heat­ing caused by har­monic cur­rents. There­fore, as a les­son learned, firstly all ef­fects such as har­monic ef­fects, etc. should be con­sid­ered and am­ple mar­gins should be ap­plied for the siz­ing of transformers with spe­cial at­ten­tion to ther­mal loads. As an in­di­ca­tion 20–35 per­cent mar­gins should be con­sid­ered on the ther­mal man­age­ment. Many suc­cess­fully op­er­ated transformers were de­signed with proper siz­ing mar­gins and have been op­er­ated most of the time be­low their rated ca­pac­i­ties (say be­low 85–90 per­cent); all these re­sulted in long trou­ble-free oper­a­tion time.


The nor­mal start pro­ce­dure of a LCI VSD is: • The VSD sys­tem is en­er­gised. • The cir­cuit breaker trans­former is

switched on. • The har­monic fil­ter gets the re­lease

from the VSD con­trol. • The cir­cuit breaker fil­ter is switched on. • The re­lease is sent to the VSD con­trol. • The elec­tric mo­tor starts the oper­a­tion.

The nor­mal stop pro­ce­dure is the in­verse se­quence of the above­men­tioned. Re­gard­ing the reg­u­la­tion of sup­ply­ing elec­tric grid of a plant, it is use­ful to switch-on one VSD after the other. The switch­ing oper­a­tion of har­monic fil­ter cir­cuits and the start-up of VSDs have some short time tran­sient ef­fects and also some static volt­age chang­ing ef­fects on the elec­tri­cal grid. All these should be eval­u­ated us­ing proper anal­y­sis and stud­ies.


LCI (Load Com­mu­tated In­verter) VSD con­vert­ers have been em­ployed for many elec­tric mo­tor driv­ers in in­dus­trial and man­u­fac­tur­ing plants. Har­monic fil­ters are im­por­tant parts of any LCI VSD sys­tem; a ma­jor is­sue is the ef­fects of the grid (plant, etc.) and par­tic­u­larly other ma­jor loads on the oper­a­tion and per­for­mance of these har­monic fil­ters. The loads and char­ac­ter­is­tics of the elec­tri­cal net­work of the plant which were known dur­ing the order of har­monic fil­ters should have been con­sid­ered on the fil­ter de­sign and its oper­a­tion. A ma­jor prob­lem is new loads in the plant; these are loads not de­fined at that time or changes in the plant after the order of ma­jor VSD sys­tems.

In many plants, gen­er­ated har­mon­ics are ex­ces­sive some­times even dur­ing low fault level con­di­tions. VSDs sys­tem op­er­at­ing pa­ram­e­ters might change from the orig­i­nal rated or de­sign ones, or the plant’s elec­tri­cal net­work might be mod­i­fied such as new con­sumers, etc.; there­fore, har­monic fil­ters are not as ef­fec­tive as they should be and con­se­quently har­mon­ics ex­ceed­ing the spec­i­fied lim­its. Con­sid­er­a­tions of pos­si­ble fu­ture ex­pan­sions, pos­si­ble degra­da­tions and risk mit­i­ga­tions are im­por­tant for re­li­a­bil­ity and long-term suc­cess­ful oper­a­tion of VSDs.

All res­o­nance cases as­so­ci­ated to op­er­at­ing modes should be con­sid­ered. This is not only about the plant; some­times, a har­monic res­o­nance case tends to oc­cur when one of the power gen­er­a­tors is off-line in the power source of the main elec­tri­cal grid.

This can change the tun­ing of the elec­tric sys­tem of the plant so that a res­o­nance could oc­cur near a spe­cific har­monic.

Rel­a­tively long length and large sizes of some ca­bles feed­ing to dif­fer­ent loads in var­i­ous units can in­crease their ef­fects on VSD sys­tems and over­all net­work char­ac­ter­is­tics. The re­sult could be a sig­nif­i­cant im­pact on nat­u­ral fre­quen­cies of the over­all elec­tri­cal sys­tem of such a plant. An­other ob­ser­va­tion is all pos­si­ble oper­a­tion modes should be con­sid­ered for all these loads. For “N” ca­bles that feed elec­tric pow­ers to “N” dif­fer­ent loads, which can be con­nected to the same bus as the VSD (s), there are “2N” com­bi­na­tions of the dif­fer­ent ca­bles (in­clud­ing the case of no ca­bles con­nected) which should be in­ves­ti­gated for the re­li­a­bil­ity and safety of the over­all elec­tri­cal sys­tem of a plant. For ex­am­ple, in case of N=5 (five ca­bles), 32 dif­fer­ent com­bi­na­tions should be con­sid­ered. For a very com­plex sys­tem, be­cause there are a huge num­ber of pos­si­ble sys­tem con­fig­u­ra­tions (for in­stance, for N=11: 2048 com­bi­na­tions), it is dif­fi­cult to de­ter­mine the lev­els of har­monic dis­tor­tion for the var­i­ous con­tin­gent con­di­tions. Con­sid­er­ing for some large plants, “N” is more than 20 or 25, and a print­out of all of har­monic re­sults for all pos­si­ble com­bi­na­tions would make such a har­monic study re­port thou­sands of pages long. Ob­vi­ously, it is im­prac­ti­cal to look at each of these cases in­di­vid­u­ally be­cause there are too many. How­ever, it is pos­si­ble to au­to­mat­i­cally scan through all cases to de­ter­mine worst case(s). Some­times, it may be de­cided to re­port only crit­i­cal cases. For ex­am­ple, crit­i­cal cases could be topolo­gies that have the mid-volt­age bus ties open, or if closed hav­ing one trans­former out of ser­vice.

Of­ten a fil­ter arm, known as ca­ble damp­ing fil­ter arm, is rec­om­mended to be added to the fil­ter pack­age to deal with these ca­ble load ef­fects in a plant. The per­for­mance of a fil­ter at lower order har­monic fre­quen­cies is usu­ally sim­i­lar for cases with or with­out ca­ble damp­ing fil­ter arm. How­ever, a ca­ble damper fil­ter arm can re­duce the depth of res­o­nance val­leys for higher fre­quen­cies. Too of­ten such a ca­ble damper fil­ter arm pro­posed by VSD sys­tem man­u­fac­tur­ers. Com­plete and ac­cu­rate ca­ble trans­mis­sion net­work is not of­ten con­sid­ered by VSD man­u­fac­tur­ers; there­fore, such a damper fil­ter arm is rec­om­mended as mit­i­ga­tion. These damper fil­ter arms are use­ful in some cases; but in many other cases, they just of­fer some slight ben­e­fi­cial ef­fects. In other words, in many cases, this man­u­fac­turer pro­posed har­monic fil­ter can­not have sen­si­ble ef­fect on the re­duc­tion of har­monic val­ues. A good ad­vice could be main­tain­ing damper high har­monic fil­ter arm, but the sys­tem logic should be spec­i­fied to con­tinue oper­a­tion even if this high-har­monic fil­ter arm stops work­ing.

Mod­i­fi­ca­tions could con­tinue to oc­cur in a plant, for in­stance, new con­sumers may be added, some oth­ers may be dis­con­nected or there might be changes in elec­tri­cal con­fig­u­ra­tions. Such mod­i­fi­ca­tions could change ex­pe­ri­enced har­monic at the elec­tric net­work or a VSD sys­tem; sub­se­quently both har­mon­ics on grid and the oper­a­tion of ma­chiner­ies could be af­fected. Ac­cord­ingly, a risk mit­i­ga­tion strat­egy is re­quired to cater for pos­si­ble changes or pos­si­ble fu­ture ex­pan­sions. Har­monic cur­rents in each of har­monic fil­ter arms should be mon­i­tored via com­mu­ni­ca­tion links. This in­for­ma­tion could be used to mon­i­tor the har­monic sit­u­a­tion and flag if ad­di­tional har­monic fil­ter­ing is re­quired. Ad­di­tional har­monic fil­ters may be in­stalled if har­monic lev­els are ex­ceeded the spec­i­fied lim­its (or val­ues rec­om­mended by en­gi­neer­ing prac­tices) for safe and re­li­able oper­a­tion.

Amin Almasi is a lead me­chan­i­cal en­gi­neer in Aus­tralia. He is char­tered pro­fes­sional en­gi­neer of En­gi­neers Aus­tralia ( MIEAust CPEng – Me­chan­i­cal) and IMechE (CEng MIMechE) in ad­di­tion to a M.Sc. and B.Sc. in me­chan­i­cal en­gi­neer­ing and RPEQ (Reg­is­tered Pro­fes­sional En­gi­neer in Queens­land). He spe­cialises in me­chan­i­cal equip­ment and ma­chiner­ies in­clud­ing cen­trifu­gal, screw and re­cip­ro­cat­ing com­pres­sors, gas tur­bines, steam tur­bines, en­gines, pumps, con­di­tion mon­i­tor­ing, re­li­a­bil­ity, as well as fire pro­tec­tion, power gen­er­a­tion, water treat­ment, ma­te­rial han­dling and oth­ers. Almasi is an ac­tive mem­ber of En­gi­neers Aus­tralia, IMechE, ASME, and SPE. He has au­thored more than 150 pa­pers and ar­ti­cles deal­ing with ro­tat­ing equip­ment, con­di­tion mon­i­tor­ing, fire pro­tec­tion, power gen­er­a­tion, water treat­ment, ma­te­rial han­dling and re­li­a­bil­ity.

Newspapers in English

Newspapers from New Zealand

© PressReader. All rights reserved.